diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5549 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 788,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0012698412698412698,
+      "grad_norm": 11.4375,
+      "learning_rate": 7e-08,
+      "loss": 2.3377,
+      "step": 1
+    },
+    {
+      "epoch": 0.0025396825396825397,
+      "grad_norm": 12.4375,
+      "learning_rate": 1.4e-07,
+      "loss": 2.4221,
+      "step": 2
+    },
+    {
+      "epoch": 0.0038095238095238095,
+      "grad_norm": 10.375,
+      "learning_rate": 2.1e-07,
+      "loss": 2.1995,
+      "step": 3
+    },
+    {
+      "epoch": 0.005079365079365079,
+      "grad_norm": 11.25,
+      "learning_rate": 2.8e-07,
+      "loss": 2.3242,
+      "step": 4
+    },
+    {
+      "epoch": 0.006349206349206349,
+      "grad_norm": 11.0625,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 2.3848,
+      "step": 5
+    },
+    {
+      "epoch": 0.007619047619047619,
+      "grad_norm": 10.875,
+      "learning_rate": 4.2e-07,
+      "loss": 2.2233,
+      "step": 6
+    },
+    {
+      "epoch": 0.008888888888888889,
+      "grad_norm": 11.125,
+      "learning_rate": 4.900000000000001e-07,
+      "loss": 2.3011,
+      "step": 7
+    },
+    {
+      "epoch": 0.010158730158730159,
+      "grad_norm": 12.25,
+      "learning_rate": 5.6e-07,
+      "loss": 2.4318,
+      "step": 8
+    },
+    {
+      "epoch": 0.011428571428571429,
+      "grad_norm": 10.8125,
+      "learning_rate": 6.3e-07,
+      "loss": 2.2978,
+      "step": 9
+    },
+    {
+      "epoch": 0.012698412698412698,
+      "grad_norm": 11.3125,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 2.4126,
+      "step": 10
+    },
+    {
+      "epoch": 0.013968253968253968,
+      "grad_norm": 10.75,
+      "learning_rate": 7.7e-07,
+      "loss": 2.2333,
+      "step": 11
+    },
+    {
+      "epoch": 0.015238095238095238,
+      "grad_norm": 11.3125,
+      "learning_rate": 8.4e-07,
+      "loss": 2.3186,
+      "step": 12
+    },
+    {
+      "epoch": 0.01650793650793651,
+      "grad_norm": 11.0,
+      "learning_rate": 9.1e-07,
+      "loss": 2.3917,
+      "step": 13
+    },
+    {
+      "epoch": 0.017777777777777778,
+      "grad_norm": 10.8125,
+      "learning_rate": 9.800000000000001e-07,
+      "loss": 2.3092,
+      "step": 14
+    },
+    {
+      "epoch": 0.01904761904761905,
+      "grad_norm": 10.25,
+      "learning_rate": 1.05e-06,
+      "loss": 2.1844,
+      "step": 15
+    },
+    {
+      "epoch": 0.020317460317460317,
+      "grad_norm": 9.625,
+      "learning_rate": 1.12e-06,
+      "loss": 2.2563,
+      "step": 16
+    },
+    {
+      "epoch": 0.02158730158730159,
+      "grad_norm": 10.3125,
+      "learning_rate": 1.19e-06,
+      "loss": 2.3583,
+      "step": 17
+    },
+    {
+      "epoch": 0.022857142857142857,
+      "grad_norm": 10.1875,
+      "learning_rate": 1.26e-06,
+      "loss": 2.2967,
+      "step": 18
+    },
+    {
+      "epoch": 0.02412698412698413,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.33e-06,
+      "loss": 2.2498,
+      "step": 19
+    },
+    {
+      "epoch": 0.025396825396825397,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 2.2613,
+      "step": 20
+    },
+    {
+      "epoch": 0.02666666666666667,
+      "grad_norm": 9.875,
+      "learning_rate": 1.47e-06,
+      "loss": 2.2819,
+      "step": 21
+    },
+    {
+      "epoch": 0.027936507936507936,
+      "grad_norm": 10.3125,
+      "learning_rate": 1.54e-06,
+      "loss": 2.2904,
+      "step": 22
+    },
+    {
+      "epoch": 0.029206349206349208,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.61e-06,
+      "loss": 2.1933,
+      "step": 23
+    },
+    {
+      "epoch": 0.030476190476190476,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.68e-06,
+      "loss": 2.225,
+      "step": 24
+    },
+    {
+      "epoch": 0.031746031746031744,
+      "grad_norm": 8.75,
+      "learning_rate": 1.75e-06,
+      "loss": 2.2458,
+      "step": 25
+    },
+    {
+      "epoch": 0.03301587301587302,
+      "grad_norm": 8.5,
+      "learning_rate": 1.82e-06,
+      "loss": 2.1693,
+      "step": 26
+    },
+    {
+      "epoch": 0.03428571428571429,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.8900000000000001e-06,
+      "loss": 2.2753,
+      "step": 27
+    },
+    {
+      "epoch": 0.035555555555555556,
+      "grad_norm": 9.5,
+      "learning_rate": 1.9600000000000003e-06,
+      "loss": 2.3356,
+      "step": 28
+    },
+    {
+      "epoch": 0.036825396825396824,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.0299999999999996e-06,
+      "loss": 2.2706,
+      "step": 29
+    },
+    {
+      "epoch": 0.0380952380952381,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.1e-06,
+      "loss": 2.1023,
+      "step": 30
+    },
+    {
+      "epoch": 0.03936507936507937,
+      "grad_norm": 8.375,
+      "learning_rate": 2.17e-06,
+      "loss": 2.2069,
+      "step": 31
+    },
+    {
+      "epoch": 0.040634920634920635,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.24e-06,
+      "loss": 2.1079,
+      "step": 32
+    },
+    {
+      "epoch": 0.0419047619047619,
+      "grad_norm": 7.375,
+      "learning_rate": 2.31e-06,
+      "loss": 2.2311,
+      "step": 33
+    },
+    {
+      "epoch": 0.04317460317460318,
+      "grad_norm": 6.28125,
+      "learning_rate": 2.38e-06,
+      "loss": 2.1627,
+      "step": 34
+    },
+    {
+      "epoch": 0.044444444444444446,
+      "grad_norm": 6.125,
+      "learning_rate": 2.45e-06,
+      "loss": 2.1843,
+      "step": 35
+    },
+    {
+      "epoch": 0.045714285714285714,
+      "grad_norm": 6.0,
+      "learning_rate": 2.52e-06,
+      "loss": 2.1101,
+      "step": 36
+    },
+    {
+      "epoch": 0.04698412698412698,
+      "grad_norm": 5.375,
+      "learning_rate": 2.5899999999999998e-06,
+      "loss": 2.0643,
+      "step": 37
+    },
+    {
+      "epoch": 0.04825396825396826,
+      "grad_norm": 5.0625,
+      "learning_rate": 2.66e-06,
+      "loss": 2.0844,
+      "step": 38
+    },
+    {
+      "epoch": 0.049523809523809526,
+      "grad_norm": 4.71875,
+      "learning_rate": 2.73e-06,
+      "loss": 2.0624,
+      "step": 39
+    },
+    {
+      "epoch": 0.050793650793650794,
+      "grad_norm": 4.6875,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 2.1232,
+      "step": 40
+    },
+    {
+      "epoch": 0.05206349206349206,
+      "grad_norm": 4.71875,
+      "learning_rate": 2.8699999999999996e-06,
+      "loss": 2.124,
+      "step": 41
+    },
+    {
+      "epoch": 0.05333333333333334,
+      "grad_norm": 4.53125,
+      "learning_rate": 2.94e-06,
+      "loss": 1.9933,
+      "step": 42
+    },
+    {
+      "epoch": 0.054603174603174605,
+      "grad_norm": 4.65625,
+      "learning_rate": 3.01e-06,
+      "loss": 2.0451,
+      "step": 43
+    },
+    {
+      "epoch": 0.05587301587301587,
+      "grad_norm": 4.65625,
+      "learning_rate": 3.08e-06,
+      "loss": 2.1581,
+      "step": 44
+    },
+    {
+      "epoch": 0.05714285714285714,
+      "grad_norm": 4.65625,
+      "learning_rate": 3.15e-06,
+      "loss": 2.1967,
+      "step": 45
+    },
+    {
+      "epoch": 0.058412698412698416,
+      "grad_norm": 4.3125,
+      "learning_rate": 3.22e-06,
+      "loss": 2.0406,
+      "step": 46
+    },
+    {
+      "epoch": 0.059682539682539684,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.29e-06,
+      "loss": 1.9816,
+      "step": 47
+    },
+    {
+      "epoch": 0.06095238095238095,
+      "grad_norm": 4.0625,
+      "learning_rate": 3.36e-06,
+      "loss": 1.9638,
+      "step": 48
+    },
+    {
+      "epoch": 0.06222222222222222,
+      "grad_norm": 3.796875,
+      "learning_rate": 3.4299999999999998e-06,
+      "loss": 1.9013,
+      "step": 49
+    },
+    {
+      "epoch": 0.06349206349206349,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.5e-06,
+      "loss": 2.0392,
+      "step": 50
+    },
+    {
+      "epoch": 0.06476190476190476,
+      "grad_norm": 3.28125,
+      "learning_rate": 3.57e-06,
+      "loss": 1.8965,
+      "step": 51
+    },
+    {
+      "epoch": 0.06603174603174604,
+      "grad_norm": 3.328125,
+      "learning_rate": 3.64e-06,
+      "loss": 1.8598,
+      "step": 52
+    },
+    {
+      "epoch": 0.0673015873015873,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.71e-06,
+      "loss": 1.9478,
+      "step": 53
+    },
+    {
+      "epoch": 0.06857142857142857,
+      "grad_norm": 3.09375,
+      "learning_rate": 3.7800000000000002e-06,
+      "loss": 1.8397,
+      "step": 54
+    },
+    {
+      "epoch": 0.06984126984126984,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.85e-06,
+      "loss": 1.8759,
+      "step": 55
+    },
+    {
+      "epoch": 0.07111111111111111,
+      "grad_norm": 3.078125,
+      "learning_rate": 3.920000000000001e-06,
+      "loss": 1.9694,
+      "step": 56
+    },
+    {
+      "epoch": 0.07238095238095238,
+      "grad_norm": 2.875,
+      "learning_rate": 3.99e-06,
+      "loss": 1.8433,
+      "step": 57
+    },
+    {
+      "epoch": 0.07365079365079365,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.059999999999999e-06,
+      "loss": 1.9218,
+      "step": 58
+    },
+    {
+      "epoch": 0.07492063492063492,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.129999999999999e-06,
+      "loss": 1.8318,
+      "step": 59
+    },
+    {
+      "epoch": 0.0761904761904762,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.2e-06,
+      "loss": 1.898,
+      "step": 60
+    },
+    {
+      "epoch": 0.07746031746031747,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.27e-06,
+      "loss": 1.8301,
+      "step": 61
+    },
+    {
+      "epoch": 0.07873015873015873,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.34e-06,
+      "loss": 1.8442,
+      "step": 62
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.41e-06,
+      "loss": 1.8299,
+      "step": 63
+    },
+    {
+      "epoch": 0.08126984126984127,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.48e-06,
+      "loss": 1.8178,
+      "step": 64
+    },
+    {
+      "epoch": 0.08253968253968254,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 1.7625,
+      "step": 65
+    },
+    {
+      "epoch": 0.0838095238095238,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.62e-06,
+      "loss": 1.7792,
+      "step": 66
+    },
+    {
+      "epoch": 0.08507936507936507,
+      "grad_norm": 2.75,
+      "learning_rate": 4.69e-06,
+      "loss": 1.9265,
+      "step": 67
+    },
+    {
+      "epoch": 0.08634920634920636,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.76e-06,
+      "loss": 1.8957,
+      "step": 68
+    },
+    {
+      "epoch": 0.08761904761904762,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8299999999999995e-06,
+      "loss": 1.7738,
+      "step": 69
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 2.625,
+      "learning_rate": 4.9e-06,
+      "loss": 1.7845,
+      "step": 70
+    },
+    {
+      "epoch": 0.09015873015873016,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.97e-06,
+      "loss": 1.6498,
+      "step": 71
+    },
+    {
+      "epoch": 0.09142857142857143,
+      "grad_norm": 2.609375,
+      "learning_rate": 5.04e-06,
+      "loss": 1.7736,
+      "step": 72
+    },
+    {
+      "epoch": 0.0926984126984127,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.11e-06,
+      "loss": 1.7067,
+      "step": 73
+    },
+    {
+      "epoch": 0.09396825396825396,
+      "grad_norm": 2.75,
+      "learning_rate": 5.1799999999999995e-06,
+      "loss": 1.8757,
+      "step": 74
+    },
+    {
+      "epoch": 0.09523809523809523,
+      "grad_norm": 2.53125,
+      "learning_rate": 5.25e-06,
+      "loss": 1.7789,
+      "step": 75
+    },
+    {
+      "epoch": 0.09650793650793651,
+      "grad_norm": 2.5,
+      "learning_rate": 5.32e-06,
+      "loss": 1.7109,
+      "step": 76
+    },
+    {
+      "epoch": 0.09777777777777778,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.39e-06,
+      "loss": 1.8135,
+      "step": 77
+    },
+    {
+      "epoch": 0.09904761904761905,
+      "grad_norm": 2.71875,
+      "learning_rate": 5.46e-06,
+      "loss": 1.8264,
+      "step": 78
+    },
+    {
+      "epoch": 0.10031746031746032,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.53e-06,
+      "loss": 1.822,
+      "step": 79
+    },
+    {
+      "epoch": 0.10158730158730159,
+      "grad_norm": 2.734375,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.7601,
+      "step": 80
+    },
+    {
+      "epoch": 0.10285714285714286,
+      "grad_norm": 2.609375,
+      "learning_rate": 5.67e-06,
+      "loss": 1.8076,
+      "step": 81
+    },
+    {
+      "epoch": 0.10412698412698412,
+      "grad_norm": 2.828125,
+      "learning_rate": 5.739999999999999e-06,
+      "loss": 1.7772,
+      "step": 82
+    },
+    {
+      "epoch": 0.10539682539682539,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.8099999999999994e-06,
+      "loss": 1.8925,
+      "step": 83
+    },
+    {
+      "epoch": 0.10666666666666667,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.88e-06,
+      "loss": 1.7806,
+      "step": 84
+    },
+    {
+      "epoch": 0.10793650793650794,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.95e-06,
+      "loss": 1.6426,
+      "step": 85
+    },
+    {
+      "epoch": 0.10920634920634921,
+      "grad_norm": 2.65625,
+      "learning_rate": 6.02e-06,
+      "loss": 1.7752,
+      "step": 86
+    },
+    {
+      "epoch": 0.11047619047619048,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.09e-06,
+      "loss": 1.7094,
+      "step": 87
+    },
+    {
+      "epoch": 0.11174603174603175,
+      "grad_norm": 2.5,
+      "learning_rate": 6.16e-06,
+      "loss": 1.671,
+      "step": 88
+    },
+    {
+      "epoch": 0.11301587301587301,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.23e-06,
+      "loss": 1.6576,
+      "step": 89
+    },
+    {
+      "epoch": 0.11428571428571428,
+      "grad_norm": 2.65625,
+      "learning_rate": 6.3e-06,
+      "loss": 1.6634,
+      "step": 90
+    },
+    {
+      "epoch": 0.11555555555555555,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.37e-06,
+      "loss": 1.6374,
+      "step": 91
+    },
+    {
+      "epoch": 0.11682539682539683,
+      "grad_norm": 2.5625,
+      "learning_rate": 6.44e-06,
+      "loss": 1.6832,
+      "step": 92
+    },
+    {
+      "epoch": 0.1180952380952381,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.51e-06,
+      "loss": 1.6738,
+      "step": 93
+    },
+    {
+      "epoch": 0.11936507936507937,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.58e-06,
+      "loss": 1.6874,
+      "step": 94
+    },
+    {
+      "epoch": 0.12063492063492064,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.65e-06,
+      "loss": 1.6468,
+      "step": 95
+    },
+    {
+      "epoch": 0.1219047619047619,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.72e-06,
+      "loss": 1.6261,
+      "step": 96
+    },
+    {
+      "epoch": 0.12317460317460317,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.789999999999999e-06,
+      "loss": 1.5797,
+      "step": 97
+    },
+    {
+      "epoch": 0.12444444444444444,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.8599999999999995e-06,
+      "loss": 1.6279,
+      "step": 98
+    },
+    {
+      "epoch": 0.12571428571428572,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.93e-06,
+      "loss": 1.6165,
+      "step": 99
+    },
+    {
+      "epoch": 0.12698412698412698,
+      "grad_norm": 2.4375,
+      "learning_rate": 7e-06,
+      "loss": 1.7051,
+      "step": 100
+    },
+    {
+      "epoch": 0.12825396825396826,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.999998140878257e-06,
+      "loss": 1.6242,
+      "step": 101
+    },
+    {
+      "epoch": 0.1295238095238095,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.999992563515e-06,
+      "loss": 1.591,
+      "step": 102
+    },
+    {
+      "epoch": 0.1307936507936508,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.999983267916156e-06,
+      "loss": 1.6395,
+      "step": 103
+    },
+    {
