diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9492 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1350,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0022222222222222222,
+      "grad_norm": 75361.5234375,
+      "learning_rate": 0.0,
+      "loss": 0.6475,
+      "step": 1
+    },
+    {
+      "epoch": 0.0044444444444444444,
+      "grad_norm": 979666.625,
+      "learning_rate": 3.703703703703704e-07,
+      "loss": 0.6324,
+      "step": 2
+    },
+    {
+      "epoch": 0.006666666666666667,
+      "grad_norm": 211083.375,
+      "learning_rate": 7.407407407407408e-07,
+      "loss": 0.6486,
+      "step": 3
+    },
+    {
+      "epoch": 0.008888888888888889,
+      "grad_norm": 398850.3125,
+      "learning_rate": 1.1111111111111112e-06,
+      "loss": 0.811,
+      "step": 4
+    },
+    {
+      "epoch": 0.011111111111111112,
+      "grad_norm": 660421.375,
+      "learning_rate": 1.4814814814814817e-06,
+      "loss": 0.6477,
+      "step": 5
+    },
+    {
+      "epoch": 0.013333333333333334,
+      "grad_norm": 1308591.75,
+      "learning_rate": 1.8518518518518519e-06,
+      "loss": 0.6637,
+      "step": 6
+    },
+    {
+      "epoch": 0.015555555555555555,
+      "grad_norm": 870002.25,
+      "learning_rate": 2.2222222222222225e-06,
+      "loss": 0.7728,
+      "step": 7
+    },
+    {
+      "epoch": 0.017777777777777778,
+      "grad_norm": 491811.40625,
+      "learning_rate": 2.5925925925925925e-06,
+      "loss": 0.7462,
+      "step": 8
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 675543.8125,
+      "learning_rate": 2.9629629629629633e-06,
+      "loss": 0.7465,
+      "step": 9
+    },
+    {
+      "epoch": 0.022222222222222223,
+      "grad_norm": 166720.375,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.799,
+      "step": 10
+    },
+    {
+      "epoch": 0.024444444444444446,
+      "grad_norm": 930302.5,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 0.8028,
+      "step": 11
+    },
+    {
+      "epoch": 0.02666666666666667,
+      "grad_norm": 1149363.875,
+      "learning_rate": 4.074074074074075e-06,
+      "loss": 0.7713,
+      "step": 12
+    },
+    {
+      "epoch": 0.028888888888888888,
+      "grad_norm": 470885.25,
+      "learning_rate": 4.444444444444445e-06,
+      "loss": 0.6513,
+      "step": 13
+    },
+    {
+      "epoch": 0.03111111111111111,
+      "grad_norm": 520825.34375,
+      "learning_rate": 4.814814814814815e-06,
+      "loss": 0.4761,
+      "step": 14
+    },
+    {
+      "epoch": 0.03333333333333333,
+      "grad_norm": 511413.9375,
+      "learning_rate": 5.185185185185185e-06,
+      "loss": 0.6312,
+      "step": 15
+    },
+    {
+      "epoch": 0.035555555555555556,
+      "grad_norm": 617626.375,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 0.7008,
+      "step": 16
+    },
+    {
+      "epoch": 0.03777777777777778,
+      "grad_norm": 175768.0625,
+      "learning_rate": 5.925925925925927e-06,
+      "loss": 0.9021,
+      "step": 17
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 541401.5625,
+      "learning_rate": 6.296296296296296e-06,
+      "loss": 0.681,
+      "step": 18
+    },
+    {
+      "epoch": 0.042222222222222223,
+      "grad_norm": 214184.921875,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.8062,
+      "step": 19
+    },
+    {
+      "epoch": 0.044444444444444446,
+      "grad_norm": 569162.3125,
+      "learning_rate": 7.0370370370370375e-06,
+      "loss": 0.7527,
+      "step": 20
+    },
+    {
+      "epoch": 0.04666666666666667,
+      "grad_norm": 640316.3125,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.6391,
+      "step": 21
+    },
+    {
+      "epoch": 0.04888888888888889,
+      "grad_norm": 229886.5625,
+      "learning_rate": 7.777777777777777e-06,
+      "loss": 0.6762,
+      "step": 22
+    },
+    {
+      "epoch": 0.051111111111111114,
+      "grad_norm": 2014340.25,
+      "learning_rate": 8.14814814814815e-06,
+      "loss": 0.5573,
+      "step": 23
+    },
+    {
+      "epoch": 0.05333333333333334,
+      "grad_norm": 680286.5625,
+      "learning_rate": 8.518518518518519e-06,
+      "loss": 0.6578,
+      "step": 24
+    },
+    {
+      "epoch": 0.05555555555555555,
+      "grad_norm": 569364.5,
+      "learning_rate": 8.88888888888889e-06,
+      "loss": 0.777,
+      "step": 25
+    },
+    {
+      "epoch": 0.057777777777777775,
+      "grad_norm": 2732369.25,
+      "learning_rate": 9.259259259259259e-06,
+      "loss": 0.7702,
+      "step": 26
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 519576.375,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 0.7192,
+      "step": 27
+    },
+    {
+      "epoch": 0.06222222222222222,
+      "grad_norm": 206192.546875,
+      "learning_rate": 1e-05,
+      "loss": 0.7174,
+      "step": 28
+    },
+    {
+      "epoch": 0.06444444444444444,
+      "grad_norm": 221324.625,
+      "learning_rate": 1.037037037037037e-05,
+      "loss": 0.6039,
+      "step": 29
+    },
+    {
+      "epoch": 0.06666666666666667,
+      "grad_norm": 1108986.0,
+      "learning_rate": 1.074074074074074e-05,
+      "loss": 0.763,
+      "step": 30
+    },
+    {
+      "epoch": 0.06888888888888889,
+      "grad_norm": 639494.5,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.7044,
+      "step": 31
+    },
+    {
+      "epoch": 0.07111111111111111,
+      "grad_norm": 627301.375,
+      "learning_rate": 1.1481481481481482e-05,
+      "loss": 0.7324,
+      "step": 32
+    },
+    {
+      "epoch": 0.07333333333333333,
+      "grad_norm": 714222.5,
+      "learning_rate": 1.1851851851851853e-05,
+      "loss": 0.6594,
+      "step": 33
+    },
+    {
+      "epoch": 0.07555555555555556,
+      "grad_norm": 639339.6875,
+      "learning_rate": 1.2222222222222222e-05,
+      "loss": 0.6238,
+      "step": 34
+    },
+    {
+      "epoch": 0.07777777777777778,
+      "grad_norm": 231330.515625,
+      "learning_rate": 1.2592592592592592e-05,
+      "loss": 0.6839,
+      "step": 35
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 511528.15625,
+      "learning_rate": 1.2962962962962962e-05,
+      "loss": 0.5042,
+      "step": 36
+    },
+    {
+      "epoch": 0.08222222222222222,
+      "grad_norm": 312614.03125,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.7532,
+      "step": 37
+    },
+    {
+      "epoch": 0.08444444444444445,
+      "grad_norm": 404826.78125,
+      "learning_rate": 1.3703703703703704e-05,
+      "loss": 0.6002,
+      "step": 38
+    },
+    {
+      "epoch": 0.08666666666666667,
+      "grad_norm": 328368.75,
+      "learning_rate": 1.4074074074074075e-05,
+      "loss": 0.7407,
+      "step": 39
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 212831.46875,
+      "learning_rate": 1.4444444444444444e-05,
+      "loss": 0.591,
+      "step": 40
+    },
+    {
+      "epoch": 0.09111111111111111,
+      "grad_norm": 945027.1875,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 0.6917,
+      "step": 41
+    },
+    {
+      "epoch": 0.09333333333333334,
+      "grad_norm": 1313241.375,
+      "learning_rate": 1.5185185185185186e-05,
+      "loss": 0.807,
+      "step": 42
+    },
+    {
+      "epoch": 0.09555555555555556,
+      "grad_norm": 635023.75,
+      "learning_rate": 1.5555555555555555e-05,
+      "loss": 0.8375,
+      "step": 43
+    },
+    {
+      "epoch": 0.09777777777777778,
+      "grad_norm": 2153337.0,
+      "learning_rate": 1.5925925925925926e-05,
+      "loss": 0.8316,
+      "step": 44
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 376420.84375,
+      "learning_rate": 1.62962962962963e-05,
+      "loss": 0.7526,
+      "step": 45
+    },
+    {
+      "epoch": 0.10222222222222223,
+      "grad_norm": 791087.625,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.6498,
+      "step": 46
+    },
+    {
+      "epoch": 0.10444444444444445,
+      "grad_norm": 656814.375,
+      "learning_rate": 1.7037037037037038e-05,
+      "loss": 0.5489,
+      "step": 47
+    },
+    {
+      "epoch": 0.10666666666666667,
+      "grad_norm": 1977482.875,
+      "learning_rate": 1.740740740740741e-05,
+      "loss": 0.6734,
+      "step": 48
+    },
+    {
+      "epoch": 0.10888888888888888,
+      "grad_norm": 744988.0,
+      "learning_rate": 1.777777777777778e-05,
+      "loss": 0.6745,
+      "step": 49
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 1026407.4375,
+      "learning_rate": 1.814814814814815e-05,
+      "loss": 0.7124,
+      "step": 50
+    },
+    {
+      "epoch": 0.11333333333333333,
+      "grad_norm": 2253132.25,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 0.6016,
+      "step": 51
+    },
+    {
+      "epoch": 0.11555555555555555,
+      "grad_norm": 2241903.25,
+      "learning_rate": 1.888888888888889e-05,
+      "loss": 0.5899,
+      "step": 52
+    },
+    {
+      "epoch": 0.11777777777777777,
+      "grad_norm": 838888.8125,
+      "learning_rate": 1.925925925925926e-05,
+      "loss": 0.7287,
+      "step": 53
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 228881.84375,
+      "learning_rate": 1.962962962962963e-05,
+      "loss": 0.5912,
+      "step": 54
+    },
+    {
+      "epoch": 0.12222222222222222,
+      "grad_norm": 1143221.875,
+      "learning_rate": 2e-05,
+      "loss": 0.7826,
+      "step": 55
+    },
+    {
+      "epoch": 0.12444444444444444,
+      "grad_norm": 851040.6875,
+      "learning_rate": 2.037037037037037e-05,
+      "loss": 0.514,
+      "step": 56
+    },
+    {
+      "epoch": 0.12666666666666668,
+      "grad_norm": 638632.0,
+      "learning_rate": 2.074074074074074e-05,
+      "loss": 0.5897,
+      "step": 57
+    },
+    {
+      "epoch": 0.1288888888888889,
+      "grad_norm": 688196.3125,
+      "learning_rate": 2.111111111111111e-05,
+      "loss": 0.634,
+      "step": 58
+    },
+    {
+      "epoch": 0.13111111111111112,
+      "grad_norm": 352046.25,
+      "learning_rate": 2.148148148148148e-05,
+      "loss": 0.5804,
+      "step": 59
+    },
+    {
+      "epoch": 0.13333333333333333,
+      "grad_norm": 272880.15625,
+      "learning_rate": 2.1851851851851852e-05,
+      "loss": 0.7051,
+      "step": 60
+    },
+    {
+      "epoch": 0.13555555555555557,
+      "grad_norm": 686219.4375,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.7302,
+      "step": 61
+    },
+    {
+      "epoch": 0.13777777777777778,
+      "grad_norm": 589322.25,
+      "learning_rate": 2.2592592592592594e-05,
+      "loss": 0.6278,
+      "step": 62
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1640305.5,
+      "learning_rate": 2.2962962962962965e-05,
+      "loss": 0.7666,
+      "step": 63
+    },
+    {
+      "epoch": 0.14222222222222222,
+      "grad_norm": 563189.0625,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 0.6705,
+      "step": 64
+    },
+    {
+      "epoch": 0.14444444444444443,
+      "grad_norm": 1177888.375,
+      "learning_rate": 2.3703703703703707e-05,
+      "loss": 0.6395,
+      "step": 65
+    },
+    {
+      "epoch": 0.14666666666666667,
+      "grad_norm": 355971.625,
+      "learning_rate": 2.4074074074074074e-05,
+      "loss": 0.6872,
+      "step": 66
+    },
+    {
+      "epoch": 0.14888888888888888,
+      "grad_norm": 1110798.625,
+      "learning_rate": 2.4444444444444445e-05,
+      "loss": 0.8045,
+      "step": 67
+    },
+    {
+      "epoch": 0.1511111111111111,
+      "grad_norm": 234297.390625,
+      "learning_rate": 2.4814814814814816e-05,
+      "loss": 0.6402,
+      "step": 68
+    },
+    {
+      "epoch": 0.15333333333333332,
+      "grad_norm": 944397.0,
+      "learning_rate": 2.5185185185185183e-05,
+      "loss": 0.7907,
+      "step": 69
+    },
+    {
+      "epoch": 0.15555555555555556,
+      "grad_norm": 595624.0625,
+      "learning_rate": 2.5555555555555554e-05,
+      "loss": 0.7191,
+      "step": 70
+    },
+    {
+      "epoch": 0.15777777777777777,
+      "grad_norm": 785496.6875,
+      "learning_rate": 2.5925925925925925e-05,
+      "loss": 0.6487,
+      "step": 71
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 731935.3125,
+      "learning_rate": 2.6296296296296296e-05,
+      "loss": 0.6053,
+      "step": 72
+    },
+    {
+      "epoch": 0.1622222222222222,
+      "grad_norm": 369573.5,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.7378,
+      "step": 73
+    },
+    {
+      "epoch": 0.16444444444444445,
+      "grad_norm": 389220.34375,
+      "learning_rate": 2.7037037037037037e-05,
+      "loss": 0.7437,
+      "step": 74
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 1437347.5,
+      "learning_rate": 2.7407407407407408e-05,
+      "loss": 0.5898,
+      "step": 75
+    },
+    {
+      "epoch": 0.1688888888888889,
+      "grad_norm": 282015.8125,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.7979,
+      "step": 76
+    },
+    {
+      "epoch": 0.1711111111111111,
+      "grad_norm": 3742327.5,
+      "learning_rate": 2.814814814814815e-05,
+      "loss": 0.6935,
+      "step": 77
+    },
+    {
+      "epoch": 0.17333333333333334,
+      "grad_norm": 634704.25,
+      "learning_rate": 2.851851851851852e-05,
+      "loss": 0.6088,
+      "step": 78
+    },
+    {
+      "epoch": 0.17555555555555555,
+      "grad_norm": 322738.6875,
+      "learning_rate": 2.8888888888888888e-05,
+      "loss": 0.7765,
+      "step": 79
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 460376.125,
+      "learning_rate": 2.925925925925926e-05,
+      "loss": 0.6523,
+      "step": 80
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 337121.34375,
+      "learning_rate": 2.962962962962963e-05,
+      "loss": 0.6013,
+      "step": 81
+    },
+    {
+      "epoch": 0.18222222222222223,
+      "grad_norm": 380048.375,
+      "learning_rate": 3e-05,
+      "loss": 0.5171,
+      "step": 82
+    },
+    {
+      "epoch": 0.18444444444444444,
+      "grad_norm": 187856.171875,
+      "learning_rate": 3.037037037037037e-05,
+      "loss": 0.6407,
+      "step": 83
+    },
+    {
+      "epoch": 0.18666666666666668,
+      "grad_norm": 424345.5625,
+      "learning_rate": 3.074074074074074e-05,
+      "loss": 0.6888,
+      "step": 84
+    },
+    {
+      "epoch": 0.18888888888888888,
+      "grad_norm": 330755.9375,
+      "learning_rate": 3.111111111111111e-05,
+      "loss": 0.7823,
+      "step": 85
+    },
+    {
+      "epoch": 0.19111111111111112,
+      "grad_norm": 757690.4375,
+      "learning_rate": 3.148148148148148e-05,
+      "loss": 0.6836,
+      "step": 86
+    },
+    {
+      "epoch": 0.19333333333333333,
+      "grad_norm": 615780.3125,
+      "learning_rate": 3.185185185185185e-05,
+      "loss": 0.7781,
+      "step": 87
+    },
+    {
+      "epoch": 0.19555555555555557,
+      "grad_norm": 322161.71875,
+      "learning_rate": 3.222222222222223e-05,
+      "loss": 0.5926,
+      "step": 88
+    },
+    {
+      "epoch": 0.19777777777777777,
+      "grad_norm": 522235.53125,
+      "learning_rate": 3.25925925925926e-05,
+      "loss": 0.6607,
+      "step": 89
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1229829.25,
+      "learning_rate": 3.2962962962962964e-05,
+      "loss": 0.5887,
+      "step": 90
+    },
+    {
+      "epoch": 0.20222222222222222,
+      "grad_norm": 882748.0625,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.6813,
+      "step": 91
+    },
+    {
+      "epoch": 0.20444444444444446,
+      "grad_norm": 889859.8125,
+      "learning_rate": 3.3703703703703706e-05,
+      "loss": 0.7057,
+      "step": 92
+    },
+    {
+      "epoch": 0.20666666666666667,
+      "grad_norm": 579578.375,
+      "learning_rate": 3.4074074074074077e-05,
+      "loss": 0.6658,
+      "step": 93
+    },
+    {
+      "epoch": 0.2088888888888889,
+      "grad_norm": 2560019.25,
+      "learning_rate": 3.444444444444445e-05,
+      "loss": 0.6808,
+      "step": 94
+    },
+    {
+      "epoch": 0.2111111111111111,
+      "grad_norm": 992348.1875,
+      "learning_rate": 3.481481481481482e-05,
+      "loss": 0.7378,
+      "step": 95
+    },
+    {
+      "epoch": 0.21333333333333335,
+      "grad_norm": 456105.78125,
+      "learning_rate": 3.518518518518519e-05,
+      "loss": 0.6837,
+      "step": 96
+    },
+    {
+      "epoch": 0.21555555555555556,
+      "grad_norm": 911418.9375,
+      "learning_rate": 3.555555555555556e-05,
+      "loss": 0.5922,
+      "step": 97
+    },
+    {
+      "epoch": 0.21777777777777776,
+      "grad_norm": 516042.90625,
+      "learning_rate": 3.592592592592593e-05,
+      "loss": 0.5892,
+      "step": 98
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 300478.25,
+      "learning_rate": 3.62962962962963e-05,
+      "loss": 0.5852,
+      "step": 99
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 216423.0,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 0.6518,
+      "step": 100
+    },
+    {
+      "epoch": 0.22444444444444445,
+      "grad_norm": 606593.8125,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 0.7084,
+      "step": 101
+    },
+    {
+      "epoch": 0.22666666666666666,
+      "grad_norm": 325548.90625,
+      "learning_rate": 3.740740740740741e-05,
+      "loss": 0.5128,
+      "step": 102
+    },
+    {
+      "epoch": 0.2288888888888889,
+      "grad_norm": 616763.25,
+      "learning_rate": 3.777777777777778e-05,
+      "loss": 0.6194,
+      "step": 103
+    },
+    {
+      "epoch": 0.2311111111111111,
+      "grad_norm": 1810430.5,
+      "learning_rate": 3.814814814814815e-05,
+      "loss": 0.7078,
+      "step": 104
+    },
+    {
+      "epoch": 0.23333333333333334,
+      "grad_norm": 708301.125,
+      "learning_rate": 3.851851851851852e-05,
+      "loss": 0.5812,
+      "step": 105
+    },
+    {
+      "epoch": 0.23555555555555555,
+      "grad_norm": 349738.53125,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 0.5519,
+      "step": 106
+    },
+    {
+      "epoch": 0.23777777777777778,
+      "grad_norm": 307563.15625,
+      "learning_rate": 3.925925925925926e-05,
+      "loss": 0.6777,
+      "step": 107
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1156388.125,
+      "learning_rate": 3.962962962962963e-05,
+      "loss": 0.6964,
+      "step": 108
+    },
+    {
+      "epoch": 0.24222222222222223,
+      "grad_norm": 1874161.625,
+      "learning_rate": 4e-05,
+      "loss": 0.6031,
+      "step": 109
+    },
+    {
+      "epoch": 0.24444444444444444,
+      "grad_norm": 648151.75,
+      "learning_rate": 4.0370370370370374e-05,
+      "loss": 0.5189,
+      "step": 110
+    },
+    {
+      "epoch": 0.24666666666666667,
+      "grad_norm": 484538.03125,
+      "learning_rate": 4.074074074074074e-05,
+      "loss": 0.5967,
+      "step": 111
+    },
+    {
+      "epoch": 0.24888888888888888,
+      "grad_norm": 549294.625,
+      "learning_rate": 4.111111111111111e-05,
+      "loss": 0.622,
+      "step": 112
+    },
+    {
+      "epoch": 0.2511111111111111,
+      "grad_norm": 422266.71875,
+      "learning_rate": 4.148148148148148e-05,
+      "loss": 0.5626,
+      "step": 113
+    },
+    {
+      "epoch": 0.25333333333333335,
+      "grad_norm": 4367868.0,
+      "learning_rate": 4.185185185185185e-05,
+      "loss": 0.7472,
+      "step": 114
+    },
+    {
+      "epoch": 0.25555555555555554,
+      "grad_norm": 692434.8125,
+      "learning_rate": 4.222222222222222e-05,
+      "loss": 0.5585,
+      "step": 115
+    },
+    {
+      "epoch": 0.2577777777777778,
+      "grad_norm": 945114.75,
+      "learning_rate": 4.259259259259259e-05,
+      "loss": 0.6109,
+      "step": 116
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 689406.625,
+      "learning_rate": 4.296296296296296e-05,
+      "loss": 0.5677,
+      "step": 117
+    },
+    {
+      "epoch": 0.26222222222222225,
+      "grad_norm": 1730677.875,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 0.6684,
+      "step": 118
+    },
+    {
+      "epoch": 0.2644444444444444,
+      "grad_norm": 813010.0625,
+      "learning_rate": 4.3703703703703705e-05,
+      "loss": 0.7408,
+      "step": 119
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 673770.125,
+      "learning_rate": 4.4074074074074076e-05,
+      "loss": 0.6679,
+      "step": 120
+    },
+    {
+      "epoch": 0.2688888888888889,
+      "grad_norm": 363243.59375,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.5323,
+      "step": 121
+    },
+    {
+      "epoch": 0.27111111111111114,
+      "grad_norm": 975494.6875,
+      "learning_rate": 4.481481481481482e-05,
+      "loss": 0.7136,
+      "step": 122
+    },
+    {
+      "epoch": 0.2733333333333333,
+      "grad_norm": 645500.375,
+      "learning_rate": 4.518518518518519e-05,
+      "loss": 0.5937,
+      "step": 123
+    },
+    {
+      "epoch": 0.27555555555555555,
+      "grad_norm": 485883.9375,
+      "learning_rate": 4.555555555555556e-05,
+      "loss": 0.7137,
+      "step": 124
+    },
+    {
+      "epoch": 0.2777777777777778,
+      "grad_norm": 1330097.0,
+      "learning_rate": 4.592592592592593e-05,
+      "loss": 0.7694,
+      "step": 125
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1731506.125,
+      "learning_rate": 4.62962962962963e-05,
+      "loss": 0.798,
+      "step": 126
+    },
+    {
+      "epoch": 0.2822222222222222,
+      "grad_norm": 2576107.75,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 0.5908,
+      "step": 127
+    },
+    {
+      "epoch": 0.28444444444444444,
+      "grad_norm": 631525.375,
+      "learning_rate": 4.703703703703704e-05,
+      "loss": 0.4795,
+      "step": 128
+    },
+    {
+      "epoch": 0.2866666666666667,
+      "grad_norm": 437369.09375,
+      "learning_rate": 4.740740740740741e-05,
+      "loss": 0.621,
+      "step": 129
+    },
+    {
+      "epoch": 0.28888888888888886,
+      "grad_norm": 522823.84375,
+      "learning_rate": 4.7777777777777784e-05,
+      "loss": 0.5231,
+      "step": 130
+    },
+    {
+      "epoch": 0.2911111111111111,
+      "grad_norm": 239169.03125,
+      "learning_rate": 4.814814814814815e-05,
+      "loss": 0.533,
+      "step": 131
+    },
+    {
+      "epoch": 0.29333333333333333,
+      "grad_norm": 1025517.25,
+      "learning_rate": 4.851851851851852e-05,
+      "loss": 0.7916,
+      "step": 132
+    },
+    {
+      "epoch": 0.29555555555555557,
+      "grad_norm": 780890.125,
+      "learning_rate": 4.888888888888889e-05,
+      "loss": 0.6586,
+      "step": 133
+    },
+    {
+      "epoch": 0.29777777777777775,
+      "grad_norm": 239262.828125,
+      "learning_rate": 4.925925925925926e-05,
+      "loss": 0.5212,
+      "step": 134
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1036039.0,
+      "learning_rate": 4.962962962962963e-05,
+      "loss": 0.5101,
+      "step": 135
+    },
+    {
+      "epoch": 0.3022222222222222,
+      "grad_norm": 576536.625,
+      "learning_rate": 5e-05,
+      "loss": 0.673,
+      "step": 136
+    },
+    {
+      "epoch": 0.30444444444444446,
+      "grad_norm": 648396.125,
+      "learning_rate": 4.995884773662552e-05,
+      "loss": 0.6622,
+      "step": 137
+    },
+    {
+      "epoch": 0.30666666666666664,
+      "grad_norm": 727469.6875,
+      "learning_rate": 4.991769547325103e-05,
+      "loss": 0.6532,
+      "step": 138
+    },
+    {
+      "epoch": 0.3088888888888889,
+      "grad_norm": 534778.5625,
+      "learning_rate": 4.987654320987655e-05,
+      "loss": 0.5044,
+      "step": 139
+    },
+    {
+      "epoch": 0.3111111111111111,
+      "grad_norm": 436606.625,
+      "learning_rate": 4.983539094650206e-05,
+      "loss": 0.588,
+      "step": 140
+    },
+    {
+      "epoch": 0.31333333333333335,
+      "grad_norm": 484253.90625,
+      "learning_rate": 4.9794238683127575e-05,
+      "loss": 0.5805,
+      "step": 141
+    },
+    {
+      "epoch": 0.31555555555555553,
+      "grad_norm": 1102731.625,
+      "learning_rate": 4.9753086419753084e-05,
+      "loss": 0.7003,
+      "step": 142
+    },
+    {
+      "epoch": 0.31777777777777777,
+      "grad_norm": 1500681.75,
+      "learning_rate": 4.971193415637861e-05,
+      "loss": 0.6748,
+      "step": 143
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3508447.5,
+      "learning_rate": 4.967078189300412e-05,
+      "loss": 0.6002,
+      "step": 144
+    },
+    {
+      "epoch": 0.32222222222222224,
+      "grad_norm": 532125.9375,
+      "learning_rate": 4.962962962962963e-05,
+      "loss": 0.4983,
+      "step": 145
+    },
+    {
+      "epoch": 0.3244444444444444,
+      "grad_norm": 1056551.25,
+      "learning_rate": 4.958847736625515e-05,
+      "loss": 0.6097,
+      "step": 146
+    },
+    {
+      "epoch": 0.32666666666666666,
+      "grad_norm": 481759.28125,
+      "learning_rate": 4.9547325102880656e-05,
+      "loss": 0.5946,
+      "step": 147
+    },
+    {
+      "epoch": 0.3288888888888889,
+      "grad_norm": 626184.0625,
+      "learning_rate": 4.950617283950618e-05,
+      "loss": 0.5802,
+      "step": 148
+    },
+    {
+      "epoch": 0.33111111111111113,
+      "grad_norm": 673809.4375,
+      "learning_rate": 4.946502057613169e-05,
+      "loss": 0.6302,
+      "step": 149
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 1093257.5,
+      "learning_rate": 4.9423868312757204e-05,
+      "loss": 0.6589,
+      "step": 150
+    },
+    {
+      "epoch": 0.33555555555555555,
+      "grad_norm": 214708.546875,
+      "learning_rate": 4.938271604938271e-05,
+      "loss": 0.5819,
+      "step": 151
+    },
+    {
+      "epoch": 0.3377777777777778,
+      "grad_norm": 873922.3125,
+      "learning_rate": 4.9341563786008236e-05,
+      "loss": 0.5499,
+      "step": 152
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 477493.90625,
+      "learning_rate": 4.930041152263375e-05,
+      "loss": 0.5954,
+      "step": 153
+    },
+    {
+      "epoch": 0.3422222222222222,
+      "grad_norm": 2321650.25,
+      "learning_rate": 4.925925925925926e-05,
+      "loss": 0.7243,
+      "step": 154
+    },
+    {
+      "epoch": 0.34444444444444444,
+      "grad_norm": 646336.1875,
+      "learning_rate": 4.9218106995884777e-05,
+      "loss": 0.5562,
+      "step": 155
+    },
+    {
+      "epoch": 0.3466666666666667,
+      "grad_norm": 664530.0,
+      "learning_rate": 4.9176954732510286e-05,
+      "loss": 0.6897,
+      "step": 156
+    },
+    {
+      "epoch": 0.3488888888888889,
+      "grad_norm": 559558.9375,
+      "learning_rate": 4.913580246913581e-05,
+      "loss": 0.7113,
+      "step": 157
+    },
+    {
+      "epoch": 0.3511111111111111,
+      "grad_norm": 533661.0625,
+      "learning_rate": 4.909465020576132e-05,
+      "loss": 0.5504,
+      "step": 158
+    },
+    {
+      "epoch": 0.35333333333333333,
+      "grad_norm": 1041493.0,
+      "learning_rate": 4.905349794238683e-05,
+      "loss": 0.7329,
+      "step": 159
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 1396587.375,
+      "learning_rate": 4.901234567901235e-05,
+      "loss": 0.6192,
+      "step": 160
+    },
+    {
+      "epoch": 0.35777777777777775,
+      "grad_norm": 1317377.625,
+      "learning_rate": 4.8971193415637865e-05,
+      "loss": 0.6089,
+      "step": 161
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1898401.25,
+      "learning_rate": 4.893004115226338e-05,
+      "loss": 0.5226,
+      "step": 162
+    },
+    {
+      "epoch": 0.3622222222222222,
+      "grad_norm": 927057.1875,
+      "learning_rate": 4.888888888888889e-05,
+      "loss": 0.5559,
+      "step": 163
+    },
+    {
+      "epoch": 0.36444444444444446,
+      "grad_norm": 699609.625,
+      "learning_rate": 4.8847736625514406e-05,
+      "loss": 0.7064,
+      "step": 164
+    },
+    {
+      "epoch": 0.36666666666666664,
+      "grad_norm": 1169845.25,
+      "learning_rate": 4.8806584362139915e-05,
+      "loss": 0.6133,
+      "step": 165
+    },
+    {
+      "epoch": 0.3688888888888889,
+      "grad_norm": 255957.734375,
+      "learning_rate": 4.876543209876544e-05,
+      "loss": 0.5894,
+      "step": 166
+    },
+    {
+      "epoch": 0.3711111111111111,
+      "grad_norm": 546818.125,
+      "learning_rate": 4.872427983539095e-05,
+      "loss": 0.6591,
+      "step": 167
+    },
+    {
+      "epoch": 0.37333333333333335,
+      "grad_norm": 439788.78125,
+      "learning_rate": 4.868312757201646e-05,
+      "loss": 0.6026,
+      "step": 168
+    },
+    {
+      "epoch": 0.37555555555555553,
+      "grad_norm": 1009843.0625,
+      "learning_rate": 4.864197530864198e-05,
+      "loss": 0.5856,
+      "step": 169
+    },
+    {
+      "epoch": 0.37777777777777777,
+      "grad_norm": 488187.84375,
+      "learning_rate": 4.860082304526749e-05,
+      "loss": 0.6526,
+      "step": 170
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1203392.0,
