diff --git "a/checkpoint-1970/trainer_state.json" "b/checkpoint-1970/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1970/trainer_state.json"
@@ -0,0 +1,13839 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.497777777777778,
+  "eval_steps": 500,
+  "global_step": 1970,
+  "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
+    },
+    {
+      "epoch": 1.0006349206349205,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.153909247219202e-06,
+      "loss": 0.6055,
+      "step": 789
+    },
+    {
+      "epoch": 1.0006349206349205,
+      "eval_loss": 1.1893014907836914,
+      "eval_runtime": 99.4944,
+      "eval_samples_per_second": 43.57,
+      "eval_steps_per_second": 5.448,
+      "step": 789
+    },
+    {
+      "epoch": 1.0012698412698413,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.151555937527625e-06,
+      "loss": 1.1263,
+      "step": 790
+    },
+    {
+      "epoch": 1.0025396825396826,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.149199810941588e-06,
+      "loss": 1.1561,
+      "step": 791
+    },
+    {
+      "epoch": 1.0038095238095237,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.146840869964141e-06,
+      "loss": 1.0985,
+      "step": 792
+    },
+    {
+      "epoch": 1.005079365079365,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.144479117101314e-06,
+      "loss": 1.1212,
+      "step": 793
+    },
+    {
+      "epoch": 1.0063492063492063,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.142114554862129e-06,
+      "loss": 1.0649,
+      "step": 794
+    },
+    {
+      "epoch": 1.0076190476190476,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.139747185758589e-06,
+      "loss": 1.0868,
+      "step": 795
+    },
+    {
+      "epoch": 1.008888888888889,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.137377012305684e-06,
+      "loss": 1.105,
+      "step": 796
+    },
+    {
+      "epoch": 1.0101587301587303,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.135004037021378e-06,
+      "loss": 1.0531,
+      "step": 797
+    },
+    {
+      "epoch": 1.0114285714285713,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.1326282624266155e-06,
+      "loss": 1.1147,
+      "step": 798
+    },
+    {
+      "epoch": 1.0126984126984127,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.130249691045313e-06,
+      "loss": 1.1736,
+      "step": 799
+    },
+    {
+      "epoch": 1.013968253968254,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.127868325404357e-06,
+      "loss": 1.1453,
+      "step": 800
+    },
+    {
+      "epoch": 1.0152380952380953,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.1254841680336046e-06,
+      "loss": 1.1401,
+      "step": 801
+    },
+    {
+      "epoch": 1.0165079365079366,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.123097221465878e-06,
+      "loss": 1.1181,
+      "step": 802
+    },
+    {
+      "epoch": 1.0177777777777777,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.120707488236962e-06,
+      "loss": 1.1782,
+      "step": 803
+    },
+    {
+      "epoch": 1.019047619047619,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.118314970885604e-06,
+      "loss": 1.1901,
+      "step": 804
+    },
+    {
+      "epoch": 1.0203174603174603,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.115919671953506e-06,
+      "loss": 1.1247,
+      "step": 805
+    },
+    {
+      "epoch": 1.0215873015873016,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.113521593985326e-06,
+      "loss": 1.2278,
+      "step": 806
+    },
+    {
+      "epoch": 1.022857142857143,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.111120739528676e-06,
+      "loss": 1.1343,
+      "step": 807
+    },
+    {
+      "epoch": 1.0241269841269842,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.108717111134116e-06,
+      "loss": 1.2013,
+      "step": 808
+    },
+    {
+      "epoch": 1.0253968253968253,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.106310711355153e-06,
+      "loss": 1.1276,
+      "step": 809
+    },
+    {
+      "epoch": 1.0266666666666666,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.10390154274824e-06,
+      "loss": 1.1422,
+      "step": 810
+    },
+    {
+      "epoch": 1.027936507936508,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.101489607872768e-06,
+      "loss": 1.0979,
+      "step": 811
+    },
+    {
+      "epoch": 1.0292063492063492,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.099074909291069e-06,
+      "loss": 1.0752,
+      "step": 812
+    },
+    {
+      "epoch": 1.0304761904761905,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.096657449568412e-06,
+      "loss": 1.1214,
+      "step": 813
+    },
+    {
+      "epoch": 1.0317460317460316,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.094237231272998e-06,
+      "loss": 1.1449,
+      "step": 814
+    },
+    {
+      "epoch": 1.033015873015873,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.091814256975957e-06,
+      "loss": 1.153,
+      "step": 815
+    },
+    {
+      "epoch": 1.0342857142857143,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.089388529251351e-06,
+      "loss": 1.159,
+      "step": 816
+    },
+    {
+      "epoch": 1.0355555555555556,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.086960050676163e-06,
+      "loss": 1.1684,
+      "step": 817
+    },
+    {
+      "epoch": 1.0368253968253969,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.084528823830299e-06,
+      "loss": 1.1217,
+      "step": 818
+    },
+    {
+      "epoch": 1.0380952380952382,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.082094851296589e-06,
+      "loss": 1.1292,
+      "step": 819
+    },
+    {
+      "epoch": 1.0393650793650793,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.079658135660774e-06,
+      "loss": 1.1013,
+      "step": 820
+    },
+    {
+      "epoch": 1.0406349206349206,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.077218679511512e-06,
+      "loss": 1.1311,
+      "step": 821
+    },
+    {
+      "epoch": 1.041904761904762,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.074776485440372e-06,
+      "loss": 1.197,
+      "step": 822
+    },
+    {
+      "epoch": 1.0431746031746032,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.072331556041833e-06,
+      "loss": 1.054,
+      "step": 823
+    },
+    {
+      "epoch": 1.0444444444444445,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.0698838939132786e-06,
+      "loss": 1.1083,
+      "step": 824
+    },
+    {
+      "epoch": 1.0457142857142858,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.0674335016549945e-06,
+      "loss": 1.1574,
+      "step": 825
+    },
+    {
+      "epoch": 1.046984126984127,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.064980381870168e-06,
+      "loss": 1.1341,
+      "step": 826
+    },
+    {
+      "epoch": 1.0482539682539682,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.062524537164885e-06,
+      "loss": 1.1482,
+      "step": 827
+    },
+    {
+      "epoch": 1.0495238095238095,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.060065970148123e-06,
+      "loss": 1.0957,
+      "step": 828
+    },
+    {
+      "epoch": 1.0507936507936508,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.057604683431756e-06,
+      "loss": 1.1756,
+      "step": 829
+    },
+    {
+      "epoch": 1.0520634920634921,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.055140679630543e-06,
+      "loss": 1.199,
+      "step": 830
+    },
+    {
+      "epoch": 1.0533333333333332,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.052673961362132e-06,
+      "loss": 1.1214,
+      "step": 831
+    },
+    {
+      "epoch": 1.0546031746031745,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.050204531247056e-06,
+      "loss": 1.1826,
+      "step": 832
+    },
+    {
+      "epoch": 1.0558730158730159,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.047732391908725e-06,
+      "loss": 1.0719,
+      "step": 833
+    },
+    {
+      "epoch": 1.0571428571428572,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.04525754597343e-06,
+      "loss": 1.1604,
+      "step": 834
+    },
+    {
+      "epoch": 1.0584126984126985,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.042779996070335e-06,
+      "loss": 1.1072,
+      "step": 835
+    },
+    {
+      "epoch": 1.0596825396825398,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.040299744831482e-06,
+      "loss": 1.1305,
+      "step": 836
+    },
+    {
+      "epoch": 1.0609523809523809,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.0378167948917755e-06,
+      "loss": 1.0534,
+      "step": 837
+    },
+    {
+      "epoch": 1.0622222222222222,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.035331148888992e-06,
+      "loss": 1.1919,
+      "step": 838
+    },
+    {
+      "epoch": 1.0634920634920635,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.032842809463771e-06,
+      "loss": 1.1266,
+      "step": 839
+    },
+    {
+      "epoch": 1.0647619047619048,
+      "grad_norm": 1.0,
+      "learning_rate": 6.030351779259612e-06,
+      "loss": 1.1957,
+      "step": 840
+    },
+    {
+      "epoch": 1.066031746031746,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.027858060922874e-06,
+      "loss": 1.1445,
+      "step": 841
+    },
+    {
+      "epoch": 1.0673015873015874,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.025361657102773e-06,
+      "loss": 1.0895,
+      "step": 842
+    },
+    {
+      "epoch": 1.0685714285714285,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.022862570451376e-06,
+      "loss": 1.2063,
+      "step": 843
+    },
+    {
+      "epoch": 1.0698412698412698,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.020360803623601e-06,
+      "loss": 1.0889,
+      "step": 844
+    },
+    {
+      "epoch": 1.0711111111111111,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.017856359277215e-06,
+      "loss": 1.138,
+      "step": 845
+    },
+    {
+      "epoch": 1.0723809523809524,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.015349240072824e-06,
+      "loss": 1.1788,
+      "step": 846
+    },
+    {
+      "epoch": 1.0736507936507937,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.012839448673882e-06,
+      "loss": 1.1004,
+      "step": 847
+    },
+    {
+      "epoch": 1.0749206349206348,
+      "grad_norm": 1.0,
+      "learning_rate": 6.010326987746679e-06,
+      "loss": 1.1197,
+      "step": 848
+    },
+    {
+      "epoch": 1.0761904761904761,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.007811859960339e-06,
+      "loss": 1.0873,
+      "step": 849
+    },
+    {
+      "epoch": 1.0774603174603175,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.005294067986824e-06,
+      "loss": 1.1637,
+      "step": 850
+    },
+    {
+      "epoch": 1.0787301587301588,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.002773614500922e-06,
+      "loss": 1.1315,
+      "step": 851
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.000250502180251e-06,
+      "loss": 1.1133,
+      "step": 852
+    },
+    {
+      "epoch": 1.0812698412698412,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.9977247337052515e-06,
+      "loss": 1.1943,
+      "step": 853
+    },
+    {
+      "epoch": 1.0825396825396825,
+      "grad_norm": 0.984375,
+      "learning_rate": 5.995196311759189e-06,
+      "loss": 1.133,
+      "step": 854
+    },
+    {
+      "epoch": 1.0838095238095238,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.992665239028144e-06,
+      "loss": 1.1914,
+      "step": 855
+    },
+    {
+      "epoch": 1.085079365079365,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.990131518201016e-06,
+      "loss": 1.1349,
+      "step": 856
+    },
+    {
+      "epoch": 1.0863492063492064,
+      "grad_norm": 0.9609375,
+      "learning_rate": 5.987595151969516e-06,
+      "loss": 1.1804,
+      "step": 857
+    },
+    {
+      "epoch": 1.0876190476190477,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.985056143028167e-06,
+      "loss": 1.1553,
+      "step": 858
+    },
+    {
+      "epoch": 1.0888888888888888,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.982514494074297e-06,
+      "loss": 1.0719,
+      "step": 859
+    },
+    {
+      "epoch": 1.09015873015873,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.979970207808042e-06,
+      "loss": 1.1532,
+      "step": 860
+    },
+    {
+      "epoch": 1.0914285714285714,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.9774232869323365e-06,
+      "loss": 1.0941,
+      "step": 861
+    },
+    {
+      "epoch": 1.0926984126984127,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.974873734152916e-06,
+      "loss": 1.1314,
+      "step": 862
+    },
+    {
+      "epoch": 1.093968253968254,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.972321552178312e-06,
+      "loss": 1.165,
+      "step": 863
+    },
+    {
+      "epoch": 1.0952380952380953,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.969766743719847e-06,
+      "loss": 1.1825,
+      "step": 864
+    },
+    {
+      "epoch": 1.0965079365079364,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.967209311491636e-06,
+      "loss": 1.1202,
+      "step": 865
+    },
+    {
+      "epoch": 1.0977777777777777,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.964649258210583e-06,
+      "loss": 1.139,
+      "step": 866
+    },
+    {
+      "epoch": 1.099047619047619,
+      "grad_norm": 0.8828125,
+      "learning_rate": 5.962086586596369e-06,
+      "loss": 1.0856,
+      "step": 867
+    },
+    {
+      "epoch": 1.1003174603174604,
+      "grad_norm": 0.9609375,
+      "learning_rate": 5.959521299371465e-06,
+      "loss": 1.1494,
+      "step": 868
+    },
+    {
+      "epoch": 1.1015873015873017,
+      "grad_norm": 0.91796875,
+      "learning_rate": 5.956953399261118e-06,
+      "loss": 1.0633,
+      "step": 869
+    },
+    {
+      "epoch": 1.1028571428571428,
+      "grad_norm": 0.90234375,
+      "learning_rate": 5.9543828889933484e-06,
+      "loss": 1.1601,
+      "step": 870
+    },
+    {
+      "epoch": 1.104126984126984,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.951809771298954e-06,
+      "loss": 1.1452,
+      "step": 871
+    },
+    {
+      "epoch": 1.1053968253968254,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.949234048911497e-06,
+      "loss": 1.0889,
+      "step": 872
+    },
+    {
+      "epoch": 1.1066666666666667,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.94665572456731e-06,
+      "loss": 1.1215,
+      "step": 873
+    },
+    {
+      "epoch": 1.107936507936508,
+      "grad_norm": 1.0,
+      "learning_rate": 5.94407480100549e-06,
+      "loss": 1.1259,
+      "step": 874
+    },
+    {
+      "epoch": 1.1092063492063493,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.941491280967898e-06,
+      "loss": 1.0588,
+      "step": 875
+    },
+    {
+      "epoch": 1.1104761904761904,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.938905167199145e-06,
+      "loss": 1.1346,
+      "step": 876
+    },
+    {
+      "epoch": 1.1117460317460317,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.936316462446605e-06,
+      "loss": 1.0855,
+      "step": 877
+    },
+    {
+      "epoch": 1.113015873015873,
+      "grad_norm": 0.93359375,
+      "learning_rate": 5.933725169460404e-06,
+      "loss": 1.1648,
+      "step": 878