+      "epoch": 0.13206349206349208,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.9999702540916e-06,
+      "loss": 1.6801,
+      "step": 104
+    },
+    {
+      "epoch": 0.13333333333333333,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.999953522055158e-06,
+      "loss": 1.6198,
+      "step": 105
+    },
+    {
+      "epoch": 0.1346031746031746,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.999933071824603e-06,
+      "loss": 1.6125,
+      "step": 106
+    },
+    {
+      "epoch": 0.13587301587301587,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.9999089034216635e-06,
+      "loss": 1.6474,
+      "step": 107
+    },
+    {
+      "epoch": 0.13714285714285715,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.999881016872011e-06,
+      "loss": 1.5838,
+      "step": 108
+    },
+    {
+      "epoch": 0.1384126984126984,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.9998494122052754e-06,
+      "loss": 1.5673,
+      "step": 109
+    },
+    {
+      "epoch": 0.13968253968253969,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.9998140894550295e-06,
+      "loss": 1.5612,
+      "step": 110
+    },
+    {
+      "epoch": 0.14095238095238094,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.999775048658799e-06,
+      "loss": 1.5632,
+      "step": 111
+    },
+    {
+      "epoch": 0.14222222222222222,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.999732289858059e-06,
+      "loss": 1.6273,
+      "step": 112
+    },
+    {
+      "epoch": 0.1434920634920635,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.999685813098235e-06,
+      "loss": 1.5568,
+      "step": 113
+    },
+    {
+      "epoch": 0.14476190476190476,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.999635618428701e-06,
+      "loss": 1.5606,
+      "step": 114
+    },
+    {
+      "epoch": 0.14603174603174604,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.999581705902782e-06,
+      "loss": 1.5716,
+      "step": 115
+    },
+    {
+      "epoch": 0.1473015873015873,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.999524075577753e-06,
+      "loss": 1.5832,
+      "step": 116
+    },
+    {
+      "epoch": 0.14857142857142858,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.9994627275148364e-06,
+      "loss": 1.6203,
+      "step": 117
+    },
+    {
+      "epoch": 0.14984126984126983,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.999397661779208e-06,
+      "loss": 1.5982,
+      "step": 118
+    },
+    {
+      "epoch": 0.1511111111111111,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.999328878439989e-06,
+      "loss": 1.4827,
+      "step": 119
+    },
+    {
+      "epoch": 0.1523809523809524,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.99925637757025e-06,
+      "loss": 1.5523,
+      "step": 120
+    },
+    {
+      "epoch": 0.15365079365079365,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.9991801592470155e-06,
+      "loss": 1.4812,
+      "step": 121
+    },
+    {
+      "epoch": 0.15492063492063493,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.999100223551257e-06,
+      "loss": 1.5039,
+      "step": 122
+    },
+    {
+      "epoch": 0.15619047619047619,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.9990165705678915e-06,
+      "loss": 1.5991,
+      "step": 123
+    },
+    {
+      "epoch": 0.15746031746031747,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.9989292003857905e-06,
+      "loss": 1.5819,
+      "step": 124
+    },
+    {
+      "epoch": 0.15873015873015872,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.998838113097772e-06,
+      "loss": 1.5072,
+      "step": 125
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.998743308800602e-06,
+      "loss": 1.539,
+      "step": 126
+    },
+    {
+      "epoch": 0.16126984126984126,
+      "grad_norm": 2.25,
+      "learning_rate": 6.998644787594997e-06,
+      "loss": 1.6346,
+      "step": 127
+    },
+    {
+      "epoch": 0.16253968253968254,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.998542549585622e-06,
+      "loss": 1.4874,
+      "step": 128
+    },
+    {
+      "epoch": 0.16380952380952382,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.998436594881089e-06,
+      "loss": 1.4606,
+      "step": 129
+    },
+    {
+      "epoch": 0.16507936507936508,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.99832692359396e-06,
+      "loss": 1.4556,
+      "step": 130
+    },
+    {
+      "epoch": 0.16634920634920636,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.998213535840745e-06,
+      "loss": 1.5251,
+      "step": 131
+    },
+    {
+      "epoch": 0.1676190476190476,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.998096431741903e-06,
+      "loss": 1.6041,
+      "step": 132
+    },
+    {
+      "epoch": 0.1688888888888889,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.997975611421838e-06,
+      "loss": 1.516,
+      "step": 133
+    },
+    {
+      "epoch": 0.17015873015873015,
+      "grad_norm": 1.890625,
+      "learning_rate": 6.997851075008906e-06,
+      "loss": 1.401,
+      "step": 134
+    },
+    {
+      "epoch": 0.17142857142857143,
+      "grad_norm": 1.984375,
+      "learning_rate": 6.997722822635408e-06,
+      "loss": 1.3963,
+      "step": 135
+    },
+    {
+      "epoch": 0.1726984126984127,
+      "grad_norm": 1.78125,
+      "learning_rate": 6.997590854437593e-06,
+      "loss": 1.4488,
+      "step": 136
+    },
+    {
+      "epoch": 0.17396825396825397,
+      "grad_norm": 1.9375,
+      "learning_rate": 6.9974551705556605e-06,
+      "loss": 1.4161,
+      "step": 137
+    },
+    {
+      "epoch": 0.17523809523809525,
+      "grad_norm": 1.8359375,
+      "learning_rate": 6.997315771133752e-06,
+      "loss": 1.4458,
+      "step": 138
+    },
+    {
+      "epoch": 0.1765079365079365,
+      "grad_norm": 1.921875,
+      "learning_rate": 6.997172656319962e-06,
+      "loss": 1.5626,
+      "step": 139
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 1.8515625,
+      "learning_rate": 6.997025826266326e-06,
+      "loss": 1.4079,
+      "step": 140
+    },
+    {
+      "epoch": 0.17904761904761904,
+      "grad_norm": 1.890625,
+      "learning_rate": 6.996875281128833e-06,
+      "loss": 1.5113,
+      "step": 141
+    },
+    {
+      "epoch": 0.18031746031746032,
+      "grad_norm": 1.765625,
+      "learning_rate": 6.996721021067415e-06,
+      "loss": 1.4129,
+      "step": 142
+    },
+    {
+      "epoch": 0.18158730158730158,
+      "grad_norm": 1.7734375,
+      "learning_rate": 6.996563046245947e-06,
+      "loss": 1.416,
+      "step": 143
+    },
+    {
+      "epoch": 0.18285714285714286,
+      "grad_norm": 1.890625,
+      "learning_rate": 6.99640135683226e-06,
+      "loss": 1.4688,
+      "step": 144
+    },
+    {
+      "epoch": 0.18412698412698414,
+      "grad_norm": 1.7890625,
+      "learning_rate": 6.9962359529981225e-06,
+      "loss": 1.4409,
+      "step": 145
+    },
+    {
+      "epoch": 0.1853968253968254,
+      "grad_norm": 1.796875,
+      "learning_rate": 6.996066834919252e-06,
+      "loss": 1.4301,
+      "step": 146
+    },
+    {
+      "epoch": 0.18666666666666668,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.995894002775314e-06,
+      "loss": 1.3702,
+      "step": 147
+    },
+    {
+      "epoch": 0.18793650793650793,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.995717456749914e-06,
+      "loss": 1.5495,
+      "step": 148
+    },
+    {
+      "epoch": 0.1892063492063492,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.99553719703061e-06,
+      "loss": 1.4909,
+      "step": 149
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.9953532238089014e-06,
+      "loss": 1.4744,
+      "step": 150
+    },
+    {
+      "epoch": 0.19174603174603175,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.995165537280231e-06,
+      "loss": 1.3965,
+      "step": 151
+    },
+    {
+      "epoch": 0.19301587301587303,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.994974137643991e-06,
+      "loss": 1.4505,
+      "step": 152
+    },
+    {
+      "epoch": 0.19428571428571428,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.994779025103515e-06,
+      "loss": 1.4028,
+      "step": 153
+    },
+    {
+      "epoch": 0.19555555555555557,
+      "grad_norm": 1.625,
+      "learning_rate": 6.994580199866081e-06,
+      "loss": 1.4503,
+      "step": 154
+    },
+    {
+      "epoch": 0.19682539682539682,
+      "grad_norm": 1.5234375,
+      "learning_rate": 6.994377662142914e-06,
+      "loss": 1.3517,
+      "step": 155
+    },
+    {
+      "epoch": 0.1980952380952381,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.9941714121491785e-06,
+      "loss": 1.4896,
+      "step": 156
+    },
+    {
+      "epoch": 0.19936507936507936,
+      "grad_norm": 1.5,
+      "learning_rate": 6.993961450103987e-06,
+      "loss": 1.4458,
+      "step": 157
+    },
+    {
+      "epoch": 0.20063492063492064,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.993747776230393e-06,
+      "loss": 1.4939,
+      "step": 158
+    },
+    {
+      "epoch": 0.2019047619047619,
+      "grad_norm": 1.4609375,
+      "learning_rate": 6.9935303907553945e-06,
+      "loss": 1.3731,
+      "step": 159
+    },
+    {
+      "epoch": 0.20317460317460317,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.993309293909931e-06,
+      "loss": 1.4482,
+      "step": 160
+    },
+    {
+      "epoch": 0.20444444444444446,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.993084485928888e-06,
+      "loss": 1.4789,
+      "step": 161
+    },
+    {
+      "epoch": 0.2057142857142857,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.992855967051091e-06,
+      "loss": 1.3584,
+      "step": 162
+    },
+    {
+      "epoch": 0.206984126984127,
+      "grad_norm": 1.484375,
+      "learning_rate": 6.9926237375193055e-06,
+      "loss": 1.3995,
+      "step": 163
+    },
+    {
+      "epoch": 0.20825396825396825,
+      "grad_norm": 1.421875,
+      "learning_rate": 6.992387797580246e-06,
+      "loss": 1.4282,
+      "step": 164
+    },
+    {
+      "epoch": 0.20952380952380953,
+      "grad_norm": 1.359375,
+      "learning_rate": 6.992148147484561e-06,
+      "loss": 1.4075,
+      "step": 165
+    },
+    {
+      "epoch": 0.21079365079365078,
+      "grad_norm": 1.4609375,
+      "learning_rate": 6.991904787486846e-06,
+      "loss": 1.4205,
+      "step": 166
+    },
+    {
+      "epoch": 0.21206349206349207,
+      "grad_norm": 1.4453125,
+      "learning_rate": 6.991657717845635e-06,
+      "loss": 1.365,
+      "step": 167
+    },
+    {
+      "epoch": 0.21333333333333335,
+      "grad_norm": 1.3984375,
+      "learning_rate": 6.991406938823403e-06,
+      "loss": 1.4606,
+      "step": 168
+    },
+    {
+      "epoch": 0.2146031746031746,
+      "grad_norm": 1.3671875,
+      "learning_rate": 6.991152450686569e-06,
+      "loss": 1.3828,
+      "step": 169
+    },
+    {
+      "epoch": 0.21587301587301588,
+      "grad_norm": 1.265625,
+      "learning_rate": 6.9908942537054875e-06,
+      "loss": 1.4357,
+      "step": 170
+    },
+    {
+      "epoch": 0.21714285714285714,
+      "grad_norm": 1.328125,
+      "learning_rate": 6.990632348154456e-06,
+      "loss": 1.3723,
+      "step": 171
+    },
+    {
+      "epoch": 0.21841269841269842,
+      "grad_norm": 1.4140625,
+      "learning_rate": 6.990366734311711e-06,
+      "loss": 1.4549,
+      "step": 172
+    },
+    {
+      "epoch": 0.21968253968253967,
+      "grad_norm": 1.2734375,
+      "learning_rate": 6.9900974124594295e-06,
+      "loss": 1.3897,
+      "step": 173
+    },
+    {
+      "epoch": 0.22095238095238096,
+      "grad_norm": 1.4375,
+      "learning_rate": 6.9898243828837265e-06,
+      "loss": 1.381,
+      "step": 174
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 1.3515625,
+      "learning_rate": 6.989547645874657e-06,
+      "loss": 1.3959,
+      "step": 175
+    },
+    {
+      "epoch": 0.2234920634920635,
+      "grad_norm": 1.3984375,
+      "learning_rate": 6.989267201726213e-06,
+      "loss": 1.3529,
+      "step": 176
+    },
+    {
+      "epoch": 0.22476190476190477,
+      "grad_norm": 1.3828125,
+      "learning_rate": 6.988983050736326e-06,
+      "loss": 1.3437,
+      "step": 177
+    },
+    {
+      "epoch": 0.22603174603174603,
+      "grad_norm": 1.4453125,
+      "learning_rate": 6.988695193206866e-06,
+      "loss": 1.3084,
+      "step": 178
+    },
+    {
+      "epoch": 0.2273015873015873,
+      "grad_norm": 1.390625,
+      "learning_rate": 6.98840362944364e-06,
+      "loss": 1.2895,
+      "step": 179
+    },
+    {
+      "epoch": 0.22857142857142856,
+      "grad_norm": 1.34375,
+      "learning_rate": 6.9881083597563915e-06,
+      "loss": 1.3205,
+      "step": 180
+    },
+    {
+      "epoch": 0.22984126984126985,
+      "grad_norm": 1.4140625,
+      "learning_rate": 6.987809384458803e-06,
+      "loss": 1.3976,
+      "step": 181
+    },
+    {
+      "epoch": 0.2311111111111111,
+      "grad_norm": 1.328125,
+      "learning_rate": 6.987506703868491e-06,
+      "loss": 1.3126,
+      "step": 182
+    },
+    {
+      "epoch": 0.23238095238095238,
+      "grad_norm": 1.359375,
+      "learning_rate": 6.987200318307011e-06,
+      "loss": 1.4032,
+      "step": 183
+    },
+    {
+      "epoch": 0.23365079365079366,
+      "grad_norm": 1.2890625,
+      "learning_rate": 6.986890228099852e-06,
+      "loss": 1.3045,
+      "step": 184
+    },
+    {
+      "epoch": 0.23492063492063492,
+      "grad_norm": 1.421875,
+      "learning_rate": 6.986576433576441e-06,
+      "loss": 1.3976,
+      "step": 185
+    },
+    {
+      "epoch": 0.2361904761904762,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.9862589350701396e-06,
+      "loss": 1.3839,
+      "step": 186
+    },
+    {
+      "epoch": 0.23746031746031745,
+      "grad_norm": 1.40625,
+      "learning_rate": 6.985937732918243e-06,
+      "loss": 1.3107,
+      "step": 187
+    },
+    {
+      "epoch": 0.23873015873015874,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.985612827461983e-06,
+      "loss": 1.4188,
+      "step": 188
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3359375,
+      "learning_rate": 6.9852842190465244e-06,
+      "loss": 1.2768,
+      "step": 189
+    },
+    {
+      "epoch": 0.24126984126984127,
+      "grad_norm": 1.359375,
+      "learning_rate": 6.984951908020966e-06,
+      "loss": 1.2394,
+      "step": 190
+    },
+    {
+      "epoch": 0.24253968253968253,
+      "grad_norm": 1.4375,
+      "learning_rate": 6.984615894738339e-06,
+      "loss": 1.3525,
+      "step": 191
+    },
+    {
+      "epoch": 0.2438095238095238,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.984276179555611e-06,
+      "loss": 1.3494,
+      "step": 192
+    },
+    {
+      "epoch": 0.2450793650793651,
+      "grad_norm": 1.3671875,
+      "learning_rate": 6.983932762833678e-06,
+      "loss": 1.3324,
+      "step": 193
+    },
+    {
+      "epoch": 0.24634920634920635,
+      "grad_norm": 1.421875,
+      "learning_rate": 6.983585644937373e-06,
+      "loss": 1.2826,
+      "step": 194
+    },
+    {
+      "epoch": 0.24761904761904763,
+      "grad_norm": 1.375,
+      "learning_rate": 6.983234826235456e-06,
+      "loss": 1.3029,
+      "step": 195
+    },
+    {
+      "epoch": 0.24888888888888888,
+      "grad_norm": 1.4453125,
+      "learning_rate": 6.982880307100624e-06,
+      "loss": 1.3598,
+      "step": 196
+    },
+    {
+      "epoch": 0.25015873015873014,
+      "grad_norm": 1.40625,
+      "learning_rate": 6.982522087909498e-06,
+      "loss": 1.3622,
+      "step": 197
+    },
+    {
+      "epoch": 0.25142857142857145,
+      "grad_norm": 1.296875,
+      "learning_rate": 6.9821601690426384e-06,
+      "loss": 1.2689,
+      "step": 198
+    },
+    {
+      "epoch": 0.2526984126984127,
+      "grad_norm": 1.4453125,
+      "learning_rate": 6.981794550884529e-06,
+      "loss": 1.3869,
+      "step": 199
+    },
+    {
+      "epoch": 0.25396825396825395,
+      "grad_norm": 1.4765625,
+      "learning_rate": 6.981425233823588e-06,
+      "loss": 1.3919,
+      "step": 200
+    },
+    {
+      "epoch": 0.25523809523809526,