+      "learning_rate": 4.855967078189301e-05,
+      "loss": 0.5753,
+      "step": 171
+    },
+    {
+      "epoch": 0.38222222222222224,
+      "grad_norm": 331781.71875,
+      "learning_rate": 4.851851851851852e-05,
+      "loss": 0.644,
+      "step": 172
+    },
+    {
+      "epoch": 0.3844444444444444,
+      "grad_norm": 957421.0,
+      "learning_rate": 4.8477366255144035e-05,
+      "loss": 0.5646,
+      "step": 173
+    },
+    {
+      "epoch": 0.38666666666666666,
+      "grad_norm": 369945.78125,
+      "learning_rate": 4.843621399176955e-05,
+      "loss": 0.5822,
+      "step": 174
+    },
+    {
+      "epoch": 0.3888888888888889,
+      "grad_norm": 703213.8125,
+      "learning_rate": 4.8395061728395067e-05,
+      "loss": 0.6133,
+      "step": 175
+    },
+    {
+      "epoch": 0.39111111111111113,
+      "grad_norm": 939171.25,
+      "learning_rate": 4.835390946502058e-05,
+      "loss": 0.6379,
+      "step": 176
+    },
+    {
+      "epoch": 0.3933333333333333,
+      "grad_norm": 705173.375,
+      "learning_rate": 4.831275720164609e-05,
+      "loss": 0.6845,
+      "step": 177
+    },
+    {
+      "epoch": 0.39555555555555555,
+      "grad_norm": 491406.125,
+      "learning_rate": 4.827160493827161e-05,
+      "loss": 0.6156,
+      "step": 178
+    },
+    {
+      "epoch": 0.3977777777777778,
+      "grad_norm": 218537.390625,
+      "learning_rate": 4.8230452674897116e-05,
+      "loss": 0.6632,
+      "step": 179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 421861.8125,
+      "learning_rate": 4.818930041152264e-05,
+      "loss": 0.5522,
+      "step": 180
+    },
+    {
+      "epoch": 0.4022222222222222,
+      "grad_norm": 754662.4375,
+      "learning_rate": 4.814814814814815e-05,
+      "loss": 0.524,
+      "step": 181
+    },
+    {
+      "epoch": 0.40444444444444444,
+      "grad_norm": 599035.3125,
+      "learning_rate": 4.8106995884773664e-05,
+      "loss": 0.6015,
+      "step": 182
+    },
+    {
+      "epoch": 0.4066666666666667,
+      "grad_norm": 567925.3125,
+      "learning_rate": 4.806584362139918e-05,
+      "loss": 0.6269,
+      "step": 183
+    },
+    {
+      "epoch": 0.4088888888888889,
+      "grad_norm": 205344.453125,
+      "learning_rate": 4.8024691358024696e-05,
+      "loss": 0.5824,
+      "step": 184
+    },
+    {
+      "epoch": 0.4111111111111111,
+      "grad_norm": 363204.75,
+      "learning_rate": 4.798353909465021e-05,
+      "loss": 0.5746,
+      "step": 185
+    },
+    {
+      "epoch": 0.41333333333333333,
+      "grad_norm": 595512.25,
+      "learning_rate": 4.794238683127572e-05,
+      "loss": 0.5905,
+      "step": 186
+    },
+    {
+      "epoch": 0.41555555555555557,
+      "grad_norm": 768830.75,
+      "learning_rate": 4.7901234567901237e-05,
+      "loss": 0.6321,
+      "step": 187
+    },
+    {
+      "epoch": 0.4177777777777778,
+      "grad_norm": 711439.1875,
+      "learning_rate": 4.7860082304526746e-05,
+      "loss": 0.7406,
+      "step": 188
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2251242.25,
+      "learning_rate": 4.781893004115227e-05,
+      "loss": 0.6063,
+      "step": 189
+    },
+    {
+      "epoch": 0.4222222222222222,
+      "grad_norm": 581344.6875,
+      "learning_rate": 4.7777777777777784e-05,
+      "loss": 0.7016,
+      "step": 190
+    },
+    {
+      "epoch": 0.42444444444444446,
+      "grad_norm": 652845.25,
+      "learning_rate": 4.773662551440329e-05,
+      "loss": 0.6249,
+      "step": 191
+    },
+    {
+      "epoch": 0.4266666666666667,
+      "grad_norm": 450982.625,
+      "learning_rate": 4.769547325102881e-05,
+      "loss": 0.6877,
+      "step": 192
+    },
+    {
+      "epoch": 0.4288888888888889,
+      "grad_norm": 2374753.75,
+      "learning_rate": 4.7654320987654325e-05,
+      "loss": 0.6585,
+      "step": 193
+    },
+    {
+      "epoch": 0.4311111111111111,
+      "grad_norm": 1042813.6875,
+      "learning_rate": 4.761316872427984e-05,
+      "loss": 0.6447,
+      "step": 194
+    },
+    {
+      "epoch": 0.43333333333333335,
+      "grad_norm": 460134.09375,
+      "learning_rate": 4.757201646090535e-05,
+      "loss": 0.6245,
+      "step": 195
+    },
+    {
+      "epoch": 0.43555555555555553,
+      "grad_norm": 1168080.5,
+      "learning_rate": 4.7530864197530866e-05,
+      "loss": 0.6067,
+      "step": 196
+    },
+    {
+      "epoch": 0.43777777777777777,
+      "grad_norm": 760828.0625,
+      "learning_rate": 4.748971193415638e-05,
+      "loss": 0.5517,
+      "step": 197
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 432214.28125,
+      "learning_rate": 4.74485596707819e-05,
+      "loss": 0.544,
+      "step": 198
+    },
+    {
+      "epoch": 0.44222222222222224,
+      "grad_norm": 1489343.0,
+      "learning_rate": 4.740740740740741e-05,
+      "loss": 0.4873,
+      "step": 199
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 848339.625,
+      "learning_rate": 4.736625514403292e-05,
+      "loss": 0.6345,
+      "step": 200
+    },
+    {
+      "epoch": 0.44666666666666666,
+      "grad_norm": 2258669.25,
+      "learning_rate": 4.732510288065844e-05,
+      "loss": 0.4431,
+      "step": 201
+    },
+    {
+      "epoch": 0.4488888888888889,
+      "grad_norm": 534571.0625,
+      "learning_rate": 4.7283950617283954e-05,
+      "loss": 0.5757,
+      "step": 202
+    },
+    {
+      "epoch": 0.45111111111111113,
+      "grad_norm": 1082905.875,
+      "learning_rate": 4.724279835390947e-05,
+      "loss": 0.6787,
+      "step": 203
+    },
+    {
+      "epoch": 0.4533333333333333,
+      "grad_norm": 1464140.0,
+      "learning_rate": 4.7201646090534986e-05,
+      "loss": 0.4967,
+      "step": 204
+    },
+    {
+      "epoch": 0.45555555555555555,
+      "grad_norm": 593933.625,
+      "learning_rate": 4.7160493827160495e-05,
+      "loss": 0.6246,
+      "step": 205
+    },
+    {
+      "epoch": 0.4577777777777778,
+      "grad_norm": 813640.75,
+      "learning_rate": 4.711934156378601e-05,
+      "loss": 0.47,
+      "step": 206
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 418782.53125,
+      "learning_rate": 4.7078189300411527e-05,
+      "loss": 0.4363,
+      "step": 207
+    },
+    {
+      "epoch": 0.4622222222222222,
+      "grad_norm": 648853.5625,
+      "learning_rate": 4.703703703703704e-05,
+      "loss": 0.582,
+      "step": 208
+    },
+    {
+      "epoch": 0.46444444444444444,
+      "grad_norm": 803393.0,
+      "learning_rate": 4.699588477366255e-05,
+      "loss": 0.6732,
+      "step": 209
+    },
+    {
+      "epoch": 0.4666666666666667,
+      "grad_norm": 553974.375,
+      "learning_rate": 4.695473251028807e-05,
+      "loss": 0.527,
+      "step": 210
+    },
+    {
+      "epoch": 0.4688888888888889,
+      "grad_norm": 1254667.0,
+      "learning_rate": 4.691358024691358e-05,
+      "loss": 0.5885,
+      "step": 211
+    },
+    {
+      "epoch": 0.4711111111111111,
+      "grad_norm": 1502191.0,
+      "learning_rate": 4.68724279835391e-05,
+      "loss": 0.5172,
+      "step": 212
+    },
+    {
+      "epoch": 0.47333333333333333,
+      "grad_norm": 310824.96875,
+      "learning_rate": 4.6831275720164615e-05,
+      "loss": 0.6461,
+      "step": 213
+    },
+    {
+      "epoch": 0.47555555555555556,
+      "grad_norm": 461764.15625,
+      "learning_rate": 4.6790123456790124e-05,
+      "loss": 0.509,
+      "step": 214
+    },
+    {
+      "epoch": 0.4777777777777778,
+      "grad_norm": 702569.8125,
+      "learning_rate": 4.674897119341564e-05,
+      "loss": 0.5263,
+      "step": 215
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 483824.375,
+      "learning_rate": 4.6707818930041156e-05,
+      "loss": 0.6557,
+      "step": 216
+    },
+    {
+      "epoch": 0.4822222222222222,
+      "grad_norm": 295729.90625,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 0.5004,
+      "step": 217
+    },
+    {
+      "epoch": 0.48444444444444446,
+      "grad_norm": 468939.78125,
+      "learning_rate": 4.662551440329218e-05,
+      "loss": 0.5232,
+      "step": 218
+    },
+    {
+      "epoch": 0.4866666666666667,
+      "grad_norm": 171850.140625,
+      "learning_rate": 4.6584362139917697e-05,
+      "loss": 0.6334,
+      "step": 219
+    },
+    {
+      "epoch": 0.4888888888888889,
+      "grad_norm": 1578345.875,
+      "learning_rate": 4.654320987654321e-05,
+      "loss": 0.7786,
+      "step": 220
+    },
+    {
+      "epoch": 0.4911111111111111,
+      "grad_norm": 1356533.75,
+      "learning_rate": 4.650205761316873e-05,
+      "loss": 0.5164,
+      "step": 221
+    },
+    {
+      "epoch": 0.49333333333333335,
+      "grad_norm": 634727.375,
+      "learning_rate": 4.6460905349794244e-05,
+      "loss": 0.4403,
+      "step": 222
+    },
+    {
+      "epoch": 0.4955555555555556,
+      "grad_norm": 239289.609375,
+      "learning_rate": 4.641975308641975e-05,
+      "loss": 0.6065,
+      "step": 223
+    },
+    {
+      "epoch": 0.49777777777777776,
+      "grad_norm": 240202.625,
+      "learning_rate": 4.637860082304527e-05,
+      "loss": 0.4874,
+      "step": 224
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1101113.75,
+      "learning_rate": 4.6337448559670785e-05,
+      "loss": 0.5963,
+      "step": 225
+    },
+    {
+      "epoch": 0.5022222222222222,
+      "grad_norm": 399643.0625,
+      "learning_rate": 4.62962962962963e-05,
+      "loss": 0.5306,
+      "step": 226
+    },
+    {
+      "epoch": 0.5044444444444445,
+      "grad_norm": 454625.84375,
+      "learning_rate": 4.625514403292182e-05,
+      "loss": 0.7048,
+      "step": 227
+    },
+    {
+      "epoch": 0.5066666666666667,
+      "grad_norm": 502732.28125,
+      "learning_rate": 4.6213991769547326e-05,
+      "loss": 0.5246,
+      "step": 228
+    },
+    {
+      "epoch": 0.5088888888888888,
+      "grad_norm": 302046.5625,
+      "learning_rate": 4.617283950617284e-05,
+      "loss": 0.6615,
+      "step": 229
+    },
+    {
+      "epoch": 0.5111111111111111,
+      "grad_norm": 1041901.5,
+      "learning_rate": 4.613168724279836e-05,
+      "loss": 0.7639,
+      "step": 230
+    },
+    {
+      "epoch": 0.5133333333333333,
+      "grad_norm": 464965.53125,
+      "learning_rate": 4.609053497942387e-05,
+      "loss": 0.6237,
+      "step": 231
+    },
+    {
+      "epoch": 0.5155555555555555,
+      "grad_norm": 751392.25,
+      "learning_rate": 4.604938271604938e-05,
+      "loss": 0.6681,
+      "step": 232
+    },
+    {
+      "epoch": 0.5177777777777778,
+      "grad_norm": 538912.8125,
+      "learning_rate": 4.60082304526749e-05,
+      "loss": 0.6773,
+      "step": 233
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 598586.375,
+      "learning_rate": 4.5967078189300414e-05,
+      "loss": 0.6382,
+      "step": 234
+    },
+    {
+      "epoch": 0.5222222222222223,
+      "grad_norm": 525750.0,
+      "learning_rate": 4.592592592592593e-05,
+      "loss": 0.5312,
+      "step": 235
+    },
+    {
+      "epoch": 0.5244444444444445,
+      "grad_norm": 563147.6875,
+      "learning_rate": 4.5884773662551446e-05,
+      "loss": 0.5385,
+      "step": 236
+    },
+    {
+      "epoch": 0.5266666666666666,
+      "grad_norm": 1363912.125,
+      "learning_rate": 4.5843621399176955e-05,
+      "loss": 0.7997,
+      "step": 237
+    },
+    {
+      "epoch": 0.5288888888888889,
+      "grad_norm": 488355.625,
+      "learning_rate": 4.580246913580247e-05,
+      "loss": 0.5144,
+      "step": 238
+    },
+    {
+      "epoch": 0.5311111111111111,
+      "grad_norm": 680616.4375,
+      "learning_rate": 4.5761316872427987e-05,
+      "loss": 0.5847,
+      "step": 239
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 627066.8125,
+      "learning_rate": 4.57201646090535e-05,
+      "loss": 0.5937,
+      "step": 240
+    },
+    {
+      "epoch": 0.5355555555555556,
+      "grad_norm": 982689.1875,
+      "learning_rate": 4.567901234567901e-05,
+      "loss": 0.4186,
+      "step": 241
+    },
+    {
+      "epoch": 0.5377777777777778,
+      "grad_norm": 726951.875,
+      "learning_rate": 4.563786008230453e-05,
+      "loss": 0.6762,
+      "step": 242
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 312047.375,
+      "learning_rate": 4.559670781893004e-05,
+      "loss": 0.6264,
+      "step": 243
+    },
+    {
+      "epoch": 0.5422222222222223,
+      "grad_norm": 786597.3125,
+      "learning_rate": 4.555555555555556e-05,
+      "loss": 0.9229,
+      "step": 244
+    },
+    {
+      "epoch": 0.5444444444444444,
+      "grad_norm": 967135.0625,
+      "learning_rate": 4.5514403292181075e-05,
+      "loss": 0.4451,
+      "step": 245
+    },
+    {
+      "epoch": 0.5466666666666666,
+      "grad_norm": 1016416.0,
+      "learning_rate": 4.5473251028806584e-05,
+      "loss": 0.6044,
+      "step": 246
+    },
+    {
+      "epoch": 0.5488888888888889,
+      "grad_norm": 175218.09375,
+      "learning_rate": 4.54320987654321e-05,
+      "loss": 0.6136,
+      "step": 247
+    },
+    {
+      "epoch": 0.5511111111111111,
+      "grad_norm": 803901.375,
+      "learning_rate": 4.5390946502057616e-05,
+      "loss": 0.6505,
+      "step": 248
+    },
+    {
+      "epoch": 0.5533333333333333,
+      "grad_norm": 352020.40625,
+      "learning_rate": 4.534979423868313e-05,
+      "loss": 0.4413,
+      "step": 249
+    },
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 2258271.25,
+      "learning_rate": 4.530864197530865e-05,
+      "loss": 0.4909,
+      "step": 250
+    },
+    {
+      "epoch": 0.5577777777777778,
+      "grad_norm": 378156.25,
+      "learning_rate": 4.5267489711934157e-05,
+      "loss": 0.6123,
+      "step": 251
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 815982.0,
+      "learning_rate": 4.522633744855967e-05,
+      "loss": 0.6559,
+      "step": 252
+    },
+    {
+      "epoch": 0.5622222222222222,
+      "grad_norm": 2519179.75,
+      "learning_rate": 4.518518518518519e-05,
+      "loss": 0.5177,
+      "step": 253
+    },
+    {
+      "epoch": 0.5644444444444444,
+      "grad_norm": 806211.6875,
+      "learning_rate": 4.5144032921810704e-05,
+      "loss": 0.5855,
+      "step": 254
+    },
+    {
+      "epoch": 0.5666666666666667,
+      "grad_norm": 458326.0625,
+      "learning_rate": 4.510288065843621e-05,
+      "loss": 0.5142,
+      "step": 255
+    },
+    {
+      "epoch": 0.5688888888888889,
+      "grad_norm": 1957645.25,
+      "learning_rate": 4.506172839506173e-05,
+      "loss": 0.5814,
+      "step": 256
+    },
+    {
+      "epoch": 0.5711111111111111,
+      "grad_norm": 15964121.0,
+      "learning_rate": 4.5020576131687245e-05,
+      "loss": 0.7108,
+      "step": 257
+    },
+    {
+      "epoch": 0.5733333333333334,
+      "grad_norm": 703737.875,
+      "learning_rate": 4.497942386831276e-05,
+      "loss": 0.5115,
+      "step": 258
+    },
+    {
+      "epoch": 0.5755555555555556,
+      "grad_norm": 1874186.375,
+      "learning_rate": 4.493827160493828e-05,
+      "loss": 0.5556,
+      "step": 259
+    },
+    {
+      "epoch": 0.5777777777777777,
+      "grad_norm": 2454680.5,
+      "learning_rate": 4.4897119341563786e-05,
+      "loss": 0.5399,
+      "step": 260
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 490995.4375,
+      "learning_rate": 4.48559670781893e-05,
+      "loss": 0.508,
+      "step": 261
+    },
+    {
+      "epoch": 0.5822222222222222,
+      "grad_norm": 935022.3125,
+      "learning_rate": 4.481481481481482e-05,
+      "loss": 0.5846,
+      "step": 262
+    },
+    {
+      "epoch": 0.5844444444444444,
+      "grad_norm": 1142683.0,
+      "learning_rate": 4.477366255144033e-05,
+      "loss": 0.7131,
+      "step": 263
+    },
+    {
+      "epoch": 0.5866666666666667,
+      "grad_norm": 422861.84375,
+      "learning_rate": 4.473251028806584e-05,
+      "loss": 0.7322,
+      "step": 264
+    },
+    {
+      "epoch": 0.5888888888888889,
+      "grad_norm": 572327.75,
+      "learning_rate": 4.469135802469136e-05,
+      "loss": 0.6128,
+      "step": 265
+    },
+    {
+      "epoch": 0.5911111111111111,
+      "grad_norm": 774320.0,
+      "learning_rate": 4.4650205761316874e-05,
+      "loss": 0.5271,
+      "step": 266
+    },
+    {
+      "epoch": 0.5933333333333334,
+      "grad_norm": 280896.9375,
+      "learning_rate": 4.460905349794239e-05,
+      "loss": 0.5551,
+      "step": 267
+    },
+    {
+      "epoch": 0.5955555555555555,
+      "grad_norm": 848952.5625,
+      "learning_rate": 4.4567901234567906e-05,
+      "loss": 0.673,
+      "step": 268
+    },
+    {
+      "epoch": 0.5977777777777777,
+      "grad_norm": 749778.4375,
+      "learning_rate": 4.4526748971193415e-05,
+      "loss": 0.6507,
+      "step": 269
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 454624.96875,
+      "learning_rate": 4.448559670781893e-05,
+      "loss": 0.5128,
+      "step": 270
+    },
+    {
+      "epoch": 0.6022222222222222,
+      "grad_norm": 252433.75,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.5532,
+      "step": 271
+    },
+    {
+      "epoch": 0.6044444444444445,
+      "grad_norm": 1700119.875,
+      "learning_rate": 4.440329218106996e-05,
+      "loss": 0.7207,
+      "step": 272
+    },
+    {
+      "epoch": 0.6066666666666667,
+      "grad_norm": 542159.4375,
+      "learning_rate": 4.436213991769548e-05,
+      "loss": 0.5605,
+      "step": 273
+    },
+    {
+      "epoch": 0.6088888888888889,
+      "grad_norm": 655691.8125,
+      "learning_rate": 4.432098765432099e-05,
+      "loss": 0.5493,
+      "step": 274
+    },
+    {
+      "epoch": 0.6111111111111112,
+      "grad_norm": 1386989.375,
+      "learning_rate": 4.42798353909465e-05,
+      "loss": 0.6047,
+      "step": 275
+    },
+    {
+      "epoch": 0.6133333333333333,
+      "grad_norm": 378723.8125,
+      "learning_rate": 4.423868312757202e-05,
+      "loss": 0.5983,
+      "step": 276
+    },
+    {
+      "epoch": 0.6155555555555555,
+      "grad_norm": 412323.6875,
+      "learning_rate": 4.4197530864197535e-05,
+      "loss": 0.4852,
+      "step": 277
+    },
+    {
+      "epoch": 0.6177777777777778,
+      "grad_norm": 949697.5625,
+      "learning_rate": 4.4156378600823044e-05,
+      "loss": 0.5116,
+      "step": 278
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 142350.5625,
+      "learning_rate": 4.411522633744856e-05,
+      "loss": 0.5351,
+      "step": 279
+    },
+    {
+      "epoch": 0.6222222222222222,
+      "grad_norm": 180120.21875,
+      "learning_rate": 4.4074074074074076e-05,
+      "loss": 0.487,
+      "step": 280
+    },
+    {
+      "epoch": 0.6244444444444445,
+      "grad_norm": 806719.6875,
+      "learning_rate": 4.403292181069959e-05,
+      "loss": 0.5948,
+      "step": 281
+    },
+    {
+      "epoch": 0.6266666666666667,
+      "grad_norm": 637918.25,
+      "learning_rate": 4.399176954732511e-05,
+      "loss": 0.5571,
+      "step": 282
+    },
+    {
+      "epoch": 0.6288888888888889,
+      "grad_norm": 218954.140625,
+      "learning_rate": 4.3950617283950617e-05,
+      "loss": 0.5268,
+      "step": 283
+    },
+    {
+      "epoch": 0.6311111111111111,
+      "grad_norm": 265361.21875,
+      "learning_rate": 4.390946502057613e-05,
+      "loss": 0.6142,
+      "step": 284
+    },
+    {
+      "epoch": 0.6333333333333333,
+      "grad_norm": 841901.4375,
+      "learning_rate": 4.386831275720165e-05,
+      "loss": 0.5919,
+      "step": 285
+    },
+    {
+      "epoch": 0.6355555555555555,
+      "grad_norm": 945528.875,
+      "learning_rate": 4.3827160493827164e-05,
+      "loss": 0.5273,
+      "step": 286
+    },
+    {
+      "epoch": 0.6377777777777778,
+      "grad_norm": 1093268.5,
+      "learning_rate": 4.378600823045268e-05,
+      "loss": 0.4842,
+      "step": 287
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 311890.375,
+      "learning_rate": 4.374485596707819e-05,
+      "loss": 0.5431,
+      "step": 288
+    },
+    {
+      "epoch": 0.6422222222222222,
+      "grad_norm": 668244.25,
+      "learning_rate": 4.3703703703703705e-05,
+      "loss": 0.5782,
+      "step": 289
+    },
+    {
+      "epoch": 0.6444444444444445,
+      "grad_norm": 199099.53125,
+      "learning_rate": 4.366255144032922e-05,
+      "loss": 0.6093,
+      "step": 290
+    },
+    {
+      "epoch": 0.6466666666666666,
+      "grad_norm": 646594.125,
+      "learning_rate": 4.3621399176954737e-05,
+      "loss": 0.5381,
+      "step": 291
+    },
+    {
+      "epoch": 0.6488888888888888,
+      "grad_norm": 801783.3125,
+      "learning_rate": 4.3580246913580246e-05,
+      "loss": 0.7125,
+      "step": 292
+    },
+    {
+      "epoch": 0.6511111111111111,
+      "grad_norm": 1490818.5,
+      "learning_rate": 4.353909465020576e-05,
+      "loss": 0.6342,
+      "step": 293
+    },
+    {
+      "epoch": 0.6533333333333333,
+      "grad_norm": 661538.0,
+      "learning_rate": 4.349794238683128e-05,
+      "loss": 0.5878,
+      "step": 294
+    },
+    {
+      "epoch": 0.6555555555555556,
+      "grad_norm": 872336.4375,
+      "learning_rate": 4.345679012345679e-05,
+      "loss": 0.477,
+      "step": 295
+    },
+    {
+      "epoch": 0.6577777777777778,
+      "grad_norm": 1284983.875,
+      "learning_rate": 4.341563786008231e-05,
+      "loss": 0.516,
+      "step": 296
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 469390.59375,
+      "learning_rate": 4.337448559670782e-05,
+      "loss": 0.4943,
+      "step": 297
+    },
+    {
+      "epoch": 0.6622222222222223,
+      "grad_norm": 892792.9375,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 0.4624,
+      "step": 298
+    },
+    {
+      "epoch": 0.6644444444444444,
+      "grad_norm": 571622.625,
+      "learning_rate": 4.329218106995885e-05,
+      "loss": 0.6525,
+      "step": 299
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 1495432.875,
+      "learning_rate": 4.3251028806584366e-05,
+      "loss": 0.4023,
+      "step": 300
+    },
+    {
+      "epoch": 0.6688888888888889,
+      "grad_norm": 758325.4375,
+      "learning_rate": 4.3209876543209875e-05,
+      "loss": 0.7577,
+      "step": 301
+    },
+    {
+      "epoch": 0.6711111111111111,
+      "grad_norm": 640554.1875,
+      "learning_rate": 4.316872427983539e-05,
+      "loss": 0.6508,
+      "step": 302
+    },
+    {
+      "epoch": 0.6733333333333333,
+      "grad_norm": 2041099.5,
+      "learning_rate": 4.3127572016460907e-05,
+      "loss": 0.6445,
+      "step": 303
+    },
+    {
+      "epoch": 0.6755555555555556,
+      "grad_norm": 278156.78125,
+      "learning_rate": 4.308641975308642e-05,
+      "loss": 0.6027,
+      "step": 304
+    },
+    {
+      "epoch": 0.6777777777777778,
+      "grad_norm": 1308681.5,
+      "learning_rate": 4.304526748971194e-05,
+      "loss": 0.4822,
+      "step": 305
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 967817.8125,
+      "learning_rate": 4.300411522633745e-05,
+      "loss": 0.5359,
+      "step": 306
+    },
+    {
+      "epoch": 0.6822222222222222,
+      "grad_norm": 223118.21875,
+      "learning_rate": 4.296296296296296e-05,
+      "loss": 0.4957,
+      "step": 307
+    },
+    {
+      "epoch": 0.6844444444444444,
+      "grad_norm": 1438647.25,
+      "learning_rate": 4.292181069958848e-05,
+      "loss": 0.6234,
+      "step": 308
+    },
+    {
+      "epoch": 0.6866666666666666,
+      "grad_norm": 1326128.0,
+      "learning_rate": 4.2880658436213995e-05,
+      "loss": 0.5423,
+      "step": 309
+    },
+    {
+      "epoch": 0.6888888888888889,
+      "grad_norm": 419145.375,
+      "learning_rate": 4.283950617283951e-05,
+      "loss": 0.559,
+      "step": 310
+    },
+    {
+      "epoch": 0.6911111111111111,
+      "grad_norm": 538809.1875,
+      "learning_rate": 4.279835390946502e-05,
+      "loss": 0.5048,
+      "step": 311
+    },
+    {
+      "epoch": 0.6933333333333334,
+      "grad_norm": 308140.90625,
+      "learning_rate": 4.2757201646090536e-05,
+      "loss": 0.5409,
+      "step": 312
+    },
+    {
+      "epoch": 0.6955555555555556,
+      "grad_norm": 901349.375,
+      "learning_rate": 4.271604938271605e-05,
+      "loss": 0.6302,
+      "step": 313
+    },
+    {
+      "epoch": 0.6977777777777778,
+      "grad_norm": 825236.5,
+      "learning_rate": 4.267489711934157e-05,
+      "loss": 0.6041,
+      "step": 314
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 550996.1875,
+      "learning_rate": 4.2633744855967077e-05,
+      "loss": 0.4659,
+      "step": 315
+    },
+    {
+      "epoch": 0.7022222222222222,
+      "grad_norm": 446906.25,
+      "learning_rate": 4.259259259259259e-05,
+      "loss": 0.4984,
+      "step": 316
+    },
+    {
+      "epoch": 0.7044444444444444,
+      "grad_norm": 653586.5625,
+      "learning_rate": 4.255144032921811e-05,
+      "loss": 0.6108,
+      "step": 317
+    },
+    {
+      "epoch": 0.7066666666666667,
+      "grad_norm": 1313717.75,
+      "learning_rate": 4.2510288065843624e-05,
+      "loss": 0.6092,
+      "step": 318
+    },
+    {
+      "epoch": 0.7088888888888889,
+      "grad_norm": 885470.5625,
+      "learning_rate": 4.246913580246914e-05,
+      "loss": 0.5388,
+      "step": 319
+    },
+    {
+      "epoch": 0.7111111111111111,
+      "grad_norm": 461374.0,
+      "learning_rate": 4.242798353909465e-05,
+      "loss": 0.4756,
+      "step": 320
+    },
+    {
+      "epoch": 0.7133333333333334,
+      "grad_norm": 203098.859375,
+      "learning_rate": 4.2386831275720165e-05,
+      "loss": 0.6083,
+      "step": 321
+    },
+    {
+      "epoch": 0.7155555555555555,
+      "grad_norm": 477118.0625,
+      "learning_rate": 4.234567901234568e-05,
+      "loss": 0.5137,
+      "step": 322
+    },
+    {
+      "epoch": 0.7177777777777777,
+      "grad_norm": 607326.5,
+      "learning_rate": 4.2304526748971197e-05,
+      "loss": 0.4768,
+      "step": 323
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 254751.46875,
+      "learning_rate": 4.2263374485596706e-05,
+      "loss": 0.6813,
+      "step": 324
+    },
+    {
+      "epoch": 0.7222222222222222,
+      "grad_norm": 574567.0,
+      "learning_rate": 4.222222222222222e-05,
+      "loss": 0.5853,
+      "step": 325
+    },
+    {
+      "epoch": 0.7244444444444444,
+      "grad_norm": 2085169.875,
+      "learning_rate": 4.2181069958847744e-05,
+      "loss": 0.5218,
+      "step": 326
+    },
+    {
+      "epoch": 0.7266666666666667,
+      "grad_norm": 1202527.75,
+      "learning_rate": 4.213991769547325e-05,
+      "loss": 0.4943,
+      "step": 327
+    },
+    {
+      "epoch": 0.7288888888888889,
+      "grad_norm": 296164.25,
+      "learning_rate": 4.209876543209877e-05,
+      "loss": 0.5333,
+      "step": 328
+    },
+    {
+      "epoch": 0.7311111111111112,
+      "grad_norm": 275263.03125,
+      "learning_rate": 4.205761316872428e-05,
+      "loss": 0.4711,
+      "step": 329
+    },
+    {
+      "epoch": 0.7333333333333333,
+      "grad_norm": 973031.8125,
+      "learning_rate": 4.2016460905349794e-05,
+      "loss": 0.5963,
+      "step": 330
+    },
+    {
+      "epoch": 0.7355555555555555,
+      "grad_norm": 230538.0625,
+      "learning_rate": 4.197530864197531e-05,
+      "loss": 0.6437,
+      "step": 331
+    },
+    {
+      "epoch": 0.7377777777777778,
+      "grad_norm": 668760.5625,
+      "learning_rate": 4.1934156378600826e-05,
+      "loss": 0.5524,
+      "step": 332
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 353451.75,
+      "learning_rate": 4.189300411522634e-05,
+      "loss": 0.498,
+      "step": 333
+    },
+    {
+      "epoch": 0.7422222222222222,
+      "grad_norm": 280192.4375,
+      "learning_rate": 4.185185185185185e-05,
+      "loss": 0.5159,
+      "step": 334
+    },
+    {
+      "epoch": 0.7444444444444445,
+      "grad_norm": 340414.5625,
+      "learning_rate": 4.181069958847737e-05,
+      "loss": 0.5469,
+      "step": 335
+    },
+    {
+      "epoch": 0.7466666666666667,
+      "grad_norm": 748978.5625,
+      "learning_rate": 4.176954732510288e-05,
+      "loss": 0.6063,
+      "step": 336
+    },
+    {
+      "epoch": 0.7488888888888889,
+      "grad_norm": 374095.09375,
+      "learning_rate": 4.17283950617284e-05,
+      "loss": 0.734,
+      "step": 337
+    },
+    {
+      "epoch": 0.7511111111111111,
+      "grad_norm": 163727.234375,
+      "learning_rate": 4.168724279835391e-05,
+      "loss": 0.4815,
+      "step": 338
+    },
+    {
+      "epoch": 0.7533333333333333,
+      "grad_norm": 336208.46875,
+      "learning_rate": 4.164609053497942e-05,
+      "loss": 0.6262,
+      "step": 339
+    },
+    {
+      "epoch": 0.7555555555555555,
+      "grad_norm": 606244.875,
+      "learning_rate": 4.1604938271604946e-05,