+    },
+    {
+      "epoch": 1.1142857142857143,
+      "grad_norm": 0.9609375,
+      "learning_rate": 5.931131290993411e-06,
+      "loss": 1.1728,
+      "step": 879
+    },
+    {
+      "epoch": 1.1155555555555556,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.928534829801252e-06,
+      "loss": 1.082,
+      "step": 880
+    },
+    {
+      "epoch": 1.116825396825397,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.9259357886422866e-06,
+      "loss": 1.1417,
+      "step": 881
+    },
+    {
+      "epoch": 1.118095238095238,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.923334170277623e-06,
+      "loss": 1.1807,
+      "step": 882
+    },
+    {
+      "epoch": 1.1193650793650793,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.9207299774711025e-06,
+      "loss": 1.1927,
+      "step": 883
+    },
+    {
+      "epoch": 1.1206349206349207,
+      "grad_norm": 0.9609375,
+      "learning_rate": 5.918123212989304e-06,
+      "loss": 1.1026,
+      "step": 884
+    },
+    {
+      "epoch": 1.121904761904762,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.915513879601538e-06,
+      "loss": 1.1803,
+      "step": 885
+    },
+    {
+      "epoch": 1.1231746031746033,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.912901980079843e-06,
+      "loss": 1.1667,
+      "step": 886
+    },
+    {
+      "epoch": 1.1244444444444444,
+      "grad_norm": 0.93359375,
+      "learning_rate": 5.910287517198983e-06,
+      "loss": 1.1347,
+      "step": 887
+    },
+    {
+      "epoch": 1.1257142857142857,
+      "grad_norm": 0.91796875,
+      "learning_rate": 5.9076704937364475e-06,
+      "loss": 1.1578,
+      "step": 888
+    },
+    {
+      "epoch": 1.126984126984127,
+      "grad_norm": 0.96875,
+      "learning_rate": 5.9050509124724456e-06,
+      "loss": 1.1774,
+      "step": 889
+    },
+    {
+      "epoch": 1.1282539682539683,
+      "grad_norm": 0.93359375,
+      "learning_rate": 5.902428776189903e-06,
+      "loss": 1.1567,
+      "step": 890
+    },
+    {
+      "epoch": 1.1295238095238096,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.899804087674461e-06,
+      "loss": 1.1014,
+      "step": 891
+    },
+    {
+      "epoch": 1.1307936507936507,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.897176849714469e-06,
+      "loss": 1.1083,
+      "step": 892
+    },
+    {
+      "epoch": 1.132063492063492,
+      "grad_norm": 0.9296875,
+      "learning_rate": 5.8945470651009875e-06,
+      "loss": 1.1097,
+      "step": 893
+    },
+    {
+      "epoch": 1.1333333333333333,
+      "grad_norm": 0.9453125,
+      "learning_rate": 5.891914736627785e-06,
+      "loss": 1.096,
+      "step": 894
+    },
+    {
+      "epoch": 1.1346031746031746,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.8892798670913275e-06,
+      "loss": 1.1589,
+      "step": 895
+    },
+    {
+      "epoch": 1.135873015873016,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.886642459290782e-06,
+      "loss": 1.1019,
+      "step": 896
+    },
+    {
+      "epoch": 1.1371428571428572,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.884002516028012e-06,
+      "loss": 1.1257,
+      "step": 897
+    },
+    {
+      "epoch": 1.1384126984126983,
+      "grad_norm": 0.9140625,
+      "learning_rate": 5.881360040107578e-06,
+      "loss": 1.1422,
+      "step": 898
+    },
+    {
+      "epoch": 1.1396825396825396,
+      "grad_norm": 0.90625,
+      "learning_rate": 5.878715034336726e-06,
+      "loss": 1.1903,
+      "step": 899
+    },
+    {
+      "epoch": 1.140952380952381,
+      "grad_norm": 0.92578125,
+      "learning_rate": 5.8760675015253935e-06,
+      "loss": 1.1277,
+      "step": 900
+    },
+    {
+      "epoch": 1.1422222222222222,
+      "grad_norm": 0.9453125,
+      "learning_rate": 5.8734174444862e-06,
+      "loss": 1.1103,
+      "step": 901
+    },
+    {
+      "epoch": 1.1434920634920636,
+      "grad_norm": 0.9296875,
+      "learning_rate": 5.8707648660344454e-06,
+      "loss": 1.1382,
+      "step": 902
+    },
+    {
+      "epoch": 1.1447619047619049,
+      "grad_norm": 0.91015625,
+      "learning_rate": 5.868109768988114e-06,
+      "loss": 1.1512,
+      "step": 903
+    },
+    {
+      "epoch": 1.146031746031746,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.865452156167861e-06,
+      "loss": 1.1509,
+      "step": 904
+    },
+    {
+      "epoch": 1.1473015873015873,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.862792030397015e-06,
+      "loss": 1.1152,
+      "step": 905
+    },
+    {
+      "epoch": 1.1485714285714286,
+      "grad_norm": 0.91015625,
+      "learning_rate": 5.8601293945015745e-06,
+      "loss": 1.142,
+      "step": 906
+    },
+    {
+      "epoch": 1.1498412698412699,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.857464251310206e-06,
+      "loss": 1.0839,
+      "step": 907
+    },
+    {
+      "epoch": 1.1511111111111112,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.854796603654236e-06,
+      "loss": 1.2105,
+      "step": 908
+    },
+    {
+      "epoch": 1.1523809523809523,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.852126454367657e-06,
+      "loss": 1.1147,
+      "step": 909
+    },
+    {
+      "epoch": 1.1536507936507936,
+      "grad_norm": 0.90234375,
+      "learning_rate": 5.8494538062871145e-06,
+      "loss": 1.1292,
+      "step": 910
+    },
+    {
+      "epoch": 1.154920634920635,
+      "grad_norm": 0.91015625,
+      "learning_rate": 5.8467786622519105e-06,
+      "loss": 1.1353,
+      "step": 911
+    },
+    {
+      "epoch": 1.1561904761904762,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.844101025103999e-06,
+      "loss": 1.0897,
+      "step": 912
+    },
+    {
+      "epoch": 1.1574603174603175,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.841420897687981e-06,
+      "loss": 1.1639,
+      "step": 913
+    },
+    {
+      "epoch": 1.1587301587301586,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.838738282851105e-06,
+      "loss": 1.1634,
+      "step": 914
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.953125,
+      "learning_rate": 5.836053183443262e-06,
+      "loss": 1.1569,
+      "step": 915
+    },
+    {
+      "epoch": 1.1612698412698412,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.833365602316978e-06,
+      "loss": 1.2137,
+      "step": 916
+    },
+    {
+      "epoch": 1.1625396825396825,
+      "grad_norm": 0.8671875,
+      "learning_rate": 5.830675542327421e-06,
+      "loss": 1.111,
+      "step": 917
+    },
+    {
+      "epoch": 1.1638095238095238,
+      "grad_norm": 0.90234375,
+      "learning_rate": 5.827983006332391e-06,
+      "loss": 1.0904,
+      "step": 918
+    },
+    {
+      "epoch": 1.1650793650793652,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.825287997192318e-06,
+      "loss": 1.1414,
+      "step": 919
+    },
+    {
+      "epoch": 1.1663492063492065,
+      "grad_norm": 0.8671875,
+      "learning_rate": 5.822590517770259e-06,
+      "loss": 1.1094,
+      "step": 920
+    },
+    {
+      "epoch": 1.1676190476190476,
+      "grad_norm": 0.92578125,
+      "learning_rate": 5.819890570931895e-06,
+      "loss": 1.128,
+      "step": 921
+    },
+    {
+      "epoch": 1.1688888888888889,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.817188159545529e-06,
+      "loss": 1.1724,
+      "step": 922
+    },
+    {
+      "epoch": 1.1701587301587302,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.814483286482081e-06,
+      "loss": 1.0582,
+      "step": 923
+    },
+    {
+      "epoch": 1.1714285714285715,
+      "grad_norm": 0.91796875,
+      "learning_rate": 5.811775954615088e-06,
+      "loss": 1.0837,
+      "step": 924
+    },
+    {
+      "epoch": 1.1726984126984128,
+      "grad_norm": 0.89453125,
+      "learning_rate": 5.809066166820699e-06,
+      "loss": 1.0938,
+      "step": 925
+    },
+    {
+      "epoch": 1.1739682539682539,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.806353925977671e-06,
+      "loss": 1.1441,
+      "step": 926
+    },
+    {
+      "epoch": 1.1752380952380952,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.803639234967367e-06,
+      "loss": 1.1908,
+      "step": 927
+    },
+    {
+      "epoch": 1.1765079365079365,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.800922096673753e-06,
+      "loss": 1.1109,
+      "step": 928
+    },
+    {
+      "epoch": 1.1777777777777778,
+      "grad_norm": 0.88671875,
+      "learning_rate": 5.798202513983397e-06,
+      "loss": 1.1131,
+      "step": 929
+    },
+    {
+      "epoch": 1.1790476190476191,
+      "grad_norm": 0.91796875,
+      "learning_rate": 5.79548048978546e-06,
+      "loss": 1.1502,
+      "step": 930
+    },
+    {
+      "epoch": 1.1803174603174602,
+      "grad_norm": 0.96875,
+      "learning_rate": 5.7927560269717e-06,
+      "loss": 1.1954,
+      "step": 931
+    },
+    {
+      "epoch": 1.1815873015873015,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.790029128436464e-06,
+      "loss": 1.1165,
+      "step": 932
+    },
+    {
+      "epoch": 1.1828571428571428,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.787299797076686e-06,
+      "loss": 1.1306,
+      "step": 933
+    },
+    {
+      "epoch": 1.1841269841269841,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.784568035791887e-06,
+      "loss": 1.1082,
+      "step": 934
+    },
+    {
+      "epoch": 1.1853968253968254,
+      "grad_norm": 0.9296875,
+      "learning_rate": 5.781833847484169e-06,
+      "loss": 1.1388,
+      "step": 935
+    },
+    {
+      "epoch": 1.1866666666666668,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.7790972350582095e-06,
+      "loss": 1.1739,
+      "step": 936
+    },
+    {
+      "epoch": 1.1879365079365078,
+      "grad_norm": 0.93359375,
+      "learning_rate": 5.776358201421263e-06,
+      "loss": 1.1154,
+      "step": 937
+    },
+    {
+      "epoch": 1.1892063492063492,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.773616749483157e-06,
+      "loss": 1.1866,
+      "step": 938
+    },
+    {
+      "epoch": 1.1904761904761905,
+      "grad_norm": 0.9609375,
+      "learning_rate": 5.770872882156288e-06,
+      "loss": 1.2014,
+      "step": 939
+    },
+    {
+      "epoch": 1.1917460317460318,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.768126602355618e-06,
+      "loss": 1.0875,
+      "step": 940
+    },
+    {
+      "epoch": 1.193015873015873,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.765377912998671e-06,
+      "loss": 1.1382,
+      "step": 941
+    },
+    {
+      "epoch": 1.1942857142857144,
+      "grad_norm": 0.890625,
+      "learning_rate": 5.762626817005533e-06,
+      "loss": 1.0788,
+      "step": 942
+    },
+    {
+      "epoch": 1.1955555555555555,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.759873317298842e-06,
+      "loss": 1.1058,
+      "step": 943
+    },
+    {
+      "epoch": 1.1968253968253968,
+      "grad_norm": 0.96875,
+      "learning_rate": 5.757117416803798e-06,
+      "loss": 1.1358,
+      "step": 944
+    },
+    {
+      "epoch": 1.198095238095238,
+      "grad_norm": 0.9609375,
+      "learning_rate": 5.754359118448141e-06,
+      "loss": 1.1381,
+      "step": 945
+    },
+    {
+      "epoch": 1.1993650793650794,
+      "grad_norm": 0.9140625,
+      "learning_rate": 5.751598425162169e-06,
+      "loss": 1.2095,
+      "step": 946
+    },
+    {
+      "epoch": 1.2006349206349207,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.7488353398787165e-06,
+      "loss": 1.1923,
+      "step": 947
+    },
+    {
+      "epoch": 1.2019047619047618,
+      "grad_norm": 0.90625,
+      "learning_rate": 5.746069865533161e-06,
+      "loss": 1.0759,
+      "step": 948
+    },
+    {
+      "epoch": 1.2031746031746031,
+      "grad_norm": 0.87109375,
+      "learning_rate": 5.743302005063419e-06,
+      "loss": 1.0627,
+      "step": 949
+    },
+    {
+      "epoch": 1.2044444444444444,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.740531761409945e-06,
+      "loss": 1.1598,
+      "step": 950
+    },
+    {
+      "epoch": 1.2057142857142857,
+      "grad_norm": 0.8671875,
+      "learning_rate": 5.7377591375157175e-06,
+      "loss": 1.1227,
+      "step": 951
+    },
+    {
+      "epoch": 1.206984126984127,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.73498413632625e-06,
+      "loss": 1.1589,
+      "step": 952
+    },
+    {
+      "epoch": 1.2082539682539681,
+      "grad_norm": 0.90234375,
+      "learning_rate": 5.73220676078958e-06,
+      "loss": 1.0832,
+      "step": 953
+    },
+    {
+      "epoch": 1.2095238095238094,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.7294270138562675e-06,
+      "loss": 1.2069,
+      "step": 954
+    },
+    {
+      "epoch": 1.2107936507936508,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.7266448984793905e-06,
+      "loss": 1.2382,
+      "step": 955
+    },
+    {
+      "epoch": 1.212063492063492,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.723860417614543e-06,
+      "loss": 1.1499,
+      "step": 956
+    },
+    {
+      "epoch": 1.2133333333333334,
+      "grad_norm": 0.92578125,
+      "learning_rate": 5.721073574219835e-06,
+      "loss": 1.1811,
+      "step": 957
+    },
+    {
+      "epoch": 1.2146031746031747,
+      "grad_norm": 0.87109375,
+      "learning_rate": 5.718284371255883e-06,
+      "loss": 1.0129,
+      "step": 958
+    },
+    {
+      "epoch": 1.215873015873016,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.715492811685813e-06,
+      "loss": 1.1847,
+      "step": 959
+    },
+    {
+      "epoch": 1.217142857142857,
+      "grad_norm": 0.85546875,
+      "learning_rate": 5.712698898475249e-06,
+      "loss": 1.0698,
+      "step": 960
+    },
+    {
+      "epoch": 1.2184126984126984,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.709902634592324e-06,
+      "loss": 1.0597,
+      "step": 961
+    },
+    {
+      "epoch": 1.2196825396825397,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.707104023007664e-06,
+      "loss": 1.0922,
+      "step": 962
+    },
+    {
+      "epoch": 1.220952380952381,
+      "grad_norm": 0.90625,
+      "learning_rate": 5.7043030666943846e-06,
+      "loss": 1.1159,
+      "step": 963
+    },
+    {
+      "epoch": 1.2222222222222223,
+      "grad_norm": 0.90625,
+      "learning_rate": 5.701499768628099e-06,
+      "loss": 1.1535,
+      "step": 964
+    },
+    {
+      "epoch": 1.2234920634920634,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.6986941317869065e-06,
+      "loss": 1.0798,
+      "step": 965
+    },
+    {
+      "epoch": 1.2247619047619047,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.695886159151388e-06,
+      "loss": 1.1546,
+      "step": 966
+    },
+    {
+      "epoch": 1.226031746031746,
+      "grad_norm": 0.9140625,
+      "learning_rate": 5.69307585370461e-06,
+      "loss": 1.0339,
+      "step": 967
+    },
+    {
+      "epoch": 1.2273015873015873,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.690263218432114e-06,
+      "loss": 1.1029,
+      "step": 968
+    },
+    {
+      "epoch": 1.2285714285714286,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.687448256321919e-06,
+      "loss": 1.0723,
+      "step": 969
+    },
+    {
+      "epoch": 1.2298412698412697,
+      "grad_norm": 0.96875,
+      "learning_rate": 5.684630970364515e-06,
+      "loss": 1.0463,
+      "step": 970
+    },