+      "grad_norm": 1.3515625,
+      "learning_rate": 6.98105221825216e-06,
+      "loss": 1.3206,
+      "step": 201
+    },
+    {
+      "epoch": 0.2565079365079365,
+      "grad_norm": 1.3671875,
+      "learning_rate": 6.98067550456652e-06,
+      "loss": 1.386,
+      "step": 202
+    },
+    {
+      "epoch": 0.2577777777777778,
+      "grad_norm": 1.421875,
+      "learning_rate": 6.980295093166873e-06,
+      "loss": 1.409,
+      "step": 203
+    },
+    {
+      "epoch": 0.259047619047619,
+      "grad_norm": 1.375,
+      "learning_rate": 6.97991098445735e-06,
+      "loss": 1.3044,
+      "step": 204
+    },
+    {
+      "epoch": 0.26031746031746034,
+      "grad_norm": 1.3984375,
+      "learning_rate": 6.979523178846011e-06,
+      "loss": 1.29,
+      "step": 205
+    },
+    {
+      "epoch": 0.2615873015873016,
+      "grad_norm": 1.3359375,
+      "learning_rate": 6.979131676744844e-06,
+      "loss": 1.4239,
+      "step": 206
+    },
+    {
+      "epoch": 0.26285714285714284,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.978736478569762e-06,
+      "loss": 1.4212,
+      "step": 207
+    },
+    {
+      "epoch": 0.26412698412698415,
+      "grad_norm": 1.5,
+      "learning_rate": 6.978337584740607e-06,
+      "loss": 1.3177,
+      "step": 208
+    },
+    {
+      "epoch": 0.2653968253968254,
+      "grad_norm": 1.3203125,
+      "learning_rate": 6.977934995681146e-06,
+      "loss": 1.3258,
+      "step": 209
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 1.390625,
+      "learning_rate": 6.977528711819072e-06,
+      "loss": 1.4282,
+      "step": 210
+    },
+    {
+      "epoch": 0.2679365079365079,
+      "grad_norm": 1.4453125,
+      "learning_rate": 6.977118733586e-06,
+      "loss": 1.4156,
+      "step": 211
+    },
+    {
+      "epoch": 0.2692063492063492,
+      "grad_norm": 1.4296875,
+      "learning_rate": 6.976705061417477e-06,
+      "loss": 1.4307,
+      "step": 212
+    },
+    {
+      "epoch": 0.2704761904761905,
+      "grad_norm": 1.3671875,
+      "learning_rate": 6.976287695752965e-06,
+      "loss": 1.3837,
+      "step": 213
+    },
+    {
+      "epoch": 0.27174603174603174,
+      "grad_norm": 1.453125,
+      "learning_rate": 6.975866637035859e-06,
+      "loss": 1.3319,
+      "step": 214
+    },
+    {
+      "epoch": 0.273015873015873,
+      "grad_norm": 1.4375,
+      "learning_rate": 6.975441885713471e-06,
+      "loss": 1.3346,
+      "step": 215
+    },
+    {
+      "epoch": 0.2742857142857143,
+      "grad_norm": 1.4296875,
+      "learning_rate": 6.975013442237037e-06,
+      "loss": 1.2907,
+      "step": 216
+    },
+    {
+      "epoch": 0.27555555555555555,
+      "grad_norm": 1.4296875,
+      "learning_rate": 6.974581307061718e-06,
+      "loss": 1.3801,
+      "step": 217
+    },
+    {
+      "epoch": 0.2768253968253968,
+      "grad_norm": 1.4765625,
+      "learning_rate": 6.974145480646593e-06,
+      "loss": 1.3904,
+      "step": 218
+    },
+    {
+      "epoch": 0.2780952380952381,
+      "grad_norm": 1.421875,
+      "learning_rate": 6.973705963454666e-06,
+      "loss": 1.3587,
+      "step": 219
+    },
+    {
+      "epoch": 0.27936507936507937,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.973262755952861e-06,
+      "loss": 1.3843,
+      "step": 220
+    },
+    {
+      "epoch": 0.2806349206349206,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.9728158586120195e-06,
+      "loss": 1.3899,
+      "step": 221
+    },
+    {
+      "epoch": 0.2819047619047619,
+      "grad_norm": 1.4296875,
+      "learning_rate": 6.9723652719069074e-06,
+      "loss": 1.3172,
+      "step": 222
+    },
+    {
+      "epoch": 0.2831746031746032,
+      "grad_norm": 1.3828125,
+      "learning_rate": 6.971910996316207e-06,
+      "loss": 1.3525,
+      "step": 223
+    },
+    {
+      "epoch": 0.28444444444444444,
+      "grad_norm": 1.3828125,
+      "learning_rate": 6.97145303232252e-06,
+      "loss": 1.2543,
+      "step": 224
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.970991380412367e-06,
+      "loss": 1.3654,
+      "step": 225
+    },
+    {
+      "epoch": 0.286984126984127,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.9705260410761876e-06,
+      "loss": 1.3518,
+      "step": 226
+    },
+    {
+      "epoch": 0.28825396825396826,
+      "grad_norm": 1.3984375,
+      "learning_rate": 6.970057014808337e-06,
+      "loss": 1.406,
+      "step": 227
+    },
+    {
+      "epoch": 0.2895238095238095,
+      "grad_norm": 1.3828125,
+      "learning_rate": 6.9695843021070855e-06,
+      "loss": 1.2749,
+      "step": 228
+    },
+    {
+      "epoch": 0.29079365079365077,
+      "grad_norm": 1.4609375,
+      "learning_rate": 6.969107903474625e-06,
+      "loss": 1.3623,
+      "step": 229
+    },
+    {
+      "epoch": 0.2920634920634921,
+      "grad_norm": 1.375,
+      "learning_rate": 6.9686278194170586e-06,
+      "loss": 1.3651,
+      "step": 230
+    },
+    {
+      "epoch": 0.29333333333333333,
+      "grad_norm": 1.40625,
+      "learning_rate": 6.968144050444407e-06,
+      "loss": 1.3215,
+      "step": 231
+    },
+    {
+      "epoch": 0.2946031746031746,
+      "grad_norm": 1.484375,
+      "learning_rate": 6.967656597070603e-06,
+      "loss": 1.3236,
+      "step": 232
+    },
+    {
+      "epoch": 0.2958730158730159,
+      "grad_norm": 1.4140625,
+      "learning_rate": 6.9671654598134965e-06,
+      "loss": 1.2878,
+      "step": 233
+    },
+    {
+      "epoch": 0.29714285714285715,
+      "grad_norm": 1.40625,
+      "learning_rate": 6.96667063919485e-06,
+      "loss": 1.257,
+      "step": 234
+    },
+    {
+      "epoch": 0.2984126984126984,
+      "grad_norm": 1.4296875,
+      "learning_rate": 6.966172135740339e-06,
+      "loss": 1.3465,
+      "step": 235
+    },
+    {
+      "epoch": 0.29968253968253966,
+      "grad_norm": 1.484375,
+      "learning_rate": 6.96566994997955e-06,
+      "loss": 1.3257,
+      "step": 236
+    },
+    {
+      "epoch": 0.30095238095238097,
+      "grad_norm": 1.421875,
+      "learning_rate": 6.965164082445983e-06,
+      "loss": 1.367,
+      "step": 237
+    },
+    {
+      "epoch": 0.3022222222222222,
+      "grad_norm": 1.4609375,
+      "learning_rate": 6.96465453367705e-06,
+      "loss": 1.3181,
+      "step": 238
+    },
+    {
+      "epoch": 0.3034920634920635,
+      "grad_norm": 1.3671875,
+      "learning_rate": 6.964141304214072e-06,
+      "loss": 1.3274,
+      "step": 239
+    },
+    {
+      "epoch": 0.3047619047619048,
+      "grad_norm": 1.328125,
+      "learning_rate": 6.963624394602281e-06,
+      "loss": 1.2907,
+      "step": 240
+    },
+    {
+      "epoch": 0.30603174603174604,
+      "grad_norm": 1.4375,
+      "learning_rate": 6.963103805390821e-06,
+      "loss": 1.3028,
+      "step": 241
+    },
+    {
+      "epoch": 0.3073015873015873,
+      "grad_norm": 1.3984375,
+      "learning_rate": 6.9625795371327375e-06,
+      "loss": 1.3416,
+      "step": 242
+    },
+    {
+      "epoch": 0.30857142857142855,
+      "grad_norm": 1.4140625,
+      "learning_rate": 6.962051590384995e-06,
+      "loss": 1.3399,
+      "step": 243
+    },
+    {
+      "epoch": 0.30984126984126986,
+      "grad_norm": 1.421875,
+      "learning_rate": 6.961519965708457e-06,
+      "loss": 1.2766,
+      "step": 244
+    },
+    {
+      "epoch": 0.3111111111111111,
+      "grad_norm": 1.265625,
+      "learning_rate": 6.9609846636679e-06,
+      "loss": 1.1575,
+      "step": 245
+    },
+    {
+      "epoch": 0.31238095238095237,
+      "grad_norm": 1.34375,
+      "learning_rate": 6.960445684832004e-06,
+      "loss": 1.276,
+      "step": 246
+    },
+    {
+      "epoch": 0.3136507936507936,
+      "grad_norm": 1.3984375,
+      "learning_rate": 6.959903029773356e-06,
+      "loss": 1.2703,
+      "step": 247
+    },
+    {
+      "epoch": 0.31492063492063493,
+      "grad_norm": 1.4375,
+      "learning_rate": 6.9593566990684474e-06,
+      "loss": 1.3219,
+      "step": 248
+    },
+    {
+      "epoch": 0.3161904761904762,
+      "grad_norm": 1.4296875,
+      "learning_rate": 6.9588066932976785e-06,
+      "loss": 1.314,
+      "step": 249
+    },
+    {
+      "epoch": 0.31746031746031744,
+      "grad_norm": 1.390625,
+      "learning_rate": 6.958253013045348e-06,
+      "loss": 1.2881,
+      "step": 250
+    },
+    {
+      "epoch": 0.31873015873015875,
+      "grad_norm": 1.3984375,
+      "learning_rate": 6.957695658899663e-06,
+      "loss": 1.336,
+      "step": 251
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.95713463145273e-06,
+      "loss": 1.3237,
+      "step": 252
+    },
+    {
+      "epoch": 0.32126984126984126,
+      "grad_norm": 1.3515625,
+      "learning_rate": 6.956569931300559e-06,
+      "loss": 1.2288,
+      "step": 253
+    },
+    {
+      "epoch": 0.3225396825396825,
+      "grad_norm": 1.375,
+      "learning_rate": 6.956001559043064e-06,
+      "loss": 1.3324,
+      "step": 254
+    },
+    {
+      "epoch": 0.3238095238095238,
+      "grad_norm": 1.40625,
+      "learning_rate": 6.955429515284058e-06,
+      "loss": 1.329,
+      "step": 255
+    },
+    {
+      "epoch": 0.3250793650793651,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.954853800631254e-06,
+      "loss": 1.2696,
+      "step": 256
+    },
+    {
+      "epoch": 0.32634920634920633,
+      "grad_norm": 1.3984375,
+      "learning_rate": 6.954274415696267e-06,
+      "loss": 1.3572,
+      "step": 257
+    },
+    {
+      "epoch": 0.32761904761904764,
+      "grad_norm": 1.34375,
+      "learning_rate": 6.953691361094606e-06,
+      "loss": 1.2582,
+      "step": 258
+    },
+    {
+      "epoch": 0.3288888888888889,
+      "grad_norm": 1.3828125,
+      "learning_rate": 6.953104637445686e-06,
+      "loss": 1.2598,
+      "step": 259
+    },
+    {
+      "epoch": 0.33015873015873015,
+      "grad_norm": 1.5390625,
+      "learning_rate": 6.952514245372815e-06,
+      "loss": 1.274,
+      "step": 260
+    },
+    {
+      "epoch": 0.3314285714285714,
+      "grad_norm": 1.390625,
+      "learning_rate": 6.951920185503199e-06,
+      "loss": 1.2684,
+      "step": 261
+    },
+    {
+      "epoch": 0.3326984126984127,
+      "grad_norm": 1.4453125,
+      "learning_rate": 6.951322458467938e-06,
+      "loss": 1.2921,
+      "step": 262
+    },
+    {
+      "epoch": 0.33396825396825397,
+      "grad_norm": 1.4609375,
+      "learning_rate": 6.950721064902034e-06,
+      "loss": 1.3409,
+      "step": 263
+    },
+    {
+      "epoch": 0.3352380952380952,
+      "grad_norm": 1.375,
+      "learning_rate": 6.95011600544438e-06,
+      "loss": 1.2827,
+      "step": 264
+    },
+    {
+      "epoch": 0.33650793650793653,
+      "grad_norm": 1.390625,
+      "learning_rate": 6.9495072807377634e-06,
+      "loss": 1.3019,
+      "step": 265
+    },
+    {
+      "epoch": 0.3377777777777778,
+      "grad_norm": 1.328125,
+      "learning_rate": 6.948894891428866e-06,
+      "loss": 1.2564,
+      "step": 266
+    },
+    {
+      "epoch": 0.33904761904761904,
+      "grad_norm": 1.28125,
+      "learning_rate": 6.948278838168263e-06,
+      "loss": 1.2545,
+      "step": 267
+    },
+    {
+      "epoch": 0.3403174603174603,
+      "grad_norm": 1.34375,
+      "learning_rate": 6.947659121610421e-06,
+      "loss": 1.29,
+      "step": 268
+    },
+    {
+      "epoch": 0.3415873015873016,
+      "grad_norm": 1.2890625,
+      "learning_rate": 6.947035742413701e-06,
+      "loss": 1.1939,
+      "step": 269
+    },
+    {
+      "epoch": 0.34285714285714286,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.9464087012403534e-06,
+      "loss": 1.2794,
+      "step": 270
+    },
+    {
+      "epoch": 0.3441269841269841,
+      "grad_norm": 1.359375,
+      "learning_rate": 6.945777998756516e-06,
+      "loss": 1.2046,
+      "step": 271
+    },
+    {
+      "epoch": 0.3453968253968254,
+      "grad_norm": 1.4140625,
+      "learning_rate": 6.94514363563222e-06,
+      "loss": 1.2897,
+      "step": 272
+    },
+    {
+      "epoch": 0.3466666666666667,
+      "grad_norm": 1.375,
+      "learning_rate": 6.944505612541386e-06,
+      "loss": 1.3137,
+      "step": 273
+    },
+    {
+      "epoch": 0.34793650793650793,
+      "grad_norm": 1.390625,
+      "learning_rate": 6.94386393016182e-06,
+      "loss": 1.2758,
+      "step": 274
+    },
+    {
+      "epoch": 0.3492063492063492,
+      "grad_norm": 1.34375,
+      "learning_rate": 6.943218589175216e-06,
+      "loss": 1.2078,
+      "step": 275
+    },
+    {
+      "epoch": 0.3504761904761905,
+      "grad_norm": 1.2890625,
+      "learning_rate": 6.942569590267157e-06,
+      "loss": 1.2597,
+      "step": 276
+    },
+    {
+      "epoch": 0.35174603174603175,
+      "grad_norm": 1.4140625,
+      "learning_rate": 6.9419169341271085e-06,
+      "loss": 1.2439,
+      "step": 277
+    },
+    {
+      "epoch": 0.353015873015873,
+      "grad_norm": 1.3359375,
+      "learning_rate": 6.9412606214484245e-06,
+      "loss": 1.2509,
+      "step": 278
+    },
+    {
+      "epoch": 0.35428571428571426,
+      "grad_norm": 1.296875,
+      "learning_rate": 6.9406006529283425e-06,
+      "loss": 1.2124,
+      "step": 279
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.939937029267983e-06,
+      "loss": 1.2922,
+      "step": 280
+    },
+    {
+      "epoch": 0.3568253968253968,
+      "grad_norm": 1.3671875,
+      "learning_rate": 6.93926975117235e-06,
+      "loss": 1.3179,
+      "step": 281
+    },
+    {
+      "epoch": 0.3580952380952381,
+      "grad_norm": 1.28125,
+      "learning_rate": 6.93859881935033e-06,
+      "loss": 1.2535,
+      "step": 282
+    },
+    {
+      "epoch": 0.3593650793650794,
+      "grad_norm": 1.3046875,
+      "learning_rate": 6.937924234514692e-06,
+      "loss": 1.2865,
+      "step": 283
+    },
+    {
+      "epoch": 0.36063492063492064,
+      "grad_norm": 1.28125,
+      "learning_rate": 6.9372459973820815e-06,
+      "loss": 1.2395,
+      "step": 284
+    },
+    {
+      "epoch": 0.3619047619047619,
+      "grad_norm": 1.4296875,
+      "learning_rate": 6.936564108673031e-06,
+      "loss": 1.2963,
+      "step": 285
+    },
+    {
+      "epoch": 0.36317460317460315,
+      "grad_norm": 1.3671875,
+      "learning_rate": 6.935878569111948e-06,
+      "loss": 1.2609,
+      "step": 286
+    },
+    {
+      "epoch": 0.36444444444444446,
+      "grad_norm": 1.25,
+      "learning_rate": 6.935189379427116e-06,
+      "loss": 1.1527,
+      "step": 287
+    },
+    {
+      "epoch": 0.3657142857142857,
+      "grad_norm": 1.34375,
+      "learning_rate": 6.934496540350704e-06,
+      "loss": 1.3306,
+      "step": 288
+    },
+    {
+      "epoch": 0.36698412698412697,
+      "grad_norm": 1.28125,
+      "learning_rate": 6.933800052618749e-06,
+      "loss": 1.1579,
+      "step": 289
+    },
+    {
+      "epoch": 0.3682539682539683,
+      "grad_norm": 1.3984375,
+      "learning_rate": 6.933099916971171e-06,
+      "loss": 1.2648,
+      "step": 290
+    },
+    {
+      "epoch": 0.36952380952380953,
+      "grad_norm": 1.375,
+      "learning_rate": 6.932396134151762e-06,
+      "loss": 1.2618,
+      "step": 291
+    },
+    {
+      "epoch": 0.3707936507936508,
+      "grad_norm": 1.265625,
+      "learning_rate": 6.9316887049081885e-06,
+      "loss": 1.2435,
+      "step": 292
+    },
+    {
+      "epoch": 0.37206349206349204,
+      "grad_norm": 1.3203125,
+      "learning_rate": 6.930977629991993e-06,
+      "loss": 1.196,
+      "step": 293
+    },
+    {
+      "epoch": 0.37333333333333335,
+      "grad_norm": 1.296875,
+      "learning_rate": 6.93026291015859e-06,
+      "loss": 1.1568,
+      "step": 294
+    },
+    {
+      "epoch": 0.3746031746031746,
+      "grad_norm": 1.234375,
+      "learning_rate": 6.929544546167265e-06,
+      "loss": 1.2868,
+      "step": 295
+    },
+    {
+      "epoch": 0.37587301587301586,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.928822538781175e-06,
+      "loss": 1.1996,
+      "step": 296
+    },
+    {
+      "epoch": 0.37714285714285717,
+      "grad_norm": 1.296875,
+      "learning_rate": 6.92809688876735e-06,
+      "loss": 1.2809,
+      "step": 297
+    },
+    {