+      "loss": 0.583,
+      "step": 340
+    },
+    {
+      "epoch": 0.7577777777777778,
+      "grad_norm": 531861.75,
+      "learning_rate": 4.1563786008230455e-05,
+      "loss": 0.5549,
+      "step": 341
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 489404.5625,
+      "learning_rate": 4.152263374485597e-05,
+      "loss": 0.5868,
+      "step": 342
+    },
+    {
+      "epoch": 0.7622222222222222,
+      "grad_norm": 770754.875,
+      "learning_rate": 4.148148148148148e-05,
+      "loss": 0.6607,
+      "step": 343
+    },
+    {
+      "epoch": 0.7644444444444445,
+      "grad_norm": 444973.46875,
+      "learning_rate": 4.1440329218106996e-05,
+      "loss": 0.7014,
+      "step": 344
+    },
+    {
+      "epoch": 0.7666666666666667,
+      "grad_norm": 828877.0625,
+      "learning_rate": 4.139917695473251e-05,
+      "loss": 0.4733,
+      "step": 345
+    },
+    {
+      "epoch": 0.7688888888888888,
+      "grad_norm": 2406807.25,
+      "learning_rate": 4.135802469135803e-05,
+      "loss": 0.6269,
+      "step": 346
+    },
+    {
+      "epoch": 0.7711111111111111,
+      "grad_norm": 1750097.375,
+      "learning_rate": 4.1316872427983537e-05,
+      "loss": 0.6198,
+      "step": 347
+    },
+    {
+      "epoch": 0.7733333333333333,
+      "grad_norm": 1762184.75,
+      "learning_rate": 4.127572016460905e-05,
+      "loss": 0.4977,
+      "step": 348
+    },
+    {
+      "epoch": 0.7755555555555556,
+      "grad_norm": 242493.71875,
+      "learning_rate": 4.1234567901234575e-05,
+      "loss": 0.5969,
+      "step": 349
+    },
+    {
+      "epoch": 0.7777777777777778,
+      "grad_norm": 3530379.75,
+      "learning_rate": 4.1193415637860084e-05,
+      "loss": 0.5063,
+      "step": 350
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 532942.0625,
+      "learning_rate": 4.11522633744856e-05,
+      "loss": 0.4462,
+      "step": 351
+    },
+    {
+      "epoch": 0.7822222222222223,
+      "grad_norm": 2386480.5,
+      "learning_rate": 4.111111111111111e-05,
+      "loss": 0.5945,
+      "step": 352
+    },
+    {
+      "epoch": 0.7844444444444445,
+      "grad_norm": 207174.203125,
+      "learning_rate": 4.1069958847736625e-05,
+      "loss": 0.4863,
+      "step": 353
+    },
+    {
+      "epoch": 0.7866666666666666,
+      "grad_norm": 550860.6875,
+      "learning_rate": 4.102880658436214e-05,
+      "loss": 0.5111,
+      "step": 354
+    },
+    {
+      "epoch": 0.7888888888888889,
+      "grad_norm": 1387017.5,
+      "learning_rate": 4.0987654320987657e-05,
+      "loss": 0.5244,
+      "step": 355
+    },
+    {
+      "epoch": 0.7911111111111111,
+      "grad_norm": 443979.40625,
+      "learning_rate": 4.094650205761317e-05,
+      "loss": 0.5259,
+      "step": 356
+    },
+    {
+      "epoch": 0.7933333333333333,
+      "grad_norm": 167479.6875,
+      "learning_rate": 4.090534979423868e-05,
+      "loss": 0.5652,
+      "step": 357
+    },
+    {
+      "epoch": 0.7955555555555556,
+      "grad_norm": 268729.6875,
+      "learning_rate": 4.0864197530864204e-05,
+      "loss": 0.4428,
+      "step": 358
+    },
+    {
+      "epoch": 0.7977777777777778,
+      "grad_norm": 510909.34375,
+      "learning_rate": 4.082304526748971e-05,
+      "loss": 0.5002,
+      "step": 359
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1276042.125,
+      "learning_rate": 4.078189300411523e-05,
+      "loss": 0.7983,
+      "step": 360
+    },
+    {
+      "epoch": 0.8022222222222222,
+      "grad_norm": 696396.8125,
+      "learning_rate": 4.074074074074074e-05,
+      "loss": 0.5355,
+      "step": 361
+    },
+    {
+      "epoch": 0.8044444444444444,
+      "grad_norm": 520619.625,
+      "learning_rate": 4.0699588477366254e-05,
+      "loss": 0.6058,
+      "step": 362
+    },
+    {
+      "epoch": 0.8066666666666666,
+      "grad_norm": 1366369.0,
+      "learning_rate": 4.065843621399178e-05,
+      "loss": 0.4865,
+      "step": 363
+    },
+    {
+      "epoch": 0.8088888888888889,
+      "grad_norm": 906512.5,
+      "learning_rate": 4.0617283950617286e-05,
+      "loss": 0.6686,
+      "step": 364
+    },
+    {
+      "epoch": 0.8111111111111111,
+      "grad_norm": 498396.90625,
+      "learning_rate": 4.05761316872428e-05,
+      "loss": 0.5721,
+      "step": 365
+    },
+    {
+      "epoch": 0.8133333333333334,
+      "grad_norm": 506303.21875,
+      "learning_rate": 4.053497942386831e-05,
+      "loss": 0.5422,
+      "step": 366
+    },
+    {
+      "epoch": 0.8155555555555556,
+      "grad_norm": 8037728.0,
+      "learning_rate": 4.049382716049383e-05,
+      "loss": 0.6566,
+      "step": 367
+    },
+    {
+      "epoch": 0.8177777777777778,
+      "grad_norm": 936880.9375,
+      "learning_rate": 4.045267489711934e-05,
+      "loss": 0.4807,
+      "step": 368
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1358766.25,
+      "learning_rate": 4.041152263374486e-05,
+      "loss": 0.4736,
+      "step": 369
+    },
+    {
+      "epoch": 0.8222222222222222,
+      "grad_norm": 1406607.5,
+      "learning_rate": 4.0370370370370374e-05,
+      "loss": 0.4751,
+      "step": 370
+    },
+    {
+      "epoch": 0.8244444444444444,
+      "grad_norm": 832850.125,
+      "learning_rate": 4.032921810699588e-05,
+      "loss": 0.5952,
+      "step": 371
+    },
+    {
+      "epoch": 0.8266666666666667,
+      "grad_norm": 500190.8125,
+      "learning_rate": 4.0288065843621406e-05,
+      "loss": 0.4909,
+      "step": 372
+    },
+    {
+      "epoch": 0.8288888888888889,
+      "grad_norm": 193279.265625,
+      "learning_rate": 4.0246913580246915e-05,
+      "loss": 0.4518,
+      "step": 373
+    },
+    {
+      "epoch": 0.8311111111111111,
+      "grad_norm": 1528230.625,
+      "learning_rate": 4.020576131687243e-05,
+      "loss": 0.5747,
+      "step": 374
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 228102.15625,
+      "learning_rate": 4.016460905349794e-05,
+      "loss": 0.481,
+      "step": 375
+    },
+    {
+      "epoch": 0.8355555555555556,
+      "grad_norm": 472035.78125,
+      "learning_rate": 4.012345679012346e-05,
+      "loss": 0.4831,
+      "step": 376
+    },
+    {
+      "epoch": 0.8377777777777777,
+      "grad_norm": 546926.125,
+      "learning_rate": 4.008230452674897e-05,
+      "loss": 0.4395,
+      "step": 377
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 949932.875,
+      "learning_rate": 4.004115226337449e-05,
+      "loss": 0.5096,
+      "step": 378
+    },
+    {
+      "epoch": 0.8422222222222222,
+      "grad_norm": 388450.03125,
+      "learning_rate": 4e-05,
+      "loss": 0.5798,
+      "step": 379
+    },
+    {
+      "epoch": 0.8444444444444444,
+      "grad_norm": 407452.59375,
+      "learning_rate": 3.995884773662551e-05,
+      "loss": 0.6373,
+      "step": 380
+    },
+    {
+      "epoch": 0.8466666666666667,
+      "grad_norm": 781730.5625,
+      "learning_rate": 3.9917695473251035e-05,
+      "loss": 0.5271,
+      "step": 381
+    },
+    {
+      "epoch": 0.8488888888888889,
+      "grad_norm": 360056.3125,
+      "learning_rate": 3.9876543209876544e-05,
+      "loss": 0.54,
+      "step": 382
+    },
+    {
+      "epoch": 0.8511111111111112,
+      "grad_norm": 557841.125,
+      "learning_rate": 3.983539094650206e-05,
+      "loss": 0.6939,
+      "step": 383
+    },
+    {
+      "epoch": 0.8533333333333334,
+      "grad_norm": 1382806.125,
+      "learning_rate": 3.979423868312757e-05,
+      "loss": 0.6029,
+      "step": 384
+    },
+    {
+      "epoch": 0.8555555555555555,
+      "grad_norm": 655178.8125,
+      "learning_rate": 3.975308641975309e-05,
+      "loss": 0.3791,
+      "step": 385
+    },
+    {
+      "epoch": 0.8577777777777778,
+      "grad_norm": 265361.96875,
+      "learning_rate": 3.971193415637861e-05,
+      "loss": 0.4686,
+      "step": 386
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 424605.125,
+      "learning_rate": 3.9670781893004117e-05,
+      "loss": 0.6816,
+      "step": 387
+    },
+    {
+      "epoch": 0.8622222222222222,
+      "grad_norm": 542365.8125,
+      "learning_rate": 3.962962962962963e-05,
+      "loss": 0.6044,
+      "step": 388
+    },
+    {
+      "epoch": 0.8644444444444445,
+      "grad_norm": 338897.46875,
+      "learning_rate": 3.958847736625514e-05,
+      "loss": 0.569,
+      "step": 389
+    },
+    {
+      "epoch": 0.8666666666666667,
+      "grad_norm": 1602508.375,
+      "learning_rate": 3.9547325102880664e-05,
+      "loss": 0.618,
+      "step": 390
+    },
+    {
+      "epoch": 0.8688888888888889,
+      "grad_norm": 181145.4375,
+      "learning_rate": 3.950617283950617e-05,
+      "loss": 0.5979,
+      "step": 391
+    },
+    {
+      "epoch": 0.8711111111111111,
+      "grad_norm": 704965.625,
+      "learning_rate": 3.946502057613169e-05,
+      "loss": 0.4592,
+      "step": 392
+    },
+    {
+      "epoch": 0.8733333333333333,
+      "grad_norm": 2342850.25,
+      "learning_rate": 3.9423868312757205e-05,
+      "loss": 0.4666,
+      "step": 393
+    },
+    {
+      "epoch": 0.8755555555555555,
+      "grad_norm": 429568.09375,
+      "learning_rate": 3.938271604938272e-05,
+      "loss": 0.5621,
+      "step": 394
+    },
+    {
+      "epoch": 0.8777777777777778,
+      "grad_norm": 787485.5,
+      "learning_rate": 3.934156378600824e-05,
+      "loss": 0.4794,
+      "step": 395
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4468053.0,
+      "learning_rate": 3.9300411522633746e-05,
+      "loss": 0.4851,
+      "step": 396
+    },
+    {
+      "epoch": 0.8822222222222222,
+      "grad_norm": 1580367.0,
+      "learning_rate": 3.925925925925926e-05,
+      "loss": 0.5972,
+      "step": 397
+    },
+    {
+      "epoch": 0.8844444444444445,
+      "grad_norm": 388921.8125,
+      "learning_rate": 3.921810699588477e-05,
+      "loss": 0.6099,
+      "step": 398
+    },
+    {
+      "epoch": 0.8866666666666667,
+      "grad_norm": 617553.5625,
+      "learning_rate": 3.917695473251029e-05,
+      "loss": 0.5251,
+      "step": 399
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 429902.09375,
+      "learning_rate": 3.91358024691358e-05,
+      "loss": 0.5187,
+      "step": 400
+    },
+    {
+      "epoch": 0.8911111111111111,
+      "grad_norm": 225324.890625,
+      "learning_rate": 3.909465020576132e-05,
+      "loss": 0.3663,
+      "step": 401
+    },
+    {
+      "epoch": 0.8933333333333333,
+      "grad_norm": 1841588.5,
+      "learning_rate": 3.9053497942386834e-05,
+      "loss": 0.5766,
+      "step": 402
+    },
+    {
+      "epoch": 0.8955555555555555,
+      "grad_norm": 1267175.125,
+      "learning_rate": 3.901234567901234e-05,
+      "loss": 0.5858,
+      "step": 403
+    },
+    {
+      "epoch": 0.8977777777777778,
+      "grad_norm": 716295.0,
+      "learning_rate": 3.8971193415637866e-05,
+      "loss": 0.4709,
+      "step": 404
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2756277.25,
+      "learning_rate": 3.8930041152263375e-05,
+      "loss": 0.681,
+      "step": 405
+    },
+    {
+      "epoch": 0.9022222222222223,
+      "grad_norm": 1748439.125,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 0.6071,
+      "step": 406
+    },
+    {
+      "epoch": 0.9044444444444445,
+      "grad_norm": 610785.875,
+      "learning_rate": 3.88477366255144e-05,
+      "loss": 0.4121,
+      "step": 407
+    },
+    {
+      "epoch": 0.9066666666666666,
+      "grad_norm": 527703.75,
+      "learning_rate": 3.880658436213992e-05,
+      "loss": 0.6429,
+      "step": 408
+    },
+    {
+      "epoch": 0.9088888888888889,
+      "grad_norm": 385713.40625,
+      "learning_rate": 3.876543209876544e-05,
+      "loss": 0.5652,
+      "step": 409
+    },
+    {
+      "epoch": 0.9111111111111111,
+      "grad_norm": 403583.625,
+      "learning_rate": 3.872427983539095e-05,
+      "loss": 0.623,
+      "step": 410
+    },
+    {
+      "epoch": 0.9133333333333333,
+      "grad_norm": 399882.75,
+      "learning_rate": 3.868312757201646e-05,
+      "loss": 0.5394,
+      "step": 411
+    },
+    {
+      "epoch": 0.9155555555555556,
+      "grad_norm": 312631.34375,
+      "learning_rate": 3.864197530864197e-05,
+      "loss": 0.566,
+      "step": 412
+    },
+    {
+      "epoch": 0.9177777777777778,
+      "grad_norm": 184664.34375,
+      "learning_rate": 3.8600823045267495e-05,
+      "loss": 0.3998,
+      "step": 413
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 885300.0625,
+      "learning_rate": 3.8559670781893004e-05,
+      "loss": 0.6295,
+      "step": 414
+    },
+    {
+      "epoch": 0.9222222222222223,
+      "grad_norm": 157462.03125,
+      "learning_rate": 3.851851851851852e-05,
+      "loss": 0.6012,
+      "step": 415
+    },
+    {
+      "epoch": 0.9244444444444444,
+      "grad_norm": 298935.9375,
+      "learning_rate": 3.8477366255144036e-05,
+      "loss": 0.5729,
+      "step": 416
+    },
+    {
+      "epoch": 0.9266666666666666,
+      "grad_norm": 290948.75,
+      "learning_rate": 3.843621399176955e-05,
+      "loss": 0.4952,
+      "step": 417
+    },
+    {
+      "epoch": 0.9288888888888889,
+      "grad_norm": 1232251.875,
+      "learning_rate": 3.839506172839507e-05,
+      "loss": 0.4999,
+      "step": 418
+    },
+    {
+      "epoch": 0.9311111111111111,
+      "grad_norm": 441190.5,
+      "learning_rate": 3.8353909465020577e-05,
+      "loss": 0.4751,
+      "step": 419
+    },
+    {
+      "epoch": 0.9333333333333333,
+      "grad_norm": 420174.40625,
+      "learning_rate": 3.831275720164609e-05,
+      "loss": 0.569,
+      "step": 420
+    },
+    {
+      "epoch": 0.9355555555555556,
+      "grad_norm": 794650.375,
+      "learning_rate": 3.82716049382716e-05,
+      "loss": 0.5852,
+      "step": 421
+    },
+    {
+      "epoch": 0.9377777777777778,
+      "grad_norm": 436417.28125,
+      "learning_rate": 3.8230452674897124e-05,
+      "loss": 0.5244,
+      "step": 422
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 390293.96875,
+      "learning_rate": 3.818930041152264e-05,
+      "loss": 0.6959,
+      "step": 423
+    },
+    {
+      "epoch": 0.9422222222222222,
+      "grad_norm": 562045.5,
+      "learning_rate": 3.814814814814815e-05,
+      "loss": 0.4438,
+      "step": 424
+    },
+    {
+      "epoch": 0.9444444444444444,
+      "grad_norm": 345788.0625,
+      "learning_rate": 3.8106995884773665e-05,
+      "loss": 0.5413,
+      "step": 425
+    },
+    {
+      "epoch": 0.9466666666666667,
+      "grad_norm": 708525.6875,
+      "learning_rate": 3.806584362139918e-05,
+      "loss": 0.5812,
+      "step": 426
+    },
+    {
+      "epoch": 0.9488888888888889,
+      "grad_norm": 319743.25,
+      "learning_rate": 3.80246913580247e-05,
+      "loss": 0.4005,
+      "step": 427
+    },
+    {
+      "epoch": 0.9511111111111111,
+      "grad_norm": 669220.0625,
+      "learning_rate": 3.7983539094650206e-05,
+      "loss": 0.565,
+      "step": 428
+    },
+    {
+      "epoch": 0.9533333333333334,
+      "grad_norm": 683483.625,
+      "learning_rate": 3.794238683127572e-05,
+      "loss": 0.5445,
+      "step": 429
+    },
+    {
+      "epoch": 0.9555555555555556,
+      "grad_norm": 1566102.0,
+      "learning_rate": 3.790123456790123e-05,
+      "loss": 0.45,
+      "step": 430
+    },
+    {
+      "epoch": 0.9577777777777777,
+      "grad_norm": 399423.53125,
+      "learning_rate": 3.786008230452675e-05,
+      "loss": 0.4618,
+      "step": 431
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 536128.0625,
+      "learning_rate": 3.781893004115227e-05,
+      "loss": 0.6237,
+      "step": 432
+    },
+    {
+      "epoch": 0.9622222222222222,
+      "grad_norm": 632462.3125,
+      "learning_rate": 3.777777777777778e-05,
+      "loss": 0.7016,
+      "step": 433
+    },
+    {
+      "epoch": 0.9644444444444444,
+      "grad_norm": 1254881.0,
+      "learning_rate": 3.7736625514403294e-05,
+      "loss": 0.649,
+      "step": 434
+    },
+    {
+      "epoch": 0.9666666666666667,
+      "grad_norm": 2207048.75,
+      "learning_rate": 3.769547325102881e-05,
+      "loss": 0.8038,
+      "step": 435
+    },
+    {
+      "epoch": 0.9688888888888889,
+      "grad_norm": 2232515.5,
+      "learning_rate": 3.7654320987654326e-05,
+      "loss": 0.534,
+      "step": 436
+    },
+    {
+      "epoch": 0.9711111111111111,
+      "grad_norm": 425982.53125,
+      "learning_rate": 3.7613168724279835e-05,
+      "loss": 0.5403,
+      "step": 437
+    },
+    {
+      "epoch": 0.9733333333333334,
+      "grad_norm": 577288.75,
+      "learning_rate": 3.757201646090535e-05,
+      "loss": 0.581,
+      "step": 438
+    },
+    {
+      "epoch": 0.9755555555555555,
+      "grad_norm": 320666.09375,
+      "learning_rate": 3.7530864197530867e-05,
+      "loss": 0.4917,
+      "step": 439
+    },
+    {
+      "epoch": 0.9777777777777777,
+      "grad_norm": 455171.28125,
+      "learning_rate": 3.748971193415638e-05,
+      "loss": 0.639,
+      "step": 440
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 349861.1875,
+      "learning_rate": 3.74485596707819e-05,
+      "loss": 0.4139,
+      "step": 441
+    },
+    {
+      "epoch": 0.9822222222222222,
+      "grad_norm": 1753616.125,
+      "learning_rate": 3.740740740740741e-05,
+      "loss": 0.6111,
+      "step": 442
+    },
+    {
+      "epoch": 0.9844444444444445,
+      "grad_norm": 380102.625,
+      "learning_rate": 3.736625514403292e-05,
+      "loss": 0.532,
+      "step": 443
+    },
+    {
+      "epoch": 0.9866666666666667,
+      "grad_norm": 227778.578125,
+      "learning_rate": 3.732510288065844e-05,
+      "loss": 0.5705,
+      "step": 444
+    },
+    {
+      "epoch": 0.9888888888888889,
+      "grad_norm": 523431.46875,
+      "learning_rate": 3.7283950617283955e-05,
+      "loss": 0.5337,
+      "step": 445
+    },
+    {
+      "epoch": 0.9911111111111112,
+      "grad_norm": 449052.21875,
+      "learning_rate": 3.724279835390947e-05,
+      "loss": 0.5723,
+      "step": 446
+    },
+    {
+      "epoch": 0.9933333333333333,
+      "grad_norm": 1037056.75,
+      "learning_rate": 3.720164609053498e-05,
+      "loss": 0.6206,
+      "step": 447
+    },
+    {
+      "epoch": 0.9955555555555555,
+      "grad_norm": 746349.1875,
+      "learning_rate": 3.7160493827160496e-05,
+      "loss": 0.433,
+      "step": 448
+    },
+    {
+      "epoch": 0.9977777777777778,
+      "grad_norm": 247129.5,
+      "learning_rate": 3.711934156378601e-05,
+      "loss": 0.5245,
+      "step": 449
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 473643.53125,
+      "learning_rate": 3.707818930041153e-05,
+      "loss": 0.4803,
+      "step": 450
+    },
+    {
+      "epoch": 1.0022222222222221,
+      "grad_norm": 3388953.25,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 0.5192,
+      "step": 451
+    },
+    {
+      "epoch": 1.0044444444444445,
+      "grad_norm": 352336.65625,
+      "learning_rate": 3.699588477366255e-05,
+      "loss": 0.5511,
+      "step": 452
+    },
+    {
+      "epoch": 1.0066666666666666,
+      "grad_norm": 537367.0,
+      "learning_rate": 3.695473251028807e-05,
+      "loss": 0.4823,
+      "step": 453
+    },
+    {
+      "epoch": 1.008888888888889,
+      "grad_norm": 463618.0625,
+      "learning_rate": 3.6913580246913584e-05,
+      "loss": 0.6219,
+      "step": 454
+    },
+    {
+      "epoch": 1.011111111111111,
+      "grad_norm": 890988.0,
+      "learning_rate": 3.68724279835391e-05,
+      "loss": 0.6637,
+      "step": 455
+    },
+    {
+      "epoch": 1.0133333333333334,
+      "grad_norm": 834668.8125,
+      "learning_rate": 3.683127572016461e-05,
+      "loss": 0.5514,
+      "step": 456
+    },
+    {
+      "epoch": 1.0155555555555555,
+      "grad_norm": 478754.59375,
+      "learning_rate": 3.6790123456790125e-05,
+      "loss": 0.4442,
+      "step": 457
+    },
+    {
+      "epoch": 1.0177777777777777,
+      "grad_norm": 364720.5,
+      "learning_rate": 3.674897119341564e-05,
+      "loss": 0.5295,
+      "step": 458
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 437859.78125,
+      "learning_rate": 3.670781893004116e-05,
+      "loss": 0.5602,
+      "step": 459
+    },
+    {
+      "epoch": 1.0222222222222221,
+      "grad_norm": 668818.375,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 0.6702,
+      "step": 460
+    },
+    {
+      "epoch": 1.0244444444444445,
+      "grad_norm": 1322955.875,
+      "learning_rate": 3.662551440329218e-05,
+      "loss": 0.5504,
+      "step": 461
+    },
+    {
+      "epoch": 1.0266666666666666,
+      "grad_norm": 912058.6875,
+      "learning_rate": 3.65843621399177e-05,
+      "loss": 0.485,
+      "step": 462
+    },
+    {
+      "epoch": 1.028888888888889,
+      "grad_norm": 1165694.75,
+      "learning_rate": 3.654320987654321e-05,
+      "loss": 0.5364,
+      "step": 463
+    },
+    {
+      "epoch": 1.031111111111111,
+      "grad_norm": 410742.59375,
+      "learning_rate": 3.650205761316873e-05,
+      "loss": 0.642,
+      "step": 464
+    },
+    {
+      "epoch": 1.0333333333333334,
+      "grad_norm": 517038.625,
+      "learning_rate": 3.646090534979424e-05,
+      "loss": 0.6515,
+      "step": 465
+    },
+    {
+      "epoch": 1.0355555555555556,
+      "grad_norm": 642495.5,
+      "learning_rate": 3.6419753086419754e-05,
+      "loss": 0.5096,
+      "step": 466
+    },
+    {
+      "epoch": 1.0377777777777777,
+      "grad_norm": 757100.125,
+      "learning_rate": 3.637860082304527e-05,
+      "loss": 0.6411,
+      "step": 467
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 865382.5,
+      "learning_rate": 3.6337448559670786e-05,
+      "loss": 0.5356,
+      "step": 468
+    },
+    {
+      "epoch": 1.0422222222222222,
+      "grad_norm": 2302121.75,
+      "learning_rate": 3.62962962962963e-05,
+      "loss": 0.6145,
+      "step": 469
+    },
+    {
+      "epoch": 1.0444444444444445,
+      "grad_norm": 410693.78125,
+      "learning_rate": 3.625514403292181e-05,
+      "loss": 0.4845,
+      "step": 470
+    },
+    {
+      "epoch": 1.0466666666666666,
+      "grad_norm": 343380.8125,
+      "learning_rate": 3.6213991769547327e-05,
+      "loss": 0.5627,
+      "step": 471
+    },
+    {
+      "epoch": 1.048888888888889,
+      "grad_norm": 811735.8125,
+      "learning_rate": 3.617283950617284e-05,
+      "loss": 0.5982,
+      "step": 472
+    },
+    {
+      "epoch": 1.051111111111111,
+      "grad_norm": 898176.5,
+      "learning_rate": 3.613168724279836e-05,
+      "loss": 0.5189,
+      "step": 473
+    },
+    {
+      "epoch": 1.0533333333333332,
+      "grad_norm": 257329.40625,
+      "learning_rate": 3.609053497942387e-05,
+      "loss": 0.5023,
+      "step": 474
+    },
+    {
+      "epoch": 1.0555555555555556,
+      "grad_norm": 262981.125,
+      "learning_rate": 3.604938271604938e-05,
+      "loss": 0.5989,
+      "step": 475
+    },
+    {
+      "epoch": 1.0577777777777777,
+      "grad_norm": 284406.6875,
+      "learning_rate": 3.60082304526749e-05,
+      "loss": 0.5692,
+      "step": 476
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 303435.46875,
+      "learning_rate": 3.5967078189300415e-05,
+      "loss": 0.4557,
+      "step": 477
+    },
+    {
+      "epoch": 1.0622222222222222,
+      "grad_norm": 512383.0625,
+      "learning_rate": 3.592592592592593e-05,
+      "loss": 0.4438,
+      "step": 478
+    },
+    {
+      "epoch": 1.0644444444444445,
+      "grad_norm": 362393.125,
+      "learning_rate": 3.588477366255144e-05,
+      "loss": 0.5733,
+      "step": 479
+    },
+    {
+      "epoch": 1.0666666666666667,
+      "grad_norm": 419564.6875,
+      "learning_rate": 3.5843621399176956e-05,
+      "loss": 0.5342,
+      "step": 480
+    },
+    {
+      "epoch": 1.068888888888889,
+      "grad_norm": 937582.8125,
+      "learning_rate": 3.580246913580247e-05,
+      "loss": 0.5937,
+      "step": 481
+    },
+    {
+      "epoch": 1.0711111111111111,
+      "grad_norm": 617804.5625,
+      "learning_rate": 3.576131687242799e-05,
+      "loss": 0.5317,
+      "step": 482
+    },
+    {
+      "epoch": 1.0733333333333333,
+      "grad_norm": 1090921.375,
+      "learning_rate": 3.5720164609053497e-05,
+      "loss": 0.5436,
+      "step": 483
+    },
+    {
+      "epoch": 1.0755555555555556,
+      "grad_norm": 393805.53125,
+      "learning_rate": 3.567901234567901e-05,
+      "loss": 0.6203,
+      "step": 484
+    },
+    {
+      "epoch": 1.0777777777777777,
+      "grad_norm": 401942.1875,
+      "learning_rate": 3.563786008230453e-05,
+      "loss": 0.5145,
+      "step": 485
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 426967.34375,
+      "learning_rate": 3.5596707818930044e-05,
+      "loss": 0.5899,
+      "step": 486
+    },
+    {
+      "epoch": 1.0822222222222222,
+      "grad_norm": 1499625.25,
+      "learning_rate": 3.555555555555556e-05,
+      "loss": 0.5185,
+      "step": 487
+    },
+    {
+      "epoch": 1.0844444444444445,
+      "grad_norm": 119777.875,
+      "learning_rate": 3.551440329218107e-05,
+      "loss": 0.4702,
+      "step": 488
+    },
+    {
+      "epoch": 1.0866666666666667,
+      "grad_norm": 325876.65625,
+      "learning_rate": 3.5473251028806585e-05,
+      "loss": 0.4839,
+      "step": 489
+    },
+    {
+      "epoch": 1.0888888888888888,
+      "grad_norm": 1764555.875,
+      "learning_rate": 3.54320987654321e-05,
+      "loss": 0.4194,
+      "step": 490
+    },
+    {
+      "epoch": 1.0911111111111111,
+      "grad_norm": 992762.625,
+      "learning_rate": 3.539094650205762e-05,
+      "loss": 0.6011,
+      "step": 491
+    },
+    {
+      "epoch": 1.0933333333333333,
+      "grad_norm": 1109132.0,
+      "learning_rate": 3.534979423868313e-05,
+      "loss": 0.578,
+      "step": 492
+    },
+    {
+      "epoch": 1.0955555555555556,
+      "grad_norm": 338012.0625,
+      "learning_rate": 3.530864197530864e-05,
+      "loss": 0.6753,
+      "step": 493
+    },
+    {
+      "epoch": 1.0977777777777777,
+      "grad_norm": 489635.125,
+      "learning_rate": 3.526748971193416e-05,
+      "loss": 0.484,
+      "step": 494
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 467980.4375,
+      "learning_rate": 3.522633744855967e-05,
+      "loss": 0.5041,
+      "step": 495
+    },
+    {
+      "epoch": 1.1022222222222222,
+      "grad_norm": 926030.25,
+      "learning_rate": 3.518518518518519e-05,
+      "loss": 0.6303,
+      "step": 496
+    },
+    {
+      "epoch": 1.1044444444444443,
+      "grad_norm": 514718.71875,
+      "learning_rate": 3.51440329218107e-05,
+      "loss": 0.5022,
+      "step": 497
+    },
+    {
+      "epoch": 1.1066666666666667,
+      "grad_norm": 1032987.9375,
+      "learning_rate": 3.5102880658436214e-05,
+      "loss": 0.5416,
+      "step": 498
+    },
+    {
+      "epoch": 1.1088888888888888,
+      "grad_norm": 743004.375,
+      "learning_rate": 3.506172839506173e-05,
+      "loss": 0.4624,
+      "step": 499
+    },
+    {
+      "epoch": 1.1111111111111112,
+      "grad_norm": 542330.0625,
+      "learning_rate": 3.5020576131687246e-05,
+      "loss": 0.6888,
+      "step": 500
+    },
+    {
+      "epoch": 1.1133333333333333,
+      "grad_norm": 956336.0,
+      "learning_rate": 3.497942386831276e-05,
+      "loss": 0.4579,
+      "step": 501
+    },
+    {
+      "epoch": 1.1155555555555556,
+      "grad_norm": 445892.0625,
+      "learning_rate": 3.493827160493827e-05,
+      "loss": 0.5064,
+      "step": 502
+    },
+    {
+      "epoch": 1.1177777777777778,
+      "grad_norm": 200526.984375,
+      "learning_rate": 3.4897119341563787e-05,
+      "loss": 0.4721,
+      "step": 503
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2516397.75,
+      "learning_rate": 3.48559670781893e-05,
+      "loss": 0.5609,
+      "step": 504
+    },
+    {
+      "epoch": 1.1222222222222222,
+      "grad_norm": 561231.625,
+      "learning_rate": 3.481481481481482e-05,
+      "loss": 0.5566,
+      "step": 505
+    },
+    {
+      "epoch": 1.1244444444444444,
+      "grad_norm": 1745718.375,
+      "learning_rate": 3.4773662551440334e-05,
+      "loss": 0.5487,
+      "step": 506
+    },
+    {
+      "epoch": 1.1266666666666667,
+      "grad_norm": 233383.96875,
+      "learning_rate": 3.473251028806584e-05,
+      "loss": 0.5369,
+      "step": 507
+    },
+    {
+      "epoch": 1.1288888888888888,
+      "grad_norm": 681536.8125,
+      "learning_rate": 3.469135802469136e-05,
+      "loss": 0.6202,
+      "step": 508
+    },
+    {
+      "epoch": 1.1311111111111112,
+      "grad_norm": 460907.3125,
+      "learning_rate": 3.4650205761316875e-05,
+      "loss": 0.6093,
+      "step": 509
+    },