+    {
+      "epoch": 1.231111111111111,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.6818113635528595e-06,
+      "loss": 1.1948,
+      "step": 971
+    },
+    {
+      "epoch": 1.2323809523809524,
+      "grad_norm": 1.0,
+      "learning_rate": 5.6789894388823764e-06,
+      "loss": 1.1977,
+      "step": 972
+    },
+    {
+      "epoch": 1.2336507936507937,
+      "grad_norm": 0.875,
+      "learning_rate": 5.676165199350954e-06,
+      "loss": 1.0944,
+      "step": 973
+    },
+    {
+      "epoch": 1.234920634920635,
+      "grad_norm": 0.91796875,
+      "learning_rate": 5.6733386479589375e-06,
+      "loss": 1.1806,
+      "step": 974
+    },
+    {
+      "epoch": 1.2361904761904763,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.670509787709128e-06,
+      "loss": 1.169,
+      "step": 975
+    },
+    {
+      "epoch": 1.2374603174603174,
+      "grad_norm": 0.92578125,
+      "learning_rate": 5.667678621606781e-06,
+      "loss": 1.161,
+      "step": 976
+    },
+    {
+      "epoch": 1.2387301587301587,
+      "grad_norm": 0.93359375,
+      "learning_rate": 5.664845152659599e-06,
+      "loss": 1.1933,
+      "step": 977
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.93359375,
+      "learning_rate": 5.6620093838777345e-06,
+      "loss": 1.1191,
+      "step": 978
+    },
+    {
+      "epoch": 1.2412698412698413,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.659171318273782e-06,
+      "loss": 1.1189,
+      "step": 979
+    },
+    {
+      "epoch": 1.2425396825396826,
+      "grad_norm": 0.96875,
+      "learning_rate": 5.656330958862774e-06,
+      "loss": 1.0603,
+      "step": 980
+    },
+    {
+      "epoch": 1.243809523809524,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.653488308662181e-06,
+      "loss": 1.1747,
+      "step": 981
+    },
+    {
+      "epoch": 1.245079365079365,
+      "grad_norm": 0.984375,
+      "learning_rate": 5.650643370691909e-06,
+      "loss": 1.1333,
+      "step": 982
+    },
+    {
+      "epoch": 1.2463492063492063,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.647796147974291e-06,
+      "loss": 1.166,
+      "step": 983
+    },
+    {
+      "epoch": 1.2476190476190476,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.644946643534091e-06,
+      "loss": 1.1551,
+      "step": 984
+    },
+    {
+      "epoch": 1.248888888888889,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.642094860398495e-06,
+      "loss": 1.1008,
+      "step": 985
+    },
+    {
+      "epoch": 1.2501587301587302,
+      "grad_norm": 0.9296875,
+      "learning_rate": 5.639240801597108e-06,
+      "loss": 1.0894,
+      "step": 986
+    },
+    {
+      "epoch": 1.2514285714285713,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.636384470161957e-06,
+      "loss": 1.1374,
+      "step": 987
+    },
+    {
+      "epoch": 1.2526984126984126,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.633525869127478e-06,
+      "loss": 1.1105,
+      "step": 988
+    },
+    {
+      "epoch": 1.253968253968254,
+      "grad_norm": 0.9296875,
+      "learning_rate": 5.630665001530522e-06,
+      "loss": 1.0807,
+      "step": 989
+    },
+    {
+      "epoch": 1.2552380952380953,
+      "grad_norm": 0.90625,
+      "learning_rate": 5.627801870410348e-06,
+      "loss": 1.1284,
+      "step": 990
+    },
+    {
+      "epoch": 1.2565079365079366,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.624936478808617e-06,
+      "loss": 1.1448,
+      "step": 991
+    },
+    {
+      "epoch": 1.2577777777777777,
+      "grad_norm": 0.87890625,
+      "learning_rate": 5.622068829769393e-06,
+      "loss": 1.0997,
+      "step": 992
+    },
+    {
+      "epoch": 1.259047619047619,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.61919892633914e-06,
+      "loss": 1.1034,
+      "step": 993
+    },
+    {
+      "epoch": 1.2603174603174603,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.616326771566714e-06,
+      "loss": 1.1347,
+      "step": 994
+    },
+    {
+      "epoch": 1.2615873015873016,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.613452368503362e-06,
+      "loss": 1.0823,
+      "step": 995
+    },
+    {
+      "epoch": 1.262857142857143,
+      "grad_norm": 0.87890625,
+      "learning_rate": 5.610575720202723e-06,
+      "loss": 1.0923,
+      "step": 996
+    },
+    {
+      "epoch": 1.2641269841269842,
+      "grad_norm": 0.90234375,
+      "learning_rate": 5.60769682972082e-06,
+      "loss": 1.0534,
+      "step": 997
+    },
+    {
+      "epoch": 1.2653968253968255,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.604815700116055e-06,
+      "loss": 1.1278,
+      "step": 998
+    },
+    {
+      "epoch": 1.2666666666666666,
+      "grad_norm": 0.8828125,
+      "learning_rate": 5.601932334449214e-06,
+      "loss": 1.037,
+      "step": 999
+    },
+    {
+      "epoch": 1.267936507936508,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.599046735783456e-06,
+      "loss": 1.087,
+      "step": 1000
+    },
+    {
+      "epoch": 1.2692063492063492,
+      "grad_norm": 0.984375,
+      "learning_rate": 5.596158907184309e-06,
+      "loss": 1.0948,
+      "step": 1001
+    },
+    {
+      "epoch": 1.2704761904761905,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.593268851719676e-06,
+      "loss": 1.1269,
+      "step": 1002
+    },
+    {
+      "epoch": 1.2717460317460318,
+      "grad_norm": 0.953125,
+      "learning_rate": 5.590376572459821e-06,
+      "loss": 1.1055,
+      "step": 1003
+    },
+    {
+      "epoch": 1.273015873015873,
+      "grad_norm": 0.93359375,
+      "learning_rate": 5.587482072477371e-06,
+      "loss": 1.0811,
+      "step": 1004
+    },
+    {
+      "epoch": 1.2742857142857142,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.584585354847317e-06,
+      "loss": 1.0866,
+      "step": 1005
+    },
+    {
+      "epoch": 1.2755555555555556,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.581686422646998e-06,
+      "loss": 1.0588,
+      "step": 1006
+    },
+    {
+      "epoch": 1.2768253968253969,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.578785278956114e-06,
+      "loss": 1.0968,
+      "step": 1007
+    },
+    {
+      "epoch": 1.2780952380952382,
+      "grad_norm": 0.90625,
+      "learning_rate": 5.575881926856708e-06,
+      "loss": 1.0973,
+      "step": 1008
+    },
+    {
+      "epoch": 1.2793650793650793,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.572976369433172e-06,
+      "loss": 1.1661,
+      "step": 1009
+    },
+    {
+      "epoch": 1.2806349206349206,
+      "grad_norm": 0.953125,
+      "learning_rate": 5.5700686097722384e-06,
+      "loss": 1.1793,
+      "step": 1010
+    },
+    {
+      "epoch": 1.2819047619047619,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.5671586509629845e-06,
+      "loss": 1.1664,
+      "step": 1011
+    },
+    {
+      "epoch": 1.2831746031746032,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.564246496096818e-06,
+      "loss": 1.1553,
+      "step": 1012
+    },
+    {
+      "epoch": 1.2844444444444445,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.5613321482674834e-06,
+      "loss": 1.0554,
+      "step": 1013
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.558415610571052e-06,
+      "loss": 1.1176,
+      "step": 1014
+    },
+    {
+      "epoch": 1.2869841269841271,
+      "grad_norm": 0.9296875,
+      "learning_rate": 5.555496886105924e-06,
+      "loss": 1.1676,
+      "step": 1015
+    },
+    {
+      "epoch": 1.2882539682539682,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.552575977972822e-06,
+      "loss": 1.1585,
+      "step": 1016
+    },
+    {
+      "epoch": 1.2895238095238095,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.549652889274788e-06,
+      "loss": 1.127,
+      "step": 1017
+    },
+    {
+      "epoch": 1.2907936507936508,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.546727623117179e-06,
+      "loss": 1.2112,
+      "step": 1018
+    },
+    {
+      "epoch": 1.2920634920634921,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.543800182607669e-06,
+      "loss": 1.1219,
+      "step": 1019
+    },
+    {
+      "epoch": 1.2933333333333334,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.54087057085624e-06,
+      "loss": 1.2062,
+      "step": 1020
+    },
+    {
+      "epoch": 1.2946031746031745,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.537938790975179e-06,
+      "loss": 1.1131,
+      "step": 1021
+    },
+    {
+      "epoch": 1.2958730158730158,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.535004846079079e-06,
+      "loss": 1.0721,
+      "step": 1022
+    },
+    {
+      "epoch": 1.2971428571428572,
+      "grad_norm": 0.98828125,
+      "learning_rate": 5.532068739284832e-06,
+      "loss": 1.1498,
+      "step": 1023
+    },
+    {
+      "epoch": 1.2984126984126985,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.5291304737116254e-06,
+      "loss": 1.1301,
+      "step": 1024
+    },
+    {
+      "epoch": 1.2996825396825398,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.526190052480942e-06,
+      "loss": 1.1593,
+      "step": 1025
+    },
+    {
+      "epoch": 1.3009523809523809,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.523247478716555e-06,
+      "loss": 1.0929,
+      "step": 1026
+    },
+    {
+      "epoch": 1.3022222222222222,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.5203027555445205e-06,
+      "loss": 1.1383,
+      "step": 1027
+    },
+    {
+      "epoch": 1.3034920634920635,
+      "grad_norm": 0.9296875,
+      "learning_rate": 5.517355886093182e-06,
+      "loss": 1.1199,
+      "step": 1028
+    },
+    {
+      "epoch": 1.3047619047619048,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.514406873493163e-06,
+      "loss": 1.0592,
+      "step": 1029
+    },
+    {
+      "epoch": 1.306031746031746,
+      "grad_norm": 0.8359375,
+      "learning_rate": 5.511455720877359e-06,
+      "loss": 1.0237,
+      "step": 1030
+    },
+    {
+      "epoch": 1.3073015873015872,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.508502431380948e-06,
+      "loss": 1.1415,
+      "step": 1031
+    },
+    {
+      "epoch": 1.3085714285714285,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.5055470081413684e-06,
+      "loss": 1.1732,
+      "step": 1032
+    },
+    {
+      "epoch": 1.3098412698412698,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.5025894542983315e-06,
+      "loss": 1.1345,
+      "step": 1033
+    },
+    {
+      "epoch": 1.3111111111111111,
+      "grad_norm": 1.0,
+      "learning_rate": 5.499629772993809e-06,
+      "loss": 1.0832,
+      "step": 1034
+    },
+    {
+      "epoch": 1.3123809523809524,
+      "grad_norm": 0.99609375,
+      "learning_rate": 5.496667967372035e-06,
+      "loss": 1.2329,
+      "step": 1035
+    },
+    {
+      "epoch": 1.3136507936507935,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.4937040405795e-06,
+      "loss": 1.1604,
+      "step": 1036
+    },
+    {
+      "epoch": 1.314920634920635,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.490737995764946e-06,
+      "loss": 1.133,
+      "step": 1037
+    },
+    {
+      "epoch": 1.3161904761904761,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.487769836079367e-06,
+      "loss": 1.0956,
+      "step": 1038
+    },
+    {
+      "epoch": 1.3174603174603174,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.484799564676002e-06,
+      "loss": 1.0835,
+      "step": 1039
+    },
+    {
+      "epoch": 1.3187301587301588,
+      "grad_norm": 0.88671875,
+      "learning_rate": 5.481827184710336e-06,
+      "loss": 1.0909,
+      "step": 1040
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.9296875,
+      "learning_rate": 5.478852699340092e-06,
+      "loss": 1.0682,
+      "step": 1041
+    },
+    {
+      "epoch": 1.3212698412698414,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.47587611172523e-06,
+      "loss": 1.1619,
+      "step": 1042
+    },
+    {
+      "epoch": 1.3225396825396825,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.4728974250279445e-06,
+      "loss": 1.1885,
+      "step": 1043
+    },
+    {
+      "epoch": 1.3238095238095238,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.4699166424126575e-06,
+      "loss": 1.1089,
+      "step": 1044
+    },
+    {
+      "epoch": 1.325079365079365,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.4669337670460205e-06,
+      "loss": 1.1159,
+      "step": 1045
+    },
+    {
+      "epoch": 1.3263492063492064,
+      "grad_norm": 0.9453125,
+      "learning_rate": 5.463948802096907e-06,
+      "loss": 1.153,
+      "step": 1046
+    },
+    {
+      "epoch": 1.3276190476190477,
+      "grad_norm": 0.99609375,
+      "learning_rate": 5.460961750736408e-06,
+      "loss": 1.0732,
+      "step": 1047
+    },
+    {
+      "epoch": 1.3288888888888888,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.457972616137836e-06,
+      "loss": 1.0756,
+      "step": 1048
+    },
+    {
+      "epoch": 1.33015873015873,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.454981401476713e-06,
+      "loss": 1.1332,
+      "step": 1049
+    },
+    {
+      "epoch": 1.3314285714285714,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.451988109930771e-06,
+      "loss": 1.1398,
+      "step": 1050
+    },
+    {
+      "epoch": 1.3326984126984127,
+      "grad_norm": 0.91796875,
+      "learning_rate": 5.44899274467995e-06,
+      "loss": 1.1576,
+      "step": 1051
+    },
+    {
+      "epoch": 1.333968253968254,
+      "grad_norm": 0.9296875,
+      "learning_rate": 5.4459953089063925e-06,
+      "loss": 1.0711,
+      "step": 1052
+    },
+    {
+      "epoch": 1.3352380952380951,
+      "grad_norm": 0.9453125,
+      "learning_rate": 5.442995805794438e-06,
+      "loss": 1.1551,
+      "step": 1053
+    },
+    {
+      "epoch": 1.3365079365079366,
+      "grad_norm": 0.9140625,
+      "learning_rate": 5.439994238530627e-06,
+      "loss": 1.0791,
+      "step": 1054
+    },
+    {
+      "epoch": 1.3377777777777777,
+      "grad_norm": 0.98828125,
+      "learning_rate": 5.4369906103036895e-06,
+      "loss": 1.2093,
+      "step": 1055
+    },
+    {
+      "epoch": 1.339047619047619,
+      "grad_norm": 0.92578125,
+      "learning_rate": 5.433984924304546e-06,
+      "loss": 1.1535,
+      "step": 1056
+    },
+    {
+      "epoch": 1.3403174603174604,
+      "grad_norm": 0.9609375,
+      "learning_rate": 5.430977183726302e-06,
+      "loss": 1.113,
+      "step": 1057
+    },
+    {
+      "epoch": 1.3415873015873017,
+      "grad_norm": 0.90625,
+      "learning_rate": 5.427967391764247e-06,
+      "loss": 1.1166,
+      "step": 1058
+    },
+    {
+      "epoch": 1.342857142857143,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.42495555161585e-06,
+      "loss": 1.2128,
+      "step": 1059
+    },
+    {
+      "epoch": 1.344126984126984,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.421941666480755e-06,
+      "loss": 1.0608,
+      "step": 1060
+    },
+    {
+      "epoch": 1.3453968253968254,
+      "grad_norm": 0.984375,
+      "learning_rate": 5.418925739560777e-06,
+      "loss": 1.1802,
+      "step": 1061
+    },
+    {
+      "epoch": 1.3466666666666667,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.415907774059904e-06,
+      "loss": 1.0853,
+      "step": 1062
+    },
+    {
+      "epoch": 1.347936507936508,
+      "grad_norm": 0.91796875,