+      "epoch": 0.3784126984126984,
+      "grad_norm": 1.34375,
+      "learning_rate": 6.9273675968966874e-06,
+      "loss": 1.204,
+      "step": 298
+    },
+    {
+      "epoch": 0.3796825396825397,
+      "grad_norm": 1.328125,
+      "learning_rate": 6.926634663943954e-06,
+      "loss": 1.1716,
+      "step": 299
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.925898090687786e-06,
+      "loss": 1.2552,
+      "step": 300
+    },
+    {
+      "epoch": 0.38222222222222224,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.9251578779106855e-06,
+      "loss": 1.179,
+      "step": 301
+    },
+    {
+      "epoch": 0.3834920634920635,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.9244140263990194e-06,
+      "loss": 1.1177,
+      "step": 302
+    },
+    {
+      "epoch": 0.38476190476190475,
+      "grad_norm": 1.2734375,
+      "learning_rate": 6.9236665369430255e-06,
+      "loss": 1.2488,
+      "step": 303
+    },
+    {
+      "epoch": 0.38603174603174606,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.9229154103368015e-06,
+      "loss": 1.208,
+      "step": 304
+    },
+    {
+      "epoch": 0.3873015873015873,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.92216064737831e-06,
+      "loss": 1.1659,
+      "step": 305
+    },
+    {
+      "epoch": 0.38857142857142857,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.9214022488693786e-06,
+      "loss": 1.2404,
+      "step": 306
+    },
+    {
+      "epoch": 0.3898412698412698,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.920640215615697e-06,
+      "loss": 1.2515,
+      "step": 307
+    },
+    {
+      "epoch": 0.39111111111111113,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.919874548426813e-06,
+      "loss": 1.2701,
+      "step": 308
+    },
+    {
+      "epoch": 0.3923809523809524,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.919105248116138e-06,
+      "loss": 1.253,
+      "step": 309
+    },
+    {
+      "epoch": 0.39365079365079364,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.918332315500942e-06,
+      "loss": 1.1332,
+      "step": 310
+    },
+    {
+      "epoch": 0.3949206349206349,
+      "grad_norm": 1.2578125,
+      "learning_rate": 6.917555751402356e-06,
+      "loss": 1.2004,
+      "step": 311
+    },
+    {
+      "epoch": 0.3961904761904762,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.916775556645364e-06,
+      "loss": 1.2051,
+      "step": 312
+    },
+    {
+      "epoch": 0.39746031746031746,
+      "grad_norm": 1.2890625,
+      "learning_rate": 6.915991732058812e-06,
+      "loss": 1.249,
+      "step": 313
+    },
+    {
+      "epoch": 0.3987301587301587,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.915204278475399e-06,
+      "loss": 1.2079,
+      "step": 314
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.25,
+      "learning_rate": 6.914413196731681e-06,
+      "loss": 1.2628,
+      "step": 315
+    },
+    {
+      "epoch": 0.4012698412698413,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.913618487668069e-06,
+      "loss": 1.3653,
+      "step": 316
+    },
+    {
+      "epoch": 0.40253968253968253,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.912820152128825e-06,
+      "loss": 1.1615,
+      "step": 317
+    },
+    {
+      "epoch": 0.4038095238095238,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.912018190962065e-06,
+      "loss": 1.1839,
+      "step": 318
+    },
+    {
+      "epoch": 0.4050793650793651,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.911212605019757e-06,
+      "loss": 1.1813,
+      "step": 319
+    },
+    {
+      "epoch": 0.40634920634920635,
+      "grad_norm": 1.234375,
+      "learning_rate": 6.910403395157719e-06,
+      "loss": 1.2426,
+      "step": 320
+    },
+    {
+      "epoch": 0.4076190476190476,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.909590562235621e-06,
+      "loss": 1.2174,
+      "step": 321
+    },
+    {
+      "epoch": 0.4088888888888889,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.908774107116979e-06,
+      "loss": 1.2622,
+      "step": 322
+    },
+    {
+      "epoch": 0.41015873015873017,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.907954030669158e-06,
+      "loss": 1.2048,
+      "step": 323
+    },
+    {
+      "epoch": 0.4114285714285714,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.907130333763371e-06,
+      "loss": 1.3177,
+      "step": 324
+    },
+    {
+      "epoch": 0.4126984126984127,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.906303017274677e-06,
+      "loss": 1.2328,
+      "step": 325
+    },
+    {
+      "epoch": 0.413968253968254,
+      "grad_norm": 1.2890625,
+      "learning_rate": 6.90547208208198e-06,
+      "loss": 1.2359,
+      "step": 326
+    },
+    {
+      "epoch": 0.41523809523809524,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.904637529068028e-06,
+      "loss": 1.1834,
+      "step": 327
+    },
+    {
+      "epoch": 0.4165079365079365,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.9037993591194145e-06,
+      "loss": 1.1702,
+      "step": 328
+    },
+    {
+      "epoch": 0.4177777777777778,
+      "grad_norm": 1.1484375,
+      "learning_rate": 6.902957573126571e-06,
+      "loss": 1.3178,
+      "step": 329
+    },
+    {
+      "epoch": 0.41904761904761906,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.902112171983775e-06,
+      "loss": 1.2117,
+      "step": 330
+    },
+    {
+      "epoch": 0.4203174603174603,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.901263156589144e-06,
+      "loss": 1.2687,
+      "step": 331
+    },
+    {
+      "epoch": 0.42158730158730157,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.90041052784463e-06,
+      "loss": 1.2933,
+      "step": 332
+    },
+    {
+      "epoch": 0.4228571428571429,
+      "grad_norm": 1.125,
+      "learning_rate": 6.899554286656032e-06,
+      "loss": 1.2205,
+      "step": 333
+    },
+    {
+      "epoch": 0.42412698412698413,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.89869443393298e-06,
+      "loss": 1.2044,
+      "step": 334
+    },
+    {
+      "epoch": 0.4253968253968254,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.897830970588943e-06,
+      "loss": 1.2005,
+      "step": 335
+    },
+    {
+      "epoch": 0.4266666666666667,
+      "grad_norm": 1.328125,
+      "learning_rate": 6.896963897541227e-06,
+      "loss": 1.2374,
+      "step": 336
+    },
+    {
+      "epoch": 0.42793650793650795,
+      "grad_norm": 1.3671875,
+      "learning_rate": 6.896093215710971e-06,
+      "loss": 1.2029,
+      "step": 337
+    },
+    {
+      "epoch": 0.4292063492063492,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.89521892602315e-06,
+      "loss": 1.1452,
+      "step": 338
+    },
+    {
+      "epoch": 0.43047619047619046,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.894341029406567e-06,
+      "loss": 1.2151,
+      "step": 339
+    },
+    {
+      "epoch": 0.43174603174603177,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.893459526793863e-06,
+      "loss": 1.225,
+      "step": 340
+    },
+    {
+      "epoch": 0.433015873015873,
+      "grad_norm": 1.125,
+      "learning_rate": 6.8925744191215055e-06,
+      "loss": 1.2256,
+      "step": 341
+    },
+    {
+      "epoch": 0.4342857142857143,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.8916857073297935e-06,
+      "loss": 1.1827,
+      "step": 342
+    },
+    {
+      "epoch": 0.43555555555555553,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.890793392362855e-06,
+      "loss": 1.2185,
+      "step": 343
+    },
+    {
+      "epoch": 0.43682539682539684,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.889897475168645e-06,
+      "loss": 1.1634,
+      "step": 344
+    },
+    {
+      "epoch": 0.4380952380952381,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.888997956698947e-06,
+      "loss": 1.2444,
+      "step": 345
+    },
+    {
+      "epoch": 0.43936507936507935,
+      "grad_norm": 1.0,
+      "learning_rate": 6.888094837909369e-06,
+      "loss": 1.1364,
+      "step": 346
+    },
+    {
+      "epoch": 0.44063492063492066,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.887188119759343e-06,
+      "loss": 1.2109,
+      "step": 347
+    },
+    {
+      "epoch": 0.4419047619047619,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.886277803212125e-06,
+      "loss": 1.2345,
+      "step": 348
+    },
+    {
+      "epoch": 0.44317460317460317,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.885363889234797e-06,
+      "loss": 1.1564,
+      "step": 349
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.884446378798258e-06,
+      "loss": 1.1894,
+      "step": 350
+    },
+    {
+      "epoch": 0.44571428571428573,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.8835252728772335e-06,
+      "loss": 1.194,
+      "step": 351
+    },
+    {
+      "epoch": 0.446984126984127,
+      "grad_norm": 1.234375,
+      "learning_rate": 6.882600572450261e-06,
+      "loss": 1.2596,
+      "step": 352
+    },
+    {
+      "epoch": 0.44825396825396824,
+      "grad_norm": 1.171875,
+      "learning_rate": 6.881672278499705e-06,
+      "loss": 1.1671,
+      "step": 353
+    },
+    {
+      "epoch": 0.44952380952380955,
+      "grad_norm": 0.9765625,
+      "learning_rate": 6.880740392011738e-06,
+      "loss": 1.0679,
+      "step": 354
+    },
+    {
+      "epoch": 0.4507936507936508,
+      "grad_norm": 1.2734375,
+      "learning_rate": 6.879804913976361e-06,
+      "loss": 1.2133,
+      "step": 355
+    },
+    {
+      "epoch": 0.45206349206349206,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.87886584538738e-06,
+      "loss": 1.2645,
+      "step": 356
+    },
+    {
+      "epoch": 0.4533333333333333,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.87792318724242e-06,
+      "loss": 1.2126,
+      "step": 357
+    },
+    {
+      "epoch": 0.4546031746031746,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.87697694054292e-06,
+      "loss": 1.2352,
+      "step": 358
+    },
+    {
+      "epoch": 0.4558730158730159,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.8760271062941286e-06,
+      "loss": 1.1688,
+      "step": 359
+    },
+    {
+      "epoch": 0.45714285714285713,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.875073685505108e-06,
+      "loss": 1.2012,
+      "step": 360
+    },
+    {
+      "epoch": 0.45841269841269844,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.874116679188731e-06,
+      "loss": 1.2021,
+      "step": 361
+    },
+    {
+      "epoch": 0.4596825396825397,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.873156088361677e-06,
+      "loss": 1.2262,
+      "step": 362
+    },
+    {
+      "epoch": 0.46095238095238095,
+      "grad_norm": 1.0,
+      "learning_rate": 6.872191914044435e-06,
+      "loss": 1.1552,
+      "step": 363
+    },
+    {
+      "epoch": 0.4622222222222222,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.8712241572613e-06,
+      "loss": 1.2148,
+      "step": 364
+    },
+    {
+      "epoch": 0.4634920634920635,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.870252819040374e-06,
+      "loss": 1.2414,
+      "step": 365
+    },
+    {
+      "epoch": 0.46476190476190476,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.869277900413564e-06,
+      "loss": 1.1179,
+      "step": 366
+    },
+    {
+      "epoch": 0.466031746031746,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.868299402416579e-06,
+      "loss": 1.0943,
+      "step": 367
+    },
+    {
+      "epoch": 0.46730158730158733,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.867317326088932e-06,
+      "loss": 1.1381,
+      "step": 368
+    },
+    {
+      "epoch": 0.4685714285714286,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.866331672473937e-06,
+      "loss": 1.174,
+      "step": 369
+    },
+    {
+      "epoch": 0.46984126984126984,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.865342442618709e-06,
+      "loss": 1.2216,
+      "step": 370
+    },
+    {
+      "epoch": 0.4711111111111111,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.86434963757416e-06,
+      "loss": 1.1455,
+      "step": 371
+    },
+    {
+      "epoch": 0.4723809523809524,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.863353258395003e-06,
+      "loss": 1.1913,
+      "step": 372
+    },
+    {
+      "epoch": 0.47365079365079366,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.8623533061397456e-06,
+      "loss": 1.2139,
+      "step": 373
+    },
+    {
+      "epoch": 0.4749206349206349,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.861349781870693e-06,
+      "loss": 1.1511,
+      "step": 374
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.8603426866539436e-06,
+      "loss": 1.15,
+      "step": 375
+    },
+    {
+      "epoch": 0.4774603174603175,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.859332021559393e-06,
+      "loss": 1.1681,
+      "step": 376
+    },
+    {
+      "epoch": 0.47873015873015873,
+      "grad_norm": 1.1484375,
+      "learning_rate": 6.858317787660723e-06,
+      "loss": 1.2131,
+      "step": 377
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.857299986035413e-06,
+      "loss": 1.1533,
+      "step": 378
+    },
+    {
+      "epoch": 0.4812698412698413,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.856278617764729e-06,
+      "loss": 1.1488,
+      "step": 379
+    },
+    {
+      "epoch": 0.48253968253968255,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.855253683933727e-06,
+      "loss": 1.12,
+      "step": 380
+    },
+    {
+      "epoch": 0.4838095238095238,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.85422518563125e-06,
+      "loss": 1.2726,
+      "step": 381
+    },
+    {
+      "epoch": 0.48507936507936505,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.85319312394993e-06,
+      "loss": 1.2218,
+      "step": 382
+    },
+    {
+      "epoch": 0.48634920634920636,
+      "grad_norm": 1.125,
+      "learning_rate": 6.852157499986183e-06,
+      "loss": 1.1984,
+      "step": 383
+    },
+    {
+      "epoch": 0.4876190476190476,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.85111831484021e-06,
+      "loss": 1.1689,
+      "step": 384
+    },
+    {
+      "epoch": 0.4888888888888889,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.8500755696159925e-06,
+      "loss": 1.1334,
+      "step": 385
+    },
+    {
+      "epoch": 0.4901587301587302,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.849029265421299e-06,
+      "loss": 1.0929,
+      "step": 386
+    },
+    {
+      "epoch": 0.49142857142857144,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.847979403367674e-06,
+      "loss": 1.2601,
+      "step": 387
+    },
+    {
+      "epoch": 0.4926984126984127,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.846925984570446e-06,
+      "loss": 1.2579,
+      "step": 388
+    },
+    {
+      "epoch": 0.49396825396825395,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.8458690101487195e-06,
+      "loss": 1.1605,
+      "step": 389
+    },
+    {
+      "epoch": 0.49523809523809526,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.844808481225377e-06,
+      "loss": 1.1947,
+      "step": 390
+    },
+    {
+      "epoch": 0.4965079365079365,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.8437443989270756e-06,
+      "loss": 1.0904,
+      "step": 391
+    },
+    {
+      "epoch": 0.49777777777777776,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.84267676438425e-06,
+      "loss": 1.1382,
+      "step": 392
+    },
+    {
+      "epoch": 0.4990476190476191,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.8416055787311076e-06,
+      "loss": 1.252,
+      "step": 393
+    },
+    {
+      "epoch": 0.5003174603174603,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.840530843105628e-06,
+      "loss": 1.1412,
+      "step": 394
+    },
+    {
+      "epoch": 0.5015873015873016,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.839452558649561e-06,
+      "loss": 1.1867,
+      "step": 395
+    },
+    {
+      "epoch": 0.5028571428571429,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.83837072650843e-06,