+    {
+      "epoch": 1.1333333333333333,
+      "grad_norm": 4109765.0,
+      "learning_rate": 3.460905349794239e-05,
+      "loss": 0.678,
+      "step": 510
+    },
+    {
+      "epoch": 1.1355555555555557,
+      "grad_norm": 306640.21875,
+      "learning_rate": 3.45679012345679e-05,
+      "loss": 0.5777,
+      "step": 511
+    },
+    {
+      "epoch": 1.1377777777777778,
+      "grad_norm": 1322950.75,
+      "learning_rate": 3.4526748971193416e-05,
+      "loss": 0.5124,
+      "step": 512
+    },
+    {
+      "epoch": 1.1400000000000001,
+      "grad_norm": 820002.3125,
+      "learning_rate": 3.448559670781893e-05,
+      "loss": 0.5038,
+      "step": 513
+    },
+    {
+      "epoch": 1.1422222222222222,
+      "grad_norm": 1089420.5,
+      "learning_rate": 3.444444444444445e-05,
+      "loss": 0.512,
+      "step": 514
+    },
+    {
+      "epoch": 1.1444444444444444,
+      "grad_norm": 954156.25,
+      "learning_rate": 3.440329218106996e-05,
+      "loss": 0.5496,
+      "step": 515
+    },
+    {
+      "epoch": 1.1466666666666667,
+      "grad_norm": 667740.8125,
+      "learning_rate": 3.436213991769547e-05,
+      "loss": 0.5691,
+      "step": 516
+    },
+    {
+      "epoch": 1.1488888888888888,
+      "grad_norm": 1452467.375,
+      "learning_rate": 3.432098765432099e-05,
+      "loss": 0.4714,
+      "step": 517
+    },
+    {
+      "epoch": 1.1511111111111112,
+      "grad_norm": 875589.75,
+      "learning_rate": 3.4279835390946504e-05,
+      "loss": 0.535,
+      "step": 518
+    },
+    {
+      "epoch": 1.1533333333333333,
+      "grad_norm": 127607.7421875,
+      "learning_rate": 3.423868312757202e-05,
+      "loss": 0.4445,
+      "step": 519
+    },
+    {
+      "epoch": 1.1555555555555554,
+      "grad_norm": 533242.5,
+      "learning_rate": 3.419753086419753e-05,
+      "loss": 0.5453,
+      "step": 520
+    },
+    {
+      "epoch": 1.1577777777777778,
+      "grad_norm": 1075770.25,
+      "learning_rate": 3.4156378600823045e-05,
+      "loss": 0.5702,
+      "step": 521
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 364981.03125,
+      "learning_rate": 3.411522633744856e-05,
+      "loss": 0.5414,
+      "step": 522
+    },
+    {
+      "epoch": 1.1622222222222223,
+      "grad_norm": 588666.875,
+      "learning_rate": 3.4074074074074077e-05,
+      "loss": 0.5894,
+      "step": 523
+    },
+    {
+      "epoch": 1.1644444444444444,
+      "grad_norm": 2711980.25,
+      "learning_rate": 3.403292181069959e-05,
+      "loss": 0.5031,
+      "step": 524
+    },
+    {
+      "epoch": 1.1666666666666667,
+      "grad_norm": 2523403.0,
+      "learning_rate": 3.39917695473251e-05,
+      "loss": 0.6124,
+      "step": 525
+    },
+    {
+      "epoch": 1.1688888888888889,
+      "grad_norm": 190597.90625,
+      "learning_rate": 3.395061728395062e-05,
+      "loss": 0.5361,
+      "step": 526
+    },
+    {
+      "epoch": 1.1711111111111112,
+      "grad_norm": 103686.34375,
+      "learning_rate": 3.390946502057613e-05,
+      "loss": 0.5721,
+      "step": 527
+    },
+    {
+      "epoch": 1.1733333333333333,
+      "grad_norm": 684715.5625,
+      "learning_rate": 3.386831275720165e-05,
+      "loss": 0.5941,
+      "step": 528
+    },
+    {
+      "epoch": 1.1755555555555555,
+      "grad_norm": 643524.8125,
+      "learning_rate": 3.3827160493827165e-05,
+      "loss": 0.5221,
+      "step": 529
+    },
+    {
+      "epoch": 1.1777777777777778,
+      "grad_norm": 951965.1875,
+      "learning_rate": 3.3786008230452674e-05,
+      "loss": 0.6495,
+      "step": 530
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1729085.0,
+      "learning_rate": 3.374485596707819e-05,
+      "loss": 0.4682,
+      "step": 531
+    },
+    {
+      "epoch": 1.1822222222222223,
+      "grad_norm": 355360.1875,
+      "learning_rate": 3.3703703703703706e-05,
+      "loss": 0.6197,
+      "step": 532
+    },
+    {
+      "epoch": 1.1844444444444444,
+      "grad_norm": 496380.0625,
+      "learning_rate": 3.366255144032922e-05,
+      "loss": 0.444,
+      "step": 533
+    },
+    {
+      "epoch": 1.1866666666666668,
+      "grad_norm": 409551.0,
+      "learning_rate": 3.362139917695473e-05,
+      "loss": 0.5369,
+      "step": 534
+    },
+    {
+      "epoch": 1.1888888888888889,
+      "grad_norm": 392581.9375,
+      "learning_rate": 3.3580246913580247e-05,
+      "loss": 0.7495,
+      "step": 535
+    },
+    {
+      "epoch": 1.1911111111111112,
+      "grad_norm": 881165.0,
+      "learning_rate": 3.353909465020576e-05,
+      "loss": 0.5686,
+      "step": 536
+    },
+    {
+      "epoch": 1.1933333333333334,
+      "grad_norm": 1053088.625,
+      "learning_rate": 3.349794238683128e-05,
+      "loss": 0.4673,
+      "step": 537
+    },
+    {
+      "epoch": 1.1955555555555555,
+      "grad_norm": 1257165.0,
+      "learning_rate": 3.3456790123456794e-05,
+      "loss": 0.4615,
+      "step": 538
+    },
+    {
+      "epoch": 1.1977777777777778,
+      "grad_norm": 2032428.375,
+      "learning_rate": 3.34156378600823e-05,
+      "loss": 0.5234,
+      "step": 539
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 397717.59375,
+      "learning_rate": 3.337448559670782e-05,
+      "loss": 0.6048,
+      "step": 540
+    },
+    {
+      "epoch": 1.2022222222222223,
+      "grad_norm": 470372.0625,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.5376,
+      "step": 541
+    },
+    {
+      "epoch": 1.2044444444444444,
+      "grad_norm": 340338.21875,
+      "learning_rate": 3.329218106995885e-05,
+      "loss": 0.5824,
+      "step": 542
+    },
+    {
+      "epoch": 1.2066666666666666,
+      "grad_norm": 1100316.0,
+      "learning_rate": 3.325102880658436e-05,
+      "loss": 0.6534,
+      "step": 543
+    },
+    {
+      "epoch": 1.208888888888889,
+      "grad_norm": 586183.625,
+      "learning_rate": 3.3209876543209876e-05,
+      "loss": 0.5192,
+      "step": 544
+    },
+    {
+      "epoch": 1.211111111111111,
+      "grad_norm": 187557.046875,
+      "learning_rate": 3.316872427983539e-05,
+      "loss": 0.5388,
+      "step": 545
+    },
+    {
+      "epoch": 1.2133333333333334,
+      "grad_norm": 821299.6875,
+      "learning_rate": 3.312757201646091e-05,
+      "loss": 0.6922,
+      "step": 546
+    },
+    {
+      "epoch": 1.2155555555555555,
+      "grad_norm": 202463.78125,
+      "learning_rate": 3.308641975308642e-05,
+      "loss": 0.4871,
+      "step": 547
+    },
+    {
+      "epoch": 1.2177777777777778,
+      "grad_norm": 827578.6875,
+      "learning_rate": 3.304526748971193e-05,
+      "loss": 0.6096,
+      "step": 548
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1085792.125,
+      "learning_rate": 3.300411522633745e-05,
+      "loss": 0.539,
+      "step": 549
+    },
+    {
+      "epoch": 1.2222222222222223,
+      "grad_norm": 313487.625,
+      "learning_rate": 3.2962962962962964e-05,
+      "loss": 0.4972,
+      "step": 550
+    },
+    {
+      "epoch": 1.2244444444444444,
+      "grad_norm": 1085682.25,
+      "learning_rate": 3.292181069958848e-05,
+      "loss": 0.4936,
+      "step": 551
+    },
+    {
+      "epoch": 1.2266666666666666,
+      "grad_norm": 796879.1875,
+      "learning_rate": 3.2880658436213996e-05,
+      "loss": 0.5272,
+      "step": 552
+    },
+    {
+      "epoch": 1.228888888888889,
+      "grad_norm": 1149724.125,
+      "learning_rate": 3.2839506172839505e-05,
+      "loss": 0.4691,
+      "step": 553
+    },
+    {
+      "epoch": 1.231111111111111,
+      "grad_norm": 887672.6875,
+      "learning_rate": 3.279835390946502e-05,
+      "loss": 0.5061,
+      "step": 554
+    },
+    {
+      "epoch": 1.2333333333333334,
+      "grad_norm": 1085029.75,
+      "learning_rate": 3.2757201646090537e-05,
+      "loss": 0.6217,
+      "step": 555
+    },
+    {
+      "epoch": 1.2355555555555555,
+      "grad_norm": 1328316.75,
+      "learning_rate": 3.271604938271605e-05,
+      "loss": 0.5258,
+      "step": 556
+    },
+    {
+      "epoch": 1.2377777777777779,
+      "grad_norm": 478540.59375,
+      "learning_rate": 3.267489711934156e-05,
+      "loss": 0.5077,
+      "step": 557
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 516354.71875,
+      "learning_rate": 3.263374485596708e-05,
+      "loss": 0.596,
+      "step": 558
+    },
+    {
+      "epoch": 1.2422222222222223,
+      "grad_norm": 654598.375,
+      "learning_rate": 3.25925925925926e-05,
+      "loss": 0.5235,
+      "step": 559
+    },
+    {
+      "epoch": 1.2444444444444445,
+      "grad_norm": 277117.6875,
+      "learning_rate": 3.255144032921811e-05,
+      "loss": 0.4945,
+      "step": 560
+    },
+    {
+      "epoch": 1.2466666666666666,
+      "grad_norm": 1819261.25,
+      "learning_rate": 3.2510288065843625e-05,
+      "loss": 0.6248,
+      "step": 561
+    },
+    {
+      "epoch": 1.248888888888889,
+      "grad_norm": 691423.875,
+      "learning_rate": 3.2469135802469134e-05,
+      "loss": 0.4982,
+      "step": 562
+    },
+    {
+      "epoch": 1.251111111111111,
+      "grad_norm": 409888.90625,
+      "learning_rate": 3.242798353909465e-05,
+      "loss": 0.5005,
+      "step": 563
+    },
+    {
+      "epoch": 1.2533333333333334,
+      "grad_norm": 337259.59375,
+      "learning_rate": 3.2386831275720166e-05,
+      "loss": 0.4968,
+      "step": 564
+    },
+    {
+      "epoch": 1.2555555555555555,
+      "grad_norm": 465535.34375,
+      "learning_rate": 3.234567901234568e-05,
+      "loss": 0.6587,
+      "step": 565
+    },
+    {
+      "epoch": 1.2577777777777777,
+      "grad_norm": 1500356.25,
+      "learning_rate": 3.230452674897119e-05,
+      "loss": 0.5014,
+      "step": 566
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 209156.453125,
+      "learning_rate": 3.2263374485596707e-05,
+      "loss": 0.4308,
+      "step": 567
+    },
+    {
+      "epoch": 1.2622222222222224,
+      "grad_norm": 329407.0625,
+      "learning_rate": 3.222222222222223e-05,
+      "loss": 0.65,
+      "step": 568
+    },
+    {
+      "epoch": 1.2644444444444445,
+      "grad_norm": 944331.375,
+      "learning_rate": 3.218106995884774e-05,
+      "loss": 0.5737,
+      "step": 569
+    },
+    {
+      "epoch": 1.2666666666666666,
+      "grad_norm": 1461638.5,
+      "learning_rate": 3.2139917695473254e-05,
+      "loss": 0.5372,
+      "step": 570
+    },
+    {
+      "epoch": 1.268888888888889,
+      "grad_norm": 629504.4375,
+      "learning_rate": 3.209876543209876e-05,
+      "loss": 0.6163,
+      "step": 571
+    },
+    {
+      "epoch": 1.271111111111111,
+      "grad_norm": 1049412.5,
+      "learning_rate": 3.205761316872428e-05,
+      "loss": 0.5535,
+      "step": 572
+    },
+    {
+      "epoch": 1.2733333333333334,
+      "grad_norm": 238209.34375,
+      "learning_rate": 3.2016460905349795e-05,
+      "loss": 0.4736,
+      "step": 573
+    },
+    {
+      "epoch": 1.2755555555555556,
+      "grad_norm": 1106967.625,
+      "learning_rate": 3.197530864197531e-05,
+      "loss": 0.5857,
+      "step": 574
+    },
+    {
+      "epoch": 1.2777777777777777,
+      "grad_norm": 258857.421875,
+      "learning_rate": 3.193415637860083e-05,
+      "loss": 0.4707,
+      "step": 575
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 653654.0625,
+      "learning_rate": 3.1893004115226336e-05,
+      "loss": 0.531,
+      "step": 576
+    },
+    {
+      "epoch": 1.2822222222222222,
+      "grad_norm": 614772.8125,
+      "learning_rate": 3.185185185185185e-05,
+      "loss": 0.4341,
+      "step": 577
+    },
+    {
+      "epoch": 1.2844444444444445,
+      "grad_norm": 867558.6875,
+      "learning_rate": 3.181069958847737e-05,
+      "loss": 0.5906,
+      "step": 578
+    },
+    {
+      "epoch": 1.2866666666666666,
+      "grad_norm": 567176.9375,
+      "learning_rate": 3.176954732510288e-05,
+      "loss": 0.5396,
+      "step": 579
+    },
+    {
+      "epoch": 1.2888888888888888,
+      "grad_norm": 438535.9375,
+      "learning_rate": 3.172839506172839e-05,
+      "loss": 0.5703,
+      "step": 580
+    },
+    {
+      "epoch": 1.291111111111111,
+      "grad_norm": 1499955.25,
+      "learning_rate": 3.168724279835391e-05,
+      "loss": 0.4399,
+      "step": 581
+    },
+    {
+      "epoch": 1.2933333333333334,
+      "grad_norm": 141409.5,
+      "learning_rate": 3.164609053497943e-05,
+      "loss": 0.6679,
+      "step": 582
+    },
+    {
+      "epoch": 1.2955555555555556,
+      "grad_norm": 398965.0625,
+      "learning_rate": 3.160493827160494e-05,
+      "loss": 0.4923,
+      "step": 583
+    },
+    {
+      "epoch": 1.2977777777777777,
+      "grad_norm": 2154293.5,
+      "learning_rate": 3.1563786008230456e-05,
+      "loss": 0.5707,
+      "step": 584
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 456048.875,
+      "learning_rate": 3.1522633744855965e-05,
+      "loss": 0.6075,
+      "step": 585
+    },
+    {
+      "epoch": 1.3022222222222222,
+      "grad_norm": 543668.875,
+      "learning_rate": 3.148148148148148e-05,
+      "loss": 0.5149,
+      "step": 586
+    },
+    {
+      "epoch": 1.3044444444444445,
+      "grad_norm": 539372.8125,
+      "learning_rate": 3.1440329218106997e-05,
+      "loss": 0.6545,
+      "step": 587
+    },
+    {
+      "epoch": 1.3066666666666666,
+      "grad_norm": 504076.53125,
+      "learning_rate": 3.139917695473251e-05,
+      "loss": 0.7466,
+      "step": 588
+    },
+    {
+      "epoch": 1.3088888888888888,
+      "grad_norm": 400669.0,
+      "learning_rate": 3.135802469135803e-05,
+      "loss": 0.6429,
+      "step": 589
+    },
+    {
+      "epoch": 1.3111111111111111,
+      "grad_norm": 614820.4375,
+      "learning_rate": 3.131687242798354e-05,
+      "loss": 0.4616,
+      "step": 590
+    },
+    {
+      "epoch": 1.3133333333333335,
+      "grad_norm": 187871.6875,
+      "learning_rate": 3.127572016460906e-05,
+      "loss": 0.5457,
+      "step": 591
+    },
+    {
+      "epoch": 1.3155555555555556,
+      "grad_norm": 322704.03125,
+      "learning_rate": 3.123456790123457e-05,
+      "loss": 0.492,
+      "step": 592
+    },
+    {
+      "epoch": 1.3177777777777777,
+      "grad_norm": 509237.46875,
+      "learning_rate": 3.1193415637860085e-05,
+      "loss": 0.5478,
+      "step": 593
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 953086.1875,
+      "learning_rate": 3.1152263374485594e-05,
+      "loss": 0.515,
+      "step": 594
+    },
+    {
+      "epoch": 1.3222222222222222,
+      "grad_norm": 607668.625,
+      "learning_rate": 3.111111111111111e-05,
+      "loss": 0.5707,
+      "step": 595
+    },
+    {
+      "epoch": 1.3244444444444445,
+      "grad_norm": 2008874.0,
+      "learning_rate": 3.1069958847736626e-05,
+      "loss": 0.6644,
+      "step": 596
+    },
+    {
+      "epoch": 1.3266666666666667,
+      "grad_norm": 439735.1875,
+      "learning_rate": 3.102880658436214e-05,
+      "loss": 0.499,
+      "step": 597
+    },
+    {
+      "epoch": 1.3288888888888888,
+      "grad_norm": 635145.625,
+      "learning_rate": 3.098765432098766e-05,
+      "loss": 0.6213,
+      "step": 598
+    },
+    {
+      "epoch": 1.3311111111111111,
+      "grad_norm": 1621658.25,
+      "learning_rate": 3.0946502057613167e-05,
+      "loss": 0.5868,
+      "step": 599
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 588194.4375,
+      "learning_rate": 3.090534979423869e-05,
+      "loss": 0.6255,
+      "step": 600
+    },
+    {
+      "epoch": 1.3355555555555556,
+      "grad_norm": 644249.375,
+      "learning_rate": 3.08641975308642e-05,
+      "loss": 0.5518,
+      "step": 601
+    },
+    {
+      "epoch": 1.3377777777777777,
+      "grad_norm": 165082.265625,
+      "learning_rate": 3.0823045267489714e-05,
+      "loss": 0.5313,
+      "step": 602
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1069945.5,
+      "learning_rate": 3.078189300411522e-05,
+      "loss": 0.5828,
+      "step": 603
+    },
+    {
+      "epoch": 1.3422222222222222,
+      "grad_norm": 504896.5,
+      "learning_rate": 3.074074074074074e-05,
+      "loss": 0.4712,
+      "step": 604
+    },
+    {
+      "epoch": 1.3444444444444446,
+      "grad_norm": 1039296.625,
+      "learning_rate": 3.069958847736626e-05,
+      "loss": 0.5231,
+      "step": 605
+    },
+    {
+      "epoch": 1.3466666666666667,
+      "grad_norm": 1078974.625,
+      "learning_rate": 3.065843621399177e-05,
+      "loss": 0.5408,
+      "step": 606
+    },
+    {
+      "epoch": 1.3488888888888888,
+      "grad_norm": 266013.1875,
+      "learning_rate": 3.061728395061729e-05,
+      "loss": 0.5208,
+      "step": 607
+    },
+    {
+      "epoch": 1.3511111111111112,
+      "grad_norm": 245803.53125,
+      "learning_rate": 3.0576131687242796e-05,
+      "loss": 0.5754,
+      "step": 608
+    },
+    {
+      "epoch": 1.3533333333333333,
+      "grad_norm": 1278539.875,
+      "learning_rate": 3.053497942386832e-05,
+      "loss": 0.6466,
+      "step": 609
+    },
+    {
+      "epoch": 1.3555555555555556,
+      "grad_norm": 686136.625,
+      "learning_rate": 3.0493827160493827e-05,
+      "loss": 0.5953,
+      "step": 610
+    },
+    {
+      "epoch": 1.3577777777777778,
+      "grad_norm": 606737.0625,
+      "learning_rate": 3.0452674897119343e-05,
+      "loss": 0.4736,
+      "step": 611
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 1258263.25,
+      "learning_rate": 3.041152263374486e-05,
+      "loss": 0.5339,
+      "step": 612
+    },
+    {
+      "epoch": 1.3622222222222222,
+      "grad_norm": 1875020.75,
+      "learning_rate": 3.037037037037037e-05,
+      "loss": 0.5166,
+      "step": 613
+    },
+    {
+      "epoch": 1.3644444444444446,
+      "grad_norm": 249718.53125,
+      "learning_rate": 3.0329218106995887e-05,
+      "loss": 0.7374,
+      "step": 614
+    },
+    {
+      "epoch": 1.3666666666666667,
+      "grad_norm": 245895.125,
+      "learning_rate": 3.02880658436214e-05,
+      "loss": 0.5157,
+      "step": 615
+    },
+    {
+      "epoch": 1.3688888888888888,
+      "grad_norm": 945831.4375,
+      "learning_rate": 3.0246913580246916e-05,
+      "loss": 0.5125,
+      "step": 616
+    },
+    {
+      "epoch": 1.3711111111111112,
+      "grad_norm": 1526112.375,
+      "learning_rate": 3.0205761316872428e-05,
+      "loss": 0.5882,
+      "step": 617
+    },
+    {
+      "epoch": 1.3733333333333333,
+      "grad_norm": 492628.59375,
+      "learning_rate": 3.0164609053497944e-05,
+      "loss": 0.7338,
+      "step": 618
+    },
+    {
+      "epoch": 1.3755555555555556,
+      "grad_norm": 1887908.0,
+      "learning_rate": 3.012345679012346e-05,
+      "loss": 0.5683,
+      "step": 619
+    },
+    {
+      "epoch": 1.3777777777777778,
+      "grad_norm": 4119723.5,
+      "learning_rate": 3.0082304526748972e-05,
+      "loss": 0.4499,
+      "step": 620
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 196303.25,
+      "learning_rate": 3.0041152263374488e-05,
+      "loss": 0.5139,
+      "step": 621
+    },
+    {
+      "epoch": 1.3822222222222222,
+      "grad_norm": 402786.03125,
+      "learning_rate": 3e-05,
+      "loss": 0.5233,
+      "step": 622
+    },
+    {
+      "epoch": 1.3844444444444444,
+      "grad_norm": 1017483.0,
+      "learning_rate": 2.9958847736625517e-05,
+      "loss": 0.5011,
+      "step": 623
+    },
+    {
+      "epoch": 1.3866666666666667,
+      "grad_norm": 488110.4375,
+      "learning_rate": 2.991769547325103e-05,
+      "loss": 0.5281,
+      "step": 624
+    },
+    {
+      "epoch": 1.3888888888888888,
+      "grad_norm": 975504.0,
+      "learning_rate": 2.9876543209876545e-05,
+      "loss": 0.4671,
+      "step": 625
+    },
+    {
+      "epoch": 1.3911111111111112,
+      "grad_norm": 373597.0625,
+      "learning_rate": 2.9835390946502057e-05,
+      "loss": 0.4774,
+      "step": 626
+    },
+    {
+      "epoch": 1.3933333333333333,
+      "grad_norm": 388159.65625,
+      "learning_rate": 2.9794238683127573e-05,
+      "loss": 0.6625,
+      "step": 627
+    },
+    {
+      "epoch": 1.3955555555555557,
+      "grad_norm": 784697.4375,
+      "learning_rate": 2.975308641975309e-05,
+      "loss": 0.5849,
+      "step": 628
+    },
+    {
+      "epoch": 1.3977777777777778,
+      "grad_norm": 379587.84375,
+      "learning_rate": 2.97119341563786e-05,
+      "loss": 0.4771,
+      "step": 629
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2868313.5,
+      "learning_rate": 2.9670781893004117e-05,
+      "loss": 0.5185,
+      "step": 630
+    },
+    {
+      "epoch": 1.4022222222222223,
+      "grad_norm": 1386558.5,
+      "learning_rate": 2.962962962962963e-05,
+      "loss": 0.5149,
+      "step": 631
+    },
+    {
+      "epoch": 1.4044444444444444,
+      "grad_norm": 368961.875,
+      "learning_rate": 2.9588477366255146e-05,
+      "loss": 0.5174,
+      "step": 632
+    },
+    {
+      "epoch": 1.4066666666666667,
+      "grad_norm": 503174.09375,
+      "learning_rate": 2.9547325102880658e-05,
+      "loss": 0.5064,
+      "step": 633
+    },
+    {
+      "epoch": 1.4088888888888889,
+      "grad_norm": 1131053.25,
+      "learning_rate": 2.9506172839506174e-05,
+      "loss": 0.7157,
+      "step": 634
+    },
+    {
+      "epoch": 1.411111111111111,
+      "grad_norm": 1685755.625,
+      "learning_rate": 2.946502057613169e-05,
+      "loss": 0.4297,
+      "step": 635
+    },
+    {
+      "epoch": 1.4133333333333333,
+      "grad_norm": 462870.75,
+      "learning_rate": 2.9423868312757202e-05,
+      "loss": 0.5453,
+      "step": 636
+    },
+    {
+      "epoch": 1.4155555555555557,
+      "grad_norm": 663891.375,
+      "learning_rate": 2.9382716049382718e-05,
+      "loss": 0.5529,
+      "step": 637
+    },
+    {
+      "epoch": 1.4177777777777778,
+      "grad_norm": 365754.65625,
+      "learning_rate": 2.934156378600823e-05,
+      "loss": 0.652,
+      "step": 638
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 218651.53125,
+      "learning_rate": 2.9300411522633747e-05,
+      "loss": 0.6578,
+      "step": 639
+    },
+    {
+      "epoch": 1.4222222222222223,
+      "grad_norm": 701713.8125,
+      "learning_rate": 2.925925925925926e-05,
+      "loss": 0.5603,
+      "step": 640
+    },
+    {
+      "epoch": 1.4244444444444444,
+      "grad_norm": 571417.1875,
+      "learning_rate": 2.9218106995884775e-05,
+      "loss": 0.4952,
+      "step": 641
+    },
+    {
+      "epoch": 1.4266666666666667,
+      "grad_norm": 3248552.75,
+      "learning_rate": 2.917695473251029e-05,
+      "loss": 0.6064,
+      "step": 642
+    },
+    {
+      "epoch": 1.4288888888888889,
+      "grad_norm": 284121.59375,
+      "learning_rate": 2.9135802469135803e-05,
+      "loss": 0.3896,
+      "step": 643
+    },
+    {
+      "epoch": 1.431111111111111,
+      "grad_norm": 1278210.125,
+      "learning_rate": 2.909465020576132e-05,
+      "loss": 0.5072,
+      "step": 644
+    },
+    {
+      "epoch": 1.4333333333333333,
+      "grad_norm": 436054.90625,
+      "learning_rate": 2.905349794238683e-05,
+      "loss": 0.6116,
+      "step": 645
+    },
+    {
+      "epoch": 1.4355555555555555,
+      "grad_norm": 1549142.5,
+      "learning_rate": 2.9012345679012347e-05,
+      "loss": 0.543,
+      "step": 646
+    },
+    {
+      "epoch": 1.4377777777777778,
+      "grad_norm": 1060773.625,
+      "learning_rate": 2.897119341563786e-05,
+      "loss": 0.6623,
+      "step": 647
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2348947.5,
+      "learning_rate": 2.8930041152263376e-05,
+      "loss": 0.5165,
+      "step": 648
+    },
+    {
+      "epoch": 1.4422222222222223,
+      "grad_norm": 732758.875,
+      "learning_rate": 2.8888888888888888e-05,
+      "loss": 0.5677,
+      "step": 649
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 2372214.75,
+      "learning_rate": 2.8847736625514404e-05,
+      "loss": 0.6804,
+      "step": 650
+    },
+    {
+      "epoch": 1.4466666666666668,
+      "grad_norm": 340780.15625,
+      "learning_rate": 2.880658436213992e-05,
+      "loss": 0.4305,
+      "step": 651
+    },
+    {
+      "epoch": 1.448888888888889,
+      "grad_norm": 168099.875,
+      "learning_rate": 2.8765432098765432e-05,
+      "loss": 0.5265,
+      "step": 652
+    },
+    {
+      "epoch": 1.451111111111111,
+      "grad_norm": 1534431.0,
+      "learning_rate": 2.8724279835390948e-05,
+      "loss": 0.6421,
+      "step": 653
+    },
+    {
+      "epoch": 1.4533333333333334,
+      "grad_norm": 447703.84375,
+      "learning_rate": 2.868312757201646e-05,
+      "loss": 0.5125,
+      "step": 654
+    },
+    {
+      "epoch": 1.4555555555555555,
+      "grad_norm": 195473.515625,
+      "learning_rate": 2.8641975308641977e-05,
+      "loss": 0.4829,
+      "step": 655
+    },
+    {
+      "epoch": 1.4577777777777778,
+      "grad_norm": 266345.78125,
+      "learning_rate": 2.860082304526749e-05,
+      "loss": 0.4782,
+      "step": 656
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 300344.65625,
+      "learning_rate": 2.8559670781893005e-05,
+      "loss": 0.4272,
+      "step": 657
+    },
+    {
+      "epoch": 1.462222222222222,
+      "grad_norm": 420080.15625,
+      "learning_rate": 2.851851851851852e-05,
+      "loss": 0.4432,
+      "step": 658
+    },
+    {
+      "epoch": 1.4644444444444444,
+      "grad_norm": 769079.1875,
+      "learning_rate": 2.8477366255144033e-05,
+      "loss": 0.4976,
+      "step": 659
+    },
+    {
+      "epoch": 1.4666666666666668,
+      "grad_norm": 625666.25,
+      "learning_rate": 2.843621399176955e-05,
+      "loss": 0.6125,
+      "step": 660
+    },
+    {
+      "epoch": 1.468888888888889,
+      "grad_norm": 772571.125,
+      "learning_rate": 2.839506172839506e-05,
+      "loss": 0.6447,
+      "step": 661
+    },
+    {
+      "epoch": 1.471111111111111,
+      "grad_norm": 601890.8125,
+      "learning_rate": 2.8353909465020577e-05,
+      "loss": 0.6098,
+      "step": 662
+    },
+    {
+      "epoch": 1.4733333333333334,
+      "grad_norm": 1934153.875,
+      "learning_rate": 2.831275720164609e-05,
+      "loss": 0.4742,
+      "step": 663
+    },
+    {
+      "epoch": 1.4755555555555555,
+      "grad_norm": 423096.5625,
+      "learning_rate": 2.8271604938271606e-05,
+      "loss": 0.6821,
+      "step": 664
+    },
+    {
+      "epoch": 1.4777777777777779,
+      "grad_norm": 312746.1875,
+      "learning_rate": 2.823045267489712e-05,
+      "loss": 0.4722,
+      "step": 665
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 931891.6875,
+      "learning_rate": 2.8189300411522634e-05,
+      "loss": 0.4514,
+      "step": 666
+    },
+    {
+      "epoch": 1.482222222222222,
+      "grad_norm": 294852.0625,
+      "learning_rate": 2.814814814814815e-05,
+      "loss": 0.5212,
+      "step": 667
+    },
+    {
+      "epoch": 1.4844444444444445,
+      "grad_norm": 1102205.625,
+      "learning_rate": 2.8106995884773662e-05,
+      "loss": 0.5996,
+      "step": 668
+    },
+    {
+      "epoch": 1.4866666666666668,
+      "grad_norm": 339469.875,
+      "learning_rate": 2.8065843621399178e-05,
+      "loss": 0.5537,
+      "step": 669
+    },
+    {
+      "epoch": 1.488888888888889,
+      "grad_norm": 285161.625,
+      "learning_rate": 2.802469135802469e-05,
+      "loss": 0.5498,
+      "step": 670
+    },
+    {
+      "epoch": 1.491111111111111,
+      "grad_norm": 613179.125,
+      "learning_rate": 2.7983539094650207e-05,
+      "loss": 0.4538,
+      "step": 671
+    },
+    {
+      "epoch": 1.4933333333333334,
+      "grad_norm": 419371.625,
+      "learning_rate": 2.7942386831275726e-05,
+      "loss": 0.5355,
+      "step": 672
+    },
+    {
+      "epoch": 1.4955555555555555,
+      "grad_norm": 1493819.25,
+      "learning_rate": 2.7901234567901235e-05,
+      "loss": 0.6273,
+      "step": 673
+    },
+    {
+      "epoch": 1.4977777777777779,
+      "grad_norm": 749782.5,
+      "learning_rate": 2.786008230452675e-05,
+      "loss": 0.5511,
+      "step": 674
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 417377.625,
+      "learning_rate": 2.7818930041152263e-05,
+      "loss": 0.496,
+      "step": 675
+    },
+    {
+      "epoch": 1.5022222222222221,
+      "grad_norm": 446522.75,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.4735,
+      "step": 676
+    },
+    {
+      "epoch": 1.5044444444444445,
+      "grad_norm": 1023378.1875,
+      "learning_rate": 2.773662551440329e-05,
+      "loss": 0.6545,
+      "step": 677
+    },
+    {
+      "epoch": 1.5066666666666668,
+      "grad_norm": 620501.0625,
+      "learning_rate": 2.7695473251028807e-05,
+      "loss": 0.5341,
+      "step": 678
+    },
+    {