+      "learning_rate": 5.412887773184288e-06,
+      "loss": 1.1111,
+      "step": 1063
+    },
+    {
+      "epoch": 1.3492063492063493,
+      "grad_norm": 0.98828125,
+      "learning_rate": 5.409865740142242e-06,
+      "loss": 1.0664,
+      "step": 1064
+    },
+    {
+      "epoch": 1.3504761904761904,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.406841678144237e-06,
+      "loss": 1.0606,
+      "step": 1065
+    },
+    {
+      "epoch": 1.3517460317460317,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.403815590402905e-06,
+      "loss": 1.1376,
+      "step": 1066
+    },
+    {
+      "epoch": 1.353015873015873,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.400787480133023e-06,
+      "loss": 1.1475,
+      "step": 1067
+    },
+    {
+      "epoch": 1.3542857142857143,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.397757350551521e-06,
+      "loss": 1.021,
+      "step": 1068
+    },
+    {
+      "epoch": 1.3555555555555556,
+      "grad_norm": 1.0,
+      "learning_rate": 5.394725204877474e-06,
+      "loss": 1.141,
+      "step": 1069
+    },
+    {
+      "epoch": 1.3568253968253967,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.391691046332097e-06,
+      "loss": 1.1207,
+      "step": 1070
+    },
+    {
+      "epoch": 1.358095238095238,
+      "grad_norm": 0.99609375,
+      "learning_rate": 5.3886548781387455e-06,
+      "loss": 1.1566,
+      "step": 1071
+    },
+    {
+      "epoch": 1.3593650793650793,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.385616703522907e-06,
+      "loss": 1.1329,
+      "step": 1072
+    },
+    {
+      "epoch": 1.3606349206349206,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.382576525712205e-06,
+      "loss": 1.1949,
+      "step": 1073
+    },
+    {
+      "epoch": 1.361904761904762,
+      "grad_norm": 0.96875,
+      "learning_rate": 5.379534347936386e-06,
+      "loss": 1.1444,
+      "step": 1074
+    },
+    {
+      "epoch": 1.363174603174603,
+      "grad_norm": 0.953125,
+      "learning_rate": 5.376490173427324e-06,
+      "loss": 1.1498,
+      "step": 1075
+    },
+    {
+      "epoch": 1.3644444444444446,
+      "grad_norm": 1.0,
+      "learning_rate": 5.373444005419014e-06,
+      "loss": 1.0996,
+      "step": 1076
+    },
+    {
+      "epoch": 1.3657142857142857,
+      "grad_norm": 0.9453125,
+      "learning_rate": 5.37039584714757e-06,
+      "loss": 1.1395,
+      "step": 1077
+    },
+    {
+      "epoch": 1.366984126984127,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.367345701851217e-06,
+      "loss": 1.1675,
+      "step": 1078
+    },
+    {
+      "epoch": 1.3682539682539683,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.364293572770295e-06,
+      "loss": 1.1776,
+      "step": 1079
+    },
+    {
+      "epoch": 1.3695238095238096,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.3612394631472475e-06,
+      "loss": 1.1552,
+      "step": 1080
+    },
+    {
+      "epoch": 1.370793650793651,
+      "grad_norm": 0.890625,
+      "learning_rate": 5.358183376226626e-06,
+      "loss": 1.0682,
+      "step": 1081
+    },
+    {
+      "epoch": 1.372063492063492,
+      "grad_norm": 0.98828125,
+      "learning_rate": 5.355125315255079e-06,
+      "loss": 1.117,
+      "step": 1082
+    },
+    {
+      "epoch": 1.3733333333333333,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.352065283481355e-06,
+      "loss": 1.2372,
+      "step": 1083
+    },
+    {
+      "epoch": 1.3746031746031746,
+      "grad_norm": 0.91796875,
+      "learning_rate": 5.349003284156292e-06,
+      "loss": 1.1573,
+      "step": 1084
+    },
+    {
+      "epoch": 1.375873015873016,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.3459393205328255e-06,
+      "loss": 1.1628,
+      "step": 1085
+    },
+    {
+      "epoch": 1.3771428571428572,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.342873395865971e-06,
+      "loss": 1.1204,
+      "step": 1086
+    },
+    {
+      "epoch": 1.3784126984126983,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.3398055134128295e-06,
+      "loss": 1.124,
+      "step": 1087
+    },
+    {
+      "epoch": 1.3796825396825396,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.336735676432583e-06,
+      "loss": 1.0391,
+      "step": 1088
+    },
+    {
+      "epoch": 1.380952380952381,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.333663888186488e-06,
+      "loss": 1.1251,
+      "step": 1089
+    },
+    {
+      "epoch": 1.3822222222222222,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.330590151937877e-06,
+      "loss": 1.1557,
+      "step": 1090
+    },
+    {
+      "epoch": 1.3834920634920636,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.327514470952147e-06,
+      "loss": 1.1114,
+      "step": 1091
+    },
+    {
+      "epoch": 1.3847619047619046,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.324436848496766e-06,
+      "loss": 1.1198,
+      "step": 1092
+    },
+    {
+      "epoch": 1.3860317460317462,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.321357287841262e-06,
+      "loss": 1.0892,
+      "step": 1093
+    },
+    {
+      "epoch": 1.3873015873015873,
+      "grad_norm": 0.91015625,
+      "learning_rate": 5.3182757922572226e-06,
+      "loss": 1.1187,
+      "step": 1094
+    },
+    {
+      "epoch": 1.3885714285714286,
+      "grad_norm": 0.90234375,
+      "learning_rate": 5.315192365018292e-06,
+      "loss": 1.0835,
+      "step": 1095
+    },
+    {
+      "epoch": 1.3898412698412699,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.312107009400163e-06,
+      "loss": 1.1224,
+      "step": 1096
+    },
+    {
+      "epoch": 1.3911111111111112,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.309019728680581e-06,
+      "loss": 1.1558,
+      "step": 1097
+    },
+    {
+      "epoch": 1.3923809523809525,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.3059305261393355e-06,
+      "loss": 1.2182,
+      "step": 1098
+    },
+    {
+      "epoch": 1.3936507936507936,
+      "grad_norm": 0.99609375,
+      "learning_rate": 5.3028394050582555e-06,
+      "loss": 1.1034,
+      "step": 1099
+    },
+    {
+      "epoch": 1.394920634920635,
+      "grad_norm": 1.0,
+      "learning_rate": 5.29974636872121e-06,
+      "loss": 1.1412,
+      "step": 1100
+    },
+    {
+      "epoch": 1.3961904761904762,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.296651420414104e-06,
+      "loss": 1.1102,
+      "step": 1101
+    },
+    {
+      "epoch": 1.3974603174603175,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.293554563424871e-06,
+      "loss": 1.1589,
+      "step": 1102
+    },
+    {
+      "epoch": 1.3987301587301588,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.2904558010434745e-06,
+      "loss": 1.1025,
+      "step": 1103
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.125,
+      "learning_rate": 5.2873551365619e-06,
+      "loss": 1.1318,
+      "step": 1104
+    },
+    {
+      "epoch": 1.4012698412698412,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.284252573274153e-06,
+      "loss": 1.125,
+      "step": 1105
+    },
+    {
+      "epoch": 1.4025396825396825,
+      "grad_norm": 0.90625,
+      "learning_rate": 5.281148114476263e-06,
+      "loss": 1.0316,
+      "step": 1106
+    },
+    {
+      "epoch": 1.4038095238095238,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.278041763466265e-06,
+      "loss": 1.1106,
+      "step": 1107
+    },
+    {
+      "epoch": 1.4050793650793652,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.274933523544206e-06,
+      "loss": 1.2083,
+      "step": 1108
+    },
+    {
+      "epoch": 1.4063492063492062,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.2718233980121435e-06,
+      "loss": 1.0879,
+      "step": 1109
+    },
+    {
+      "epoch": 1.4076190476190475,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.268711390174135e-06,
+      "loss": 1.2132,
+      "step": 1110
+    },
+    {
+      "epoch": 1.4088888888888889,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.265597503336238e-06,
+      "loss": 1.1204,
+      "step": 1111
+    },
+    {
+      "epoch": 1.4101587301587302,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.262481740806509e-06,
+      "loss": 1.048,
+      "step": 1112
+    },
+    {
+      "epoch": 1.4114285714285715,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.259364105894991e-06,
+      "loss": 1.1192,
+      "step": 1113
+    },
+    {
+      "epoch": 1.4126984126984126,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.2562446019137225e-06,
+      "loss": 1.1135,
+      "step": 1114
+    },
+    {
+      "epoch": 1.413968253968254,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.253123232176724e-06,
+      "loss": 1.1061,
+      "step": 1115
+    },
+    {
+      "epoch": 1.4152380952380952,
+      "grad_norm": 0.96875,
+      "learning_rate": 5.25e-06,
+      "loss": 1.0388,
+      "step": 1116
+    },
+    {
+      "epoch": 1.4165079365079365,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.246874908701532e-06,
+      "loss": 1.0881,
+      "step": 1117
+    },
+    {
+      "epoch": 1.4177777777777778,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.2437479616012775e-06,
+      "loss": 1.0944,
+      "step": 1118
+    },
+    {
+      "epoch": 1.4190476190476191,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.240619162021165e-06,
+      "loss": 1.119,
+      "step": 1119
+    },
+    {
+      "epoch": 1.4203174603174604,
+      "grad_norm": 0.99609375,
+      "learning_rate": 5.237488513285092e-06,
+      "loss": 1.1366,
+      "step": 1120
+    },
+    {
+      "epoch": 1.4215873015873015,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.234356018718919e-06,
+      "loss": 1.0363,
+      "step": 1121
+    },
+    {
+      "epoch": 1.4228571428571428,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.231221681650469e-06,
+      "loss": 1.1057,
+      "step": 1122
+    },
+    {
+      "epoch": 1.4241269841269841,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.228085505409519e-06,
+      "loss": 1.1738,
+      "step": 1123
+    },
+    {
+      "epoch": 1.4253968253968254,
+      "grad_norm": 0.875,
+      "learning_rate": 5.224947493327806e-06,
+      "loss": 1.0843,
+      "step": 1124
+    },
+    {
+      "epoch": 1.4266666666666667,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.221807648739012e-06,
+      "loss": 1.2114,
+      "step": 1125
+    },
+    {
+      "epoch": 1.4279365079365078,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.218665974978767e-06,
+      "loss": 1.1338,
+      "step": 1126
+    },
+    {
+      "epoch": 1.4292063492063491,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.215522475384645e-06,
+      "loss": 1.0466,
+      "step": 1127
+    },
+    {
+      "epoch": 1.4304761904761905,
+      "grad_norm": 0.953125,
+      "learning_rate": 5.21237715329616e-06,
+      "loss": 1.1675,
+      "step": 1128
+    },
+    {
+      "epoch": 1.4317460317460318,
+      "grad_norm": 0.984375,
+      "learning_rate": 5.209230012054761e-06,
+      "loss": 1.2035,
+      "step": 1129
+    },
+    {
+      "epoch": 1.433015873015873,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.20608105500383e-06,
+      "loss": 1.0926,
+      "step": 1130
+    },
+    {
+      "epoch": 1.4342857142857142,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.202930285488679e-06,
+      "loss": 1.0956,
+      "step": 1131
+    },
+    {
+      "epoch": 1.4355555555555555,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.1997777068565426e-06,
+      "loss": 1.1305,
+      "step": 1132
+    },
+    {
+      "epoch": 1.4368253968253968,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.196623322456582e-06,
+      "loss": 1.1132,
+      "step": 1133
+    },
+    {
+      "epoch": 1.438095238095238,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.193467135639873e-06,
+      "loss": 1.0628,
+      "step": 1134
+    },
+    {
+      "epoch": 1.4393650793650794,
+      "grad_norm": 0.9296875,
+      "learning_rate": 5.1903091497594075e-06,
+      "loss": 1.2005,
+      "step": 1135
+    },
+    {
+      "epoch": 1.4406349206349207,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.187149368170087e-06,
+      "loss": 1.1237,
+      "step": 1136
+    },
+    {
+      "epoch": 1.441904761904762,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.1839877942287245e-06,
+      "loss": 1.099,
+      "step": 1137
+    },
+    {
+      "epoch": 1.443174603174603,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.180824431294034e-06,
+      "loss": 1.1586,
+      "step": 1138
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 0.96875,
+      "learning_rate": 5.17765928272663e-06,
+      "loss": 1.1243,
+      "step": 1139
+    },
+    {
+      "epoch": 1.4457142857142857,
+      "grad_norm": 0.875,
+      "learning_rate": 5.1744923518890255e-06,
+      "loss": 1.0623,
+      "step": 1140
+    },
+    {
+      "epoch": 1.446984126984127,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.171323642145625e-06,
+      "loss": 1.111,
+      "step": 1141
+    },
+    {
+      "epoch": 1.4482539682539683,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.168153156862726e-06,
+      "loss": 1.0733,
+      "step": 1142
+    },
+    {
+      "epoch": 1.4495238095238094,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.164980899408509e-06,
+      "loss": 1.0906,
+      "step": 1143
+    },
+    {
+      "epoch": 1.4507936507936507,
+      "grad_norm": 0.96875,
+      "learning_rate": 5.1618068731530395e-06,
+      "loss": 1.1173,
+      "step": 1144
+    },
+    {
+      "epoch": 1.452063492063492,
+      "grad_norm": 1.0,
+      "learning_rate": 5.15863108146826e-06,
+      "loss": 1.1727,
+      "step": 1145
+    },
+    {
+      "epoch": 1.4533333333333334,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.155453527727989e-06,
+      "loss": 1.1193,
+      "step": 1146
+    },
+    {
+      "epoch": 1.4546031746031747,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.15227421530792e-06,
+      "loss": 1.1275,
+      "step": 1147
+    },
+    {
+      "epoch": 1.4558730158730158,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.14909314758561e-06,
+      "loss": 1.1938,
+      "step": 1148
+    },
+    {
+      "epoch": 1.457142857142857,
+      "grad_norm": 0.953125,
+      "learning_rate": 5.145910327940484e-06,
+      "loss": 1.1307,
+      "step": 1149
+    },
+    {
+      "epoch": 1.4584126984126984,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.142725759753826e-06,
+      "loss": 1.2017,
+      "step": 1150
+    },
+    {
+      "epoch": 1.4596825396825397,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.139539446408781e-06,
+      "loss": 1.1306,
+      "step": 1151
+    },
+    {
+      "epoch": 1.460952380952381,
+      "grad_norm": 0.9453125,
+      "learning_rate": 5.136351391290346e-06,
+      "loss": 1.102,
+      "step": 1152
+    },
+    {
+      "epoch": 1.462222222222222,
+      "grad_norm": 0.859375,
+      "learning_rate": 5.133161597785365e-06,
+      "loss": 1.1168,
+      "step": 1153
+    },
+    {
+      "epoch": 1.4634920634920636,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.129970069282535e-06,
+      "loss": 1.1488,
+      "step": 1154
+    },
+    {
+      "epoch": 1.4647619047619047,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.126776809172392e-06,
+      "loss": 1.0992,