+      "loss": 1.1628,
+      "step": 396
+    },
+    {
+      "epoch": 0.5041269841269841,
+      "grad_norm": 0.9765625,
+      "learning_rate": 6.837285347831523e-06,
+      "loss": 1.1583,
+      "step": 397
+    },
+    {
+      "epoch": 0.5053968253968254,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.836196423771898e-06,
+      "loss": 1.2381,
+      "step": 398
+    },
+    {
+      "epoch": 0.5066666666666667,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.835103955486381e-06,
+      "loss": 1.2314,
+      "step": 399
+    },
+    {
+      "epoch": 0.5079365079365079,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.83400794413556e-06,
+      "loss": 1.2734,
+      "step": 400
+    },
+    {
+      "epoch": 0.5092063492063492,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.832908390883789e-06,
+      "loss": 1.2223,
+      "step": 401
+    },
+    {
+      "epoch": 0.5104761904761905,
+      "grad_norm": 1.0,
+      "learning_rate": 6.831805296899183e-06,
+      "loss": 1.1663,
+      "step": 402
+    },
+    {
+      "epoch": 0.5117460317460317,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.8306986633536214e-06,
+      "loss": 1.1652,
+      "step": 403
+    },
+    {
+      "epoch": 0.513015873015873,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.829588491422743e-06,
+      "loss": 1.121,
+      "step": 404
+    },
+    {
+      "epoch": 0.5142857142857142,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.8284747822859415e-06,
+      "loss": 1.2003,
+      "step": 405
+    },
+    {
+      "epoch": 0.5155555555555555,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.827357537126373e-06,
+      "loss": 1.2256,
+      "step": 406
+    },
+    {
+      "epoch": 0.5168253968253969,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.826236757130951e-06,
+      "loss": 1.1817,
+      "step": 407
+    },
+    {
+      "epoch": 0.518095238095238,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.8251124434903385e-06,
+      "loss": 1.1242,
+      "step": 408
+    },
+    {
+      "epoch": 0.5193650793650794,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.823984597398959e-06,
+      "loss": 1.1524,
+      "step": 409
+    },
+    {
+      "epoch": 0.5206349206349207,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.822853220054984e-06,
+      "loss": 1.2177,
+      "step": 410
+    },
+    {
+      "epoch": 0.5219047619047619,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.821718312660339e-06,
+      "loss": 1.3058,
+      "step": 411
+    },
+    {
+      "epoch": 0.5231746031746032,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.820579876420699e-06,
+      "loss": 1.1484,
+      "step": 412
+    },
+    {
+      "epoch": 0.5244444444444445,
+      "grad_norm": 1.0,
+      "learning_rate": 6.819437912545487e-06,
+      "loss": 1.1845,
+      "step": 413
+    },
+    {
+      "epoch": 0.5257142857142857,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.818292422247875e-06,
+      "loss": 1.1738,
+      "step": 414
+    },
+    {
+      "epoch": 0.526984126984127,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.8171434067447804e-06,
+      "loss": 1.2016,
+      "step": 415
+    },
+    {
+      "epoch": 0.5282539682539683,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.815990867256866e-06,
+      "loss": 1.1561,
+      "step": 416
+    },
+    {
+      "epoch": 0.5295238095238095,
+      "grad_norm": 0.9765625,
+      "learning_rate": 6.814834805008538e-06,
+      "loss": 1.1634,
+      "step": 417
+    },
+    {
+      "epoch": 0.5307936507936508,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.813675221227945e-06,
+      "loss": 1.1828,
+      "step": 418
+    },
+    {
+      "epoch": 0.532063492063492,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.812512117146979e-06,
+      "loss": 1.1899,
+      "step": 419
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.811345494001267e-06,
+      "loss": 1.2415,
+      "step": 420
+    },
+    {
+      "epoch": 0.5346031746031746,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.810175353030178e-06,
+      "loss": 1.2068,
+      "step": 421
+    },
+    {
+      "epoch": 0.5358730158730158,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.809001695476819e-06,
+      "loss": 1.2132,
+      "step": 422
+    },
+    {
+      "epoch": 0.5371428571428571,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.807824522588029e-06,
+      "loss": 1.235,
+      "step": 423
+    },
+    {
+      "epoch": 0.5384126984126985,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.806643835614384e-06,
+      "loss": 1.1771,
+      "step": 424
+    },
+    {
+      "epoch": 0.5396825396825397,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.805459635810196e-06,
+      "loss": 1.1837,
+      "step": 425
+    },
+    {
+      "epoch": 0.540952380952381,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.804271924433503e-06,
+      "loss": 1.2026,
+      "step": 426
+    },
+    {
+      "epoch": 0.5422222222222223,
+      "grad_norm": 1.0,
+      "learning_rate": 6.8030807027460774e-06,
+      "loss": 1.194,
+      "step": 427
+    },
+    {
+      "epoch": 0.5434920634920635,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.80188597201342e-06,
+      "loss": 1.2045,
+      "step": 428
+    },
+    {
+      "epoch": 0.5447619047619048,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.800687733504759e-06,
+      "loss": 1.1725,
+      "step": 429
+    },
+    {
+      "epoch": 0.546031746031746,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.79948598849305e-06,
+      "loss": 1.2031,
+      "step": 430
+    },
+    {
+      "epoch": 0.5473015873015873,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.798280738254972e-06,
+      "loss": 1.2332,
+      "step": 431
+    },
+    {
+      "epoch": 0.5485714285714286,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.797071984070929e-06,
+      "loss": 1.1543,
+      "step": 432
+    },
+    {
+      "epoch": 0.5498412698412698,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.7958597272250504e-06,
+      "loss": 1.2151,
+      "step": 433
+    },
+    {
+      "epoch": 0.5511111111111111,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.794643969005179e-06,
+      "loss": 1.236,
+      "step": 434
+    },
+    {
+      "epoch": 0.5523809523809524,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.793424710702885e-06,
+      "loss": 1.2502,
+      "step": 435
+    },
+    {
+      "epoch": 0.5536507936507936,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.792201953613454e-06,
+      "loss": 1.1059,
+      "step": 436
+    },
+    {
+      "epoch": 0.5549206349206349,
+      "grad_norm": 1.0,
+      "learning_rate": 6.790975699035886e-06,
+      "loss": 1.1807,
+      "step": 437
+    },
+    {
+      "epoch": 0.5561904761904762,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.789745948272902e-06,
+      "loss": 1.1415,
+      "step": 438
+    },
+    {
+      "epoch": 0.5574603174603174,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.788512702630932e-06,
+      "loss": 1.1708,
+      "step": 439
+    },
+    {
+      "epoch": 0.5587301587301587,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.787275963420123e-06,
+      "loss": 1.147,
+      "step": 440
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.786035731954329e-06,
+      "loss": 1.1925,
+      "step": 441
+    },
+    {
+      "epoch": 0.5612698412698413,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.7847920095511185e-06,
+      "loss": 1.1599,
+      "step": 442
+    },
+    {
+      "epoch": 0.5625396825396826,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.783544797531766e-06,
+      "loss": 1.1751,
+      "step": 443
+    },
+    {
+      "epoch": 0.5638095238095238,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.782294097221253e-06,
+      "loss": 1.2241,
+      "step": 444
+    },
+    {
+      "epoch": 0.5650793650793651,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.781039909948268e-06,
+      "loss": 1.2607,
+      "step": 445
+    },
+    {
+      "epoch": 0.5663492063492064,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.779782237045204e-06,
+      "loss": 1.1699,
+      "step": 446
+    },
+    {
+      "epoch": 0.5676190476190476,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.778521079848157e-06,
+      "loss": 1.2567,
+      "step": 447
+    },
+    {
+      "epoch": 0.5688888888888889,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.777256439696922e-06,
+      "loss": 1.168,
+      "step": 448
+    },
+    {
+      "epoch": 0.5701587301587302,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.775988317934998e-06,
+      "loss": 1.2496,
+      "step": 449
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.774716715909579e-06,
+      "loss": 1.2178,
+      "step": 450
+    },
+    {
+      "epoch": 0.5726984126984127,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.773441634971561e-06,
+      "loss": 1.1594,
+      "step": 451
+    },
+    {
+      "epoch": 0.573968253968254,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.772163076475531e-06,
+      "loss": 1.1178,
+      "step": 452
+    },
+    {
+      "epoch": 0.5752380952380952,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.770881041779771e-06,
+      "loss": 1.1934,
+      "step": 453
+    },
+    {
+      "epoch": 0.5765079365079365,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.769595532246259e-06,
+      "loss": 1.1575,
+      "step": 454
+    },
+    {
+      "epoch": 0.5777777777777777,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.768306549240664e-06,
+      "loss": 1.1751,
+      "step": 455
+    },
+    {
+      "epoch": 0.579047619047619,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.767014094132342e-06,
+      "loss": 1.2337,
+      "step": 456
+    },
+    {
+      "epoch": 0.5803174603174603,
+      "grad_norm": 0.859375,
+      "learning_rate": 6.7657181682943415e-06,
+      "loss": 1.1441,
+      "step": 457
+    },
+    {
+      "epoch": 0.5815873015873015,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.764418773103393e-06,
+      "loss": 1.2132,
+      "step": 458
+    },
+    {
+      "epoch": 0.5828571428571429,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.763115909939919e-06,
+      "loss": 1.171,
+      "step": 459
+    },
+    {
+      "epoch": 0.5841269841269842,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.761809580188021e-06,
+      "loss": 1.1907,
+      "step": 460
+    },
+    {
+      "epoch": 0.5853968253968254,
+      "grad_norm": 0.953125,
+      "learning_rate": 6.760499785235486e-06,
+      "loss": 1.055,
+      "step": 461
+    },
+    {
+      "epoch": 0.5866666666666667,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.759186526473782e-06,
+      "loss": 1.2006,
+      "step": 462
+    },
+    {
+      "epoch": 0.587936507936508,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.757869805298055e-06,
+      "loss": 1.1229,
+      "step": 463
+    },
+    {
+      "epoch": 0.5892063492063492,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.756549623107134e-06,
+      "loss": 1.1498,
+      "step": 464
+    },
+    {
+      "epoch": 0.5904761904761905,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.755225981303518e-06,
+      "loss": 1.1409,
+      "step": 465
+    },
+    {
+      "epoch": 0.5917460317460318,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.7538988812933854e-06,
+      "loss": 1.121,
+      "step": 466
+    },
+    {
+      "epoch": 0.593015873015873,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.752568324486589e-06,
+      "loss": 1.2086,
+      "step": 467
+    },
+    {
+      "epoch": 0.5942857142857143,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.751234312296654e-06,
+      "loss": 1.1454,
+      "step": 468
+    },
+    {
+      "epoch": 0.5955555555555555,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.749896846140772e-06,
+      "loss": 1.242,
+      "step": 469
+    },
+    {
+      "epoch": 0.5968253968253968,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.74855592743981e-06,
+      "loss": 1.1985,
+      "step": 470
+    },
+    {
+      "epoch": 0.5980952380952381,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.7472115576182975e-06,
+      "loss": 1.1654,
+      "step": 471
+    },
+    {
+      "epoch": 0.5993650793650793,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.745863738104435e-06,
+      "loss": 1.2013,
+      "step": 472
+    },
+    {
+      "epoch": 0.6006349206349206,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.744512470330087e-06,
+      "loss": 1.2208,
+      "step": 473
+    },
+    {
+      "epoch": 0.6019047619047619,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.743157755730776e-06,
+      "loss": 1.183,
+      "step": 474
+    },
+    {
+      "epoch": 0.6031746031746031,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.741799595745694e-06,
+      "loss": 1.1281,
+      "step": 475
+    },
+    {
+      "epoch": 0.6044444444444445,
+      "grad_norm": 0.9765625,
+      "learning_rate": 6.740437991817686e-06,
+      "loss": 1.2161,
+      "step": 476
+    },
+    {
+      "epoch": 0.6057142857142858,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.739072945393262e-06,
+      "loss": 1.1631,
+      "step": 477
+    },
+    {
+      "epoch": 0.606984126984127,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.737704457922585e-06,
+      "loss": 1.1185,
+      "step": 478
+    },
+    {
+      "epoch": 0.6082539682539683,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.736332530859474e-06,
+      "loss": 1.1549,
+      "step": 479
+    },
+    {
+      "epoch": 0.6095238095238096,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.734957165661405e-06,
+      "loss": 1.248,
+      "step": 480
+    },
+    {
+      "epoch": 0.6107936507936508,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.733578363789503e-06,
+      "loss": 1.213,
+      "step": 481
+    },
+    {
+      "epoch": 0.6120634920634921,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.732196126708547e-06,
+      "loss": 1.2044,
+      "step": 482
+    },
+    {
+      "epoch": 0.6133333333333333,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.7308104558869616e-06,
+      "loss": 1.1523,
+      "step": 483
+    },
+    {
+      "epoch": 0.6146031746031746,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.729421352796823e-06,
+      "loss": 1.2592,
+      "step": 484
+    },
+    {
+      "epoch": 0.6158730158730159,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.728028818913852e-06,
+      "loss": 1.1671,
+      "step": 485
+    },
+    {
+      "epoch": 0.6171428571428571,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.726632855717414e-06,
+      "loss": 1.1276,
+      "step": 486
+    },
+    {
+      "epoch": 0.6184126984126984,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.725233464690519e-06,
+      "loss": 1.1712,
+      "step": 487
+    },
+    {
+      "epoch": 0.6196825396825397,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.723830647319816e-06,
+      "loss": 1.1428,
+      "step": 488
+    },
+    {
+      "epoch": 0.6209523809523809,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.722424405095596e-06,
+      "loss": 1.1649,
+      "step": 489
+    },
+    {
+      "epoch": 0.6222222222222222,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.721014739511787e-06,
+      "loss": 1.1549,
+      "step": 490
+    },
+    {
+      "epoch": 0.6234920634920635,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.719601652065957e-06,
+      "loss": 1.0888,
+      "step": 491
+    },
+    {
+      "epoch": 0.6247619047619047,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.7181851442593045e-06,
+      "loss": 1.1915,
+      "step": 492
+    },
+    {
+      "epoch": 0.626031746031746,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.716765217596666e-06,
+      "loss": 1.0786,
+      "step": 493
+    },
+    {
+      "epoch": 0.6273015873015872,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.715341873586506e-06,
+      "loss": 1.209,
+      "step": 494