+      "epoch": 1.508888888888889,
+      "grad_norm": 409442.4375,
+      "learning_rate": 2.765432098765432e-05,
+      "loss": 0.6423,
+      "step": 679
+    },
+    {
+      "epoch": 1.511111111111111,
+      "grad_norm": 1429912.75,
+      "learning_rate": 2.7613168724279836e-05,
+      "loss": 0.5688,
+      "step": 680
+    },
+    {
+      "epoch": 1.5133333333333332,
+      "grad_norm": 851064.125,
+      "learning_rate": 2.757201646090535e-05,
+      "loss": 0.6428,
+      "step": 681
+    },
+    {
+      "epoch": 1.5155555555555555,
+      "grad_norm": 222915.484375,
+      "learning_rate": 2.7530864197530864e-05,
+      "loss": 0.5159,
+      "step": 682
+    },
+    {
+      "epoch": 1.517777777777778,
+      "grad_norm": 327829.65625,
+      "learning_rate": 2.748971193415638e-05,
+      "loss": 0.4942,
+      "step": 683
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 694012.4375,
+      "learning_rate": 2.7448559670781892e-05,
+      "loss": 0.6507,
+      "step": 684
+    },
+    {
+      "epoch": 1.5222222222222221,
+      "grad_norm": 1516590.375,
+      "learning_rate": 2.7407407407407408e-05,
+      "loss": 0.5754,
+      "step": 685
+    },
+    {
+      "epoch": 1.5244444444444445,
+      "grad_norm": 960898.75,
+      "learning_rate": 2.736625514403292e-05,
+      "loss": 0.5405,
+      "step": 686
+    },
+    {
+      "epoch": 1.5266666666666666,
+      "grad_norm": 1006265.8125,
+      "learning_rate": 2.7325102880658437e-05,
+      "loss": 0.5338,
+      "step": 687
+    },
+    {
+      "epoch": 1.528888888888889,
+      "grad_norm": 329611.46875,
+      "learning_rate": 2.7283950617283956e-05,
+      "loss": 0.4846,
+      "step": 688
+    },
+    {
+      "epoch": 1.531111111111111,
+      "grad_norm": 314864.375,
+      "learning_rate": 2.7242798353909465e-05,
+      "loss": 0.4881,
+      "step": 689
+    },
+    {
+      "epoch": 1.5333333333333332,
+      "grad_norm": 555823.1875,
+      "learning_rate": 2.720164609053498e-05,
+      "loss": 0.5142,
+      "step": 690
+    },
+    {
+      "epoch": 1.5355555555555556,
+      "grad_norm": 975172.6875,
+      "learning_rate": 2.7160493827160493e-05,
+      "loss": 0.4287,
+      "step": 691
+    },
+    {
+      "epoch": 1.537777777777778,
+      "grad_norm": 698330.1875,
+      "learning_rate": 2.711934156378601e-05,
+      "loss": 0.5588,
+      "step": 692
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 911089.25,
+      "learning_rate": 2.707818930041152e-05,
+      "loss": 0.5083,
+      "step": 693
+    },
+    {
+      "epoch": 1.5422222222222222,
+      "grad_norm": 264078.0625,
+      "learning_rate": 2.7037037037037037e-05,
+      "loss": 0.5387,
+      "step": 694
+    },
+    {
+      "epoch": 1.5444444444444443,
+      "grad_norm": 852672.875,
+      "learning_rate": 2.6995884773662557e-05,
+      "loss": 0.5044,
+      "step": 695
+    },
+    {
+      "epoch": 1.5466666666666666,
+      "grad_norm": 455226.28125,
+      "learning_rate": 2.6954732510288066e-05,
+      "loss": 0.4886,
+      "step": 696
+    },
+    {
+      "epoch": 1.548888888888889,
+      "grad_norm": 522844.03125,
+      "learning_rate": 2.6913580246913585e-05,
+      "loss": 0.5363,
+      "step": 697
+    },
+    {
+      "epoch": 1.551111111111111,
+      "grad_norm": 129947.859375,
+      "learning_rate": 2.6872427983539094e-05,
+      "loss": 0.6613,
+      "step": 698
+    },
+    {
+      "epoch": 1.5533333333333332,
+      "grad_norm": 890144.8125,
+      "learning_rate": 2.683127572016461e-05,
+      "loss": 0.6588,
+      "step": 699
+    },
+    {
+      "epoch": 1.5555555555555556,
+      "grad_norm": 371911.78125,
+      "learning_rate": 2.6790123456790122e-05,
+      "loss": 0.4451,
+      "step": 700
+    },
+    {
+      "epoch": 1.557777777777778,
+      "grad_norm": 130950.0078125,
+      "learning_rate": 2.6748971193415638e-05,
+      "loss": 0.4966,
+      "step": 701
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 142102.953125,
+      "learning_rate": 2.6707818930041158e-05,
+      "loss": 0.4307,
+      "step": 702
+    },
+    {
+      "epoch": 1.5622222222222222,
+      "grad_norm": 359522.21875,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.5764,
+      "step": 703
+    },
+    {
+      "epoch": 1.5644444444444443,
+      "grad_norm": 154281.8125,
+      "learning_rate": 2.6625514403292186e-05,
+      "loss": 0.4768,
+      "step": 704
+    },
+    {
+      "epoch": 1.5666666666666667,
+      "grad_norm": 806375.1875,
+      "learning_rate": 2.6584362139917695e-05,
+      "loss": 0.6178,
+      "step": 705
+    },
+    {
+      "epoch": 1.568888888888889,
+      "grad_norm": 648758.4375,
+      "learning_rate": 2.654320987654321e-05,
+      "loss": 0.6353,
+      "step": 706
+    },
+    {
+      "epoch": 1.5711111111111111,
+      "grad_norm": 582554.125,
+      "learning_rate": 2.6502057613168723e-05,
+      "loss": 0.5469,
+      "step": 707
+    },
+    {
+      "epoch": 1.5733333333333333,
+      "grad_norm": 1223310.5,
+      "learning_rate": 2.646090534979424e-05,
+      "loss": 0.6492,
+      "step": 708
+    },
+    {
+      "epoch": 1.5755555555555556,
+      "grad_norm": 991635.6875,
+      "learning_rate": 2.641975308641975e-05,
+      "loss": 0.468,
+      "step": 709
+    },
+    {
+      "epoch": 1.5777777777777777,
+      "grad_norm": 1111512.25,
+      "learning_rate": 2.6378600823045267e-05,
+      "loss": 0.5682,
+      "step": 710
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 179395.46875,
+      "learning_rate": 2.6337448559670787e-05,
+      "loss": 0.562,
+      "step": 711
+    },
+    {
+      "epoch": 1.5822222222222222,
+      "grad_norm": 380546.03125,
+      "learning_rate": 2.6296296296296296e-05,
+      "loss": 0.5148,
+      "step": 712
+    },
+    {
+      "epoch": 1.5844444444444443,
+      "grad_norm": 2229058.75,
+      "learning_rate": 2.6255144032921815e-05,
+      "loss": 0.6432,
+      "step": 713
+    },
+    {
+      "epoch": 1.5866666666666667,
+      "grad_norm": 765920.0625,
+      "learning_rate": 2.6213991769547324e-05,
+      "loss": 0.4749,
+      "step": 714
+    },
+    {
+      "epoch": 1.588888888888889,
+      "grad_norm": 585911.75,
+      "learning_rate": 2.617283950617284e-05,
+      "loss": 0.5296,
+      "step": 715
+    },
+    {
+      "epoch": 1.5911111111111111,
+      "grad_norm": 2377990.25,
+      "learning_rate": 2.6131687242798352e-05,
+      "loss": 0.5784,
+      "step": 716
+    },
+    {
+      "epoch": 1.5933333333333333,
+      "grad_norm": 1983418.125,
+      "learning_rate": 2.6090534979423868e-05,
+      "loss": 0.6722,
+      "step": 717
+    },
+    {
+      "epoch": 1.5955555555555554,
+      "grad_norm": 636937.1875,
+      "learning_rate": 2.6049382716049388e-05,
+      "loss": 0.5298,
+      "step": 718
+    },
+    {
+      "epoch": 1.5977777777777777,
+      "grad_norm": 819429.75,
+      "learning_rate": 2.6008230452674897e-05,
+      "loss": 0.5188,
+      "step": 719
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 346863.21875,
+      "learning_rate": 2.5967078189300416e-05,
+      "loss": 0.5294,
+      "step": 720
+    },
+    {
+      "epoch": 1.6022222222222222,
+      "grad_norm": 1708457.25,
+      "learning_rate": 2.5925925925925925e-05,
+      "loss": 0.4534,
+      "step": 721
+    },
+    {
+      "epoch": 1.6044444444444443,
+      "grad_norm": 264422.75,
+      "learning_rate": 2.5884773662551444e-05,
+      "loss": 0.5927,
+      "step": 722
+    },
+    {
+      "epoch": 1.6066666666666667,
+      "grad_norm": 782417.5625,
+      "learning_rate": 2.5843621399176953e-05,
+      "loss": 0.5819,
+      "step": 723
+    },
+    {
+      "epoch": 1.608888888888889,
+      "grad_norm": 541727.5,
+      "learning_rate": 2.580246913580247e-05,
+      "loss": 0.5413,
+      "step": 724
+    },
+    {
+      "epoch": 1.6111111111111112,
+      "grad_norm": 362651.84375,
+      "learning_rate": 2.576131687242799e-05,
+      "loss": 0.5311,
+      "step": 725
+    },
+    {
+      "epoch": 1.6133333333333333,
+      "grad_norm": 302592.6875,
+      "learning_rate": 2.5720164609053497e-05,
+      "loss": 0.6317,
+      "step": 726
+    },
+    {
+      "epoch": 1.6155555555555554,
+      "grad_norm": 4282987.0,
+      "learning_rate": 2.5679012345679017e-05,
+      "loss": 0.5404,
+      "step": 727
+    },
+    {
+      "epoch": 1.6177777777777778,
+      "grad_norm": 388880.84375,
+      "learning_rate": 2.5637860082304526e-05,
+      "loss": 0.5138,
+      "step": 728
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 915515.4375,
+      "learning_rate": 2.5596707818930045e-05,
+      "loss": 0.5676,
+      "step": 729
+    },
+    {
+      "epoch": 1.6222222222222222,
+      "grad_norm": 1237556.5,
+      "learning_rate": 2.5555555555555554e-05,
+      "loss": 0.5387,
+      "step": 730
+    },
+    {
+      "epoch": 1.6244444444444444,
+      "grad_norm": 648158.125,
+      "learning_rate": 2.551440329218107e-05,
+      "loss": 0.362,
+      "step": 731
+    },
+    {
+      "epoch": 1.6266666666666667,
+      "grad_norm": 4585280.0,
+      "learning_rate": 2.5473251028806582e-05,
+      "loss": 0.6261,
+      "step": 732
+    },
+    {
+      "epoch": 1.628888888888889,
+      "grad_norm": 812792.9375,
+      "learning_rate": 2.5432098765432098e-05,
+      "loss": 0.5417,
+      "step": 733
+    },
+    {
+      "epoch": 1.6311111111111112,
+      "grad_norm": 600218.625,
+      "learning_rate": 2.5390946502057617e-05,
+      "loss": 0.579,
+      "step": 734
+    },
+    {
+      "epoch": 1.6333333333333333,
+      "grad_norm": 125346.09375,
+      "learning_rate": 2.5349794238683127e-05,
+      "loss": 0.542,
+      "step": 735
+    },
+    {
+      "epoch": 1.6355555555555554,
+      "grad_norm": 688033.3125,
+      "learning_rate": 2.5308641975308646e-05,
+      "loss": 0.5469,
+      "step": 736
+    },
+    {
+      "epoch": 1.6377777777777778,
+      "grad_norm": 816828.375,
+      "learning_rate": 2.5267489711934155e-05,
+      "loss": 0.5175,
+      "step": 737
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 1660627.625,
+      "learning_rate": 2.5226337448559674e-05,
+      "loss": 0.5644,
+      "step": 738
+    },
+    {
+      "epoch": 1.6422222222222222,
+      "grad_norm": 727887.4375,
+      "learning_rate": 2.5185185185185183e-05,
+      "loss": 0.689,
+      "step": 739
+    },
+    {
+      "epoch": 1.6444444444444444,
+      "grad_norm": 637634.625,
+      "learning_rate": 2.51440329218107e-05,
+      "loss": 0.5897,
+      "step": 740
+    },
+    {
+      "epoch": 1.6466666666666665,
+      "grad_norm": 457325.75,
+      "learning_rate": 2.510288065843622e-05,
+      "loss": 0.6766,
+      "step": 741
+    },
+    {
+      "epoch": 1.6488888888888888,
+      "grad_norm": 250656.8125,
+      "learning_rate": 2.5061728395061727e-05,
+      "loss": 0.5797,
+      "step": 742
+    },
+    {
+      "epoch": 1.6511111111111112,
+      "grad_norm": 667376.6875,
+      "learning_rate": 2.5020576131687247e-05,
+      "loss": 0.4846,
+      "step": 743
+    },
+    {
+      "epoch": 1.6533333333333333,
+      "grad_norm": 203570.921875,
+      "learning_rate": 2.497942386831276e-05,
+      "loss": 0.3637,
+      "step": 744
+    },
+    {
+      "epoch": 1.6555555555555554,
+      "grad_norm": 757491.5625,
+      "learning_rate": 2.4938271604938275e-05,
+      "loss": 0.5933,
+      "step": 745
+    },
+    {
+      "epoch": 1.6577777777777778,
+      "grad_norm": 258822.21875,
+      "learning_rate": 2.4897119341563787e-05,
+      "loss": 0.4402,
+      "step": 746
+    },
+    {
+      "epoch": 1.6600000000000001,
+      "grad_norm": 404580.1875,
+      "learning_rate": 2.4855967078189303e-05,
+      "loss": 0.4108,
+      "step": 747
+    },
+    {
+      "epoch": 1.6622222222222223,
+      "grad_norm": 189135.4375,
+      "learning_rate": 2.4814814814814816e-05,
+      "loss": 0.481,
+      "step": 748
+    },
+    {
+      "epoch": 1.6644444444444444,
+      "grad_norm": 365586.9375,
+      "learning_rate": 2.4773662551440328e-05,
+      "loss": 0.5955,
+      "step": 749
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 637579.75,
+      "learning_rate": 2.4732510288065844e-05,
+      "loss": 0.6268,
+      "step": 750
+    },
+    {
+      "epoch": 1.6688888888888889,
+      "grad_norm": 234886.171875,
+      "learning_rate": 2.4691358024691357e-05,
+      "loss": 0.431,
+      "step": 751
+    },
+    {
+      "epoch": 1.6711111111111112,
+      "grad_norm": 641196.1875,
+      "learning_rate": 2.4650205761316876e-05,
+      "loss": 0.4341,
+      "step": 752
+    },
+    {
+      "epoch": 1.6733333333333333,
+      "grad_norm": 269056.125,
+      "learning_rate": 2.4609053497942388e-05,
+      "loss": 0.4744,
+      "step": 753
+    },
+    {
+      "epoch": 1.6755555555555555,
+      "grad_norm": 2218215.0,
+      "learning_rate": 2.4567901234567904e-05,
+      "loss": 0.604,
+      "step": 754
+    },
+    {
+      "epoch": 1.6777777777777778,
+      "grad_norm": 695057.75,
+      "learning_rate": 2.4526748971193417e-05,
+      "loss": 0.6412,
+      "step": 755
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 810964.0625,
+      "learning_rate": 2.4485596707818932e-05,
+      "loss": 0.5174,
+      "step": 756
+    },
+    {
+      "epoch": 1.6822222222222223,
+      "grad_norm": 1622732.875,
+      "learning_rate": 2.4444444444444445e-05,
+      "loss": 0.4582,
+      "step": 757
+    },
+    {
+      "epoch": 1.6844444444444444,
+      "grad_norm": 362912.90625,
+      "learning_rate": 2.4403292181069957e-05,
+      "loss": 0.583,
+      "step": 758
+    },
+    {
+      "epoch": 1.6866666666666665,
+      "grad_norm": 2453790.75,
+      "learning_rate": 2.4362139917695477e-05,
+      "loss": 0.5999,
+      "step": 759
+    },
+    {
+      "epoch": 1.6888888888888889,
+      "grad_norm": 334428.9375,
+      "learning_rate": 2.432098765432099e-05,
+      "loss": 0.5055,
+      "step": 760
+    },
+    {
+      "epoch": 1.6911111111111112,
+      "grad_norm": 993397.0,
+      "learning_rate": 2.4279835390946505e-05,
+      "loss": 0.6,
+      "step": 761
+    },
+    {
+      "epoch": 1.6933333333333334,
+      "grad_norm": 387189.09375,
+      "learning_rate": 2.4238683127572017e-05,
+      "loss": 0.6245,
+      "step": 762
+    },
+    {
+      "epoch": 1.6955555555555555,
+      "grad_norm": 1040618.1875,
+      "learning_rate": 2.4197530864197533e-05,
+      "loss": 0.4439,
+      "step": 763
+    },
+    {
+      "epoch": 1.6977777777777778,
+      "grad_norm": 855744.9375,
+      "learning_rate": 2.4156378600823046e-05,
+      "loss": 0.4944,
+      "step": 764
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1253714.5,
+      "learning_rate": 2.4115226337448558e-05,
+      "loss": 0.7158,
+      "step": 765
+    },
+    {
+      "epoch": 1.7022222222222223,
+      "grad_norm": 692958.8125,
+      "learning_rate": 2.4074074074074074e-05,
+      "loss": 0.4465,
+      "step": 766
+    },
+    {
+      "epoch": 1.7044444444444444,
+      "grad_norm": 1175609.875,
+      "learning_rate": 2.403292181069959e-05,
+      "loss": 0.5385,
+      "step": 767
+    },
+    {
+      "epoch": 1.7066666666666666,
+      "grad_norm": 649715.375,
+      "learning_rate": 2.3991769547325106e-05,
+      "loss": 0.4884,
+      "step": 768
+    },
+    {
+      "epoch": 1.708888888888889,
+      "grad_norm": 309797.25,
+      "learning_rate": 2.3950617283950618e-05,
+      "loss": 0.4184,
+      "step": 769
+    },
+    {
+      "epoch": 1.7111111111111112,
+      "grad_norm": 302643.53125,
+      "learning_rate": 2.3909465020576134e-05,
+      "loss": 0.5094,
+      "step": 770
+    },
+    {
+      "epoch": 1.7133333333333334,
+      "grad_norm": 679658.9375,
+      "learning_rate": 2.3868312757201647e-05,
+      "loss": 0.4653,
+      "step": 771
+    },
+    {
+      "epoch": 1.7155555555555555,
+      "grad_norm": 383316.59375,
+      "learning_rate": 2.3827160493827162e-05,
+      "loss": 0.5941,
+      "step": 772
+    },
+    {
+      "epoch": 1.7177777777777776,
+      "grad_norm": 450449.125,
+      "learning_rate": 2.3786008230452675e-05,
+      "loss": 0.6031,
+      "step": 773
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 193325.40625,
+      "learning_rate": 2.374485596707819e-05,
+      "loss": 0.5443,
+      "step": 774
+    },
+    {
+      "epoch": 1.7222222222222223,
+      "grad_norm": 1077144.75,
+      "learning_rate": 2.3703703703703707e-05,
+      "loss": 0.5123,
+      "step": 775
+    },
+    {
+      "epoch": 1.7244444444444444,
+      "grad_norm": 449495.15625,
+      "learning_rate": 2.366255144032922e-05,
+      "loss": 0.5594,
+      "step": 776
+    },
+    {
+      "epoch": 1.7266666666666666,
+      "grad_norm": 319506.9375,
+      "learning_rate": 2.3621399176954735e-05,
+      "loss": 0.5426,
+      "step": 777
+    },
+    {
+      "epoch": 1.728888888888889,
+      "grad_norm": 1278679.125,
+      "learning_rate": 2.3580246913580247e-05,
+      "loss": 0.6726,
+      "step": 778
+    },
+    {
+      "epoch": 1.7311111111111113,
+      "grad_norm": 897134.25,
+      "learning_rate": 2.3539094650205763e-05,
+      "loss": 0.5594,
+      "step": 779
+    },
+    {
+      "epoch": 1.7333333333333334,
+      "grad_norm": 427793.71875,
+      "learning_rate": 2.3497942386831276e-05,
+      "loss": 0.5476,
+      "step": 780
+    },
+    {
+      "epoch": 1.7355555555555555,
+      "grad_norm": 359693.71875,
+      "learning_rate": 2.345679012345679e-05,
+      "loss": 0.4278,
+      "step": 781
+    },
+    {
+      "epoch": 1.7377777777777776,
+      "grad_norm": 678628.125,
+      "learning_rate": 2.3415637860082307e-05,
+      "loss": 0.5121,
+      "step": 782
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 823664.75,
+      "learning_rate": 2.337448559670782e-05,
+      "loss": 0.4467,
+      "step": 783
+    },
+    {
+      "epoch": 1.7422222222222223,
+      "grad_norm": 556514.875,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 0.6415,
+      "step": 784
+    },
+    {
+      "epoch": 1.7444444444444445,
+      "grad_norm": 844012.6875,
+      "learning_rate": 2.3292181069958848e-05,
+      "loss": 0.5389,
+      "step": 785
+    },
+    {
+      "epoch": 1.7466666666666666,
+      "grad_norm": 731721.375,
+      "learning_rate": 2.3251028806584364e-05,
+      "loss": 0.4813,
+      "step": 786
+    },
+    {
+      "epoch": 1.748888888888889,
+      "grad_norm": 657326.875,
+      "learning_rate": 2.3209876543209877e-05,
+      "loss": 0.6037,
+      "step": 787
+    },
+    {
+      "epoch": 1.751111111111111,
+      "grad_norm": 1010242.3125,
+      "learning_rate": 2.3168724279835392e-05,
+      "loss": 0.5776,
+      "step": 788
+    },
+    {
+      "epoch": 1.7533333333333334,
+      "grad_norm": 1569845.625,
+      "learning_rate": 2.312757201646091e-05,
+      "loss": 0.5562,
+      "step": 789
+    },
+    {
+      "epoch": 1.7555555555555555,
+      "grad_norm": 446353.375,
+      "learning_rate": 2.308641975308642e-05,
+      "loss": 0.6496,
+      "step": 790
+    },
+    {
+      "epoch": 1.7577777777777777,
+      "grad_norm": 860934.6875,
+      "learning_rate": 2.3045267489711937e-05,
+      "loss": 0.5168,
+      "step": 791
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2472300.25,
+      "learning_rate": 2.300411522633745e-05,
+      "loss": 0.4787,
+      "step": 792
+    },
+    {
+      "epoch": 1.7622222222222224,
+      "grad_norm": 1831251.25,
+      "learning_rate": 2.2962962962962965e-05,
+      "loss": 0.5387,
+      "step": 793
+    },
+    {
+      "epoch": 1.7644444444444445,
+      "grad_norm": 1542738.5,
+      "learning_rate": 2.2921810699588477e-05,
+      "loss": 0.5983,
+      "step": 794
+    },
+    {
+      "epoch": 1.7666666666666666,
+      "grad_norm": 1019469.1875,
+      "learning_rate": 2.2880658436213993e-05,
+      "loss": 0.6107,
+      "step": 795
+    },
+    {
+      "epoch": 1.7688888888888887,
+      "grad_norm": 392084.78125,
+      "learning_rate": 2.2839506172839506e-05,
+      "loss": 0.4184,
+      "step": 796
+    },
+    {
+      "epoch": 1.771111111111111,
+      "grad_norm": 1160610.25,
+      "learning_rate": 2.279835390946502e-05,
+      "loss": 0.5625,
+      "step": 797
+    },
+    {
+      "epoch": 1.7733333333333334,
+      "grad_norm": 229940.953125,
+      "learning_rate": 2.2757201646090537e-05,
+      "loss": 0.6152,
+      "step": 798
+    },
+    {
+      "epoch": 1.7755555555555556,
+      "grad_norm": 356666.3125,
+      "learning_rate": 2.271604938271605e-05,
+      "loss": 0.4933,
+      "step": 799
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "grad_norm": 130159.8125,
+      "learning_rate": 2.2674897119341566e-05,
+      "loss": 0.5619,
+      "step": 800
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 148995.453125,
+      "learning_rate": 2.2633744855967078e-05,
+      "loss": 0.4679,
+      "step": 801
+    },
+    {
+      "epoch": 1.7822222222222224,
+      "grad_norm": 690270.125,
+      "learning_rate": 2.2592592592592594e-05,
+      "loss": 0.4568,
+      "step": 802
+    },
+    {
+      "epoch": 1.7844444444444445,
+      "grad_norm": 1709161.25,
+      "learning_rate": 2.2551440329218107e-05,
+      "loss": 0.4541,
+      "step": 803
+    },
+    {
+      "epoch": 1.7866666666666666,
+      "grad_norm": 1714657.25,
+      "learning_rate": 2.2510288065843622e-05,
+      "loss": 0.4326,
+      "step": 804
+    },
+    {
+      "epoch": 1.7888888888888888,
+      "grad_norm": 1946265.25,
+      "learning_rate": 2.246913580246914e-05,
+      "loss": 0.5646,
+      "step": 805
+    },
+    {
+      "epoch": 1.791111111111111,
+      "grad_norm": 872982.875,
+      "learning_rate": 2.242798353909465e-05,
+      "loss": 0.4512,
+      "step": 806
+    },
+    {
+      "epoch": 1.7933333333333334,
+      "grad_norm": 929096.4375,
+      "learning_rate": 2.2386831275720167e-05,
+      "loss": 0.4812,
+      "step": 807
+    },
+    {
+      "epoch": 1.7955555555555556,
+      "grad_norm": 431499.1875,
+      "learning_rate": 2.234567901234568e-05,
+      "loss": 0.5798,
+      "step": 808
+    },
+    {
+      "epoch": 1.7977777777777777,
+      "grad_norm": 837512.3125,
+      "learning_rate": 2.2304526748971195e-05,
+      "loss": 0.5095,
+      "step": 809
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 446406.6875,
+      "learning_rate": 2.2263374485596707e-05,
+      "loss": 0.5401,
+      "step": 810
+    },
+    {
+      "epoch": 1.8022222222222222,
+      "grad_norm": 358855.78125,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.5959,
+      "step": 811
+    },
+    {
+      "epoch": 1.8044444444444445,
+      "grad_norm": 326300.875,
+      "learning_rate": 2.218106995884774e-05,
+      "loss": 0.4101,
+      "step": 812
+    },
+    {
+      "epoch": 1.8066666666666666,
+      "grad_norm": 910762.8125,
+      "learning_rate": 2.213991769547325e-05,
+      "loss": 0.5932,
+      "step": 813
+    },
+    {
+      "epoch": 1.8088888888888888,
+      "grad_norm": 1639550.125,
+      "learning_rate": 2.2098765432098767e-05,
+      "loss": 0.437,
+      "step": 814
+    },
+    {
+      "epoch": 1.8111111111111111,
+      "grad_norm": 443086.21875,
+      "learning_rate": 2.205761316872428e-05,
+      "loss": 0.5197,
+      "step": 815
+    },
+    {
+      "epoch": 1.8133333333333335,
+      "grad_norm": 481764.21875,
+      "learning_rate": 2.2016460905349796e-05,
+      "loss": 0.4687,
+      "step": 816
+    },
+    {
+      "epoch": 1.8155555555555556,
+      "grad_norm": 395750.53125,
+      "learning_rate": 2.1975308641975308e-05,
+      "loss": 0.5638,
+      "step": 817
+    },
+    {
+      "epoch": 1.8177777777777777,
+      "grad_norm": 427531.65625,
+      "learning_rate": 2.1934156378600824e-05,
+      "loss": 0.5704,
+      "step": 818
+    },
+    {
+      "epoch": 1.8199999999999998,
+      "grad_norm": 1041394.3125,
+      "learning_rate": 2.189300411522634e-05,
+      "loss": 0.5553,
+      "step": 819
+    },
+    {
+      "epoch": 1.8222222222222222,
+      "grad_norm": 368493.375,
+      "learning_rate": 2.1851851851851852e-05,
+      "loss": 0.4636,
+      "step": 820
+    },
+    {
+      "epoch": 1.8244444444444445,
+      "grad_norm": 911481.6875,
+      "learning_rate": 2.1810699588477368e-05,
+      "loss": 0.4623,
+      "step": 821
+    },
+    {
+      "epoch": 1.8266666666666667,
+      "grad_norm": 772321.5,
+      "learning_rate": 2.176954732510288e-05,
+      "loss": 0.619,
+      "step": 822
+    },
+    {
+      "epoch": 1.8288888888888888,
+      "grad_norm": 2411277.0,
+      "learning_rate": 2.1728395061728397e-05,
+      "loss": 0.4113,
+      "step": 823
+    },
+    {
+      "epoch": 1.8311111111111111,
+      "grad_norm": 282978.75,
+      "learning_rate": 2.168724279835391e-05,
+      "loss": 0.5235,
+      "step": 824
+    },
+    {
+      "epoch": 1.8333333333333335,
+      "grad_norm": 369288.125,
+      "learning_rate": 2.1646090534979425e-05,
+      "loss": 0.4668,
+      "step": 825
+    },
+    {
+      "epoch": 1.8355555555555556,
+      "grad_norm": 427154.0625,
+      "learning_rate": 2.1604938271604937e-05,
+      "loss": 0.5174,
+      "step": 826
+    },
+    {
+      "epoch": 1.8377777777777777,
+      "grad_norm": 590274.0625,
+      "learning_rate": 2.1563786008230453e-05,
+      "loss": 0.6154,
+      "step": 827
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 479509.1875,
+      "learning_rate": 2.152263374485597e-05,
+      "loss": 0.4884,
+      "step": 828
+    },
+    {
+      "epoch": 1.8422222222222222,
+      "grad_norm": 627632.5625,
+      "learning_rate": 2.148148148148148e-05,
+      "loss": 0.5925,
+      "step": 829
+    },
+    {
+      "epoch": 1.8444444444444446,
+      "grad_norm": 126343.203125,
+      "learning_rate": 2.1440329218106997e-05,
+      "loss": 0.4947,
+      "step": 830
+    },
+    {
+      "epoch": 1.8466666666666667,
+      "grad_norm": 662519.25,
+      "learning_rate": 2.139917695473251e-05,
+      "loss": 0.4442,
+      "step": 831
+    },
+    {
+      "epoch": 1.8488888888888888,
+      "grad_norm": 1551904.25,
+      "learning_rate": 2.1358024691358026e-05,
+      "loss": 0.4669,
+      "step": 832
+    },
+    {
+      "epoch": 1.8511111111111112,
+      "grad_norm": 495174.375,
+      "learning_rate": 2.1316872427983538e-05,
+      "loss": 0.4312,
+      "step": 833
+    },
+    {
+      "epoch": 1.8533333333333335,
+      "grad_norm": 500009.3125,
+      "learning_rate": 2.1275720164609054e-05,
+      "loss": 0.4621,
+      "step": 834
+    },
+    {
+      "epoch": 1.8555555555555556,
+      "grad_norm": 1010434.4375,
+      "learning_rate": 2.123456790123457e-05,
+      "loss": 0.5229,
+      "step": 835
+    },
+    {
+      "epoch": 1.8577777777777778,
+      "grad_norm": 265641.6875,
+      "learning_rate": 2.1193415637860082e-05,
+      "loss": 0.5296,
+      "step": 836
+    },
+    {
+      "epoch": 1.8599999999999999,
+      "grad_norm": 795082.4375,
+      "learning_rate": 2.1152263374485598e-05,
+      "loss": 0.5857,
+      "step": 837
+    },
+    {
+      "epoch": 1.8622222222222222,
+      "grad_norm": 838311.375,
+      "learning_rate": 2.111111111111111e-05,
+      "loss": 0.4881,
+      "step": 838
+    },
+    {
+      "epoch": 1.8644444444444446,
+      "grad_norm": 978832.3125,
+      "learning_rate": 2.1069958847736627e-05,
+      "loss": 0.5223,
+      "step": 839
+    },
+    {
+      "epoch": 1.8666666666666667,
+      "grad_norm": 1003945.3125,
+      "learning_rate": 2.102880658436214e-05,
+      "loss": 0.4662,
+      "step": 840
+    },
+    {
+      "epoch": 1.8688888888888888,
+      "grad_norm": 3484293.5,
+      "learning_rate": 2.0987654320987655e-05,
+      "loss": 0.7073,
+      "step": 841
+    },
+    {
+      "epoch": 1.871111111111111,
+      "grad_norm": 544340.5,
+      "learning_rate": 2.094650205761317e-05,
+      "loss": 0.7396,
+      "step": 842
+    },
+    {
+      "epoch": 1.8733333333333333,
+      "grad_norm": 954412.9375,
+      "learning_rate": 2.0905349794238687e-05,
+      "loss": 0.5633,
+      "step": 843
+    },
+    {
+      "epoch": 1.8755555555555556,
+      "grad_norm": 1018313.875,
+      "learning_rate": 2.08641975308642e-05,
+      "loss": 0.5066,
+      "step": 844
+    },
+    {
+      "epoch": 1.8777777777777778,
+      "grad_norm": 2038616.875,
+      "learning_rate": 2.082304526748971e-05,
+      "loss": 0.7643,
+      "step": 845
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 788535.3125,
+      "learning_rate": 2.0781893004115227e-05,
+      "loss": 0.5479,
+      "step": 846
+    },
+    {
+      "epoch": 1.8822222222222222,
+      "grad_norm": 1077129.875,