+      "step": 1155
+    },
+    {
+      "epoch": 1.466031746031746,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.123581820847313e-06,
+      "loss": 1.1016,
+      "step": 1156
+    },
+    {
+      "epoch": 1.4673015873015873,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.1203851077015106e-06,
+      "loss": 1.12,
+      "step": 1157
+    },
+    {
+      "epoch": 1.4685714285714286,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.1171866731310285e-06,
+      "loss": 1.087,
+      "step": 1158
+    },
+    {
+      "epoch": 1.46984126984127,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.113986520533743e-06,
+      "loss": 1.1197,
+      "step": 1159
+    },
+    {
+      "epoch": 1.471111111111111,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.110784653309353e-06,
+      "loss": 1.1768,
+      "step": 1160
+    },
+    {
+      "epoch": 1.4723809523809523,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.1075810748593765e-06,
+      "loss": 1.1136,
+      "step": 1161
+    },
+    {
+      "epoch": 1.4736507936507937,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.104375788587154e-06,
+      "loss": 1.1602,
+      "step": 1162
+    },
+    {
+      "epoch": 1.474920634920635,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.1011687978978394e-06,
+      "loss": 1.1295,
+      "step": 1163
+    },
+    {
+      "epoch": 1.4761904761904763,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.097960106198393e-06,
+      "loss": 1.2086,
+      "step": 1164
+    },
+    {
+      "epoch": 1.4774603174603174,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.094749716897588e-06,
+      "loss": 1.1108,
+      "step": 1165
+    },
+    {
+      "epoch": 1.4787301587301587,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.091537633405996e-06,
+      "loss": 1.1007,
+      "step": 1166
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.088323859135995e-06,
+      "loss": 1.0905,
+      "step": 1167
+    },
+    {
+      "epoch": 1.4812698412698413,
+      "grad_norm": 0.953125,
+      "learning_rate": 5.0851083975017515e-06,
+      "loss": 1.1302,
+      "step": 1168
+    },
+    {
+      "epoch": 1.4825396825396826,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.081891251919227e-06,
+      "loss": 1.104,
+      "step": 1169
+    },
+    {
+      "epoch": 1.4838095238095237,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.078672425806178e-06,
+      "loss": 1.0844,
+      "step": 1170
+    },
+    {
+      "epoch": 1.485079365079365,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.075451922582138e-06,
+      "loss": 1.0839,
+      "step": 1171
+    },
+    {
+      "epoch": 1.4863492063492063,
+      "grad_norm": 0.9609375,
+      "learning_rate": 5.072229745668424e-06,
+      "loss": 1.1437,
+      "step": 1172
+    },
+    {
+      "epoch": 1.4876190476190476,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.069005898488137e-06,
+      "loss": 1.0498,
+      "step": 1173
+    },
+    {
+      "epoch": 1.488888888888889,
+      "grad_norm": 0.99609375,
+      "learning_rate": 5.065780384466147e-06,
+      "loss": 1.1558,
+      "step": 1174
+    },
+    {
+      "epoch": 1.4901587301587302,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.062553207029093e-06,
+      "loss": 1.1752,
+      "step": 1175
+    },
+    {
+      "epoch": 1.4914285714285715,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.059324369605386e-06,
+      "loss": 1.1394,
+      "step": 1176
+    },
+    {
+      "epoch": 1.4926984126984126,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.056093875625199e-06,
+      "loss": 1.1602,
+      "step": 1177
+    },
+    {
+      "epoch": 1.493968253968254,
+      "grad_norm": 0.953125,
+      "learning_rate": 5.052861728520462e-06,
+      "loss": 1.1289,
+      "step": 1178
+    },
+    {
+      "epoch": 1.4952380952380953,
+      "grad_norm": 0.93359375,
+      "learning_rate": 5.049627931724868e-06,
+      "loss": 1.1,
+      "step": 1179
+    },
+    {
+      "epoch": 1.4965079365079366,
+      "grad_norm": 0.9140625,
+      "learning_rate": 5.046392488673853e-06,
+      "loss": 1.1183,
+      "step": 1180
+    },
+    {
+      "epoch": 1.4977777777777779,
+      "grad_norm": 0.9296875,
+      "learning_rate": 5.043155402804611e-06,
+      "loss": 1.2041,
+      "step": 1181
+    },
+    {
+      "epoch": 1.499047619047619,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.039916677556075e-06,
+      "loss": 1.1673,
+      "step": 1182
+    },
+    {
+      "epoch": 1.5003174603174603,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.036676316368923e-06,
+      "loss": 1.1223,
+      "step": 1183
+    },
+    {
+      "epoch": 1.5015873015873016,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.033434322685568e-06,
+      "loss": 1.111,
+      "step": 1184
+    },
+    {
+      "epoch": 1.502857142857143,
+      "grad_norm": 0.96875,
+      "learning_rate": 5.030190699950161e-06,
+      "loss": 1.0827,
+      "step": 1185
+    },
+    {
+      "epoch": 1.5041269841269842,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.026945451608582e-06,
+      "loss": 1.1214,
+      "step": 1186
+    },
+    {
+      "epoch": 1.5053968253968253,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.023698581108435e-06,
+      "loss": 1.1387,
+      "step": 1187
+    },
+    {
+      "epoch": 1.5066666666666668,
+      "grad_norm": 0.92578125,
+      "learning_rate": 5.020450091899054e-06,
+      "loss": 1.0856,
+      "step": 1188
+    },
+    {
+      "epoch": 1.507936507936508,
+      "grad_norm": 0.984375,
+      "learning_rate": 5.017199987431486e-06,
+      "loss": 1.1512,
+      "step": 1189
+    },
+    {
+      "epoch": 1.5092063492063492,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.013948271158496e-06,
+      "loss": 1.1007,
+      "step": 1190
+    },
+    {
+      "epoch": 1.5104761904761905,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.010694946534565e-06,
+      "loss": 1.1165,
+      "step": 1191
+    },
+    {
+      "epoch": 1.5117460317460316,
+      "grad_norm": 1.0,
+      "learning_rate": 5.007440017015877e-06,
+      "loss": 1.1697,
+      "step": 1192
+    },
+    {
+      "epoch": 1.5130158730158731,
+      "grad_norm": 0.984375,
+      "learning_rate": 5.004183486060324e-06,
+      "loss": 1.148,
+      "step": 1193
+    },
+    {
+      "epoch": 1.5142857142857142,
+      "grad_norm": 0.9140625,
+      "learning_rate": 5.0009253571275e-06,
+      "loss": 1.1619,
+      "step": 1194
+    },
+    {
+      "epoch": 1.5155555555555555,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.997665633678694e-06,
+      "loss": 1.1353,
+      "step": 1195
+    },
+    {
+      "epoch": 1.5168253968253969,
+      "grad_norm": 1.0,
+      "learning_rate": 4.994404319176892e-06,
+      "loss": 1.1593,
+      "step": 1196
+    },
+    {
+      "epoch": 1.518095238095238,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.991141417086767e-06,
+      "loss": 1.1655,
+      "step": 1197
+    },
+    {
+      "epoch": 1.5193650793650795,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.987876930874681e-06,
+      "loss": 1.0915,
+      "step": 1198
+    },
+    {
+      "epoch": 1.5206349206349206,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.984610864008679e-06,
+      "loss": 1.0815,
+      "step": 1199
+    },
+    {
+      "epoch": 1.5219047619047619,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.981343219958484e-06,
+      "loss": 1.0946,
+      "step": 1200
+    },
+    {
+      "epoch": 1.5231746031746032,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.978074002195494e-06,
+      "loss": 1.1965,
+      "step": 1201
+    },
+    {
+      "epoch": 1.5244444444444445,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.97480321419278e-06,
+      "loss": 1.0397,
+      "step": 1202
+    },
+    {
+      "epoch": 1.5257142857142858,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.971530859425082e-06,
+      "loss": 1.1509,
+      "step": 1203
+    },
+    {
+      "epoch": 1.5269841269841269,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.9682569413688014e-06,
+      "loss": 1.2044,
+      "step": 1204
+    },
+    {
+      "epoch": 1.5282539682539684,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.964981463502005e-06,
+      "loss": 1.1998,
+      "step": 1205
+    },
+    {
+      "epoch": 1.5295238095238095,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.961704429304411e-06,
+      "loss": 1.1526,
+      "step": 1206
+    },
+    {
+      "epoch": 1.5307936507936508,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.9584258422573966e-06,
+      "loss": 1.1067,
+      "step": 1207
+    },
+    {
+      "epoch": 1.5320634920634921,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.955145705843985e-06,
+      "loss": 1.0507,
+      "step": 1208
+    },
+    {
+      "epoch": 1.5333333333333332,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.9518640235488464e-06,
+      "loss": 1.1291,
+      "step": 1209
+    },
+    {
+      "epoch": 1.5346031746031747,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.948580798858294e-06,
+      "loss": 1.1125,
+      "step": 1210
+    },
+    {
+      "epoch": 1.5358730158730158,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.945296035260276e-06,
+      "loss": 1.1414,
+      "step": 1211
+    },
+    {
+      "epoch": 1.5371428571428571,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.9420097362443834e-06,
+      "loss": 1.096,
+      "step": 1212
+    },
+    {
+      "epoch": 1.5384126984126985,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.938721905301829e-06,
+      "loss": 1.1666,
+      "step": 1213
+    },
+    {
+      "epoch": 1.5396825396825395,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.93543254592546e-06,
+      "loss": 1.1204,
+      "step": 1214
+    },
+    {
+      "epoch": 1.540952380952381,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.932141661609743e-06,
+      "loss": 1.0977,
+      "step": 1215
+    },
+    {
+      "epoch": 1.5422222222222222,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.9288492558507664e-06,
+      "loss": 1.1075,
+      "step": 1216
+    },
+    {
+      "epoch": 1.5434920634920635,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.925555332146236e-06,
+      "loss": 1.0444,
+      "step": 1217
+    },
+    {
+      "epoch": 1.5447619047619048,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.922259893995469e-06,
+      "loss": 1.0876,
+      "step": 1218
+    },
+    {
+      "epoch": 1.5460317460317459,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.918962944899391e-06,
+      "loss": 1.1154,
+      "step": 1219
+    },
+    {
+      "epoch": 1.5473015873015874,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.915664488360532e-06,
+      "loss": 1.161,
+      "step": 1220
+    },
+    {
+      "epoch": 1.5485714285714285,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.912364527883026e-06,
+      "loss": 1.1021,
+      "step": 1221
+    },
+    {
+      "epoch": 1.5498412698412698,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.9090630669726045e-06,
+      "loss": 1.115,
+      "step": 1222
+    },
+    {
+      "epoch": 1.551111111111111,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.905760109136591e-06,
+      "loss": 1.0934,
+      "step": 1223
+    },
+    {
+      "epoch": 1.5523809523809524,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.9024556578839e-06,
+      "loss": 1.0865,
+      "step": 1224
+    },
+    {
+      "epoch": 1.5536507936507937,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.899149716725033e-06,
+      "loss": 1.164,
+      "step": 1225
+    },
+    {
+      "epoch": 1.5549206349206348,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.895842289172073e-06,
+      "loss": 1.1473,
+      "step": 1226
+    },
+    {
+      "epoch": 1.5561904761904763,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.892533378738685e-06,
+      "loss": 1.1032,
+      "step": 1227
+    },
+    {
+      "epoch": 1.5574603174603174,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.8892229889401054e-06,
+      "loss": 1.0927,
+      "step": 1228
+    },
+    {
+      "epoch": 1.5587301587301587,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.885911123293146e-06,
+      "loss": 1.0953,
+      "step": 1229
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.882597785316183e-06,
+      "loss": 1.1072,
+      "step": 1230
+    },
+    {
+      "epoch": 1.5612698412698411,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.879282978529162e-06,
+      "loss": 1.1429,
+      "step": 1231
+    },
+    {
+      "epoch": 1.5625396825396827,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.875966706453582e-06,
+      "loss": 1.0791,
+      "step": 1232
+    },
+    {
+      "epoch": 1.5638095238095238,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.872648972612502e-06,
+      "loss": 1.1152,
+      "step": 1233
+    },
+    {
+      "epoch": 1.565079365079365,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.8693297805305375e-06,
+      "loss": 1.0872,
+      "step": 1234
+    },
+    {
+      "epoch": 1.5663492063492064,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.866009133733848e-06,
+      "loss": 1.0842,
+      "step": 1235
+    },
+    {
+      "epoch": 1.5676190476190475,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.862687035750137e-06,
+      "loss": 1.1449,
+      "step": 1236
+    },
+    {
+      "epoch": 1.568888888888889,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.859363490108657e-06,
+      "loss": 1.1102,
+      "step": 1237
+    },
+    {
+      "epoch": 1.57015873015873,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.856038500340194e-06,
+      "loss": 1.0849,
+      "step": 1238
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.852712069977067e-06,
+      "loss": 1.1458,
+      "step": 1239
+    },
+    {
+      "epoch": 1.5726984126984127,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.849384202553127e-06,
+      "loss": 1.1222,
+      "step": 1240
+    },
+    {
+      "epoch": 1.573968253968254,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.84605490160375e-06,
+      "loss": 1.123,
+      "step": 1241
+    },
+    {
+      "epoch": 1.5752380952380953,
+      "grad_norm": 1.125,
+      "learning_rate": 4.8427241706658395e-06,
+      "loss": 1.1793,
+      "step": 1242
+    },
+    {
+      "epoch": 1.5765079365079364,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.8393920132778144e-06,
+      "loss": 1.0034,
+      "step": 1243
+    },
+    {
+      "epoch": 1.5777777777777777,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.836058432979608e-06,
+      "loss": 1.1062,
+      "step": 1244
+    },
+    {
+      "epoch": 1.579047619047619,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.83272343331267e-06,
+      "loss": 1.045,
+      "step": 1245
+    },
+    {
+      "epoch": 1.5803174603174603,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.829387017819951e-06,
+      "loss": 1.1662,
+      "step": 1246
+    },
+    {
+      "epoch": 1.5815873015873017,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.826049190045912e-06,
+      "loss": 1.1569,
+      "step": 1247
+    },
+    {
+      "epoch": 1.5828571428571427,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.822709953536511e-06,
+      "loss": 1.1727,
+      "step": 1248
+    },
+    {