+    },
+    {
+      "epoch": 0.6285714285714286,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.713915113740924e-06,
+      "loss": 1.2102,
+      "step": 495
+    },
+    {
+      "epoch": 0.6298412698412699,
+      "grad_norm": 0.9765625,
+      "learning_rate": 6.712484939575643e-06,
+      "loss": 1.2003,
+      "step": 496
+    },
+    {
+      "epoch": 0.6311111111111111,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.7110513526100184e-06,
+      "loss": 1.2036,
+      "step": 497
+    },
+    {
+      "epoch": 0.6323809523809524,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.7096143543670264e-06,
+      "loss": 1.1949,
+      "step": 498
+    },
+    {
+      "epoch": 0.6336507936507937,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.708173946373274e-06,
+      "loss": 1.1559,
+      "step": 499
+    },
+    {
+      "epoch": 0.6349206349206349,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.706730130158981e-06,
+      "loss": 1.2509,
+      "step": 500
+    },
+    {
+      "epoch": 0.6361904761904762,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.705282907257997e-06,
+      "loss": 1.1922,
+      "step": 501
+    },
+    {
+      "epoch": 0.6374603174603175,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.703832279207786e-06,
+      "loss": 1.1691,
+      "step": 502
+    },
+    {
+      "epoch": 0.6387301587301587,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.7023782475494296e-06,
+      "loss": 1.1493,
+      "step": 503
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.700920813827626e-06,
+      "loss": 1.1078,
+      "step": 504
+    },
+    {
+      "epoch": 0.6412698412698413,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.699459979590689e-06,
+      "loss": 1.1494,
+      "step": 505
+    },
+    {
+      "epoch": 0.6425396825396825,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.697995746390542e-06,
+      "loss": 1.2077,
+      "step": 506
+    },
+    {
+      "epoch": 0.6438095238095238,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.696528115782721e-06,
+      "loss": 1.1758,
+      "step": 507
+    },
+    {
+      "epoch": 0.645079365079365,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.695057089326372e-06,
+      "loss": 1.1779,
+      "step": 508
+    },
+    {
+      "epoch": 0.6463492063492063,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.693582668584248e-06,
+      "loss": 1.2281,
+      "step": 509
+    },
+    {
+      "epoch": 0.6476190476190476,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.6921048551227065e-06,
+      "loss": 1.1934,
+      "step": 510
+    },
+    {
+      "epoch": 0.6488888888888888,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.690623650511711e-06,
+      "loss": 1.24,
+      "step": 511
+    },
+    {
+      "epoch": 0.6501587301587302,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.689139056324826e-06,
+      "loss": 1.1979,
+      "step": 512
+    },
+    {
+      "epoch": 0.6514285714285715,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.68765107413922e-06,
+      "loss": 1.2034,
+      "step": 513
+    },
+    {
+      "epoch": 0.6526984126984127,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.686159705535657e-06,
+      "loss": 1.1913,
+      "step": 514
+    },
+    {
+      "epoch": 0.653968253968254,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.684664952098503e-06,
+      "loss": 1.2158,
+      "step": 515
+    },
+    {
+      "epoch": 0.6552380952380953,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.683166815415712e-06,
+      "loss": 1.1612,
+      "step": 516
+    },
+    {
+      "epoch": 0.6565079365079365,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.681665297078842e-06,
+      "loss": 1.1911,
+      "step": 517
+    },
+    {
+      "epoch": 0.6577777777777778,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.6801603986830395e-06,
+      "loss": 1.1902,
+      "step": 518
+    },
+    {
+      "epoch": 0.659047619047619,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.6786521218270374e-06,
+      "loss": 1.1424,
+      "step": 519
+    },
+    {
+      "epoch": 0.6603174603174603,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.6771404681131634e-06,
+      "loss": 1.1362,
+      "step": 520
+    },
+    {
+      "epoch": 0.6615873015873016,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.675625439147331e-06,
+      "loss": 1.1517,
+      "step": 521
+    },
+    {
+      "epoch": 0.6628571428571428,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.67410703653904e-06,
+      "loss": 1.1232,
+      "step": 522
+    },
+    {
+      "epoch": 0.6641269841269841,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.672585261901371e-06,
+      "loss": 1.1583,
+      "step": 523
+    },
+    {
+      "epoch": 0.6653968253968254,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.671060116850993e-06,
+      "loss": 1.1405,
+      "step": 524
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.669531603008148e-06,
+      "loss": 1.1955,
+      "step": 525
+    },
+    {
+      "epoch": 0.6679365079365079,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.6679997219966635e-06,
+      "loss": 1.1239,
+      "step": 526
+    },
+    {
+      "epoch": 0.6692063492063492,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.66646447544394e-06,
+      "loss": 1.2009,
+      "step": 527
+    },
+    {
+      "epoch": 0.6704761904761904,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.664925864980956e-06,
+      "loss": 1.1367,
+      "step": 528
+    },
+    {
+      "epoch": 0.6717460317460318,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.663383892242261e-06,
+      "loss": 1.1926,
+      "step": 529
+    },
+    {
+      "epoch": 0.6730158730158731,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.661838558865978e-06,
+      "loss": 1.1656,
+      "step": 530
+    },
+    {
+      "epoch": 0.6742857142857143,
+      "grad_norm": 0.953125,
+      "learning_rate": 6.6602898664938014e-06,
+      "loss": 1.2526,
+      "step": 531
+    },
+    {
+      "epoch": 0.6755555555555556,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.658737816770992e-06,
+      "loss": 1.2039,
+      "step": 532
+    },
+    {
+      "epoch": 0.6768253968253968,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.657182411346377e-06,
+      "loss": 1.1791,
+      "step": 533
+    },
+    {
+      "epoch": 0.6780952380952381,
+      "grad_norm": 0.87890625,
+      "learning_rate": 6.655623651872352e-06,
+      "loss": 1.1603,
+      "step": 534
+    },
+    {
+      "epoch": 0.6793650793650794,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.654061540004871e-06,
+      "loss": 1.1164,
+      "step": 535
+    },
+    {
+      "epoch": 0.6806349206349206,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.652496077403453e-06,
+      "loss": 1.2685,
+      "step": 536
+    },
+    {
+      "epoch": 0.6819047619047619,
+      "grad_norm": 0.953125,
+      "learning_rate": 6.650927265731175e-06,
+      "loss": 1.1049,
+      "step": 537
+    },
+    {
+      "epoch": 0.6831746031746032,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.649355106654672e-06,
+      "loss": 1.1492,
+      "step": 538
+    },
+    {
+      "epoch": 0.6844444444444444,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.647779601844137e-06,
+      "loss": 1.2699,
+      "step": 539
+    },
+    {
+      "epoch": 0.6857142857142857,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.646200752973315e-06,
+      "loss": 1.1014,
+      "step": 540
+    },
+    {
+      "epoch": 0.686984126984127,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.6446185617195034e-06,
+      "loss": 1.2045,
+      "step": 541
+    },
+    {
+      "epoch": 0.6882539682539682,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.643033029763553e-06,
+      "loss": 1.1571,
+      "step": 542
+    },
+    {
+      "epoch": 0.6895238095238095,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.641444158789861e-06,
+      "loss": 1.1805,
+      "step": 543
+    },
+    {
+      "epoch": 0.6907936507936508,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.639851950486374e-06,
+      "loss": 1.28,
+      "step": 544
+    },
+    {
+      "epoch": 0.692063492063492,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.638256406544582e-06,
+      "loss": 1.2255,
+      "step": 545
+    },
+    {
+      "epoch": 0.6933333333333334,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.636657528659518e-06,
+      "loss": 1.1858,
+      "step": 546
+    },
+    {
+      "epoch": 0.6946031746031746,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.6350553185297625e-06,
+      "loss": 1.1587,
+      "step": 547
+    },
+    {
+      "epoch": 0.6958730158730159,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.633449777857429e-06,
+      "loss": 1.1482,
+      "step": 548
+    },
+    {
+      "epoch": 0.6971428571428572,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.631840908348171e-06,
+      "loss": 1.232,
+      "step": 549
+    },
+    {
+      "epoch": 0.6984126984126984,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.6302287117111825e-06,
+      "loss": 1.0773,
+      "step": 550
+    },
+    {
+      "epoch": 0.6996825396825397,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.628613189659187e-06,
+      "loss": 1.1754,
+      "step": 551
+    },
+    {
+      "epoch": 0.700952380952381,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.626994343908444e-06,
+      "loss": 1.1066,
+      "step": 552
+    },
+    {
+      "epoch": 0.7022222222222222,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.625372176178743e-06,
+      "loss": 1.2023,
+      "step": 553
+    },
+    {
+      "epoch": 0.7034920634920635,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.6237466881934e-06,
+      "loss": 1.1735,
+      "step": 554
+    },
+    {
+      "epoch": 0.7047619047619048,
+      "grad_norm": 0.875,
+      "learning_rate": 6.622117881679265e-06,
+      "loss": 1.0454,
+      "step": 555
+    },
+    {
+      "epoch": 0.706031746031746,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.620485758366705e-06,
+      "loss": 1.1948,
+      "step": 556
+    },
+    {
+      "epoch": 0.7073015873015873,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.6188503199896175e-06,
+      "loss": 1.1723,
+      "step": 557
+    },
+    {
+      "epoch": 0.7085714285714285,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.617211568285419e-06,
+      "loss": 1.2298,
+      "step": 558
+    },
+    {
+      "epoch": 0.7098412698412698,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.615569504995044e-06,
+      "loss": 1.1308,
+      "step": 559
+    },
+    {
+      "epoch": 0.7111111111111111,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.6139241318629486e-06,
+      "loss": 1.1852,
+      "step": 560
+    },
+    {
+      "epoch": 0.7123809523809523,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.612275450637103e-06,
+      "loss": 1.1247,
+      "step": 561
+    },
+    {
+      "epoch": 0.7136507936507936,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.610623463068995e-06,
+      "loss": 1.1734,
+      "step": 562
+    },
+    {
+      "epoch": 0.714920634920635,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.608968170913618e-06,
+      "loss": 1.1356,
+      "step": 563
+    },
+    {
+      "epoch": 0.7161904761904762,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.607309575929483e-06,
+      "loss": 1.1212,
+      "step": 564
+    },
+    {
+      "epoch": 0.7174603174603175,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.6056476798786075e-06,
+      "loss": 1.1004,
+      "step": 565
+    },
+    {
+      "epoch": 0.7187301587301588,
+      "grad_norm": 0.953125,
+      "learning_rate": 6.603982484526515e-06,
+      "loss": 1.1515,
+      "step": 566
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.602313991642234e-06,
+      "loss": 1.1864,
+      "step": 567
+    },
+    {
+      "epoch": 0.7212698412698413,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.600642202998296e-06,
+      "loss": 1.2114,
+      "step": 568
+    },
+    {
+      "epoch": 0.7225396825396826,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.598967120370735e-06,
+      "loss": 1.2265,
+      "step": 569
+    },
+    {
+      "epoch": 0.7238095238095238,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.597288745539086e-06,
+      "loss": 1.2089,
+      "step": 570
+    },
+    {
+      "epoch": 0.7250793650793651,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.595607080286378e-06,
+      "loss": 1.1903,
+      "step": 571
+    },
+    {
+      "epoch": 0.7263492063492063,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.593922126399136e-06,
+      "loss": 1.1926,
+      "step": 572
+    },
+    {
+      "epoch": 0.7276190476190476,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.59223388566738e-06,
+      "loss": 1.1675,
+      "step": 573
+    },
+    {
+      "epoch": 0.7288888888888889,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.5905423598846224e-06,
+      "loss": 1.1888,
+      "step": 574
+    },
+    {
+      "epoch": 0.7301587301587301,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.588847550847864e-06,
+      "loss": 1.1543,
+      "step": 575
+    },
+    {
+      "epoch": 0.7314285714285714,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.587149460357595e-06,
+      "loss": 1.198,
+      "step": 576
+    },
+    {
+      "epoch": 0.7326984126984127,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.5854480902177885e-06,
+      "loss": 1.1214,
+      "step": 577
+    },
+    {
+      "epoch": 0.7339682539682539,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.583743442235906e-06,
+      "loss": 1.1976,
+      "step": 578
+    },
+    {
+      "epoch": 0.7352380952380952,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.582035518222888e-06,
+      "loss": 1.1888,
+      "step": 579
+    },
+    {
+      "epoch": 0.7365079365079366,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.580324319993159e-06,
+      "loss": 1.1821,
+      "step": 580
+    },
+    {
+      "epoch": 0.7377777777777778,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.578609849364617e-06,
+      "loss": 1.1231,
+      "step": 581
+    },
+    {
+      "epoch": 0.7390476190476191,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.57689210815864e-06,
+      "loss": 1.2334,
+      "step": 582
+    },
+    {
+      "epoch": 0.7403174603174603,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.5751710982000795e-06,
+      "loss": 1.1552,
+      "step": 583
+    },
+    {
+      "epoch": 0.7415873015873016,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.573446821317259e-06,
+      "loss": 1.1631,
+      "step": 584
+    },
+    {
+      "epoch": 0.7428571428571429,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.571719279341972e-06,
+      "loss": 1.153,
+      "step": 585
+    },
+    {
+      "epoch": 0.7441269841269841,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.569988474109486e-06,
+      "loss": 1.1754,
+      "step": 586
+    },
+    {
+      "epoch": 0.7453968253968254,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.568254407458526e-06,
+      "loss": 1.1606,
+      "step": 587
+    },
+    {
+      "epoch": 0.7466666666666667,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.56651708123129e-06,
+      "loss": 1.1695,
+      "step": 588
+    },
+    {
+      "epoch": 0.7479365079365079,
+      "grad_norm": 1.0,
+      "learning_rate": 6.564776497273436e-06,
+      "loss": 1.2078,
+      "step": 589
+    },
+    {
+      "epoch": 0.7492063492063492,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.56303265743408e-06,
+      "loss": 1.125,
+      "step": 590
+    },
+    {
+      "epoch": 0.7504761904761905,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.5612855635658015e-06,
+      "loss": 1.1365,
+      "step": 591
+    },
+    {
+      "epoch": 0.7517460317460317,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.559535217524634e-06,
+      "loss": 1.2071,
+      "step": 592
+    },
+    {