+      "learning_rate": 2.074074074074074e-05,
+      "loss": 0.5288,
+      "step": 847
+    },
+    {
+      "epoch": 1.8844444444444446,
+      "grad_norm": 944732.9375,
+      "learning_rate": 2.0699588477366256e-05,
+      "loss": 0.5459,
+      "step": 848
+    },
+    {
+      "epoch": 1.8866666666666667,
+      "grad_norm": 692355.9375,
+      "learning_rate": 2.0658436213991768e-05,
+      "loss": 0.5624,
+      "step": 849
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 788133.9375,
+      "learning_rate": 2.0617283950617287e-05,
+      "loss": 0.5595,
+      "step": 850
+    },
+    {
+      "epoch": 1.891111111111111,
+      "grad_norm": 1503811.75,
+      "learning_rate": 2.05761316872428e-05,
+      "loss": 0.4503,
+      "step": 851
+    },
+    {
+      "epoch": 1.8933333333333333,
+      "grad_norm": 608034.5625,
+      "learning_rate": 2.0534979423868312e-05,
+      "loss": 0.597,
+      "step": 852
+    },
+    {
+      "epoch": 1.8955555555555557,
+      "grad_norm": 432831.4375,
+      "learning_rate": 2.0493827160493828e-05,
+      "loss": 0.5731,
+      "step": 853
+    },
+    {
+      "epoch": 1.8977777777777778,
+      "grad_norm": 2556177.75,
+      "learning_rate": 2.045267489711934e-05,
+      "loss": 0.455,
+      "step": 854
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 818600.25,
+      "learning_rate": 2.0411522633744857e-05,
+      "loss": 0.5477,
+      "step": 855
+    },
+    {
+      "epoch": 1.9022222222222223,
+      "grad_norm": 702948.9375,
+      "learning_rate": 2.037037037037037e-05,
+      "loss": 0.5564,
+      "step": 856
+    },
+    {
+      "epoch": 1.9044444444444446,
+      "grad_norm": 352416.96875,
+      "learning_rate": 2.032921810699589e-05,
+      "loss": 0.479,
+      "step": 857
+    },
+    {
+      "epoch": 1.9066666666666667,
+      "grad_norm": 641082.25,
+      "learning_rate": 2.02880658436214e-05,
+      "loss": 0.5326,
+      "step": 858
+    },
+    {
+      "epoch": 1.9088888888888889,
+      "grad_norm": 1330086.625,
+      "learning_rate": 2.0246913580246917e-05,
+      "loss": 0.5112,
+      "step": 859
+    },
+    {
+      "epoch": 1.911111111111111,
+      "grad_norm": 402839.9375,
+      "learning_rate": 2.020576131687243e-05,
+      "loss": 0.4653,
+      "step": 860
+    },
+    {
+      "epoch": 1.9133333333333333,
+      "grad_norm": 1415979.0,
+      "learning_rate": 2.016460905349794e-05,
+      "loss": 0.5702,
+      "step": 861
+    },
+    {
+      "epoch": 1.9155555555555557,
+      "grad_norm": 582900.9375,
+      "learning_rate": 2.0123456790123457e-05,
+      "loss": 0.5496,
+      "step": 862
+    },
+    {
+      "epoch": 1.9177777777777778,
+      "grad_norm": 1254026.625,
+      "learning_rate": 2.008230452674897e-05,
+      "loss": 0.5884,
+      "step": 863
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 681267.3125,
+      "learning_rate": 2.0041152263374486e-05,
+      "loss": 0.5028,
+      "step": 864
+    },
+    {
+      "epoch": 1.9222222222222223,
+      "grad_norm": 639010.25,
+      "learning_rate": 2e-05,
+      "loss": 0.5437,
+      "step": 865
+    },
+    {
+      "epoch": 1.9244444444444444,
+      "grad_norm": 750335.0,
+      "learning_rate": 1.9958847736625517e-05,
+      "loss": 0.5413,
+      "step": 866
+    },
+    {
+      "epoch": 1.9266666666666667,
+      "grad_norm": 395516.53125,
+      "learning_rate": 1.991769547325103e-05,
+      "loss": 0.453,
+      "step": 867
+    },
+    {
+      "epoch": 1.9288888888888889,
+      "grad_norm": 496803.71875,
+      "learning_rate": 1.9876543209876546e-05,
+      "loss": 0.487,
+      "step": 868
+    },
+    {
+      "epoch": 1.931111111111111,
+      "grad_norm": 1806078.25,
+      "learning_rate": 1.9835390946502058e-05,
+      "loss": 0.5195,
+      "step": 869
+    },
+    {
+      "epoch": 1.9333333333333333,
+      "grad_norm": 424641.3125,
+      "learning_rate": 1.979423868312757e-05,
+      "loss": 0.4678,
+      "step": 870
+    },
+    {
+      "epoch": 1.9355555555555557,
+      "grad_norm": 1400120.0,
+      "learning_rate": 1.9753086419753087e-05,
+      "loss": 0.4432,
+      "step": 871
+    },
+    {
+      "epoch": 1.9377777777777778,
+      "grad_norm": 887819.125,
+      "learning_rate": 1.9711934156378602e-05,
+      "loss": 0.5575,
+      "step": 872
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 216944.609375,
+      "learning_rate": 1.967078189300412e-05,
+      "loss": 0.5241,
+      "step": 873
+    },
+    {
+      "epoch": 1.942222222222222,
+      "grad_norm": 963737.4375,
+      "learning_rate": 1.962962962962963e-05,
+      "loss": 0.5449,
+      "step": 874
+    },
+    {
+      "epoch": 1.9444444444444444,
+      "grad_norm": 656178.6875,
+      "learning_rate": 1.9588477366255147e-05,
+      "loss": 0.5254,
+      "step": 875
+    },
+    {
+      "epoch": 1.9466666666666668,
+      "grad_norm": 601821.625,
+      "learning_rate": 1.954732510288066e-05,
+      "loss": 0.4921,
+      "step": 876
+    },
+    {
+      "epoch": 1.948888888888889,
+      "grad_norm": 2348228.75,
+      "learning_rate": 1.950617283950617e-05,
+      "loss": 0.4908,
+      "step": 877
+    },
+    {
+      "epoch": 1.951111111111111,
+      "grad_norm": 777158.4375,
+      "learning_rate": 1.9465020576131687e-05,
+      "loss": 0.4783,
+      "step": 878
+    },
+    {
+      "epoch": 1.9533333333333334,
+      "grad_norm": 565114.375,
+      "learning_rate": 1.94238683127572e-05,
+      "loss": 0.6376,
+      "step": 879
+    },
+    {
+      "epoch": 1.9555555555555557,
+      "grad_norm": 549776.4375,
+      "learning_rate": 1.938271604938272e-05,
+      "loss": 0.5003,
+      "step": 880
+    },
+    {
+      "epoch": 1.9577777777777778,
+      "grad_norm": 470974.875,
+      "learning_rate": 1.934156378600823e-05,
+      "loss": 0.5577,
+      "step": 881
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 988247.875,
+      "learning_rate": 1.9300411522633747e-05,
+      "loss": 0.5346,
+      "step": 882
+    },
+    {
+      "epoch": 1.962222222222222,
+      "grad_norm": 555580.3125,
+      "learning_rate": 1.925925925925926e-05,
+      "loss": 0.5311,
+      "step": 883
+    },
+    {
+      "epoch": 1.9644444444444444,
+      "grad_norm": 371100.03125,
+      "learning_rate": 1.9218106995884776e-05,
+      "loss": 0.5582,
+      "step": 884
+    },
+    {
+      "epoch": 1.9666666666666668,
+      "grad_norm": 426462.53125,
+      "learning_rate": 1.9176954732510288e-05,
+      "loss": 0.5114,
+      "step": 885
+    },
+    {
+      "epoch": 1.968888888888889,
+      "grad_norm": 560246.5625,
+      "learning_rate": 1.91358024691358e-05,
+      "loss": 0.5406,
+      "step": 886
+    },
+    {
+      "epoch": 1.971111111111111,
+      "grad_norm": 676496.4375,
+      "learning_rate": 1.909465020576132e-05,
+      "loss": 0.636,
+      "step": 887
+    },
+    {
+      "epoch": 1.9733333333333334,
+      "grad_norm": 942200.4375,
+      "learning_rate": 1.9053497942386832e-05,
+      "loss": 0.6124,
+      "step": 888
+    },
+    {
+      "epoch": 1.9755555555555555,
+      "grad_norm": 986006.3125,
+      "learning_rate": 1.901234567901235e-05,
+      "loss": 0.6272,
+      "step": 889
+    },
+    {
+      "epoch": 1.9777777777777779,
+      "grad_norm": 415790.6875,
+      "learning_rate": 1.897119341563786e-05,
+      "loss": 0.5431,
+      "step": 890
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 415508.59375,
+      "learning_rate": 1.8930041152263377e-05,
+      "loss": 0.5771,
+      "step": 891
+    },
+    {
+      "epoch": 1.982222222222222,
+      "grad_norm": 364349.125,
+      "learning_rate": 1.888888888888889e-05,
+      "loss": 0.4451,
+      "step": 892
+    },
+    {
+      "epoch": 1.9844444444444445,
+      "grad_norm": 357926.125,
+      "learning_rate": 1.8847736625514405e-05,
+      "loss": 0.5708,
+      "step": 893
+    },
+    {
+      "epoch": 1.9866666666666668,
+      "grad_norm": 648236.625,
+      "learning_rate": 1.8806584362139917e-05,
+      "loss": 0.6643,
+      "step": 894
+    },
+    {
+      "epoch": 1.988888888888889,
+      "grad_norm": 442882.90625,
+      "learning_rate": 1.8765432098765433e-05,
+      "loss": 0.4524,
+      "step": 895
+    },
+    {
+      "epoch": 1.991111111111111,
+      "grad_norm": 661978.375,
+      "learning_rate": 1.872427983539095e-05,
+      "loss": 0.5812,
+      "step": 896
+    },
+    {
+      "epoch": 1.9933333333333332,
+      "grad_norm": 1080674.0,
+      "learning_rate": 1.868312757201646e-05,
+      "loss": 0.6168,
+      "step": 897
+    },
+    {
+      "epoch": 1.9955555555555555,
+      "grad_norm": 2142185.5,
+      "learning_rate": 1.8641975308641977e-05,
+      "loss": 0.4752,
+      "step": 898
+    },
+    {
+      "epoch": 1.9977777777777779,
+      "grad_norm": 487222.90625,
+      "learning_rate": 1.860082304526749e-05,
+      "loss": 0.6078,
+      "step": 899
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 399853.9375,
+      "learning_rate": 1.8559670781893006e-05,
+      "loss": 0.4525,
+      "step": 900
+    },
+    {
+      "epoch": 2.002222222222222,
+      "grad_norm": 442879.53125,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 0.48,
+      "step": 901
+    },
+    {
+      "epoch": 2.0044444444444443,
+      "grad_norm": 463064.875,
+      "learning_rate": 1.8477366255144034e-05,
+      "loss": 0.5115,
+      "step": 902
+    },
+    {
+      "epoch": 2.006666666666667,
+      "grad_norm": 452375.40625,
+      "learning_rate": 1.843621399176955e-05,
+      "loss": 0.6397,
+      "step": 903
+    },
+    {
+      "epoch": 2.008888888888889,
+      "grad_norm": 694365.6875,
+      "learning_rate": 1.8395061728395062e-05,
+      "loss": 0.5129,
+      "step": 904
+    },
+    {
+      "epoch": 2.011111111111111,
+      "grad_norm": 380162.875,
+      "learning_rate": 1.835390946502058e-05,
+      "loss": 0.486,
+      "step": 905
+    },
+    {
+      "epoch": 2.013333333333333,
+      "grad_norm": 563086.0625,
+      "learning_rate": 1.831275720164609e-05,
+      "loss": 0.5581,
+      "step": 906
+    },
+    {
+      "epoch": 2.0155555555555558,
+      "grad_norm": 344576.40625,
+      "learning_rate": 1.8271604938271607e-05,
+      "loss": 0.57,
+      "step": 907
+    },
+    {
+      "epoch": 2.017777777777778,
+      "grad_norm": 398236.9375,
+      "learning_rate": 1.823045267489712e-05,
+      "loss": 0.5709,
+      "step": 908
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 814398.9375,
+      "learning_rate": 1.8189300411522635e-05,
+      "loss": 0.4834,
+      "step": 909
+    },
+    {
+      "epoch": 2.022222222222222,
+      "grad_norm": 492153.21875,
+      "learning_rate": 1.814814814814815e-05,
+      "loss": 0.7421,
+      "step": 910
+    },
+    {
+      "epoch": 2.0244444444444443,
+      "grad_norm": 164438.71875,
+      "learning_rate": 1.8106995884773663e-05,
+      "loss": 0.4633,
+      "step": 911
+    },
+    {
+      "epoch": 2.026666666666667,
+      "grad_norm": 460060.9375,
+      "learning_rate": 1.806584362139918e-05,
+      "loss": 0.4765,
+      "step": 912
+    },
+    {
+      "epoch": 2.028888888888889,
+      "grad_norm": 1290432.625,
+      "learning_rate": 1.802469135802469e-05,
+      "loss": 0.6418,
+      "step": 913
+    },
+    {
+      "epoch": 2.031111111111111,
+      "grad_norm": 704266.6875,
+      "learning_rate": 1.7983539094650207e-05,
+      "loss": 0.4835,
+      "step": 914
+    },
+    {
+      "epoch": 2.033333333333333,
+      "grad_norm": 505275.875,
+      "learning_rate": 1.794238683127572e-05,
+      "loss": 0.5829,
+      "step": 915
+    },
+    {
+      "epoch": 2.0355555555555553,
+      "grad_norm": 1274834.25,
+      "learning_rate": 1.7901234567901236e-05,
+      "loss": 0.5759,
+      "step": 916
+    },
+    {
+      "epoch": 2.037777777777778,
+      "grad_norm": 615392.875,
+      "learning_rate": 1.7860082304526748e-05,
+      "loss": 0.5188,
+      "step": 917
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 198756.796875,
+      "learning_rate": 1.7818930041152264e-05,
+      "loss": 0.5418,
+      "step": 918
+    },
+    {
+      "epoch": 2.042222222222222,
+      "grad_norm": 687675.1875,
+      "learning_rate": 1.777777777777778e-05,
+      "loss": 0.6929,
+      "step": 919
+    },
+    {
+      "epoch": 2.0444444444444443,
+      "grad_norm": 552092.4375,
+      "learning_rate": 1.7736625514403292e-05,
+      "loss": 0.4773,
+      "step": 920
+    },
+    {
+      "epoch": 2.046666666666667,
+      "grad_norm": 497010.9375,
+      "learning_rate": 1.769547325102881e-05,
+      "loss": 0.4841,
+      "step": 921
+    },
+    {
+      "epoch": 2.048888888888889,
+      "grad_norm": 435989.84375,
+      "learning_rate": 1.765432098765432e-05,
+      "loss": 0.4522,
+      "step": 922
+    },
+    {
+      "epoch": 2.051111111111111,
+      "grad_norm": 869393.125,
+      "learning_rate": 1.7613168724279837e-05,
+      "loss": 0.6985,
+      "step": 923
+    },
+    {
+      "epoch": 2.0533333333333332,
+      "grad_norm": 385168.8125,
+      "learning_rate": 1.757201646090535e-05,
+      "loss": 0.5334,
+      "step": 924
+    },
+    {
+      "epoch": 2.0555555555555554,
+      "grad_norm": 831693.0,
+      "learning_rate": 1.7530864197530865e-05,
+      "loss": 0.4533,
+      "step": 925
+    },
+    {
+      "epoch": 2.057777777777778,
+      "grad_norm": 1018997.9375,
+      "learning_rate": 1.748971193415638e-05,
+      "loss": 0.5729,
+      "step": 926
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 691982.25,
+      "learning_rate": 1.7448559670781893e-05,
+      "loss": 0.5815,
+      "step": 927
+    },
+    {
+      "epoch": 2.062222222222222,
+      "grad_norm": 280133.25,
+      "learning_rate": 1.740740740740741e-05,
+      "loss": 0.4563,
+      "step": 928
+    },
+    {
+      "epoch": 2.0644444444444443,
+      "grad_norm": 340764.8125,
+      "learning_rate": 1.736625514403292e-05,
+      "loss": 0.6178,
+      "step": 929
+    },
+    {
+      "epoch": 2.066666666666667,
+      "grad_norm": 629493.5,
+      "learning_rate": 1.7325102880658437e-05,
+      "loss": 0.6427,
+      "step": 930
+    },
+    {
+      "epoch": 2.068888888888889,
+      "grad_norm": 781840.5,
+      "learning_rate": 1.728395061728395e-05,
+      "loss": 0.7296,
+      "step": 931
+    },
+    {
+      "epoch": 2.071111111111111,
+      "grad_norm": 267660.0625,
+      "learning_rate": 1.7242798353909466e-05,
+      "loss": 0.6227,
+      "step": 932
+    },
+    {
+      "epoch": 2.0733333333333333,
+      "grad_norm": 436358.3125,
+      "learning_rate": 1.720164609053498e-05,
+      "loss": 0.5042,
+      "step": 933
+    },
+    {
+      "epoch": 2.0755555555555554,
+      "grad_norm": 1105647.875,
+      "learning_rate": 1.7160493827160494e-05,
+      "loss": 0.5156,
+      "step": 934
+    },
+    {
+      "epoch": 2.077777777777778,
+      "grad_norm": 451249.875,
+      "learning_rate": 1.711934156378601e-05,
+      "loss": 0.6336,
+      "step": 935
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 826279.0625,
+      "learning_rate": 1.7078189300411522e-05,
+      "loss": 0.462,
+      "step": 936
+    },
+    {
+      "epoch": 2.082222222222222,
+      "grad_norm": 433239.28125,
+      "learning_rate": 1.7037037037037038e-05,
+      "loss": 0.5149,
+      "step": 937
+    },
+    {
+      "epoch": 2.0844444444444443,
+      "grad_norm": 610665.125,
+      "learning_rate": 1.699588477366255e-05,
+      "loss": 0.6512,
+      "step": 938
+    },
+    {
+      "epoch": 2.086666666666667,
+      "grad_norm": 784074.75,
+      "learning_rate": 1.6954732510288067e-05,
+      "loss": 0.5048,
+      "step": 939
+    },
+    {
+      "epoch": 2.088888888888889,
+      "grad_norm": 440249.28125,
+      "learning_rate": 1.6913580246913582e-05,
+      "loss": 0.5385,
+      "step": 940
+    },
+    {
+      "epoch": 2.091111111111111,
+      "grad_norm": 248265.953125,
+      "learning_rate": 1.6872427983539095e-05,
+      "loss": 0.4792,
+      "step": 941
+    },
+    {
+      "epoch": 2.0933333333333333,
+      "grad_norm": 4933805.5,
+      "learning_rate": 1.683127572016461e-05,
+      "loss": 0.5055,
+      "step": 942
+    },
+    {
+      "epoch": 2.0955555555555554,
+      "grad_norm": 368027.625,
+      "learning_rate": 1.6790123456790123e-05,
+      "loss": 0.4709,
+      "step": 943
+    },
+    {
+      "epoch": 2.097777777777778,
+      "grad_norm": 792823.0625,
+      "learning_rate": 1.674897119341564e-05,
+      "loss": 0.6333,
+      "step": 944
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2230348.75,
+      "learning_rate": 1.670781893004115e-05,
+      "loss": 0.4773,
+      "step": 945
+    },
+    {
+      "epoch": 2.102222222222222,
+      "grad_norm": 1184062.25,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.6342,
+      "step": 946
+    },
+    {
+      "epoch": 2.1044444444444443,
+      "grad_norm": 1087129.75,
+      "learning_rate": 1.662551440329218e-05,
+      "loss": 0.479,
+      "step": 947
+    },
+    {
+      "epoch": 2.1066666666666665,
+      "grad_norm": 400438.1875,
+      "learning_rate": 1.6584362139917696e-05,
+      "loss": 0.7179,
+      "step": 948
+    },
+    {
+      "epoch": 2.108888888888889,
+      "grad_norm": 1183529.5,
+      "learning_rate": 1.654320987654321e-05,
+      "loss": 0.5167,
+      "step": 949
+    },
+    {
+      "epoch": 2.111111111111111,
+      "grad_norm": 721013.0,
+      "learning_rate": 1.6502057613168724e-05,
+      "loss": 0.3956,
+      "step": 950
+    },
+    {
+      "epoch": 2.1133333333333333,
+      "grad_norm": 785808.9375,
+      "learning_rate": 1.646090534979424e-05,
+      "loss": 0.4442,
+      "step": 951
+    },
+    {
+      "epoch": 2.1155555555555554,
+      "grad_norm": 224943.34375,
+      "learning_rate": 1.6419753086419752e-05,
+      "loss": 0.5059,
+      "step": 952
+    },
+    {
+      "epoch": 2.117777777777778,
+      "grad_norm": 646250.3125,
+      "learning_rate": 1.6378600823045268e-05,
+      "loss": 0.6695,
+      "step": 953
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1041619.8125,
+      "learning_rate": 1.633744855967078e-05,
+      "loss": 0.5669,
+      "step": 954
+    },
+    {
+      "epoch": 2.1222222222222222,
+      "grad_norm": 610161.875,
+      "learning_rate": 1.62962962962963e-05,
+      "loss": 0.679,
+      "step": 955
+    },
+    {
+      "epoch": 2.1244444444444444,
+      "grad_norm": 175245.484375,
+      "learning_rate": 1.6255144032921812e-05,
+      "loss": 0.6749,
+      "step": 956
+    },
+    {
+      "epoch": 2.1266666666666665,
+      "grad_norm": 616233.6875,
+      "learning_rate": 1.6213991769547325e-05,
+      "loss": 0.5229,
+      "step": 957
+    },
+    {
+      "epoch": 2.128888888888889,
+      "grad_norm": 449539.5,
+      "learning_rate": 1.617283950617284e-05,
+      "loss": 0.6578,
+      "step": 958
+    },
+    {
+      "epoch": 2.131111111111111,
+      "grad_norm": 789648.6875,
+      "learning_rate": 1.6131687242798353e-05,
+      "loss": 0.4188,
+      "step": 959
+    },
+    {
+      "epoch": 2.1333333333333333,
+      "grad_norm": 2248975.5,
+      "learning_rate": 1.609053497942387e-05,
+      "loss": 0.6399,
+      "step": 960
+    },
+    {
+      "epoch": 2.1355555555555554,
+      "grad_norm": 468038.8125,
+      "learning_rate": 1.604938271604938e-05,
+      "loss": 0.4748,
+      "step": 961
+    },
+    {
+      "epoch": 2.137777777777778,
+      "grad_norm": 2706175.25,
+      "learning_rate": 1.6008230452674897e-05,
+      "loss": 0.5013,
+      "step": 962
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 545295.0625,
+      "learning_rate": 1.5967078189300413e-05,
+      "loss": 0.5944,
+      "step": 963
+    },
+    {
+      "epoch": 2.1422222222222222,
+      "grad_norm": 466568.6875,
+      "learning_rate": 1.5925925925925926e-05,
+      "loss": 0.5908,
+      "step": 964
+    },
+    {
+      "epoch": 2.1444444444444444,
+      "grad_norm": 237403.34375,
+      "learning_rate": 1.588477366255144e-05,
+      "loss": 0.6484,
+      "step": 965
+    },
+    {
+      "epoch": 2.1466666666666665,
+      "grad_norm": 637861.4375,
+      "learning_rate": 1.5843621399176954e-05,
+      "loss": 0.4325,
+      "step": 966
+    },
+    {
+      "epoch": 2.148888888888889,
+      "grad_norm": 232760.671875,
+      "learning_rate": 1.580246913580247e-05,
+      "loss": 0.4771,
+      "step": 967
+    },
+    {
+      "epoch": 2.151111111111111,
+      "grad_norm": 811513.5625,
+      "learning_rate": 1.5761316872427982e-05,
+      "loss": 0.552,
+      "step": 968
+    },
+    {
+      "epoch": 2.1533333333333333,
+      "grad_norm": 318511.15625,
+      "learning_rate": 1.5720164609053498e-05,
+      "loss": 0.5806,
+      "step": 969
+    },
+    {
+      "epoch": 2.1555555555555554,
+      "grad_norm": 183807.328125,
+      "learning_rate": 1.5679012345679014e-05,
+      "loss": 0.5115,
+      "step": 970
+    },
+    {
+      "epoch": 2.1577777777777776,
+      "grad_norm": 785010.8125,
+      "learning_rate": 1.563786008230453e-05,
+      "loss": 0.6294,
+      "step": 971
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3169989.25,
+      "learning_rate": 1.5596707818930042e-05,
+      "loss": 0.5082,
+      "step": 972
+    },
+    {
+      "epoch": 2.1622222222222223,
+      "grad_norm": 652374.625,
+      "learning_rate": 1.5555555555555555e-05,
+      "loss": 0.6602,
+      "step": 973
+    },
+    {
+      "epoch": 2.1644444444444444,
+      "grad_norm": 672441.8125,
+      "learning_rate": 1.551440329218107e-05,
+      "loss": 0.4875,
+      "step": 974
+    },
+    {
+      "epoch": 2.1666666666666665,
+      "grad_norm": 424183.3125,
+      "learning_rate": 1.5473251028806583e-05,
+      "loss": 0.5186,
+      "step": 975
+    },
+    {
+      "epoch": 2.168888888888889,
+      "grad_norm": 660900.0625,
+      "learning_rate": 1.54320987654321e-05,
+      "loss": 0.5495,
+      "step": 976
+    },
+    {
+      "epoch": 2.171111111111111,
+      "grad_norm": 607010.9375,
+      "learning_rate": 1.539094650205761e-05,
+      "loss": 0.4558,
+      "step": 977
+    },
+    {
+      "epoch": 2.1733333333333333,
+      "grad_norm": 542253.1875,
+      "learning_rate": 1.534979423868313e-05,
+      "loss": 0.5178,
+      "step": 978
+    },
+    {
+      "epoch": 2.1755555555555555,
+      "grad_norm": 733078.125,
+      "learning_rate": 1.5308641975308643e-05,
+      "loss": 0.4317,
+      "step": 979
+    },
+    {
+      "epoch": 2.1777777777777776,
+      "grad_norm": 745732.3125,
+      "learning_rate": 1.526748971193416e-05,
+      "loss": 0.5786,
+      "step": 980
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 173377.171875,
+      "learning_rate": 1.5226337448559672e-05,
+      "loss": 0.4642,
+      "step": 981
+    },
+    {
+      "epoch": 2.1822222222222223,
+      "grad_norm": 439588.0,
+      "learning_rate": 1.5185185185185186e-05,
+      "loss": 0.6901,
+      "step": 982
+    },
+    {
+      "epoch": 2.1844444444444444,
+      "grad_norm": 1793941.5,
+      "learning_rate": 1.51440329218107e-05,
+      "loss": 0.5587,
+      "step": 983
+    },
+    {
+      "epoch": 2.1866666666666665,
+      "grad_norm": 197852.171875,
+      "learning_rate": 1.5102880658436214e-05,
+      "loss": 0.5243,
+      "step": 984
+    },
+    {
+      "epoch": 2.188888888888889,
+      "grad_norm": 781908.25,
+      "learning_rate": 1.506172839506173e-05,
+      "loss": 0.6291,
+      "step": 985
+    },
+    {
+      "epoch": 2.1911111111111112,
+      "grad_norm": 426974.3125,
+      "learning_rate": 1.5020576131687244e-05,
+      "loss": 0.3932,
+      "step": 986
+    },
+    {
+      "epoch": 2.1933333333333334,
+      "grad_norm": 2040763.875,
+      "learning_rate": 1.4979423868312758e-05,
+      "loss": 0.5105,
+      "step": 987
+    },
+    {
+      "epoch": 2.1955555555555555,
+      "grad_norm": 178407.359375,
+      "learning_rate": 1.4938271604938272e-05,
+      "loss": 0.3597,
+      "step": 988
+    },
+    {
+      "epoch": 2.1977777777777776,
+      "grad_norm": 3175741.75,
+      "learning_rate": 1.4897119341563787e-05,
+      "loss": 0.5331,
+      "step": 989
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 854287.3125,
+      "learning_rate": 1.48559670781893e-05,
+      "loss": 0.5446,
+      "step": 990
+    },
+    {
+      "epoch": 2.2022222222222223,
+      "grad_norm": 164806.8125,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 0.6404,
+      "step": 991
+    },
+    {
+      "epoch": 2.2044444444444444,
+      "grad_norm": 770559.1875,
+      "learning_rate": 1.4773662551440329e-05,
+      "loss": 0.4534,
+      "step": 992
+    },
+    {
+      "epoch": 2.2066666666666666,
+      "grad_norm": 3397058.5,
+      "learning_rate": 1.4732510288065845e-05,
+      "loss": 0.6368,
+      "step": 993
+    },
+    {
+      "epoch": 2.2088888888888887,
+      "grad_norm": 628163.8125,
+      "learning_rate": 1.4691358024691359e-05,
+      "loss": 0.5398,
+      "step": 994
+    },
+    {
+      "epoch": 2.2111111111111112,
+      "grad_norm": 451703.46875,
+      "learning_rate": 1.4650205761316873e-05,
+      "loss": 0.5462,
+      "step": 995
+    },
+    {
+      "epoch": 2.2133333333333334,
+      "grad_norm": 375908.25,
+      "learning_rate": 1.4609053497942387e-05,
+      "loss": 0.4582,
+      "step": 996
+    },
+    {
+      "epoch": 2.2155555555555555,
+      "grad_norm": 450731.4375,
+      "learning_rate": 1.4567901234567902e-05,
+      "loss": 0.4855,
+      "step": 997
+    },
+    {
+      "epoch": 2.2177777777777776,
+      "grad_norm": 436833.4375,
+      "learning_rate": 1.4526748971193416e-05,
+      "loss": 0.544,
+      "step": 998
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 214384.796875,
+      "learning_rate": 1.448559670781893e-05,
+      "loss": 0.5323,
+      "step": 999
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 813175.4375,
+      "learning_rate": 1.4444444444444444e-05,
+      "loss": 0.5391,
+      "step": 1000
+    },
+    {
+      "epoch": 2.2244444444444444,
+      "grad_norm": 252968.859375,
+      "learning_rate": 1.440329218106996e-05,
+      "loss": 0.4484,
+      "step": 1001
+    },
+    {
+      "epoch": 2.2266666666666666,
+      "grad_norm": 698381.125,
+      "learning_rate": 1.4362139917695474e-05,
+      "loss": 0.4685,
+      "step": 1002
+    },
+    {
+      "epoch": 2.2288888888888887,
+      "grad_norm": 1146888.125,
+      "learning_rate": 1.4320987654320988e-05,
+      "loss": 0.4942,
+      "step": 1003
+    },
+    {
+      "epoch": 2.2311111111111113,
+      "grad_norm": 562644.9375,
+      "learning_rate": 1.4279835390946502e-05,
+      "loss": 0.5762,
+      "step": 1004
+    },
+    {
+      "epoch": 2.2333333333333334,
+      "grad_norm": 259112.0,
+      "learning_rate": 1.4238683127572017e-05,
+      "loss": 0.4417,
+      "step": 1005
+    },
+    {
+      "epoch": 2.2355555555555555,
+      "grad_norm": 440842.65625,
+      "learning_rate": 1.419753086419753e-05,
+      "loss": 0.5042,
+      "step": 1006
+    },
+    {
+      "epoch": 2.2377777777777776,
+      "grad_norm": 1526809.625,
+      "learning_rate": 1.4156378600823045e-05,
+      "loss": 0.6759,
+      "step": 1007
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 344080.34375,
+      "learning_rate": 1.411522633744856e-05,
+      "loss": 0.4194,
+      "step": 1008
+    },
+    {
+      "epoch": 2.2422222222222223,
+      "grad_norm": 1327518.875,
+      "learning_rate": 1.4074074074074075e-05,
+      "loss": 0.5189,
+      "step": 1009
+    },
+    {
+      "epoch": 2.2444444444444445,
+      "grad_norm": 496668.375,
+      "learning_rate": 1.4032921810699589e-05,
+      "loss": 0.4864,
+      "step": 1010
+    },
+    {
+      "epoch": 2.2466666666666666,
+      "grad_norm": 341703.5,
+      "learning_rate": 1.3991769547325103e-05,
+      "loss": 0.4511,
+      "step": 1011
+    },
+    {
+      "epoch": 2.2488888888888887,
+      "grad_norm": 1272398.0,
+      "learning_rate": 1.3950617283950617e-05,
+      "loss": 0.4773,
+      "step": 1012
+    },
+    {
+      "epoch": 2.2511111111111113,
+      "grad_norm": 1034172.625,
+      "learning_rate": 1.3909465020576132e-05,
+      "loss": 0.5444,
+      "step": 1013
+    },
+    {
+      "epoch": 2.2533333333333334,
+      "grad_norm": 538651.0625,
+      "learning_rate": 1.3868312757201646e-05,
+      "loss": 0.4896,
+      "step": 1014
+    },
+    {
+      "epoch": 2.2555555555555555,
+      "grad_norm": 759162.3125,
+      "learning_rate": 1.382716049382716e-05,
+      "loss": 0.4478,
+      "step": 1015
+    },
+    {
+      "epoch": 2.2577777777777777,
+      "grad_norm": 1303284.375,