+      "epoch": 1.5841269841269843,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.819369311839206e-06,
+      "loss": 1.0916,
+      "step": 1249
+    },
+    {
+      "epoch": 1.5853968253968254,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.8160272685029414e-06,
+      "loss": 1.0817,
+      "step": 1250
+    },
+    {
+      "epoch": 1.5866666666666667,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.812683827078157e-06,
+      "loss": 1.1218,
+      "step": 1251
+    },
+    {
+      "epoch": 1.587936507936508,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.8093389911167766e-06,
+      "loss": 1.1506,
+      "step": 1252
+    },
+    {
+      "epoch": 1.589206349206349,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.805992764172202e-06,
+      "loss": 1.1086,
+      "step": 1253
+    },
+    {
+      "epoch": 1.5904761904761906,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.802645149799315e-06,
+      "loss": 1.1163,
+      "step": 1254
+    },
+    {
+      "epoch": 1.5917460317460317,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.799296151554472e-06,
+      "loss": 1.1357,
+      "step": 1255
+    },
+    {
+      "epoch": 1.593015873015873,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.795945772995502e-06,
+      "loss": 1.0616,
+      "step": 1256
+    },
+    {
+      "epoch": 1.5942857142857143,
+      "grad_norm": 1.0,
+      "learning_rate": 4.792594017681692e-06,
+      "loss": 1.1926,
+      "step": 1257
+    },
+    {
+      "epoch": 1.5955555555555554,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.789240889173801e-06,
+      "loss": 1.1069,
+      "step": 1258
+    },
+    {
+      "epoch": 1.596825396825397,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.785886391034041e-06,
+      "loss": 1.1115,
+      "step": 1259
+    },
+    {
+      "epoch": 1.598095238095238,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.78253052682608e-06,
+      "loss": 1.1455,
+      "step": 1260
+    },
+    {
+      "epoch": 1.5993650793650793,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.7791733001150415e-06,
+      "loss": 1.0381,
+      "step": 1261
+    },
+    {
+      "epoch": 1.6006349206349206,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.775814714467489e-06,
+      "loss": 1.1659,
+      "step": 1262
+    },
+    {
+      "epoch": 1.601904761904762,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.772454773451437e-06,
+      "loss": 1.0763,
+      "step": 1263
+    },
+    {
+      "epoch": 1.6031746031746033,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.769093480636334e-06,
+      "loss": 1.1287,
+      "step": 1264
+    },
+    {
+      "epoch": 1.6044444444444443,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.7657308395930685e-06,
+      "loss": 1.0678,
+      "step": 1265
+    },
+    {
+      "epoch": 1.6057142857142859,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.7623668538939605e-06,
+      "loss": 1.1395,
+      "step": 1266
+    },
+    {
+      "epoch": 1.606984126984127,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.759001527112757e-06,
+      "loss": 1.1187,
+      "step": 1267
+    },
+    {
+      "epoch": 1.6082539682539683,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.75563486282463e-06,
+      "loss": 1.058,
+      "step": 1268
+    },
+    {
+      "epoch": 1.6095238095238096,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.752266864606174e-06,
+      "loss": 1.1288,
+      "step": 1269
+    },
+    {
+      "epoch": 1.6107936507936507,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.748897536035402e-06,
+      "loss": 1.0998,
+      "step": 1270
+    },
+    {
+      "epoch": 1.6120634920634922,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.745526880691734e-06,
+      "loss": 1.0686,
+      "step": 1271
+    },
+    {
+      "epoch": 1.6133333333333333,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.742154902156005e-06,
+      "loss": 1.0893,
+      "step": 1272
+    },
+    {
+      "epoch": 1.6146031746031746,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.738781604010454e-06,
+      "loss": 1.1056,
+      "step": 1273
+    },
+    {
+      "epoch": 1.615873015873016,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.735406989838723e-06,
+      "loss": 1.0968,
+      "step": 1274
+    },
+    {
+      "epoch": 1.617142857142857,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.7320310632258515e-06,
+      "loss": 1.0764,
+      "step": 1275
+    },
+    {
+      "epoch": 1.6184126984126985,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.728653827758271e-06,
+      "loss": 1.0824,
+      "step": 1276
+    },
+    {
+      "epoch": 1.6196825396825396,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.725275287023808e-06,
+      "loss": 1.0422,
+      "step": 1277
+    },
+    {
+      "epoch": 1.620952380952381,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.721895444611671e-06,
+      "loss": 1.1258,
+      "step": 1278
+    },
+    {
+      "epoch": 1.6222222222222222,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.718514304112455e-06,
+      "loss": 1.1382,
+      "step": 1279
+    },
+    {
+      "epoch": 1.6234920634920635,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.71513186911813e-06,
+      "loss": 1.1232,
+      "step": 1280
+    },
+    {
+      "epoch": 1.6247619047619049,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.711748143222045e-06,
+      "loss": 1.0685,
+      "step": 1281
+    },
+    {
+      "epoch": 1.626031746031746,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.708363130018921e-06,
+      "loss": 1.1688,
+      "step": 1282
+    },
+    {
+      "epoch": 1.6273015873015872,
+      "grad_norm": 1.0,
+      "learning_rate": 4.704976833104843e-06,
+      "loss": 1.1588,
+      "step": 1283
+    },
+    {
+      "epoch": 1.6285714285714286,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.701589256077261e-06,
+      "loss": 1.1249,
+      "step": 1284
+    },
+    {
+      "epoch": 1.6298412698412699,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.698200402534986e-06,
+      "loss": 1.1255,
+      "step": 1285
+    },
+    {
+      "epoch": 1.6311111111111112,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.694810276078184e-06,
+      "loss": 1.1134,
+      "step": 1286
+    },
+    {
+      "epoch": 1.6323809523809523,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.691418880308375e-06,
+      "loss": 1.087,
+      "step": 1287
+    },
+    {
+      "epoch": 1.6336507936507938,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.6880262188284235e-06,
+      "loss": 1.1309,
+      "step": 1288
+    },
+    {
+      "epoch": 1.6349206349206349,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.684632295242544e-06,
+      "loss": 1.1354,
+      "step": 1289
+    },
+    {
+      "epoch": 1.6361904761904762,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.681237113156287e-06,
+      "loss": 1.0666,
+      "step": 1290
+    },
+    {
+      "epoch": 1.6374603174603175,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.677840676176545e-06,
+      "loss": 1.0729,
+      "step": 1291
+    },
+    {
+      "epoch": 1.6387301587301586,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.674442987911539e-06,
+      "loss": 1.0809,
+      "step": 1292
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 1.0,
+      "learning_rate": 4.6710440519708205e-06,
+      "loss": 1.1709,
+      "step": 1293
+    },
+    {
+      "epoch": 1.6412698412698412,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.667643871965267e-06,
+      "loss": 1.1142,
+      "step": 1294
+    },
+    {
+      "epoch": 1.6425396825396825,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.66424245150708e-06,
+      "loss": 1.0243,
+      "step": 1295
+    },
+    {
+      "epoch": 1.6438095238095238,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.660839794209774e-06,
+      "loss": 1.0941,
+      "step": 1296
+    },
+    {
+      "epoch": 1.645079365079365,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.657435903688182e-06,
+      "loss": 1.1224,
+      "step": 1297
+    },
+    {
+      "epoch": 1.6463492063492065,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.654030783558442e-06,
+      "loss": 1.2249,
+      "step": 1298
+    },
+    {
+      "epoch": 1.6476190476190475,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.650624437438005e-06,
+      "loss": 1.1019,
+      "step": 1299
+    },
+    {
+      "epoch": 1.6488888888888888,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.647216868945618e-06,
+      "loss": 1.0937,
+      "step": 1300
+    },
+    {
+      "epoch": 1.6501587301587302,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.643808081701332e-06,
+      "loss": 1.1725,
+      "step": 1301
+    },
+    {
+      "epoch": 1.6514285714285715,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.640398079326487e-06,
+      "loss": 1.0583,
+      "step": 1302
+    },
+    {
+      "epoch": 1.6526984126984128,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.636986865443719e-06,
+      "loss": 1.1474,
+      "step": 1303
+    },
+    {
+      "epoch": 1.6539682539682539,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.63357444367695e-06,
+      "loss": 1.1133,
+      "step": 1304
+    },
+    {
+      "epoch": 1.6552380952380954,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.6301608176513825e-06,
+      "loss": 1.0726,
+      "step": 1305
+    },
+    {
+      "epoch": 1.6565079365079365,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.6267459909935e-06,
+      "loss": 1.1477,
+      "step": 1306
+    },
+    {
+      "epoch": 1.6577777777777778,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.623329967331062e-06,
+      "loss": 1.0958,
+      "step": 1307
+    },
+    {
+      "epoch": 1.659047619047619,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.6199127502931e-06,
+      "loss": 1.0743,
+      "step": 1308
+    },
+    {
+      "epoch": 1.6603174603174602,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.616494343509912e-06,
+      "loss": 1.1027,
+      "step": 1309
+    },
+    {
+      "epoch": 1.6615873015873017,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.613074750613062e-06,
+      "loss": 1.0481,
+      "step": 1310
+    },
+    {
+      "epoch": 1.6628571428571428,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.60965397523537e-06,
+      "loss": 1.1293,
+      "step": 1311
+    },
+    {
+      "epoch": 1.6641269841269841,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.606232021010916e-06,
+      "loss": 1.132,
+      "step": 1312
+    },
+    {
+      "epoch": 1.6653968253968254,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.6028088915750316e-06,
+      "loss": 1.1219,
+      "step": 1313
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.5993845905642965e-06,
+      "loss": 1.1903,
+      "step": 1314
+    },
+    {
+      "epoch": 1.667936507936508,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.595959121616532e-06,
+      "loss": 1.0881,
+      "step": 1315
+    },
+    {
+      "epoch": 1.6692063492063491,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.592532488370808e-06,
+      "loss": 1.1348,
+      "step": 1316
+    },
+    {
+      "epoch": 1.6704761904761904,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.589104694467424e-06,
+      "loss": 1.0883,
+      "step": 1317
+    },
+    {
+      "epoch": 1.6717460317460318,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.585675743547915e-06,
+      "loss": 1.1143,
+      "step": 1318
+    },
+    {
+      "epoch": 1.673015873015873,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.582245639255044e-06,
+      "loss": 1.1538,
+      "step": 1319
+    },
+    {
+      "epoch": 1.6742857142857144,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.578814385232802e-06,
+      "loss": 1.096,
+      "step": 1320
+    },
+    {
+      "epoch": 1.6755555555555555,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.5753819851264e-06,
+      "loss": 1.1107,
+      "step": 1321
+    },
+    {
+      "epoch": 1.6768253968253968,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.571948442582265e-06,
+      "loss": 1.1652,
+      "step": 1322
+    },
+    {
+      "epoch": 1.678095238095238,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.56851376124804e-06,
+      "loss": 1.076,
+      "step": 1323
+    },
+    {
+      "epoch": 1.6793650793650794,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.565077944772576e-06,
+      "loss": 1.1744,
+      "step": 1324
+    },
+    {
+      "epoch": 1.6806349206349207,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.561640996805933e-06,
+      "loss": 1.1625,
+      "step": 1325
+    },
+    {
+      "epoch": 1.6819047619047618,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.558202920999368e-06,
+      "loss": 1.137,
+      "step": 1326
+    },
+    {
+      "epoch": 1.6831746031746033,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.5547637210053405e-06,
+      "loss": 1.0616,
+      "step": 1327
+    },
+    {
+      "epoch": 1.6844444444444444,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.551323400477504e-06,
+      "loss": 1.1175,
+      "step": 1328
+    },
+    {
+      "epoch": 1.6857142857142857,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.547881963070698e-06,
+      "loss": 1.0922,
+      "step": 1329
+    },
+    {
+      "epoch": 1.686984126984127,
+      "grad_norm": 1.0,
+      "learning_rate": 4.544439412440954e-06,
+      "loss": 1.0534,
+      "step": 1330
+    },
+    {
+      "epoch": 1.6882539682539681,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.540995752245483e-06,
+      "loss": 1.1033,
+      "step": 1331
+    },
+    {
+      "epoch": 1.6895238095238096,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.537550986142676e-06,
+      "loss": 1.067,
+      "step": 1332
+    },
+    {
+      "epoch": 1.6907936507936507,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.534105117792098e-06,
+      "loss": 1.1344,
+      "step": 1333
+    },
+    {
+      "epoch": 1.692063492063492,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.5306581508544865e-06,
+      "loss": 1.1352,
+      "step": 1334
+    },
+    {
+      "epoch": 1.6933333333333334,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.527210088991744e-06,
+      "loss": 1.0163,
+      "step": 1335
+    },
+    {
+      "epoch": 1.6946031746031744,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.523760935866939e-06,
+      "loss": 1.0919,
+      "step": 1336
+    },
+    {
+      "epoch": 1.695873015873016,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.520310695144294e-06,
+      "loss": 1.097,
+      "step": 1337
+    },
+    {
+      "epoch": 1.697142857142857,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.5168593704891926e-06,
+      "loss": 1.1038,
+      "step": 1338
+    },
+    {
+      "epoch": 1.6984126984126984,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.513406965568169e-06,
+      "loss": 1.1644,
+      "step": 1339
+    },
+    {
+      "epoch": 1.6996825396825397,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.509953484048903e-06,
+      "loss": 1.0151,
+      "step": 1340
+    },
+    {
+      "epoch": 1.700952380952381,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.506498929600217e-06,
+      "loss": 1.146,
+      "step": 1341
+    },
+    {