+      "epoch": 0.753015873015873,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.557781621170068e-06,
+      "loss": 1.1188,
+      "step": 593
+    },
+    {
+      "epoch": 0.7542857142857143,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.556024776365044e-06,
+      "loss": 1.1334,
+      "step": 594
+    },
+    {
+      "epoch": 0.7555555555555555,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.554264684975958e-06,
+      "loss": 1.2258,
+      "step": 595
+    },
+    {
+      "epoch": 0.7568253968253968,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.552501348872649e-06,
+      "loss": 1.1216,
+      "step": 596
+    },
+    {
+      "epoch": 0.758095238095238,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.550734769928409e-06,
+      "loss": 1.1999,
+      "step": 597
+    },
+    {
+      "epoch": 0.7593650793650794,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.548964950019971e-06,
+      "loss": 1.2001,
+      "step": 598
+    },
+    {
+      "epoch": 0.7606349206349207,
+      "grad_norm": 0.87890625,
+      "learning_rate": 6.547191891027514e-06,
+      "loss": 1.1323,
+      "step": 599
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.545415594834656e-06,
+      "loss": 1.0832,
+      "step": 600
+    },
+    {
+      "epoch": 0.7631746031746032,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.5436360633284535e-06,
+      "loss": 1.145,
+      "step": 601
+    },
+    {
+      "epoch": 0.7644444444444445,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.541853298399402e-06,
+      "loss": 1.1393,
+      "step": 602
+    },
+    {
+      "epoch": 0.7657142857142857,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.540067301941432e-06,
+      "loss": 1.1743,
+      "step": 603
+    },
+    {
+      "epoch": 0.766984126984127,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.538278075851906e-06,
+      "loss": 1.1077,
+      "step": 604
+    },
+    {
+      "epoch": 0.7682539682539683,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.536485622031617e-06,
+      "loss": 1.1973,
+      "step": 605
+    },
+    {
+      "epoch": 0.7695238095238095,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.534689942384788e-06,
+      "loss": 1.2011,
+      "step": 606
+    },
+    {
+      "epoch": 0.7707936507936508,
+      "grad_norm": 1.0,
+      "learning_rate": 6.532891038819069e-06,
+      "loss": 1.2117,
+      "step": 607
+    },
+    {
+      "epoch": 0.7720634920634921,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.531088913245536e-06,
+      "loss": 1.1794,
+      "step": 608
+    },
+    {
+      "epoch": 0.7733333333333333,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.529283567578683e-06,
+      "loss": 1.1103,
+      "step": 609
+    },
+    {
+      "epoch": 0.7746031746031746,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.527475003736433e-06,
+      "loss": 1.1573,
+      "step": 610
+    },
+    {
+      "epoch": 0.7758730158730158,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.52566322364012e-06,
+      "loss": 1.2098,
+      "step": 611
+    },
+    {
+      "epoch": 0.7771428571428571,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.523848229214499e-06,
+      "loss": 1.1054,
+      "step": 612
+    },
+    {
+      "epoch": 0.7784126984126984,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.52203002238774e-06,
+      "loss": 1.1287,
+      "step": 613
+    },
+    {
+      "epoch": 0.7796825396825396,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.520208605091422e-06,
+      "loss": 1.1954,
+      "step": 614
+    },
+    {
+      "epoch": 0.780952380952381,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.518383979260542e-06,
+      "loss": 1.1641,
+      "step": 615
+    },
+    {
+      "epoch": 0.7822222222222223,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.516556146833496e-06,
+      "loss": 1.1232,
+      "step": 616
+    },
+    {
+      "epoch": 0.7834920634920635,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.514725109752094e-06,
+      "loss": 1.0694,
+      "step": 617
+    },
+    {
+      "epoch": 0.7847619047619048,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.512890869961548e-06,
+      "loss": 1.2341,
+      "step": 618
+    },
+    {
+      "epoch": 0.7860317460317461,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.51105342941047e-06,
+      "loss": 1.1597,
+      "step": 619
+    },
+    {
+      "epoch": 0.7873015873015873,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.509212790050879e-06,
+      "loss": 1.2067,
+      "step": 620
+    },
+    {
+      "epoch": 0.7885714285714286,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.507368953838183e-06,
+      "loss": 1.1783,
+      "step": 621
+    },
+    {
+      "epoch": 0.7898412698412698,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.505521922731195e-06,
+      "loss": 1.1478,
+      "step": 622
+    },
+    {
+      "epoch": 0.7911111111111111,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.503671698692115e-06,
+      "loss": 1.1335,
+      "step": 623
+    },
+    {
+      "epoch": 0.7923809523809524,
+      "grad_norm": 0.953125,
+      "learning_rate": 6.501818283686542e-06,
+      "loss": 1.1269,
+      "step": 624
+    },
+    {
+      "epoch": 0.7936507936507936,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.499961679683458e-06,
+      "loss": 1.165,
+      "step": 625
+    },
+    {
+      "epoch": 0.7949206349206349,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.498101888655238e-06,
+      "loss": 1.2347,
+      "step": 626
+    },
+    {
+      "epoch": 0.7961904761904762,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.4962389125776395e-06,
+      "loss": 1.1218,
+      "step": 627
+    },
+    {
+      "epoch": 0.7974603174603174,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.494372753429806e-06,
+      "loss": 1.1768,
+      "step": 628
+    },
+    {
+      "epoch": 0.7987301587301587,
+      "grad_norm": 0.9765625,
+      "learning_rate": 6.492503413194261e-06,
+      "loss": 1.1636,
+      "step": 629
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.490630893856908e-06,
+      "loss": 1.1189,
+      "step": 630
+    },
+    {
+      "epoch": 0.8012698412698412,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.488755197407028e-06,
+      "loss": 1.1812,
+      "step": 631
+    },
+    {
+      "epoch": 0.8025396825396826,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.486876325837276e-06,
+      "loss": 1.1498,
+      "step": 632
+    },
+    {
+      "epoch": 0.8038095238095239,
+      "grad_norm": 0.953125,
+      "learning_rate": 6.4849942811436835e-06,
+      "loss": 1.1018,
+      "step": 633
+    },
+    {
+      "epoch": 0.8050793650793651,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.483109065325648e-06,
+      "loss": 1.1646,
+      "step": 634
+    },
+    {
+      "epoch": 0.8063492063492064,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.481220680385941e-06,
+      "loss": 1.2117,
+      "step": 635
+    },
+    {
+      "epoch": 0.8076190476190476,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.479329128330696e-06,
+      "loss": 1.1717,
+      "step": 636
+    },
+    {
+      "epoch": 0.8088888888888889,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.477434411169414e-06,
+      "loss": 1.1423,
+      "step": 637
+    },
+    {
+      "epoch": 0.8101587301587302,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.475536530914958e-06,
+      "loss": 1.204,
+      "step": 638
+    },
+    {
+      "epoch": 0.8114285714285714,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.473635489583552e-06,
+      "loss": 1.1459,
+      "step": 639
+    },
+    {
+      "epoch": 0.8126984126984127,
+      "grad_norm": 0.8359375,
+      "learning_rate": 6.471731289194776e-06,
+      "loss": 1.1183,
+      "step": 640
+    },
+    {
+      "epoch": 0.813968253968254,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.469823931771567e-06,
+      "loss": 1.1436,
+      "step": 641
+    },
+    {
+      "epoch": 0.8152380952380952,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.467913419340218e-06,
+      "loss": 1.2367,
+      "step": 642
+    },
+    {
+      "epoch": 0.8165079365079365,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.465999753930371e-06,
+      "loss": 1.0817,
+      "step": 643
+    },
+    {
+      "epoch": 0.8177777777777778,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.4640829375750175e-06,
+      "loss": 1.168,
+      "step": 644
+    },
+    {
+      "epoch": 0.819047619047619,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.462162972310498e-06,
+      "loss": 1.1851,
+      "step": 645
+    },
+    {
+      "epoch": 0.8203174603174603,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.4602398601765e-06,
+      "loss": 1.1486,
+      "step": 646
+    },
+    {
+      "epoch": 0.8215873015873015,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.458313603216049e-06,
+      "loss": 1.1594,
+      "step": 647
+    },
+    {
+      "epoch": 0.8228571428571428,
+      "grad_norm": 0.953125,
+      "learning_rate": 6.456384203475516e-06,
+      "loss": 1.1407,
+      "step": 648
+    },
+    {
+      "epoch": 0.8241269841269842,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.454451663004607e-06,
+      "loss": 1.1415,
+      "step": 649
+    },
+    {
+      "epoch": 0.8253968253968254,
+      "grad_norm": 0.87890625,
+      "learning_rate": 6.4525159838563695e-06,
+      "loss": 1.084,
+      "step": 650
+    },
+    {
+      "epoch": 0.8266666666666667,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.45057716808718e-06,
+      "loss": 1.1939,
+      "step": 651
+    },
+    {
+      "epoch": 0.827936507936508,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.448635217756751e-06,
+      "loss": 1.1927,
+      "step": 652
+    },
+    {
+      "epoch": 0.8292063492063492,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.446690134928123e-06,
+      "loss": 1.1945,
+      "step": 653
+    },
+    {
+      "epoch": 0.8304761904761905,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.444741921667665e-06,
+      "loss": 1.1275,
+      "step": 654
+    },
+    {
+      "epoch": 0.8317460317460318,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.442790580045072e-06,
+      "loss": 1.1588,
+      "step": 655
+    },
+    {
+      "epoch": 0.833015873015873,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.440836112133362e-06,
+      "loss": 1.2396,
+      "step": 656
+    },
+    {
+      "epoch": 0.8342857142857143,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.438878520008875e-06,
+      "loss": 1.12,
+      "step": 657
+    },
+    {
+      "epoch": 0.8355555555555556,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.436917805751268e-06,
+      "loss": 1.1548,
+      "step": 658
+    },
+    {
+      "epoch": 0.8368253968253968,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.434953971443517e-06,
+      "loss": 1.0921,
+      "step": 659
+    },
+    {
+      "epoch": 0.8380952380952381,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.432987019171911e-06,
+      "loss": 1.102,
+      "step": 660
+    },
+    {
+      "epoch": 0.8393650793650793,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.431016951026052e-06,
+      "loss": 1.2026,
+      "step": 661
+    },
+    {
+      "epoch": 0.8406349206349206,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.429043769098855e-06,
+      "loss": 1.1753,
+      "step": 662
+    },
+    {
+      "epoch": 0.8419047619047619,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.427067475486537e-06,
+      "loss": 1.1094,
+      "step": 663
+    },
+    {
+      "epoch": 0.8431746031746031,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.425088072288624e-06,
+      "loss": 1.1124,
+      "step": 664
+    },
+    {
+      "epoch": 0.8444444444444444,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.423105561607948e-06,
+      "loss": 1.2325,
+      "step": 665
+    },
+    {
+      "epoch": 0.8457142857142858,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.421119945550639e-06,
+      "loss": 1.175,
+      "step": 666
+    },
+    {
+      "epoch": 0.846984126984127,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.419131226226125e-06,
+      "loss": 1.0951,
+      "step": 667
+    },
+    {
+      "epoch": 0.8482539682539683,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.4171394057471335e-06,
+      "loss": 1.1837,
+      "step": 668
+    },
+    {
+      "epoch": 0.8495238095238096,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.415144486229686e-06,
+      "loss": 1.1108,
+      "step": 669
+    },
+    {
+      "epoch": 0.8507936507936508,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.413146469793095e-06,
+      "loss": 1.2185,
+      "step": 670
+    },
+    {
+      "epoch": 0.8520634920634921,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.411145358559964e-06,
+      "loss": 1.1745,
+      "step": 671
+    },
+    {
+      "epoch": 0.8533333333333334,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.409141154656184e-06,
+      "loss": 1.1682,
+      "step": 672
+    },
+    {
+      "epoch": 0.8546031746031746,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.407133860210932e-06,
+      "loss": 1.1913,
+      "step": 673
+    },
+    {
+      "epoch": 0.8558730158730159,
+      "grad_norm": 1.0,
+      "learning_rate": 6.405123477356667e-06,
+      "loss": 1.1316,
+      "step": 674
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.403110008229131e-06,
+      "loss": 1.1423,
+      "step": 675
+    },
+    {
+      "epoch": 0.8584126984126984,
+      "grad_norm": 0.875,
+      "learning_rate": 6.401093454967343e-06,
+      "loss": 1.1023,
+      "step": 676
+    },
+    {
+      "epoch": 0.8596825396825397,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.3990738197136e-06,
+      "loss": 1.2511,
+      "step": 677
+    },
+    {
+      "epoch": 0.8609523809523809,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.39705110461347e-06,
+      "loss": 1.1134,
+      "step": 678
+    },
+    {
+      "epoch": 0.8622222222222222,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.395025311815797e-06,
+      "loss": 1.2409,
+      "step": 679
+    },
+    {
+      "epoch": 0.8634920634920635,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.392996443472691e-06,
+      "loss": 1.1694,
+      "step": 680
+    },
+    {
+      "epoch": 0.8647619047619047,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.390964501739533e-06,
+      "loss": 1.1082,
+      "step": 681
+    },
+    {
+      "epoch": 0.866031746031746,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.388929488774966e-06,
+      "loss": 1.1303,
+      "step": 682
+    },
+    {
+      "epoch": 0.8673015873015874,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.386891406740896e-06,
+      "loss": 1.1679,
+      "step": 683
+    },
+    {
+      "epoch": 0.8685714285714285,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.384850257802492e-06,
+      "loss": 1.1124,
+      "step": 684
+    },
+    {
+      "epoch": 0.8698412698412699,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.382806044128178e-06,
+      "loss": 1.1452,
+      "step": 685
+    },
+    {
+      "epoch": 0.8711111111111111,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.3807587678896345e-06,
+      "loss": 1.1504,
+      "step": 686
+    },
+    {
+      "epoch": 0.8723809523809524,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.378708431261799e-06,
+      "loss": 1.1171,
+      "step": 687
+    },
+    {
+      "epoch": 0.8736507936507937,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.376655036422854e-06,
+      "loss": 1.0953,
+      "step": 688
+    },
+    {
+      "epoch": 0.8749206349206349,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.374598585554238e-06,
+      "loss": 1.1465,
+      "step": 689
+    },
+    {
+      "epoch": 0.8761904761904762,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.372539080840629e-06,
+      "loss": 1.1303,
+      "step": 690
+    },
+    {
+      "epoch": 0.8774603174603175,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.370476524469955e-06,