+      "learning_rate": 1.3786008230452676e-05,
+      "loss": 0.6972,
+      "step": 1016
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1262487.5,
+      "learning_rate": 1.374485596707819e-05,
+      "loss": 0.5338,
+      "step": 1017
+    },
+    {
+      "epoch": 2.2622222222222224,
+      "grad_norm": 916309.125,
+      "learning_rate": 1.3703703703703704e-05,
+      "loss": 0.6122,
+      "step": 1018
+    },
+    {
+      "epoch": 2.2644444444444445,
+      "grad_norm": 459222.71875,
+      "learning_rate": 1.3662551440329218e-05,
+      "loss": 0.4757,
+      "step": 1019
+    },
+    {
+      "epoch": 2.2666666666666666,
+      "grad_norm": 517379.3125,
+      "learning_rate": 1.3621399176954732e-05,
+      "loss": 0.6726,
+      "step": 1020
+    },
+    {
+      "epoch": 2.2688888888888887,
+      "grad_norm": 391503.15625,
+      "learning_rate": 1.3580246913580247e-05,
+      "loss": 0.6403,
+      "step": 1021
+    },
+    {
+      "epoch": 2.2711111111111113,
+      "grad_norm": 1065075.875,
+      "learning_rate": 1.353909465020576e-05,
+      "loss": 0.5819,
+      "step": 1022
+    },
+    {
+      "epoch": 2.2733333333333334,
+      "grad_norm": 819979.5,
+      "learning_rate": 1.3497942386831278e-05,
+      "loss": 0.5818,
+      "step": 1023
+    },
+    {
+      "epoch": 2.2755555555555556,
+      "grad_norm": 604300.375,
+      "learning_rate": 1.3456790123456793e-05,
+      "loss": 0.5714,
+      "step": 1024
+    },
+    {
+      "epoch": 2.2777777777777777,
+      "grad_norm": 298492.9375,
+      "learning_rate": 1.3415637860082305e-05,
+      "loss": 0.4724,
+      "step": 1025
+    },
+    {
+      "epoch": 2.2800000000000002,
+      "grad_norm": 535608.0,
+      "learning_rate": 1.3374485596707819e-05,
+      "loss": 0.4827,
+      "step": 1026
+    },
+    {
+      "epoch": 2.2822222222222224,
+      "grad_norm": 2979703.75,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.5448,
+      "step": 1027
+    },
+    {
+      "epoch": 2.2844444444444445,
+      "grad_norm": 876960.875,
+      "learning_rate": 1.3292181069958847e-05,
+      "loss": 0.5854,
+      "step": 1028
+    },
+    {
+      "epoch": 2.2866666666666666,
+      "grad_norm": 350376.8125,
+      "learning_rate": 1.3251028806584362e-05,
+      "loss": 0.5619,
+      "step": 1029
+    },
+    {
+      "epoch": 2.2888888888888888,
+      "grad_norm": 1407370.875,
+      "learning_rate": 1.3209876543209876e-05,
+      "loss": 0.4644,
+      "step": 1030
+    },
+    {
+      "epoch": 2.2911111111111113,
+      "grad_norm": 431142.75,
+      "learning_rate": 1.3168724279835393e-05,
+      "loss": 0.577,
+      "step": 1031
+    },
+    {
+      "epoch": 2.2933333333333334,
+      "grad_norm": 1626133.875,
+      "learning_rate": 1.3127572016460907e-05,
+      "loss": 0.5798,
+      "step": 1032
+    },
+    {
+      "epoch": 2.2955555555555556,
+      "grad_norm": 731867.6875,
+      "learning_rate": 1.308641975308642e-05,
+      "loss": 0.5515,
+      "step": 1033
+    },
+    {
+      "epoch": 2.2977777777777777,
+      "grad_norm": 519474.03125,
+      "learning_rate": 1.3045267489711934e-05,
+      "loss": 0.5661,
+      "step": 1034
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1156532.125,
+      "learning_rate": 1.3004115226337448e-05,
+      "loss": 0.5095,
+      "step": 1035
+    },
+    {
+      "epoch": 2.3022222222222224,
+      "grad_norm": 566316.4375,
+      "learning_rate": 1.2962962962962962e-05,
+      "loss": 0.5757,
+      "step": 1036
+    },
+    {
+      "epoch": 2.3044444444444445,
+      "grad_norm": 1570133.75,
+      "learning_rate": 1.2921810699588477e-05,
+      "loss": 0.5725,
+      "step": 1037
+    },
+    {
+      "epoch": 2.3066666666666666,
+      "grad_norm": 233635.09375,
+      "learning_rate": 1.2880658436213994e-05,
+      "loss": 0.4753,
+      "step": 1038
+    },
+    {
+      "epoch": 2.3088888888888888,
+      "grad_norm": 667559.75,
+      "learning_rate": 1.2839506172839508e-05,
+      "loss": 0.5486,
+      "step": 1039
+    },
+    {
+      "epoch": 2.311111111111111,
+      "grad_norm": 259317.078125,
+      "learning_rate": 1.2798353909465022e-05,
+      "loss": 0.5512,
+      "step": 1040
+    },
+    {
+      "epoch": 2.3133333333333335,
+      "grad_norm": 334346.96875,
+      "learning_rate": 1.2757201646090535e-05,
+      "loss": 0.6806,
+      "step": 1041
+    },
+    {
+      "epoch": 2.3155555555555556,
+      "grad_norm": 380433.90625,
+      "learning_rate": 1.2716049382716049e-05,
+      "loss": 0.4861,
+      "step": 1042
+    },
+    {
+      "epoch": 2.3177777777777777,
+      "grad_norm": 591011.1875,
+      "learning_rate": 1.2674897119341563e-05,
+      "loss": 0.5437,
+      "step": 1043
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 488778.875,
+      "learning_rate": 1.2633744855967077e-05,
+      "loss": 0.528,
+      "step": 1044
+    },
+    {
+      "epoch": 2.3222222222222224,
+      "grad_norm": 1632864.625,
+      "learning_rate": 1.2592592592592592e-05,
+      "loss": 0.5726,
+      "step": 1045
+    },
+    {
+      "epoch": 2.3244444444444445,
+      "grad_norm": 222425.59375,
+      "learning_rate": 1.255144032921811e-05,
+      "loss": 0.5392,
+      "step": 1046
+    },
+    {
+      "epoch": 2.3266666666666667,
+      "grad_norm": 402851.4375,
+      "learning_rate": 1.2510288065843623e-05,
+      "loss": 0.4061,
+      "step": 1047
+    },
+    {
+      "epoch": 2.328888888888889,
+      "grad_norm": 253659.3125,
+      "learning_rate": 1.2469135802469137e-05,
+      "loss": 0.5208,
+      "step": 1048
+    },
+    {
+      "epoch": 2.3311111111111114,
+      "grad_norm": 753892.75,
+      "learning_rate": 1.2427983539094652e-05,
+      "loss": 0.5614,
+      "step": 1049
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 760216.625,
+      "learning_rate": 1.2386831275720164e-05,
+      "loss": 0.4855,
+      "step": 1050
+    },
+    {
+      "epoch": 2.3355555555555556,
+      "grad_norm": 1279448.5,
+      "learning_rate": 1.2345679012345678e-05,
+      "loss": 0.5871,
+      "step": 1051
+    },
+    {
+      "epoch": 2.3377777777777777,
+      "grad_norm": 878854.0625,
+      "learning_rate": 1.2304526748971194e-05,
+      "loss": 0.469,
+      "step": 1052
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1088848.0,
+      "learning_rate": 1.2263374485596708e-05,
+      "loss": 0.4336,
+      "step": 1053
+    },
+    {
+      "epoch": 2.3422222222222224,
+      "grad_norm": 986841.875,
+      "learning_rate": 1.2222222222222222e-05,
+      "loss": 0.4855,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3444444444444446,
+      "grad_norm": 534454.5,
+      "learning_rate": 1.2181069958847738e-05,
+      "loss": 0.5174,
+      "step": 1055
+    },
+    {
+      "epoch": 2.3466666666666667,
+      "grad_norm": 1499670.625,
+      "learning_rate": 1.2139917695473252e-05,
+      "loss": 0.4638,
+      "step": 1056
+    },
+    {
+      "epoch": 2.348888888888889,
+      "grad_norm": 1340620.375,
+      "learning_rate": 1.2098765432098767e-05,
+      "loss": 0.4267,
+      "step": 1057
+    },
+    {
+      "epoch": 2.351111111111111,
+      "grad_norm": 1036810.5625,
+      "learning_rate": 1.2057613168724279e-05,
+      "loss": 0.4333,
+      "step": 1058
+    },
+    {
+      "epoch": 2.3533333333333335,
+      "grad_norm": 415893.0,
+      "learning_rate": 1.2016460905349795e-05,
+      "loss": 0.5185,
+      "step": 1059
+    },
+    {
+      "epoch": 2.3555555555555556,
+      "grad_norm": 490719.5,
+      "learning_rate": 1.1975308641975309e-05,
+      "loss": 0.4477,
+      "step": 1060
+    },
+    {
+      "epoch": 2.3577777777777778,
+      "grad_norm": 342900.34375,
+      "learning_rate": 1.1934156378600823e-05,
+      "loss": 0.4744,
+      "step": 1061
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 299380.8125,
+      "learning_rate": 1.1893004115226337e-05,
+      "loss": 0.5143,
+      "step": 1062
+    },
+    {
+      "epoch": 2.362222222222222,
+      "grad_norm": 789056.3125,
+      "learning_rate": 1.1851851851851853e-05,
+      "loss": 0.5449,
+      "step": 1063
+    },
+    {
+      "epoch": 2.3644444444444446,
+      "grad_norm": 476553.375,
+      "learning_rate": 1.1810699588477367e-05,
+      "loss": 0.5639,
+      "step": 1064
+    },
+    {
+      "epoch": 2.3666666666666667,
+      "grad_norm": 607017.375,
+      "learning_rate": 1.1769547325102882e-05,
+      "loss": 0.4924,
+      "step": 1065
+    },
+    {
+      "epoch": 2.368888888888889,
+      "grad_norm": 197236.609375,
+      "learning_rate": 1.1728395061728396e-05,
+      "loss": 0.5817,
+      "step": 1066
+    },
+    {
+      "epoch": 2.371111111111111,
+      "grad_norm": 438141.1875,
+      "learning_rate": 1.168724279835391e-05,
+      "loss": 0.5657,
+      "step": 1067
+    },
+    {
+      "epoch": 2.3733333333333335,
+      "grad_norm": 351896.28125,
+      "learning_rate": 1.1646090534979424e-05,
+      "loss": 0.5776,
+      "step": 1068
+    },
+    {
+      "epoch": 2.3755555555555556,
+      "grad_norm": 486697.25,
+      "learning_rate": 1.1604938271604938e-05,
+      "loss": 0.4206,
+      "step": 1069
+    },
+    {
+      "epoch": 2.3777777777777778,
+      "grad_norm": 895090.4375,
+      "learning_rate": 1.1563786008230454e-05,
+      "loss": 0.4537,
+      "step": 1070
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 385076.90625,
+      "learning_rate": 1.1522633744855968e-05,
+      "loss": 0.5011,
+      "step": 1071
+    },
+    {
+      "epoch": 2.3822222222222225,
+      "grad_norm": 422288.5625,
+      "learning_rate": 1.1481481481481482e-05,
+      "loss": 0.5286,
+      "step": 1072
+    },
+    {
+      "epoch": 2.3844444444444446,
+      "grad_norm": 1136325.5,
+      "learning_rate": 1.1440329218106997e-05,
+      "loss": 0.5168,
+      "step": 1073
+    },
+    {
+      "epoch": 2.3866666666666667,
+      "grad_norm": 385298.5625,
+      "learning_rate": 1.139917695473251e-05,
+      "loss": 0.5362,
+      "step": 1074
+    },
+    {
+      "epoch": 2.388888888888889,
+      "grad_norm": 982441.0,
+      "learning_rate": 1.1358024691358025e-05,
+      "loss": 0.5212,
+      "step": 1075
+    },
+    {
+      "epoch": 2.391111111111111,
+      "grad_norm": 119587.1484375,
+      "learning_rate": 1.1316872427983539e-05,
+      "loss": 0.5002,
+      "step": 1076
+    },
+    {
+      "epoch": 2.3933333333333335,
+      "grad_norm": 656648.4375,
+      "learning_rate": 1.1275720164609053e-05,
+      "loss": 0.6891,
+      "step": 1077
+    },
+    {
+      "epoch": 2.3955555555555557,
+      "grad_norm": 761881.3125,
+      "learning_rate": 1.123456790123457e-05,
+      "loss": 0.3717,
+      "step": 1078
+    },
+    {
+      "epoch": 2.397777777777778,
+      "grad_norm": 1027932.5,
+      "learning_rate": 1.1193415637860083e-05,
+      "loss": 0.6497,
+      "step": 1079
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 405516.03125,
+      "learning_rate": 1.1152263374485597e-05,
+      "loss": 0.4851,
+      "step": 1080
+    },
+    {
+      "epoch": 2.402222222222222,
+      "grad_norm": 1639993.875,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.5472,
+      "step": 1081
+    },
+    {
+      "epoch": 2.4044444444444446,
+      "grad_norm": 212303.59375,
+      "learning_rate": 1.1069958847736626e-05,
+      "loss": 0.5818,
+      "step": 1082
+    },
+    {
+      "epoch": 2.4066666666666667,
+      "grad_norm": 152677.265625,
+      "learning_rate": 1.102880658436214e-05,
+      "loss": 0.5628,
+      "step": 1083
+    },
+    {
+      "epoch": 2.408888888888889,
+      "grad_norm": 957934.75,
+      "learning_rate": 1.0987654320987654e-05,
+      "loss": 0.4918,
+      "step": 1084
+    },
+    {
+      "epoch": 2.411111111111111,
+      "grad_norm": 318580.0625,
+      "learning_rate": 1.094650205761317e-05,
+      "loss": 0.5454,
+      "step": 1085
+    },
+    {
+      "epoch": 2.413333333333333,
+      "grad_norm": 335783.8125,
+      "learning_rate": 1.0905349794238684e-05,
+      "loss": 0.3913,
+      "step": 1086
+    },
+    {
+      "epoch": 2.4155555555555557,
+      "grad_norm": 606167.6875,
+      "learning_rate": 1.0864197530864198e-05,
+      "loss": 0.4603,
+      "step": 1087
+    },
+    {
+      "epoch": 2.417777777777778,
+      "grad_norm": 1082866.5,
+      "learning_rate": 1.0823045267489712e-05,
+      "loss": 0.556,
+      "step": 1088
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 439366.1875,
+      "learning_rate": 1.0781893004115227e-05,
+      "loss": 0.5366,
+      "step": 1089
+    },
+    {
+      "epoch": 2.422222222222222,
+      "grad_norm": 859306.625,
+      "learning_rate": 1.074074074074074e-05,
+      "loss": 0.557,
+      "step": 1090
+    },
+    {
+      "epoch": 2.4244444444444446,
+      "grad_norm": 356610.53125,
+      "learning_rate": 1.0699588477366255e-05,
+      "loss": 0.6092,
+      "step": 1091
+    },
+    {
+      "epoch": 2.4266666666666667,
+      "grad_norm": 862626.875,
+      "learning_rate": 1.0658436213991769e-05,
+      "loss": 0.5368,
+      "step": 1092
+    },
+    {
+      "epoch": 2.428888888888889,
+      "grad_norm": 477033.71875,
+      "learning_rate": 1.0617283950617285e-05,
+      "loss": 0.6038,
+      "step": 1093
+    },
+    {
+      "epoch": 2.431111111111111,
+      "grad_norm": 184494.3125,
+      "learning_rate": 1.0576131687242799e-05,
+      "loss": 0.4473,
+      "step": 1094
+    },
+    {
+      "epoch": 2.4333333333333336,
+      "grad_norm": 704048.75,
+      "learning_rate": 1.0534979423868313e-05,
+      "loss": 0.5575,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4355555555555557,
+      "grad_norm": 1460752.5,
+      "learning_rate": 1.0493827160493827e-05,
+      "loss": 0.6811,
+      "step": 1096
+    },
+    {
+      "epoch": 2.437777777777778,
+      "grad_norm": 1865951.0,
+      "learning_rate": 1.0452674897119343e-05,
+      "loss": 0.7688,
+      "step": 1097
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 316543.03125,
+      "learning_rate": 1.0411522633744856e-05,
+      "loss": 0.4987,
+      "step": 1098
+    },
+    {
+      "epoch": 2.442222222222222,
+      "grad_norm": 392440.625,
+      "learning_rate": 1.037037037037037e-05,
+      "loss": 0.6592,
+      "step": 1099
+    },
+    {
+      "epoch": 2.4444444444444446,
+      "grad_norm": 286228.875,
+      "learning_rate": 1.0329218106995884e-05,
+      "loss": 0.6525,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4466666666666668,
+      "grad_norm": 436523.78125,
+      "learning_rate": 1.02880658436214e-05,
+      "loss": 0.4301,
+      "step": 1101
+    },
+    {
+      "epoch": 2.448888888888889,
+      "grad_norm": 501997.0625,
+      "learning_rate": 1.0246913580246914e-05,
+      "loss": 0.4842,
+      "step": 1102
+    },
+    {
+      "epoch": 2.451111111111111,
+      "grad_norm": 902289.5625,
+      "learning_rate": 1.0205761316872428e-05,
+      "loss": 0.5599,
+      "step": 1103
+    },
+    {
+      "epoch": 2.453333333333333,
+      "grad_norm": 2513825.25,
+      "learning_rate": 1.0164609053497944e-05,
+      "loss": 0.6615,
+      "step": 1104
+    },
+    {
+      "epoch": 2.4555555555555557,
+      "grad_norm": 345013.1875,
+      "learning_rate": 1.0123456790123458e-05,
+      "loss": 0.5646,
+      "step": 1105
+    },
+    {
+      "epoch": 2.457777777777778,
+      "grad_norm": 651783.6875,
+      "learning_rate": 1.008230452674897e-05,
+      "loss": 0.6124,
+      "step": 1106
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2414475.75,
+      "learning_rate": 1.0041152263374485e-05,
+      "loss": 0.4809,
+      "step": 1107
+    },
+    {
+      "epoch": 2.462222222222222,
+      "grad_norm": 312356.15625,
+      "learning_rate": 1e-05,
+      "loss": 0.5917,
+      "step": 1108
+    },
+    {
+      "epoch": 2.464444444444444,
+      "grad_norm": 1360208.625,
+      "learning_rate": 9.958847736625515e-06,
+      "loss": 0.7328,
+      "step": 1109
+    },
+    {
+      "epoch": 2.466666666666667,
+      "grad_norm": 564814.125,
+      "learning_rate": 9.917695473251029e-06,
+      "loss": 0.5012,
+      "step": 1110
+    },
+    {
+      "epoch": 2.468888888888889,
+      "grad_norm": 783253.1875,
+      "learning_rate": 9.876543209876543e-06,
+      "loss": 0.5959,
+      "step": 1111
+    },
+    {
+      "epoch": 2.471111111111111,
+      "grad_norm": 1069690.125,
+      "learning_rate": 9.83539094650206e-06,
+      "loss": 0.4775,
+      "step": 1112
+    },
+    {
+      "epoch": 2.473333333333333,
+      "grad_norm": 698115.0,
+      "learning_rate": 9.794238683127573e-06,
+      "loss": 0.6113,
+      "step": 1113
+    },
+    {
+      "epoch": 2.4755555555555557,
+      "grad_norm": 595121.625,
+      "learning_rate": 9.753086419753086e-06,
+      "loss": 0.4327,
+      "step": 1114
+    },
+    {
+      "epoch": 2.477777777777778,
+      "grad_norm": 686340.625,
+      "learning_rate": 9.7119341563786e-06,
+      "loss": 0.5782,
+      "step": 1115
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 562637.4375,
+      "learning_rate": 9.670781893004116e-06,
+      "loss": 0.572,
+      "step": 1116
+    },
+    {
+      "epoch": 2.482222222222222,
+      "grad_norm": 700603.75,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 0.4896,
+      "step": 1117
+    },
+    {
+      "epoch": 2.4844444444444447,
+      "grad_norm": 8574117.0,
+      "learning_rate": 9.588477366255144e-06,
+      "loss": 0.5126,
+      "step": 1118
+    },
+    {
+      "epoch": 2.486666666666667,
+      "grad_norm": 1065802.0,
+      "learning_rate": 9.54732510288066e-06,
+      "loss": 0.5362,
+      "step": 1119
+    },
+    {
+      "epoch": 2.488888888888889,
+      "grad_norm": 333646.25,
+      "learning_rate": 9.506172839506174e-06,
+      "loss": 0.5609,
+      "step": 1120
+    },
+    {
+      "epoch": 2.491111111111111,
+      "grad_norm": 804433.9375,
+      "learning_rate": 9.465020576131688e-06,
+      "loss": 0.5512,
+      "step": 1121
+    },
+    {
+      "epoch": 2.493333333333333,
+      "grad_norm": 391463.375,
+      "learning_rate": 9.423868312757202e-06,
+      "loss": 0.4817,
+      "step": 1122
+    },
+    {
+      "epoch": 2.4955555555555557,
+      "grad_norm": 1378079.375,
+      "learning_rate": 9.382716049382717e-06,
+      "loss": 0.3697,
+      "step": 1123
+    },
+    {
+      "epoch": 2.497777777777778,
+      "grad_norm": 505990.03125,
+      "learning_rate": 9.34156378600823e-06,
+      "loss": 0.5476,
+      "step": 1124
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3523881.5,
+      "learning_rate": 9.300411522633745e-06,
+      "loss": 0.5519,
+      "step": 1125
+    },
+    {
+      "epoch": 2.502222222222222,
+      "grad_norm": 341878.0625,
+      "learning_rate": 9.259259259259259e-06,
+      "loss": 0.6628,
+      "step": 1126
+    },
+    {
+      "epoch": 2.5044444444444443,
+      "grad_norm": 391592.46875,
+      "learning_rate": 9.218106995884775e-06,
+      "loss": 0.5047,
+      "step": 1127
+    },
+    {
+      "epoch": 2.506666666666667,
+      "grad_norm": 1288312.375,
+      "learning_rate": 9.17695473251029e-06,
+      "loss": 0.5387,
+      "step": 1128
+    },
+    {
+      "epoch": 2.508888888888889,
+      "grad_norm": 965933.8125,
+      "learning_rate": 9.135802469135803e-06,
+      "loss": 0.6002,
+      "step": 1129
+    },
+    {
+      "epoch": 2.511111111111111,
+      "grad_norm": 575313.9375,
+      "learning_rate": 9.094650205761317e-06,
+      "loss": 0.5218,
+      "step": 1130
+    },
+    {
+      "epoch": 2.513333333333333,
+      "grad_norm": 508500.59375,
+      "learning_rate": 9.053497942386832e-06,
+      "loss": 0.5815,
+      "step": 1131
+    },
+    {
+      "epoch": 2.5155555555555553,
+      "grad_norm": 388008.0,
+      "learning_rate": 9.012345679012346e-06,
+      "loss": 0.4585,
+      "step": 1132
+    },
+    {
+      "epoch": 2.517777777777778,
+      "grad_norm": 927888.25,
+      "learning_rate": 8.97119341563786e-06,
+      "loss": 0.5115,
+      "step": 1133
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1426170.375,
+      "learning_rate": 8.930041152263374e-06,
+      "loss": 0.4249,
+      "step": 1134
+    },
+    {
+      "epoch": 2.522222222222222,
+      "grad_norm": 710862.0625,
+      "learning_rate": 8.88888888888889e-06,
+      "loss": 0.4375,
+      "step": 1135
+    },
+    {
+      "epoch": 2.5244444444444447,
+      "grad_norm": 900323.4375,
+      "learning_rate": 8.847736625514404e-06,
+      "loss": 0.5462,
+      "step": 1136
+    },
+    {
+      "epoch": 2.5266666666666664,
+      "grad_norm": 444571.1875,
+      "learning_rate": 8.806584362139918e-06,
+      "loss": 0.4567,
+      "step": 1137
+    },
+    {
+      "epoch": 2.528888888888889,
+      "grad_norm": 567729.4375,
+      "learning_rate": 8.765432098765432e-06,
+      "loss": 0.5864,
+      "step": 1138
+    },
+    {
+      "epoch": 2.531111111111111,
+      "grad_norm": 215833.796875,
+      "learning_rate": 8.724279835390947e-06,
+      "loss": 0.4469,
+      "step": 1139
+    },
+    {
+      "epoch": 2.533333333333333,
+      "grad_norm": 569421.875,
+      "learning_rate": 8.68312757201646e-06,
+      "loss": 0.5494,
+      "step": 1140
+    },
+    {
+      "epoch": 2.535555555555556,
+      "grad_norm": 558661.5,
+      "learning_rate": 8.641975308641975e-06,
+      "loss": 0.4522,
+      "step": 1141
+    },
+    {
+      "epoch": 2.537777777777778,
+      "grad_norm": 2251505.0,
+      "learning_rate": 8.60082304526749e-06,
+      "loss": 0.4683,
+      "step": 1142
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 796107.0625,
+      "learning_rate": 8.559670781893005e-06,
+      "loss": 0.6036,
+      "step": 1143
+    },
+    {
+      "epoch": 2.542222222222222,
+      "grad_norm": 360527.28125,
+      "learning_rate": 8.518518518518519e-06,
+      "loss": 0.4764,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5444444444444443,
+      "grad_norm": 412231.875,
+      "learning_rate": 8.477366255144033e-06,
+      "loss": 0.4871,
+      "step": 1145
+    },
+    {
+      "epoch": 2.546666666666667,
+      "grad_norm": 331903.4375,
+      "learning_rate": 8.436213991769547e-06,
+      "loss": 0.5037,
+      "step": 1146
+    },
+    {
+      "epoch": 2.548888888888889,
+      "grad_norm": 1922785.125,
+      "learning_rate": 8.395061728395062e-06,
+      "loss": 0.5288,
+      "step": 1147
+    },
+    {
+      "epoch": 2.551111111111111,
+      "grad_norm": 391779.03125,
+      "learning_rate": 8.353909465020576e-06,
+      "loss": 0.5619,
+      "step": 1148
+    },
+    {
+      "epoch": 2.5533333333333332,
+      "grad_norm": 590551.1875,
+      "learning_rate": 8.31275720164609e-06,
+      "loss": 0.539,
+      "step": 1149
+    },
+    {
+      "epoch": 2.5555555555555554,
+      "grad_norm": 1713623.75,
+      "learning_rate": 8.271604938271606e-06,
+      "loss": 0.5789,
+      "step": 1150
+    },
+    {
+      "epoch": 2.557777777777778,
+      "grad_norm": 550370.375,
+      "learning_rate": 8.23045267489712e-06,
+      "loss": 0.5584,
+      "step": 1151
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 662291.75,
+      "learning_rate": 8.189300411522634e-06,
+      "loss": 0.5375,
+      "step": 1152
+    },
+    {
+      "epoch": 2.562222222222222,
+      "grad_norm": 457474.65625,
+      "learning_rate": 8.14814814814815e-06,
+      "loss": 0.5263,
+      "step": 1153
+    },
+    {
+      "epoch": 2.5644444444444443,
+      "grad_norm": 955969.6875,
+      "learning_rate": 8.106995884773662e-06,
+      "loss": 0.589,
+      "step": 1154
+    },
+    {
+      "epoch": 2.5666666666666664,
+      "grad_norm": 712221.1875,
+      "learning_rate": 8.065843621399177e-06,
+      "loss": 0.4752,
+      "step": 1155
+    },
+    {
+      "epoch": 2.568888888888889,
+      "grad_norm": 2151926.25,
+      "learning_rate": 8.02469135802469e-06,
+      "loss": 0.4964,
+      "step": 1156
+    },
+    {
+      "epoch": 2.571111111111111,
+      "grad_norm": 749660.25,
+      "learning_rate": 7.983539094650207e-06,
+      "loss": 0.7246,
+      "step": 1157
+    },
+    {
+      "epoch": 2.5733333333333333,
+      "grad_norm": 1508001.0,
+      "learning_rate": 7.94238683127572e-06,
+      "loss": 0.4743,
+      "step": 1158
+    },
+    {
+      "epoch": 2.575555555555556,
+      "grad_norm": 749034.625,
+      "learning_rate": 7.901234567901235e-06,
+      "loss": 0.5987,
+      "step": 1159
+    },
+    {
+      "epoch": 2.5777777777777775,
+      "grad_norm": 288706.375,
+      "learning_rate": 7.860082304526749e-06,
+      "loss": 0.4942,
+      "step": 1160
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 256755.78125,
+      "learning_rate": 7.818930041152265e-06,
+      "loss": 0.569,
+      "step": 1161
+    },
+    {
+      "epoch": 2.582222222222222,
+      "grad_norm": 871312.375,
+      "learning_rate": 7.777777777777777e-06,
+      "loss": 0.4861,
+      "step": 1162
+    },
+    {
+      "epoch": 2.5844444444444443,
+      "grad_norm": 900119.3125,
+      "learning_rate": 7.736625514403292e-06,
+      "loss": 0.6361,
+      "step": 1163
+    },
+    {
+      "epoch": 2.586666666666667,
+      "grad_norm": 1303723.375,
+      "learning_rate": 7.695473251028806e-06,
+      "loss": 0.6025,
+      "step": 1164
+    },
+    {
+      "epoch": 2.588888888888889,
+      "grad_norm": 1132907.75,
+      "learning_rate": 7.654320987654322e-06,
+      "loss": 0.6582,
+      "step": 1165
+    },
+    {
+      "epoch": 2.591111111111111,
+      "grad_norm": 928938.375,
+      "learning_rate": 7.613168724279836e-06,
+      "loss": 0.6281,
+      "step": 1166
+    },
+    {
+      "epoch": 2.5933333333333333,
+      "grad_norm": 289348.6875,
+      "learning_rate": 7.57201646090535e-06,
+      "loss": 0.7755,
+      "step": 1167
+    },
+    {
+      "epoch": 2.5955555555555554,
+      "grad_norm": 447535.09375,
+      "learning_rate": 7.530864197530865e-06,
+      "loss": 0.5153,
+      "step": 1168
+    },
+    {
+      "epoch": 2.597777777777778,
+      "grad_norm": 886854.625,
+      "learning_rate": 7.489711934156379e-06,
+      "loss": 0.5239,
+      "step": 1169
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1512038.0,
+      "learning_rate": 7.448559670781893e-06,
+      "loss": 0.7509,
+      "step": 1170
+    },
+    {
+      "epoch": 2.602222222222222,
+      "grad_norm": 322297.5625,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.5442,
+      "step": 1171
+    },
+    {
+      "epoch": 2.6044444444444443,
+      "grad_norm": 359192.5,
+      "learning_rate": 7.3662551440329225e-06,
+      "loss": 0.5081,
+      "step": 1172
+    },
+    {
+      "epoch": 2.6066666666666665,
+      "grad_norm": 1025895.5,
+      "learning_rate": 7.325102880658437e-06,
+      "loss": 0.5026,
+      "step": 1173
+    },
+    {
+      "epoch": 2.608888888888889,
+      "grad_norm": 358256.5625,
+      "learning_rate": 7.283950617283951e-06,
+      "loss": 0.5034,
+      "step": 1174
+    },
+    {
+      "epoch": 2.611111111111111,
+      "grad_norm": 276032.125,
+      "learning_rate": 7.242798353909465e-06,
+      "loss": 0.5133,
+      "step": 1175
+    },
+    {
+      "epoch": 2.6133333333333333,
+      "grad_norm": 566315.75,
+      "learning_rate": 7.20164609053498e-06,
+      "loss": 0.5736,
+      "step": 1176
+    },
+    {
+      "epoch": 2.6155555555555554,
+      "grad_norm": 442241.8125,
+      "learning_rate": 7.160493827160494e-06,
+      "loss": 0.5482,
+      "step": 1177
+    },
+    {
+      "epoch": 2.6177777777777775,
+      "grad_norm": 414541.0625,
+      "learning_rate": 7.119341563786008e-06,
+      "loss": 0.4391,
+      "step": 1178
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 462216.65625,
+      "learning_rate": 7.0781893004115225e-06,
+      "loss": 0.512,
+      "step": 1179
+    },
+    {
+      "epoch": 2.6222222222222222,
+      "grad_norm": 352849.78125,
+      "learning_rate": 7.0370370370370375e-06,
+      "loss": 0.5742,
+      "step": 1180
+    },
+    {
+      "epoch": 2.6244444444444444,
+      "grad_norm": 1022206.0,
+      "learning_rate": 6.995884773662552e-06,
+      "loss": 0.5395,
+      "step": 1181
+    },
+    {
+      "epoch": 2.626666666666667,
+      "grad_norm": 1133120.0,
+      "learning_rate": 6.954732510288066e-06,
+      "loss": 0.591,
+      "step": 1182
+    },
+    {
+      "epoch": 2.628888888888889,
+      "grad_norm": 1288370.625,
+      "learning_rate": 6.91358024691358e-06,
+      "loss": 0.5779,
+      "step": 1183
+    },
+    {
+      "epoch": 2.631111111111111,
+      "grad_norm": 271517.6875,
+      "learning_rate": 6.872427983539095e-06,
+      "loss": 0.5678,
+      "step": 1184
+    },
+    {