+      "epoch": 1.7022222222222223,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.503043305892077e-06,
+      "loss": 1.0788,
+      "step": 1342
+    },
+    {
+      "epoch": 1.7034920634920634,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.499586616595583e-06,
+      "loss": 1.0936,
+      "step": 1343
+    },
+    {
+      "epoch": 1.704761904761905,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.4961288653829665e-06,
+      "loss": 1.1669,
+      "step": 1344
+    },
+    {
+      "epoch": 1.706031746031746,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.492670055927589e-06,
+      "loss": 1.0815,
+      "step": 1345
+    },
+    {
+      "epoch": 1.7073015873015873,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.489210191903933e-06,
+      "loss": 1.1458,
+      "step": 1346
+    },
+    {
+      "epoch": 1.7085714285714286,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.4857492769876045e-06,
+      "loss": 1.1193,
+      "step": 1347
+    },
+    {
+      "epoch": 1.7098412698412697,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.482287314855326e-06,
+      "loss": 1.0507,
+      "step": 1348
+    },
+    {
+      "epoch": 1.7111111111111112,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.4788243091849295e-06,
+      "loss": 1.201,
+      "step": 1349
+    },
+    {
+      "epoch": 1.7123809523809523,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.475360263655357e-06,
+      "loss": 1.0748,
+      "step": 1350
+    },
+    {
+      "epoch": 1.7136507936507936,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.471895181946656e-06,
+      "loss": 1.1482,
+      "step": 1351
+    },
+    {
+      "epoch": 1.714920634920635,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.468429067739976e-06,
+      "loss": 1.098,
+      "step": 1352
+    },
+    {
+      "epoch": 1.716190476190476,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.4649619247175605e-06,
+      "loss": 1.135,
+      "step": 1353
+    },
+    {
+      "epoch": 1.7174603174603176,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.461493756562748e-06,
+      "loss": 1.0311,
+      "step": 1354
+    },
+    {
+      "epoch": 1.7187301587301587,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.458024566959962e-06,
+      "loss": 1.1186,
+      "step": 1355
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.454554359594718e-06,
+      "loss": 1.096,
+      "step": 1356
+    },
+    {
+      "epoch": 1.7212698412698413,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.451083138153609e-06,
+      "loss": 1.0987,
+      "step": 1357
+    },
+    {
+      "epoch": 1.7225396825396826,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.447610906324303e-06,
+      "loss": 1.0812,
+      "step": 1358
+    },
+    {
+      "epoch": 1.723809523809524,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.444137667795546e-06,
+      "loss": 1.1388,
+      "step": 1359
+    },
+    {
+      "epoch": 1.725079365079365,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.44066342625715e-06,
+      "loss": 1.1375,
+      "step": 1360
+    },
+    {
+      "epoch": 1.7263492063492063,
+      "grad_norm": 1.0,
+      "learning_rate": 4.437188185399994e-06,
+      "loss": 1.1185,
+      "step": 1361
+    },
+    {
+      "epoch": 1.7276190476190476,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.43371194891602e-06,
+      "loss": 1.1783,
+      "step": 1362
+    },
+    {
+      "epoch": 1.728888888888889,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.430234720498224e-06,
+      "loss": 1.0944,
+      "step": 1363
+    },
+    {
+      "epoch": 1.7301587301587302,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.426756503840659e-06,
+      "loss": 1.1701,
+      "step": 1364
+    },
+    {
+      "epoch": 1.7314285714285713,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.423277302638428e-06,
+      "loss": 1.1493,
+      "step": 1365
+    },
+    {
+      "epoch": 1.7326984126984128,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.4197971205876775e-06,
+      "loss": 1.1577,
+      "step": 1366
+    },
+    {
+      "epoch": 1.733968253968254,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.416315961385597e-06,
+      "loss": 1.1811,
+      "step": 1367
+    },
+    {
+      "epoch": 1.7352380952380952,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.412833828730414e-06,
+      "loss": 1.115,
+      "step": 1368
+    },
+    {
+      "epoch": 1.7365079365079366,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.409350726321392e-06,
+      "loss": 1.1033,
+      "step": 1369
+    },
+    {
+      "epoch": 1.7377777777777776,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.405866657858823e-06,
+      "loss": 1.1589,
+      "step": 1370
+    },
+    {
+      "epoch": 1.7390476190476192,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.4023816270440245e-06,
+      "loss": 1.0999,
+      "step": 1371
+    },
+    {
+      "epoch": 1.7403174603174603,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.398895637579338e-06,
+      "loss": 1.0698,
+      "step": 1372
+    },
+    {
+      "epoch": 1.7415873015873016,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.395408693168123e-06,
+      "loss": 1.1485,
+      "step": 1373
+    },
+    {
+      "epoch": 1.7428571428571429,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.391920797514753e-06,
+      "loss": 1.0692,
+      "step": 1374
+    },
+    {
+      "epoch": 1.744126984126984,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.3884319543246125e-06,
+      "loss": 1.1225,
+      "step": 1375
+    },
+    {
+      "epoch": 1.7453968253968255,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.384942167304093e-06,
+      "loss": 1.1042,
+      "step": 1376
+    },
+    {
+      "epoch": 1.7466666666666666,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.381451440160587e-06,
+      "loss": 1.1989,
+      "step": 1377
+    },
+    {
+      "epoch": 1.747936507936508,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.377959776602488e-06,
+      "loss": 1.1128,
+      "step": 1378
+    },
+    {
+      "epoch": 1.7492063492063492,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.374467180339182e-06,
+      "loss": 0.9888,
+      "step": 1379
+    },
+    {
+      "epoch": 1.7504761904761905,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.370973655081048e-06,
+      "loss": 1.1188,
+      "step": 1380
+    },
+    {
+      "epoch": 1.7517460317460318,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.367479204539452e-06,
+      "loss": 1.1025,
+      "step": 1381
+    },
+    {
+      "epoch": 1.753015873015873,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.363983832426739e-06,
+      "loss": 1.1934,
+      "step": 1382
+    },
+    {
+      "epoch": 1.7542857142857144,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.36048754245624e-06,
+      "loss": 0.9798,
+      "step": 1383
+    },
+    {
+      "epoch": 1.7555555555555555,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.356990338342255e-06,
+      "loss": 1.0822,
+      "step": 1384
+    },
+    {
+      "epoch": 1.7568253968253968,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.353492223800056e-06,
+      "loss": 1.069,
+      "step": 1385
+    },
+    {
+      "epoch": 1.7580952380952382,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.349993202545884e-06,
+      "loss": 1.1006,
+      "step": 1386
+    },
+    {
+      "epoch": 1.7593650793650792,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.346493278296945e-06,
+      "loss": 1.1396,
+      "step": 1387
+    },
+    {
+      "epoch": 1.7606349206349208,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.342992454771401e-06,
+      "loss": 1.0967,
+      "step": 1388
+    },
+    {
+      "epoch": 1.7619047619047619,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.33949073568837e-06,
+      "loss": 1.172,
+      "step": 1389
+    },
+    {
+      "epoch": 1.7631746031746032,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.335988124767919e-06,
+      "loss": 1.1301,
+      "step": 1390
+    },
+    {
+      "epoch": 1.7644444444444445,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.332484625731069e-06,
+      "loss": 1.1052,
+      "step": 1391
+    },
+    {
+      "epoch": 1.7657142857142856,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.328980242299781e-06,
+      "loss": 1.1134,
+      "step": 1392
+    },
+    {
+      "epoch": 1.766984126984127,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.325474978196952e-06,
+      "loss": 1.1061,
+      "step": 1393
+    },
+    {
+      "epoch": 1.7682539682539682,
+      "grad_norm": 1.0,
+      "learning_rate": 4.321968837146421e-06,
+      "loss": 1.0766,
+      "step": 1394
+    },
+    {
+      "epoch": 1.7695238095238095,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.318461822872954e-06,
+      "loss": 1.1707,
+      "step": 1395
+    },
+    {
+      "epoch": 1.7707936507936508,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.314953939102245e-06,
+      "loss": 1.0879,
+      "step": 1396
+    },
+    {
+      "epoch": 1.7720634920634921,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.311445189560915e-06,
+      "loss": 1.0676,
+      "step": 1397
+    },
+    {
+      "epoch": 1.7733333333333334,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.307935577976502e-06,
+      "loss": 1.1716,
+      "step": 1398
+    },
+    {
+      "epoch": 1.7746031746031745,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.3044251080774585e-06,
+      "loss": 1.0704,
+      "step": 1399
+    },
+    {
+      "epoch": 1.7758730158730158,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.300913783593154e-06,
+      "loss": 1.1177,
+      "step": 1400
+    },
+    {
+      "epoch": 1.7771428571428571,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.2974016082538605e-06,
+      "loss": 1.1566,
+      "step": 1401
+    },
+    {
+      "epoch": 1.7784126984126984,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.293888585790755e-06,
+      "loss": 1.1727,
+      "step": 1402
+    },
+    {
+      "epoch": 1.7796825396825398,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.290374719935918e-06,
+      "loss": 1.102,
+      "step": 1403
+    },
+    {
+      "epoch": 1.7809523809523808,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.286860014422323e-06,
+      "loss": 1.0966,
+      "step": 1404
+    },
+    {
+      "epoch": 1.7822222222222224,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.283344472983835e-06,
+      "loss": 1.1441,
+      "step": 1405
+    },
+    {
+      "epoch": 1.7834920634920635,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.279828099355209e-06,
+      "loss": 1.1356,
+      "step": 1406
+    },
+    {
+      "epoch": 1.7847619047619048,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.2763108972720815e-06,
+      "loss": 1.0482,
+      "step": 1407
+    },
+    {
+      "epoch": 1.786031746031746,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.272792870470971e-06,
+      "loss": 1.0668,
+      "step": 1408
+    },
+    {
+      "epoch": 1.7873015873015872,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.269274022689275e-06,
+      "loss": 1.0832,
+      "step": 1409
+    },
+    {
+      "epoch": 1.7885714285714287,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.265754357665256e-06,
+      "loss": 1.0726,
+      "step": 1410
+    },
+    {
+      "epoch": 1.7898412698412698,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.26223387913805e-06,
+      "loss": 1.106,
+      "step": 1411
+    },
+    {
+      "epoch": 1.791111111111111,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.258712590847656e-06,
+      "loss": 1.0797,
+      "step": 1412
+    },
+    {
+      "epoch": 1.7923809523809524,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.255190496534934e-06,
+      "loss": 1.173,
+      "step": 1413
+    },
+    {
+      "epoch": 1.7936507936507935,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.251667599941599e-06,
+      "loss": 1.1476,
+      "step": 1414
+    },
+    {
+      "epoch": 1.794920634920635,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.248143904810219e-06,
+      "loss": 1.0581,
+      "step": 1415
+    },
+    {
+      "epoch": 1.7961904761904761,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.2446194148842095e-06,
+      "loss": 1.142,
+      "step": 1416
+    },
+    {
+      "epoch": 1.7974603174603174,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.241094133907832e-06,
+      "loss": 1.1658,
+      "step": 1417
+    },
+    {
+      "epoch": 1.7987301587301587,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.2375680656261864e-06,
+      "loss": 1.0855,
+      "step": 1418
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.23404121378521e-06,
+      "loss": 1.0656,
+      "step": 1419
+    },
+    {
+      "epoch": 1.8012698412698414,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.230513582131674e-06,
+      "loss": 1.1448,
+      "step": 1420
+    },
+    {
+      "epoch": 1.8025396825396824,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.2269851744131735e-06,
+      "loss": 1.1289,
+      "step": 1421
+    },
+    {
+      "epoch": 1.803809523809524,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.223455994378133e-06,
+      "loss": 1.1863,
+      "step": 1422
+    },
+    {
+      "epoch": 1.805079365079365,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.219926045775797e-06,
+      "loss": 1.0682,
+      "step": 1423
+    },
+    {
+      "epoch": 1.8063492063492064,
+      "grad_norm": 1.0,
+      "learning_rate": 4.216395332356223e-06,
+      "loss": 1.1456,
+      "step": 1424
+    },
+    {
+      "epoch": 1.8076190476190477,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.212863857870283e-06,
+      "loss": 1.0884,
+      "step": 1425
+    },
+    {
+      "epoch": 1.8088888888888888,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.209331626069658e-06,
+      "loss": 1.2046,
+      "step": 1426
+    },
+    {
+      "epoch": 1.8101587301587303,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.2057986407068335e-06,
+      "loss": 1.113,
+      "step": 1427
+    },
+    {
+      "epoch": 1.8114285714285714,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.202264905535095e-06,
+      "loss": 1.0876,
+      "step": 1428
+    },
+    {
+      "epoch": 1.8126984126984127,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.198730424308525e-06,
+      "loss": 1.0892,
+      "step": 1429
+    },
+    {
+      "epoch": 1.813968253968254,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.195195200781997e-06,
+      "loss": 1.1297,
+      "step": 1430
+    },
+    {
+      "epoch": 1.815238095238095,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.191659238711176e-06,
+      "loss": 1.0367,
+      "step": 1431
+    },
+    {
+      "epoch": 1.8165079365079366,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.18812254185251e-06,
+      "loss": 1.0728,
+      "step": 1432
+    },
+    {
+      "epoch": 1.8177777777777777,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.184585113963227e-06,
+      "loss": 1.0955,
+      "step": 1433
+    },
+    {
+      "epoch": 1.819047619047619,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.181046958801329e-06,
+      "loss": 1.0745,
+      "step": 1434
+    },
+    {
+      "epoch": 1.8203174603174603,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.1775080801256005e-06,