+      "loss": 1.1897,
+      "step": 691
+    },
+    {
+      "epoch": 0.8787301587301587,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.368410918633383e-06,
+      "loss": 1.165,
+      "step": 692
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.3663422655253195e-06,
+      "loss": 1.1836,
+      "step": 693
+    },
+    {
+      "epoch": 0.8812698412698413,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.36427056734341e-06,
+      "loss": 1.1677,
+      "step": 694
+    },
+    {
+      "epoch": 0.8825396825396825,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.362195826288534e-06,
+      "loss": 1.1874,
+      "step": 695
+    },
+    {
+      "epoch": 0.8838095238095238,
+      "grad_norm": 0.953125,
+      "learning_rate": 6.3601180445648025e-06,
+      "loss": 1.1586,
+      "step": 696
+    },
+    {
+      "epoch": 0.8850793650793651,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.358037224379559e-06,
+      "loss": 1.1884,
+      "step": 697
+    },
+    {
+      "epoch": 0.8863492063492063,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.355953367943375e-06,
+      "loss": 1.1844,
+      "step": 698
+    },
+    {
+      "epoch": 0.8876190476190476,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.353866477470043e-06,
+      "loss": 1.1958,
+      "step": 699
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.351776555176585e-06,
+      "loss": 1.1412,
+      "step": 700
+    },
+    {
+      "epoch": 0.8901587301587301,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.34968360328324e-06,
+      "loss": 1.0831,
+      "step": 701
+    },
+    {
+      "epoch": 0.8914285714285715,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.347587624013467e-06,
+      "loss": 1.2741,
+      "step": 702
+    },
+    {
+      "epoch": 0.8926984126984127,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.34548861959394e-06,
+      "loss": 1.184,
+      "step": 703
+    },
+    {
+      "epoch": 0.893968253968254,
+      "grad_norm": 0.9765625,
+      "learning_rate": 6.343386592254547e-06,
+      "loss": 1.1357,
+      "step": 704
+    },
+    {
+      "epoch": 0.8952380952380953,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.341281544228389e-06,
+      "loss": 1.147,
+      "step": 705
+    },
+    {
+      "epoch": 0.8965079365079365,
+      "grad_norm": 1.0,
+      "learning_rate": 6.339173477751774e-06,
+      "loss": 1.1581,
+      "step": 706
+    },
+    {
+      "epoch": 0.8977777777777778,
+      "grad_norm": 0.875,
+      "learning_rate": 6.337062395064218e-06,
+      "loss": 1.0782,
+      "step": 707
+    },
+    {
+      "epoch": 0.8990476190476191,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.334948298408441e-06,
+      "loss": 1.0997,
+      "step": 708
+    },
+    {
+      "epoch": 0.9003174603174603,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.332831190030364e-06,
+      "loss": 1.1336,
+      "step": 709
+    },
+    {
+      "epoch": 0.9015873015873016,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.33071107217911e-06,
+      "loss": 1.2151,
+      "step": 710
+    },
+    {
+      "epoch": 0.9028571428571428,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.328587947106995e-06,
+      "loss": 1.2093,
+      "step": 711
+    },
+    {
+      "epoch": 0.9041269841269841,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.326461817069534e-06,
+      "loss": 1.2357,
+      "step": 712
+    },
+    {
+      "epoch": 0.9053968253968254,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.324332684325431e-06,
+      "loss": 1.1374,
+      "step": 713
+    },
+    {
+      "epoch": 0.9066666666666666,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.3222005511365826e-06,
+      "loss": 1.1561,
+      "step": 714
+    },
+    {
+      "epoch": 0.9079365079365079,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.3200654197680716e-06,
+      "loss": 1.1071,
+      "step": 715
+    },
+    {
+      "epoch": 0.9092063492063492,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.317927292488165e-06,
+      "loss": 1.167,
+      "step": 716
+    },
+    {
+      "epoch": 0.9104761904761904,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.3157861715683156e-06,
+      "loss": 1.1746,
+      "step": 717
+    },
+    {
+      "epoch": 0.9117460317460317,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.313642059283153e-06,
+      "loss": 1.1708,
+      "step": 718
+    },
+    {
+      "epoch": 0.9130158730158731,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.311494957910487e-06,
+      "loss": 1.1456,
+      "step": 719
+    },
+    {
+      "epoch": 0.9142857142857143,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.309344869731301e-06,
+      "loss": 1.1039,
+      "step": 720
+    },
+    {
+      "epoch": 0.9155555555555556,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.307191797029753e-06,
+      "loss": 1.2034,
+      "step": 721
+    },
+    {
+      "epoch": 0.9168253968253969,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.305035742093172e-06,
+      "loss": 1.2167,
+      "step": 722
+    },
+    {
+      "epoch": 0.9180952380952381,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.3028767072120536e-06,
+      "loss": 1.1112,
+      "step": 723
+    },
+    {
+      "epoch": 0.9193650793650794,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.3007146946800605e-06,
+      "loss": 1.1532,
+      "step": 724
+    },
+    {
+      "epoch": 0.9206349206349206,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.298549706794017e-06,
+      "loss": 1.128,
+      "step": 725
+    },
+    {
+      "epoch": 0.9219047619047619,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.29638174585391e-06,
+      "loss": 1.132,
+      "step": 726
+    },
+    {
+      "epoch": 0.9231746031746032,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.294210814162884e-06,
+      "loss": 1.1487,
+      "step": 727
+    },
+    {
+      "epoch": 0.9244444444444444,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.29203691402724e-06,
+      "loss": 1.1577,
+      "step": 728
+    },
+    {
+      "epoch": 0.9257142857142857,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.289860047756434e-06,
+      "loss": 1.128,
+      "step": 729
+    },
+    {
+      "epoch": 0.926984126984127,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.287680217663068e-06,
+      "loss": 1.1734,
+      "step": 730
+    },
+    {
+      "epoch": 0.9282539682539682,
+      "grad_norm": 0.87890625,
+      "learning_rate": 6.285497426062898e-06,
+      "loss": 1.1226,
+      "step": 731
+    },
+    {
+      "epoch": 0.9295238095238095,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.2833116752748255e-06,
+      "loss": 1.0581,
+      "step": 732
+    },
+    {
+      "epoch": 0.9307936507936508,
+      "grad_norm": 0.953125,
+      "learning_rate": 6.281122967620892e-06,
+      "loss": 1.1644,
+      "step": 733
+    },
+    {
+      "epoch": 0.932063492063492,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.2789313054262826e-06,
+      "loss": 1.1771,
+      "step": 734
+    },
+    {
+      "epoch": 0.9333333333333333,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.276736691019323e-06,
+      "loss": 1.1277,
+      "step": 735
+    },
+    {
+      "epoch": 0.9346031746031747,
+      "grad_norm": 0.875,
+      "learning_rate": 6.274539126731473e-06,
+      "loss": 1.0941,
+      "step": 736
+    },
+    {
+      "epoch": 0.9358730158730159,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.2723386148973244e-06,
+      "loss": 1.1645,
+      "step": 737
+    },
+    {
+      "epoch": 0.9371428571428572,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.270135157854606e-06,
+      "loss": 1.1453,
+      "step": 738
+    },
+    {
+      "epoch": 0.9384126984126984,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.267928757944169e-06,
+      "loss": 1.1461,
+      "step": 739
+    },
+    {
+      "epoch": 0.9396825396825397,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.2657194175099964e-06,
+      "loss": 1.0925,
+      "step": 740
+    },
+    {
+      "epoch": 0.940952380952381,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.263507138899191e-06,
+      "loss": 1.1974,
+      "step": 741
+    },
+    {
+      "epoch": 0.9422222222222222,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.26129192446198e-06,
+      "loss": 1.1136,
+      "step": 742
+    },
+    {
+      "epoch": 0.9434920634920635,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.259073776551707e-06,
+      "loss": 1.1404,
+      "step": 743
+    },
+    {
+      "epoch": 0.9447619047619048,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.2568526975248335e-06,
+      "loss": 1.1954,
+      "step": 744
+    },
+    {
+      "epoch": 0.946031746031746,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.254628689740935e-06,
+      "loss": 1.1038,
+      "step": 745
+    },
+    {
+      "epoch": 0.9473015873015873,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.252401755562697e-06,
+      "loss": 1.1808,
+      "step": 746
+    },
+    {
+      "epoch": 0.9485714285714286,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.2501718973559155e-06,
+      "loss": 1.1993,
+      "step": 747
+    },
+    {
+      "epoch": 0.9498412698412698,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.247939117489492e-06,
+      "loss": 1.2538,
+      "step": 748
+    },
+    {
+      "epoch": 0.9511111111111111,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.245703418335432e-06,
+      "loss": 1.1869,
+      "step": 749
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.243464802268843e-06,
+      "loss": 1.0689,
+      "step": 750
+    },
+    {
+      "epoch": 0.9536507936507936,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.2412232716679286e-06,
+      "loss": 1.1005,
+      "step": 751
+    },
+    {
+      "epoch": 0.954920634920635,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.238978828913992e-06,
+      "loss": 1.092,
+      "step": 752
+    },
+    {
+      "epoch": 0.9561904761904761,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.236731476391429e-06,
+      "loss": 1.0574,
+      "step": 753
+    },
+    {
+      "epoch": 0.9574603174603175,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.234481216487727e-06,
+      "loss": 1.1339,
+      "step": 754
+    },
+    {
+      "epoch": 0.9587301587301588,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.2322280515934605e-06,
+      "loss": 1.1577,
+      "step": 755
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.229971984102291e-06,
+      "loss": 1.2549,
+      "step": 756
+    },
+    {
+      "epoch": 0.9612698412698413,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.227713016410963e-06,
+      "loss": 1.1451,
+      "step": 757
+    },
+    {
+      "epoch": 0.9625396825396826,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.225451150919305e-06,
+      "loss": 1.191,
+      "step": 758
+    },
+    {
+      "epoch": 0.9638095238095238,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.22318639003022e-06,
+      "loss": 1.1361,
+      "step": 759
+    },
+    {
+      "epoch": 0.9650793650793651,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.2209187361496904e-06,
+      "loss": 1.122,
+      "step": 760
+    },
+    {
+      "epoch": 0.9663492063492064,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.2186481916867684e-06,
+      "loss": 1.0857,
+      "step": 761
+    },
+    {
+      "epoch": 0.9676190476190476,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.216374759053579e-06,
+      "loss": 1.1183,
+      "step": 762
+    },
+    {
+      "epoch": 0.9688888888888889,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.214098440665316e-06,
+      "loss": 1.234,
+      "step": 763
+    },
+    {
+      "epoch": 0.9701587301587301,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.211819238940239e-06,
+      "loss": 1.168,
+      "step": 764
+    },
+    {
+      "epoch": 0.9714285714285714,
+      "grad_norm": 0.953125,
+      "learning_rate": 6.209537156299669e-06,
+      "loss": 1.1411,
+      "step": 765
+    },
+    {
+      "epoch": 0.9726984126984127,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.207252195167988e-06,
+      "loss": 1.1865,
+      "step": 766
+    },
+    {
+      "epoch": 0.9739682539682539,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.204964357972639e-06,
+      "loss": 1.0926,
+      "step": 767
+    },
+    {
+      "epoch": 0.9752380952380952,
+      "grad_norm": 0.953125,
+      "learning_rate": 6.202673647144115e-06,
+      "loss": 1.2023,
+      "step": 768
+    },
+    {
+      "epoch": 0.9765079365079365,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.2003800651159655e-06,
+      "loss": 1.1062,
+      "step": 769
+    },
+    {
+      "epoch": 0.9777777777777777,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.198083614324791e-06,
+      "loss": 1.2015,
+      "step": 770
+    },
+    {
+      "epoch": 0.979047619047619,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.195784297210236e-06,
+      "loss": 1.1814,
+      "step": 771
+    },
+    {
+      "epoch": 0.9803174603174604,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.193482116214995e-06,
+      "loss": 1.155,
+      "step": 772
+    },
+    {
+      "epoch": 0.9815873015873016,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.191177073784799e-06,
+      "loss": 1.103,
+      "step": 773
+    },
+    {
+      "epoch": 0.9828571428571429,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.188869172368424e-06,
+      "loss": 1.0483,
+      "step": 774
+    },
+    {
+      "epoch": 0.9841269841269841,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.18655841441768e-06,
+      "loss": 1.1273,
+      "step": 775
+    },
+    {
+      "epoch": 0.9853968253968254,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.184244802387415e-06,
+      "loss": 1.1635,
+      "step": 776
+    },
+    {
+      "epoch": 0.9866666666666667,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.181928338735503e-06,
+      "loss": 1.1616,
+      "step": 777
+    },
+    {
+      "epoch": 0.9879365079365079,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.179609025922857e-06,
+      "loss": 1.1586,
+      "step": 778
+    },
+    {
+      "epoch": 0.9892063492063492,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.177286866413406e-06,
+      "loss": 1.1268,
+      "step": 779
+    },
+    {
+      "epoch": 0.9904761904761905,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.174961862674111e-06,
+      "loss": 1.1264,
+      "step": 780
+    },
+    {
+      "epoch": 0.9917460317460317,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.172634017174953e-06,
+      "loss": 1.2188,
+      "step": 781
+    },
+    {
+      "epoch": 0.993015873015873,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.1703033323889286e-06,
+      "loss": 1.0856,
+      "step": 782
+    },
+    {
+      "epoch": 0.9942857142857143,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.167969810792053e-06,
+      "loss": 1.142,
+      "step": 783
+    },
+    {
+      "epoch": 0.9955555555555555,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.165633454863357e-06,
+      "loss": 1.1106,
+      "step": 784
+    },
+    {
+      "epoch": 0.9968253968253968,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.163294267084881e-06,
+      "loss": 1.0554,
+      "step": 785
+    },
+    {
+      "epoch": 0.9980952380952381,
+      "grad_norm": 0.875,
+      "learning_rate": 6.160952249941671e-06,
+      "loss": 1.1311,
+      "step": 786
+    },
+    {
+      "epoch": 0.9993650793650793,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.158607405921784e-06,
+      "loss": 1.0738,
+      "step": 787
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.265625,
+      "learning_rate": 6.156259737516275e-06,
+      "loss": 0.5108,
+      "step": 788
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3148,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 394,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.7456931048259584e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}