+      "epoch": 2.6333333333333333,
+      "grad_norm": 429049.8125,
+      "learning_rate": 6.831275720164609e-06,
+      "loss": 0.4092,
+      "step": 1185
+    },
+    {
+      "epoch": 2.6355555555555554,
+      "grad_norm": 456271.40625,
+      "learning_rate": 6.790123456790123e-06,
+      "loss": 0.4815,
+      "step": 1186
+    },
+    {
+      "epoch": 2.637777777777778,
+      "grad_norm": 1050220.75,
+      "learning_rate": 6.748971193415639e-06,
+      "loss": 0.5069,
+      "step": 1187
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 350885.25,
+      "learning_rate": 6.7078189300411525e-06,
+      "loss": 0.5406,
+      "step": 1188
+    },
+    {
+      "epoch": 2.6422222222222222,
+      "grad_norm": 195271.328125,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.39,
+      "step": 1189
+    },
+    {
+      "epoch": 2.6444444444444444,
+      "grad_norm": 583369.375,
+      "learning_rate": 6.625514403292181e-06,
+      "loss": 0.601,
+      "step": 1190
+    },
+    {
+      "epoch": 2.6466666666666665,
+      "grad_norm": 644399.125,
+      "learning_rate": 6.584362139917697e-06,
+      "loss": 0.5584,
+      "step": 1191
+    },
+    {
+      "epoch": 2.648888888888889,
+      "grad_norm": 396948.625,
+      "learning_rate": 6.54320987654321e-06,
+      "loss": 0.4902,
+      "step": 1192
+    },
+    {
+      "epoch": 2.651111111111111,
+      "grad_norm": 538472.9375,
+      "learning_rate": 6.502057613168724e-06,
+      "loss": 0.5069,
+      "step": 1193
+    },
+    {
+      "epoch": 2.6533333333333333,
+      "grad_norm": 734710.625,
+      "learning_rate": 6.460905349794238e-06,
+      "loss": 0.5317,
+      "step": 1194
+    },
+    {
+      "epoch": 2.6555555555555554,
+      "grad_norm": 932180.3125,
+      "learning_rate": 6.419753086419754e-06,
+      "loss": 0.5767,
+      "step": 1195
+    },
+    {
+      "epoch": 2.6577777777777776,
+      "grad_norm": 534092.5,
+      "learning_rate": 6.3786008230452675e-06,
+      "loss": 0.5063,
+      "step": 1196
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 756892.3125,
+      "learning_rate": 6.337448559670782e-06,
+      "loss": 0.491,
+      "step": 1197
+    },
+    {
+      "epoch": 2.6622222222222223,
+      "grad_norm": 469926.84375,
+      "learning_rate": 6.296296296296296e-06,
+      "loss": 0.4764,
+      "step": 1198
+    },
+    {
+      "epoch": 2.6644444444444444,
+      "grad_norm": 853601.875,
+      "learning_rate": 6.255144032921812e-06,
+      "loss": 0.6428,
+      "step": 1199
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 134657.203125,
+      "learning_rate": 6.213991769547326e-06,
+      "loss": 0.3901,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6688888888888886,
+      "grad_norm": 912945.125,
+      "learning_rate": 6.172839506172839e-06,
+      "loss": 0.8071,
+      "step": 1201
+    },
+    {
+      "epoch": 2.671111111111111,
+      "grad_norm": 359723.96875,
+      "learning_rate": 6.131687242798354e-06,
+      "loss": 0.4296,
+      "step": 1202
+    },
+    {
+      "epoch": 2.6733333333333333,
+      "grad_norm": 336840.28125,
+      "learning_rate": 6.090534979423869e-06,
+      "loss": 0.5657,
+      "step": 1203
+    },
+    {
+      "epoch": 2.6755555555555555,
+      "grad_norm": 381600.5625,
+      "learning_rate": 6.049382716049383e-06,
+      "loss": 0.6522,
+      "step": 1204
+    },
+    {
+      "epoch": 2.677777777777778,
+      "grad_norm": 159033.796875,
+      "learning_rate": 6.0082304526748975e-06,
+      "loss": 0.5458,
+      "step": 1205
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1330061.5,
+      "learning_rate": 5.967078189300412e-06,
+      "loss": 0.5794,
+      "step": 1206
+    },
+    {
+      "epoch": 2.6822222222222223,
+      "grad_norm": 631698.125,
+      "learning_rate": 5.925925925925927e-06,
+      "loss": 0.5271,
+      "step": 1207
+    },
+    {
+      "epoch": 2.6844444444444444,
+      "grad_norm": 357874.53125,
+      "learning_rate": 5.884773662551441e-06,
+      "loss": 0.3977,
+      "step": 1208
+    },
+    {
+      "epoch": 2.6866666666666665,
+      "grad_norm": 3945204.0,
+      "learning_rate": 5.843621399176955e-06,
+      "loss": 0.4593,
+      "step": 1209
+    },
+    {
+      "epoch": 2.688888888888889,
+      "grad_norm": 1239144.25,
+      "learning_rate": 5.802469135802469e-06,
+      "loss": 0.4449,
+      "step": 1210
+    },
+    {
+      "epoch": 2.6911111111111112,
+      "grad_norm": 1069190.125,
+      "learning_rate": 5.761316872427984e-06,
+      "loss": 0.526,
+      "step": 1211
+    },
+    {
+      "epoch": 2.6933333333333334,
+      "grad_norm": 1473693.75,
+      "learning_rate": 5.720164609053498e-06,
+      "loss": 0.6264,
+      "step": 1212
+    },
+    {
+      "epoch": 2.6955555555555555,
+      "grad_norm": 660884.1875,
+      "learning_rate": 5.6790123456790125e-06,
+      "loss": 0.5883,
+      "step": 1213
+    },
+    {
+      "epoch": 2.6977777777777776,
+      "grad_norm": 450712.21875,
+      "learning_rate": 5.637860082304527e-06,
+      "loss": 0.5217,
+      "step": 1214
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1028915.375,
+      "learning_rate": 5.596707818930042e-06,
+      "loss": 0.5244,
+      "step": 1215
+    },
+    {
+      "epoch": 2.7022222222222223,
+      "grad_norm": 1851113.75,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 0.5434,
+      "step": 1216
+    },
+    {
+      "epoch": 2.7044444444444444,
+      "grad_norm": 1106682.5,
+      "learning_rate": 5.51440329218107e-06,
+      "loss": 0.5526,
+      "step": 1217
+    },
+    {
+      "epoch": 2.7066666666666666,
+      "grad_norm": 295857.375,
+      "learning_rate": 5.473251028806585e-06,
+      "loss": 0.595,
+      "step": 1218
+    },
+    {
+      "epoch": 2.7088888888888887,
+      "grad_norm": 563381.375,
+      "learning_rate": 5.432098765432099e-06,
+      "loss": 0.5262,
+      "step": 1219
+    },
+    {
+      "epoch": 2.7111111111111112,
+      "grad_norm": 454781.5625,
+      "learning_rate": 5.390946502057613e-06,
+      "loss": 0.4509,
+      "step": 1220
+    },
+    {
+      "epoch": 2.7133333333333334,
+      "grad_norm": 932880.8125,
+      "learning_rate": 5.3497942386831275e-06,
+      "loss": 0.6851,
+      "step": 1221
+    },
+    {
+      "epoch": 2.7155555555555555,
+      "grad_norm": 1671730.75,
+      "learning_rate": 5.3086419753086425e-06,
+      "loss": 0.6121,
+      "step": 1222
+    },
+    {
+      "epoch": 2.7177777777777776,
+      "grad_norm": 4467779.0,
+      "learning_rate": 5.267489711934157e-06,
+      "loss": 0.6277,
+      "step": 1223
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 900557.4375,
+      "learning_rate": 5.226337448559672e-06,
+      "loss": 0.4294,
+      "step": 1224
+    },
+    {
+      "epoch": 2.7222222222222223,
+      "grad_norm": 160277.46875,
+      "learning_rate": 5.185185185185185e-06,
+      "loss": 0.6811,
+      "step": 1225
+    },
+    {
+      "epoch": 2.7244444444444444,
+      "grad_norm": 149784.03125,
+      "learning_rate": 5.1440329218107e-06,
+      "loss": 0.5307,
+      "step": 1226
+    },
+    {
+      "epoch": 2.7266666666666666,
+      "grad_norm": 244266.984375,
+      "learning_rate": 5.102880658436214e-06,
+      "loss": 0.5173,
+      "step": 1227
+    },
+    {
+      "epoch": 2.728888888888889,
+      "grad_norm": 401191.625,
+      "learning_rate": 5.061728395061729e-06,
+      "loss": 0.4737,
+      "step": 1228
+    },
+    {
+      "epoch": 2.7311111111111113,
+      "grad_norm": 1498860.25,
+      "learning_rate": 5.0205761316872425e-06,
+      "loss": 0.4934,
+      "step": 1229
+    },
+    {
+      "epoch": 2.7333333333333334,
+      "grad_norm": 145583.71875,
+      "learning_rate": 4.9794238683127575e-06,
+      "loss": 0.5605,
+      "step": 1230
+    },
+    {
+      "epoch": 2.7355555555555555,
+      "grad_norm": 740040.5,
+      "learning_rate": 4.938271604938272e-06,
+      "loss": 0.5821,
+      "step": 1231
+    },
+    {
+      "epoch": 2.7377777777777776,
+      "grad_norm": 2460071.25,
+      "learning_rate": 4.897119341563787e-06,
+      "loss": 0.7144,
+      "step": 1232
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 803059.0,
+      "learning_rate": 4.8559670781893e-06,
+      "loss": 0.6062,
+      "step": 1233
+    },
+    {
+      "epoch": 2.7422222222222223,
+      "grad_norm": 657412.25,
+      "learning_rate": 4.814814814814815e-06,
+      "loss": 0.4763,
+      "step": 1234
+    },
+    {
+      "epoch": 2.7444444444444445,
+      "grad_norm": 307322.96875,
+      "learning_rate": 4.77366255144033e-06,
+      "loss": 0.4573,
+      "step": 1235
+    },
+    {
+      "epoch": 2.7466666666666666,
+      "grad_norm": 1087119.5,
+      "learning_rate": 4.732510288065844e-06,
+      "loss": 0.5798,
+      "step": 1236
+    },
+    {
+      "epoch": 2.7488888888888887,
+      "grad_norm": 476222.65625,
+      "learning_rate": 4.691358024691358e-06,
+      "loss": 0.4826,
+      "step": 1237
+    },
+    {
+      "epoch": 2.7511111111111113,
+      "grad_norm": 769200.875,
+      "learning_rate": 4.6502057613168725e-06,
+      "loss": 0.477,
+      "step": 1238
+    },
+    {
+      "epoch": 2.7533333333333334,
+      "grad_norm": 604819.4375,
+      "learning_rate": 4.6090534979423875e-06,
+      "loss": 0.4938,
+      "step": 1239
+    },
+    {
+      "epoch": 2.7555555555555555,
+      "grad_norm": 1535238.25,
+      "learning_rate": 4.567901234567902e-06,
+      "loss": 0.3676,
+      "step": 1240
+    },
+    {
+      "epoch": 2.7577777777777777,
+      "grad_norm": 841712.125,
+      "learning_rate": 4.526748971193416e-06,
+      "loss": 0.5449,
+      "step": 1241
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 586374.5,
+      "learning_rate": 4.48559670781893e-06,
+      "loss": 0.4049,
+      "step": 1242
+    },
+    {
+      "epoch": 2.7622222222222224,
+      "grad_norm": 1455824.125,
+      "learning_rate": 4.444444444444445e-06,
+      "loss": 0.528,
+      "step": 1243
+    },
+    {
+      "epoch": 2.7644444444444445,
+      "grad_norm": 1251346.75,
+      "learning_rate": 4.403292181069959e-06,
+      "loss": 0.4563,
+      "step": 1244
+    },
+    {
+      "epoch": 2.7666666666666666,
+      "grad_norm": 253219.390625,
+      "learning_rate": 4.362139917695473e-06,
+      "loss": 0.4896,
+      "step": 1245
+    },
+    {
+      "epoch": 2.7688888888888887,
+      "grad_norm": 348807.59375,
+      "learning_rate": 4.3209876543209875e-06,
+      "loss": 0.4689,
+      "step": 1246
+    },
+    {
+      "epoch": 2.771111111111111,
+      "grad_norm": 355768.21875,
+      "learning_rate": 4.2798353909465025e-06,
+      "loss": 0.5001,
+      "step": 1247
+    },
+    {
+      "epoch": 2.7733333333333334,
+      "grad_norm": 222135.0,
+      "learning_rate": 4.238683127572017e-06,
+      "loss": 0.533,
+      "step": 1248
+    },
+    {
+      "epoch": 2.7755555555555556,
+      "grad_norm": 638996.125,
+      "learning_rate": 4.197530864197531e-06,
+      "loss": 0.4011,
+      "step": 1249
+    },
+    {
+      "epoch": 2.7777777777777777,
+      "grad_norm": 493758.21875,
+      "learning_rate": 4.156378600823045e-06,
+      "loss": 0.549,
+      "step": 1250
+    },
+    {
+      "epoch": 2.7800000000000002,
+      "grad_norm": 468891.9375,
+      "learning_rate": 4.11522633744856e-06,
+      "loss": 0.5394,
+      "step": 1251
+    },
+    {
+      "epoch": 2.7822222222222224,
+      "grad_norm": 793925.3125,
+      "learning_rate": 4.074074074074075e-06,
+      "loss": 0.7011,
+      "step": 1252
+    },
+    {
+      "epoch": 2.7844444444444445,
+      "grad_norm": 572089.5,
+      "learning_rate": 4.032921810699588e-06,
+      "loss": 0.4487,
+      "step": 1253
+    },
+    {
+      "epoch": 2.7866666666666666,
+      "grad_norm": 559628.75,
+      "learning_rate": 3.991769547325103e-06,
+      "loss": 0.5446,
+      "step": 1254
+    },
+    {
+      "epoch": 2.7888888888888888,
+      "grad_norm": 947452.4375,
+      "learning_rate": 3.9506172839506175e-06,
+      "loss": 0.6898,
+      "step": 1255
+    },
+    {
+      "epoch": 2.7911111111111113,
+      "grad_norm": 402679.625,
+      "learning_rate": 3.9094650205761325e-06,
+      "loss": 0.5524,
+      "step": 1256
+    },
+    {
+      "epoch": 2.7933333333333334,
+      "grad_norm": 299420.40625,
+      "learning_rate": 3.868312757201646e-06,
+      "loss": 0.5117,
+      "step": 1257
+    },
+    {
+      "epoch": 2.7955555555555556,
+      "grad_norm": 611257.6875,
+      "learning_rate": 3.827160493827161e-06,
+      "loss": 0.6882,
+      "step": 1258
+    },
+    {
+      "epoch": 2.7977777777777777,
+      "grad_norm": 574839.9375,
+      "learning_rate": 3.786008230452675e-06,
+      "loss": 0.5716,
+      "step": 1259
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 492008.53125,
+      "learning_rate": 3.7448559670781896e-06,
+      "loss": 0.6211,
+      "step": 1260
+    },
+    {
+      "epoch": 2.8022222222222224,
+      "grad_norm": 794452.0,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 0.5312,
+      "step": 1261
+    },
+    {
+      "epoch": 2.8044444444444445,
+      "grad_norm": 208639.828125,
+      "learning_rate": 3.6625514403292183e-06,
+      "loss": 0.5969,
+      "step": 1262
+    },
+    {
+      "epoch": 2.8066666666666666,
+      "grad_norm": 2034471.75,
+      "learning_rate": 3.6213991769547325e-06,
+      "loss": 0.7033,
+      "step": 1263
+    },
+    {
+      "epoch": 2.8088888888888888,
+      "grad_norm": 357638.625,
+      "learning_rate": 3.580246913580247e-06,
+      "loss": 0.5115,
+      "step": 1264
+    },
+    {
+      "epoch": 2.811111111111111,
+      "grad_norm": 398690.84375,
+      "learning_rate": 3.5390946502057612e-06,
+      "loss": 0.3857,
+      "step": 1265
+    },
+    {
+      "epoch": 2.8133333333333335,
+      "grad_norm": 376240.9375,
+      "learning_rate": 3.497942386831276e-06,
+      "loss": 0.3986,
+      "step": 1266
+    },
+    {
+      "epoch": 2.8155555555555556,
+      "grad_norm": 974289.6875,
+      "learning_rate": 3.45679012345679e-06,
+      "loss": 0.4911,
+      "step": 1267
+    },
+    {
+      "epoch": 2.8177777777777777,
+      "grad_norm": 711186.6875,
+      "learning_rate": 3.4156378600823046e-06,
+      "loss": 0.4323,
+      "step": 1268
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 765747.1875,
+      "learning_rate": 3.3744855967078196e-06,
+      "loss": 0.4529,
+      "step": 1269
+    },
+    {
+      "epoch": 2.822222222222222,
+      "grad_norm": 587382.0625,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.5034,
+      "step": 1270
+    },
+    {
+      "epoch": 2.8244444444444445,
+      "grad_norm": 284979.75,
+      "learning_rate": 3.2921810699588483e-06,
+      "loss": 0.4863,
+      "step": 1271
+    },
+    {
+      "epoch": 2.8266666666666667,
+      "grad_norm": 718397.25,
+      "learning_rate": 3.251028806584362e-06,
+      "loss": 0.5261,
+      "step": 1272
+    },
+    {
+      "epoch": 2.828888888888889,
+      "grad_norm": 367435.875,
+      "learning_rate": 3.209876543209877e-06,
+      "loss": 0.5775,
+      "step": 1273
+    },
+    {
+      "epoch": 2.8311111111111114,
+      "grad_norm": 804869.0,
+      "learning_rate": 3.168724279835391e-06,
+      "loss": 0.4268,
+      "step": 1274
+    },
+    {
+      "epoch": 2.8333333333333335,
+      "grad_norm": 893933.8125,
+      "learning_rate": 3.127572016460906e-06,
+      "loss": 0.4797,
+      "step": 1275
+    },
+    {
+      "epoch": 2.8355555555555556,
+      "grad_norm": 601643.5625,
+      "learning_rate": 3.0864197530864196e-06,
+      "loss": 0.64,
+      "step": 1276
+    },
+    {
+      "epoch": 2.8377777777777777,
+      "grad_norm": 812132.75,
+      "learning_rate": 3.0452674897119346e-06,
+      "loss": 0.4334,
+      "step": 1277
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 6934080.0,
+      "learning_rate": 3.0041152263374487e-06,
+      "loss": 0.5341,
+      "step": 1278
+    },
+    {
+      "epoch": 2.8422222222222224,
+      "grad_norm": 1309447.25,
+      "learning_rate": 2.9629629629629633e-06,
+      "loss": 0.491,
+      "step": 1279
+    },
+    {
+      "epoch": 2.8444444444444446,
+      "grad_norm": 298212.8125,
+      "learning_rate": 2.9218106995884775e-06,
+      "loss": 0.561,
+      "step": 1280
+    },
+    {
+      "epoch": 2.8466666666666667,
+      "grad_norm": 2192948.5,
+      "learning_rate": 2.880658436213992e-06,
+      "loss": 0.603,
+      "step": 1281
+    },
+    {
+      "epoch": 2.848888888888889,
+      "grad_norm": 673454.4375,
+      "learning_rate": 2.8395061728395062e-06,
+      "loss": 0.4945,
+      "step": 1282
+    },
+    {
+      "epoch": 2.851111111111111,
+      "grad_norm": 419024.84375,
+      "learning_rate": 2.798353909465021e-06,
+      "loss": 0.5274,
+      "step": 1283
+    },
+    {
+      "epoch": 2.8533333333333335,
+      "grad_norm": 1039461.1875,
+      "learning_rate": 2.757201646090535e-06,
+      "loss": 0.5511,
+      "step": 1284
+    },
+    {
+      "epoch": 2.8555555555555556,
+      "grad_norm": 670416.1875,
+      "learning_rate": 2.7160493827160496e-06,
+      "loss": 0.5403,
+      "step": 1285
+    },
+    {
+      "epoch": 2.8577777777777778,
+      "grad_norm": 835062.6875,
+      "learning_rate": 2.6748971193415637e-06,
+      "loss": 0.4858,
+      "step": 1286
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 959062.8125,
+      "learning_rate": 2.6337448559670783e-06,
+      "loss": 0.4005,
+      "step": 1287
+    },
+    {
+      "epoch": 2.862222222222222,
+      "grad_norm": 1145211.875,
+      "learning_rate": 2.5925925925925925e-06,
+      "loss": 0.4198,
+      "step": 1288
+    },
+    {
+      "epoch": 2.8644444444444446,
+      "grad_norm": 1677770.375,
+      "learning_rate": 2.551440329218107e-06,
+      "loss": 0.4583,
+      "step": 1289
+    },
+    {
+      "epoch": 2.8666666666666667,
+      "grad_norm": 1481635.875,
+      "learning_rate": 2.5102880658436212e-06,
+      "loss": 0.4812,
+      "step": 1290
+    },
+    {
+      "epoch": 2.868888888888889,
+      "grad_norm": 83220.25,
+      "learning_rate": 2.469135802469136e-06,
+      "loss": 0.4624,
+      "step": 1291
+    },
+    {
+      "epoch": 2.871111111111111,
+      "grad_norm": 414993.90625,
+      "learning_rate": 2.42798353909465e-06,
+      "loss": 0.3868,
+      "step": 1292
+    },
+    {
+      "epoch": 2.873333333333333,
+      "grad_norm": 627476.8125,
+      "learning_rate": 2.386831275720165e-06,
+      "loss": 0.4232,
+      "step": 1293
+    },
+    {
+      "epoch": 2.8755555555555556,
+      "grad_norm": 650311.125,
+      "learning_rate": 2.345679012345679e-06,
+      "loss": 0.5251,
+      "step": 1294
+    },
+    {
+      "epoch": 2.8777777777777778,
+      "grad_norm": 392837.4375,
+      "learning_rate": 2.3045267489711937e-06,
+      "loss": 0.4596,
+      "step": 1295
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1472083.375,
+      "learning_rate": 2.263374485596708e-06,
+      "loss": 0.5652,
+      "step": 1296
+    },
+    {
+      "epoch": 2.8822222222222225,
+      "grad_norm": 197789.484375,
+      "learning_rate": 2.2222222222222225e-06,
+      "loss": 0.5791,
+      "step": 1297
+    },
+    {
+      "epoch": 2.8844444444444446,
+      "grad_norm": 416251.53125,
+      "learning_rate": 2.1810699588477367e-06,
+      "loss": 0.4691,
+      "step": 1298
+    },
+    {
+      "epoch": 2.8866666666666667,
+      "grad_norm": 332912.5,
+      "learning_rate": 2.1399176954732512e-06,
+      "loss": 0.4653,
+      "step": 1299
+    },
+    {
+      "epoch": 2.888888888888889,
+      "grad_norm": 1015003.5,
+      "learning_rate": 2.0987654320987654e-06,
+      "loss": 0.5523,
+      "step": 1300
+    },
+    {
+      "epoch": 2.891111111111111,
+      "grad_norm": 560585.5625,
+      "learning_rate": 2.05761316872428e-06,
+      "loss": 0.6112,
+      "step": 1301
+    },
+    {
+      "epoch": 2.8933333333333335,
+      "grad_norm": 1809974.5,
+      "learning_rate": 2.016460905349794e-06,
+      "loss": 0.5622,
+      "step": 1302
+    },
+    {
+      "epoch": 2.8955555555555557,
+      "grad_norm": 449177.0625,
+      "learning_rate": 1.9753086419753087e-06,
+      "loss": 0.5725,
+      "step": 1303
+    },
+    {
+      "epoch": 2.897777777777778,
+      "grad_norm": 317168.25,
+      "learning_rate": 1.934156378600823e-06,
+      "loss": 0.5962,
+      "step": 1304
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 456991.1875,
+      "learning_rate": 1.8930041152263375e-06,
+      "loss": 0.563,
+      "step": 1305
+    },
+    {
+      "epoch": 2.902222222222222,
+      "grad_norm": 301806.40625,
+      "learning_rate": 1.8518518518518519e-06,
+      "loss": 0.5332,
+      "step": 1306
+    },
+    {
+      "epoch": 2.9044444444444446,
+      "grad_norm": 334963.5,
+      "learning_rate": 1.8106995884773662e-06,
+      "loss": 0.5427,
+      "step": 1307
+    },
+    {
+      "epoch": 2.9066666666666667,
+      "grad_norm": 1178584.125,
+      "learning_rate": 1.7695473251028806e-06,
+      "loss": 0.4654,
+      "step": 1308
+    },
+    {
+      "epoch": 2.908888888888889,
+      "grad_norm": 538667.8125,
+      "learning_rate": 1.728395061728395e-06,
+      "loss": 0.5941,
+      "step": 1309
+    },
+    {
+      "epoch": 2.911111111111111,
+      "grad_norm": 251747.875,
+      "learning_rate": 1.6872427983539098e-06,
+      "loss": 0.6021,
+      "step": 1310
+    },
+    {
+      "epoch": 2.913333333333333,
+      "grad_norm": 1805239.0,
+      "learning_rate": 1.6460905349794242e-06,
+      "loss": 0.5335,
+      "step": 1311
+    },
+    {
+      "epoch": 2.9155555555555557,
+      "grad_norm": 142086.625,
+      "learning_rate": 1.6049382716049385e-06,
+      "loss": 0.4195,
+      "step": 1312
+    },
+    {
+      "epoch": 2.917777777777778,
+      "grad_norm": 805839.0,
+      "learning_rate": 1.563786008230453e-06,
+      "loss": 0.4475,
+      "step": 1313
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 714949.875,
+      "learning_rate": 1.5226337448559673e-06,
+      "loss": 0.5176,
+      "step": 1314
+    },
+    {
+      "epoch": 2.9222222222222225,
+      "grad_norm": 359683.90625,
+      "learning_rate": 1.4814814814814817e-06,
+      "loss": 0.4859,
+      "step": 1315
+    },
+    {
+      "epoch": 2.924444444444444,
+      "grad_norm": 775748.875,
+      "learning_rate": 1.440329218106996e-06,
+      "loss": 0.6062,
+      "step": 1316
+    },
+    {
+      "epoch": 2.9266666666666667,
+      "grad_norm": 826315.4375,
+      "learning_rate": 1.3991769547325104e-06,
+      "loss": 0.6138,
+      "step": 1317
+    },
+    {
+      "epoch": 2.928888888888889,
+      "grad_norm": 377858.5625,
+      "learning_rate": 1.3580246913580248e-06,
+      "loss": 0.5518,
+      "step": 1318
+    },
+    {
+      "epoch": 2.931111111111111,
+      "grad_norm": 3151661.0,
+      "learning_rate": 1.3168724279835392e-06,
+      "loss": 0.5037,
+      "step": 1319
+    },
+    {
+      "epoch": 2.9333333333333336,
+      "grad_norm": 1737743.75,
+      "learning_rate": 1.2757201646090535e-06,
+      "loss": 0.5886,
+      "step": 1320
+    },
+    {
+      "epoch": 2.9355555555555557,
+      "grad_norm": 281431.375,
+      "learning_rate": 1.234567901234568e-06,
+      "loss": 0.4492,
+      "step": 1321
+    },
+    {
+      "epoch": 2.937777777777778,
+      "grad_norm": 572928.0,
+      "learning_rate": 1.1934156378600825e-06,
+      "loss": 0.7053,
+      "step": 1322
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 2890382.0,
+      "learning_rate": 1.1522633744855969e-06,
+      "loss": 0.7321,
+      "step": 1323
+    },
+    {
+      "epoch": 2.942222222222222,
+      "grad_norm": 365583.78125,
+      "learning_rate": 1.1111111111111112e-06,
+      "loss": 0.5227,
+      "step": 1324
+    },
+    {
+      "epoch": 2.9444444444444446,
+      "grad_norm": 586235.625,
+      "learning_rate": 1.0699588477366256e-06,
+      "loss": 0.4605,
+      "step": 1325
+    },
+    {
+      "epoch": 2.9466666666666668,
+      "grad_norm": 525312.6875,
+      "learning_rate": 1.02880658436214e-06,
+      "loss": 0.5955,
+      "step": 1326
+    },
+    {
+      "epoch": 2.948888888888889,
+      "grad_norm": 1755663.5,
+      "learning_rate": 9.876543209876544e-07,
+      "loss": 0.6344,
+      "step": 1327
+    },
+    {
+      "epoch": 2.951111111111111,
+      "grad_norm": 235478.671875,
+      "learning_rate": 9.465020576131687e-07,
+      "loss": 0.617,
+      "step": 1328
+    },
+    {
+      "epoch": 2.953333333333333,
+      "grad_norm": 1451552.75,
+      "learning_rate": 9.053497942386831e-07,
+      "loss": 0.5483,
+      "step": 1329
+    },
+    {
+      "epoch": 2.9555555555555557,
+      "grad_norm": 696397.1875,
+      "learning_rate": 8.641975308641975e-07,
+      "loss": 0.4778,
+      "step": 1330
+    },
+    {
+      "epoch": 2.957777777777778,
+      "grad_norm": 415077.21875,
+      "learning_rate": 8.230452674897121e-07,
+      "loss": 0.632,
+      "step": 1331
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 345612.71875,
+      "learning_rate": 7.818930041152265e-07,
+      "loss": 0.4817,
+      "step": 1332
+    },
+    {
+      "epoch": 2.962222222222222,
+      "grad_norm": 1004227.3125,
+      "learning_rate": 7.407407407407408e-07,
+      "loss": 0.4939,
+      "step": 1333
+    },
+    {
+      "epoch": 2.964444444444444,
+      "grad_norm": 647075.375,
+      "learning_rate": 6.995884773662552e-07,
+      "loss": 0.5492,
+      "step": 1334
+    },
+    {
+      "epoch": 2.966666666666667,
+      "grad_norm": 877828.8125,
+      "learning_rate": 6.584362139917696e-07,
+      "loss": 0.6042,
+      "step": 1335
+    },
+    {
+      "epoch": 2.968888888888889,
+      "grad_norm": 938837.3125,
+      "learning_rate": 6.17283950617284e-07,
+      "loss": 0.4177,
+      "step": 1336
+    },
+    {
+      "epoch": 2.971111111111111,
+      "grad_norm": 793827.875,
+      "learning_rate": 5.761316872427984e-07,
+      "loss": 0.3873,
+      "step": 1337
+    },
+    {
+      "epoch": 2.9733333333333336,
+      "grad_norm": 594568.5,
+      "learning_rate": 5.349794238683128e-07,
+      "loss": 0.4783,
+      "step": 1338
+    },
+    {
+      "epoch": 2.9755555555555553,
+      "grad_norm": 496043.03125,
+      "learning_rate": 4.938271604938272e-07,
+      "loss": 0.6735,
+      "step": 1339
+    },
+    {
+      "epoch": 2.977777777777778,
+      "grad_norm": 662431.375,
+      "learning_rate": 4.5267489711934156e-07,
+      "loss": 0.5808,
+      "step": 1340
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 308026.96875,
+      "learning_rate": 4.1152263374485604e-07,
+      "loss": 0.5529,
+      "step": 1341
+    },
+    {
+      "epoch": 2.982222222222222,
+      "grad_norm": 1318756.75,
+      "learning_rate": 3.703703703703704e-07,
+      "loss": 0.4589,
+      "step": 1342
+    },
+    {
+      "epoch": 2.9844444444444447,
+      "grad_norm": 401215.625,
+      "learning_rate": 3.292181069958848e-07,
+      "loss": 0.4334,
+      "step": 1343
+    },
+    {
+      "epoch": 2.986666666666667,
+      "grad_norm": 398647.75,
+      "learning_rate": 2.880658436213992e-07,
+      "loss": 0.4161,
+      "step": 1344
+    },
+    {
+      "epoch": 2.988888888888889,
+      "grad_norm": 701269.9375,
+      "learning_rate": 2.469135802469136e-07,
+      "loss": 0.5187,
+      "step": 1345
+    },
+    {
+      "epoch": 2.991111111111111,
+      "grad_norm": 816582.25,
+      "learning_rate": 2.0576131687242802e-07,
+      "loss": 0.7391,
+      "step": 1346
+    },
+    {
+      "epoch": 2.993333333333333,
+      "grad_norm": 566105.1875,
+      "learning_rate": 1.646090534979424e-07,
+      "loss": 0.5438,
+      "step": 1347
+    },
+    {
+      "epoch": 2.9955555555555557,
+      "grad_norm": 345409.1875,
+      "learning_rate": 1.234567901234568e-07,
+      "loss": 0.5929,
+      "step": 1348
+    },
+    {
+      "epoch": 2.997777777777778,
+      "grad_norm": 576835.75,
+      "learning_rate": 8.23045267489712e-08,
+      "loss": 0.3647,
+      "step": 1349
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 242567.28125,
+      "learning_rate": 4.11522633744856e-08,
+      "loss": 0.465,
+      "step": 1350
+    },
+    {
+      "epoch": 3.0,
+      "step": 1350,
+      "total_flos": 5.087834939588608e+17,
+      "train_loss": 0.5604817596409056,
+      "train_runtime": 51124.8319,
+      "train_samples_per_second": 0.422,
+      "train_steps_per_second": 0.026
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1350,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.087834939588608e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}