+      "loss": 1.107,
+      "step": 1435
+    },
+    {
+      "epoch": 1.8215873015873014,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.173968481695583e-06,
+      "loss": 1.1076,
+      "step": 1436
+    },
+    {
+      "epoch": 1.822857142857143,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.1704281672715885e-06,
+      "loss": 1.0636,
+      "step": 1437
+    },
+    {
+      "epoch": 1.824126984126984,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.1668871406146895e-06,
+      "loss": 1.1531,
+      "step": 1438
+    },
+    {
+      "epoch": 1.8253968253968254,
+      "grad_norm": 1.0,
+      "learning_rate": 4.163345405486714e-06,
+      "loss": 1.0968,
+      "step": 1439
+    },
+    {
+      "epoch": 1.8266666666666667,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.1598029656502435e-06,
+      "loss": 1.1005,
+      "step": 1440
+    },
+    {
+      "epoch": 1.827936507936508,
+      "grad_norm": 1.0,
+      "learning_rate": 4.156259824868607e-06,
+      "loss": 1.1411,
+      "step": 1441
+    },
+    {
+      "epoch": 1.8292063492063493,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.15271598690588e-06,
+      "loss": 1.1068,
+      "step": 1442
+    },
+    {
+      "epoch": 1.8304761904761904,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.149171455526874e-06,
+      "loss": 1.1175,
+      "step": 1443
+    },
+    {
+      "epoch": 1.831746031746032,
+      "grad_norm": 1.0,
+      "learning_rate": 4.1456262344971455e-06,
+      "loss": 1.102,
+      "step": 1444
+    },
+    {
+      "epoch": 1.833015873015873,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.142080327582976e-06,
+      "loss": 1.2087,
+      "step": 1445
+    },
+    {
+      "epoch": 1.8342857142857143,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.138533738551379e-06,
+      "loss": 1.1595,
+      "step": 1446
+    },
+    {
+      "epoch": 1.8355555555555556,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.134986471170092e-06,
+      "loss": 1.0973,
+      "step": 1447
+    },
+    {
+      "epoch": 1.8368253968253967,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.1314385292075725e-06,
+      "loss": 1.0418,
+      "step": 1448
+    },
+    {
+      "epoch": 1.8380952380952382,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.127889916432997e-06,
+      "loss": 1.1054,
+      "step": 1449
+    },
+    {
+      "epoch": 1.8393650793650793,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.124340636616252e-06,
+      "loss": 1.0538,
+      "step": 1450
+    },
+    {
+      "epoch": 1.8406349206349206,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.120790693527933e-06,
+      "loss": 1.1456,
+      "step": 1451
+    },
+    {
+      "epoch": 1.841904761904762,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.1172400909393414e-06,
+      "loss": 1.1267,
+      "step": 1452
+    },
+    {
+      "epoch": 1.843174603174603,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.1136888326224805e-06,
+      "loss": 1.1251,
+      "step": 1453
+    },
+    {
+      "epoch": 1.8444444444444446,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.110136922350045e-06,
+      "loss": 1.1643,
+      "step": 1454
+    },
+    {
+      "epoch": 1.8457142857142856,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.106584363895429e-06,
+      "loss": 1.0467,
+      "step": 1455
+    },
+    {
+      "epoch": 1.846984126984127,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.103031161032708e-06,
+      "loss": 1.0191,
+      "step": 1456
+    },
+    {
+      "epoch": 1.8482539682539683,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.099477317536649e-06,
+      "loss": 1.1017,
+      "step": 1457
+    },
+    {
+      "epoch": 1.8495238095238096,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.0959228371826955e-06,
+      "loss": 1.1175,
+      "step": 1458
+    },
+    {
+      "epoch": 1.8507936507936509,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.092367723746968e-06,
+      "loss": 1.0939,
+      "step": 1459
+    },
+    {
+      "epoch": 1.852063492063492,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.0888119810062594e-06,
+      "loss": 1.1079,
+      "step": 1460
+    },
+    {
+      "epoch": 1.8533333333333335,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.085255612738032e-06,
+      "loss": 1.1314,
+      "step": 1461
+    },
+    {
+      "epoch": 1.8546031746031746,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.081698622720414e-06,
+      "loss": 1.0679,
+      "step": 1462
+    },
+    {
+      "epoch": 1.855873015873016,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.078141014732191e-06,
+      "loss": 1.0438,
+      "step": 1463
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.074582792552808e-06,
+      "loss": 1.0791,
+      "step": 1464
+    },
+    {
+      "epoch": 1.8584126984126983,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.0710239599623606e-06,
+      "loss": 1.0715,
+      "step": 1465
+    },
+    {
+      "epoch": 1.8596825396825398,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.067464520741591e-06,
+      "loss": 1.1739,
+      "step": 1466
+    },
+    {
+      "epoch": 1.860952380952381,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.0639044786718925e-06,
+      "loss": 1.0591,
+      "step": 1467
+    },
+    {
+      "epoch": 1.8622222222222222,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.060343837535292e-06,
+      "loss": 1.1822,
+      "step": 1468
+    },
+    {
+      "epoch": 1.8634920634920635,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.056782601114454e-06,
+      "loss": 1.1192,
+      "step": 1469
+    },
+    {
+      "epoch": 1.8647619047619046,
+      "grad_norm": 1.0,
+      "learning_rate": 4.053220773192681e-06,
+      "loss": 1.1677,
+      "step": 1470
+    },
+    {
+      "epoch": 1.8660317460317462,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.049658357553897e-06,
+      "loss": 1.0656,
+      "step": 1471
+    },
+    {
+      "epoch": 1.8673015873015872,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.046095357982652e-06,
+      "loss": 1.0669,
+      "step": 1472
+    },
+    {
+      "epoch": 1.8685714285714285,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.042531778264121e-06,
+      "loss": 1.1329,
+      "step": 1473
+    },
+    {
+      "epoch": 1.8698412698412699,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.038967622184088e-06,
+      "loss": 1.0856,
+      "step": 1474
+    },
+    {
+      "epoch": 1.871111111111111,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.035402893528955e-06,
+      "loss": 1.1338,
+      "step": 1475
+    },
+    {
+      "epoch": 1.8723809523809525,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.03183759608573e-06,
+      "loss": 1.0468,
+      "step": 1476
+    },
+    {
+      "epoch": 1.8736507936507936,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.0282717336420254e-06,
+      "loss": 1.1713,
+      "step": 1477
+    },
+    {
+      "epoch": 1.8749206349206349,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.024705309986054e-06,
+      "loss": 1.1247,
+      "step": 1478
+    },
+    {
+      "epoch": 1.8761904761904762,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.021138328906626e-06,
+      "loss": 1.0776,
+      "step": 1479
+    },
+    {
+      "epoch": 1.8774603174603175,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.017570794193141e-06,
+      "loss": 1.0879,
+      "step": 1480
+    },
+    {
+      "epoch": 1.8787301587301588,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.014002709635589e-06,
+      "loss": 1.13,
+      "step": 1481
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.010434079024543e-06,
+      "loss": 1.1174,
+      "step": 1482
+    },
+    {
+      "epoch": 1.8812698412698414,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.006864906151157e-06,
+      "loss": 1.1313,
+      "step": 1483
+    },
+    {
+      "epoch": 1.8825396825396825,
+      "grad_norm": 1.0,
+      "learning_rate": 4.003295194807162e-06,
+      "loss": 1.1686,
+      "step": 1484
+    },
+    {
+      "epoch": 1.8838095238095238,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.999724948784857e-06,
+      "loss": 1.0579,
+      "step": 1485
+    },
+    {
+      "epoch": 1.8850793650793651,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.996154171877113e-06,
+      "loss": 1.1215,
+      "step": 1486
+    },
+    {
+      "epoch": 1.8863492063492062,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.992582867877363e-06,
+      "loss": 1.1345,
+      "step": 1487
+    },
+    {
+      "epoch": 1.8876190476190478,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.989011040579602e-06,
+      "loss": 1.0882,
+      "step": 1488
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.98543869377838e-06,
+      "loss": 1.1399,
+      "step": 1489
+    },
+    {
+      "epoch": 1.8901587301587301,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.981865831268797e-06,
+      "loss": 1.1774,
+      "step": 1490
+    },
+    {
+      "epoch": 1.8914285714285715,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.978292456846502e-06,
+      "loss": 1.1639,
+      "step": 1491
+    },
+    {
+      "epoch": 1.8926984126984125,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.974718574307689e-06,
+      "loss": 1.1198,
+      "step": 1492
+    },
+    {
+      "epoch": 1.893968253968254,
+      "grad_norm": 1.1171875,
+      "learning_rate": 3.971144187449093e-06,
+      "loss": 1.0223,
+      "step": 1493
+    },
+    {
+      "epoch": 1.8952380952380952,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.967569300067979e-06,
+      "loss": 1.1347,
+      "step": 1494
+    },
+    {
+      "epoch": 1.8965079365079365,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.96399391596215e-06,
+      "loss": 1.1409,
+      "step": 1495
+    },
+    {
+      "epoch": 1.8977777777777778,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.960418038929933e-06,
+      "loss": 1.0625,
+      "step": 1496
+    },
+    {
+      "epoch": 1.899047619047619,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.956841672770181e-06,
+      "loss": 1.0833,
+      "step": 1497
+    },
+    {
+      "epoch": 1.9003174603174604,
+      "grad_norm": 1.1015625,
+      "learning_rate": 3.953264821282264e-06,
+      "loss": 1.182,
+      "step": 1498
+    },
+    {
+      "epoch": 1.9015873015873015,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.949687488266071e-06,
+      "loss": 1.1809,
+      "step": 1499
+    },
+    {
+      "epoch": 1.9028571428571428,
+      "grad_norm": 1.1640625,
+      "learning_rate": 3.946109677521999e-06,
+      "loss": 1.2017,
+      "step": 1500
+    },
+    {
+      "epoch": 1.9041269841269841,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.942531392850953e-06,
+      "loss": 1.1001,
+      "step": 1501
+    },
+    {
+      "epoch": 1.9053968253968254,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.938952638054347e-06,
+      "loss": 1.1442,
+      "step": 1502
+    },
+    {
+      "epoch": 1.9066666666666667,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.935373416934086e-06,
+      "loss": 1.1296,
+      "step": 1503
+    },
+    {
+      "epoch": 1.9079365079365078,
+      "grad_norm": 1.15625,
+      "learning_rate": 3.931793733292575e-06,
+      "loss": 1.1029,
+      "step": 1504
+    },
+    {
+      "epoch": 1.9092063492063494,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.928213590932712e-06,
+      "loss": 1.1349,
+      "step": 1505
+    },
+    {
+      "epoch": 1.9104761904761904,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.924632993657879e-06,
+      "loss": 1.0712,
+      "step": 1506
+    },
+    {
+      "epoch": 1.9117460317460317,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.921051945271941e-06,
+      "loss": 1.074,
+      "step": 1507
+    },
+    {
+      "epoch": 1.913015873015873,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.917470449579244e-06,
+      "loss": 1.1432,
+      "step": 1508
+    },
+    {
+      "epoch": 1.9142857142857141,
+      "grad_norm": 1.1640625,
+      "learning_rate": 3.913888510384612e-06,
+      "loss": 1.1314,
+      "step": 1509
+    },
+    {
+      "epoch": 1.9155555555555557,
+      "grad_norm": 1.25,
+      "learning_rate": 3.910306131493335e-06,
+      "loss": 1.0772,
+      "step": 1510
+    },
+    {
+      "epoch": 1.9168253968253968,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.90672331671117e-06,
+      "loss": 1.1526,
+      "step": 1511
+    },
+    {
+      "epoch": 1.918095238095238,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.9031400698443414e-06,
+      "loss": 1.0993,
+      "step": 1512
+    },
+    {
+      "epoch": 1.9193650793650794,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.8995563946995305e-06,
+      "loss": 1.0668,
+      "step": 1513
+    },
+    {
+      "epoch": 1.9206349206349205,
+      "grad_norm": 1.296875,
+      "learning_rate": 3.8959722950838745e-06,
+      "loss": 1.1875,
+      "step": 1514
+    },
+    {
+      "epoch": 1.921904761904762,
+      "grad_norm": 1.25,
+      "learning_rate": 3.8923877748049575e-06,
+      "loss": 1.1899,
+      "step": 1515
+    },
+    {
+      "epoch": 1.923174603174603,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.888802837670816e-06,
+      "loss": 1.1052,
+      "step": 1516
+    },
+    {
+      "epoch": 1.9244444444444444,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.885217487489926e-06,
+      "loss": 1.0402,
+      "step": 1517
+    },
+    {
+      "epoch": 1.9257142857142857,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.8816317280712026e-06,
+      "loss": 1.0479,
+      "step": 1518
+    },
+    {
+      "epoch": 1.926984126984127,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.878045563223999e-06,
+      "loss": 1.1563,
+      "step": 1519
+    },
+    {
+      "epoch": 1.9282539682539683,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.8744589967580935e-06,
+      "loss": 1.1338,
+      "step": 1520
+    },
+    {
+      "epoch": 1.9295238095238094,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.870872032483695e-06,
+      "loss": 1.0558,
+      "step": 1521
+    },
+    {
+      "epoch": 1.930793650793651,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.8672846742114325e-06,
+      "loss": 1.0867,
+      "step": 1522
+    },
+    {
+      "epoch": 1.932063492063492,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.863696925752357e-06,
+      "loss": 1.1178,
+      "step": 1523
+    },
+    {
+      "epoch": 1.9333333333333333,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.860108790917931e-06,
+      "loss": 1.1162,
+      "step": 1524
+    },
+    {
+      "epoch": 1.9346031746031747,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.85652027352003e-06,
+      "loss": 1.1278,
+      "step": 1525
+    },
+    {
+      "epoch": 1.9358730158730157,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.8529313773709324e-06,
+      "loss": 1.0379,
+      "step": 1526
+    },
+    {
+      "epoch": 1.9371428571428573,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.849342106283322e-06,
+      "loss": 1.0832,
+      "step": 1527
+    },
+    {
+      "epoch": 1.9384126984126984,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.84575246407028e-06,
+      "loss": 1.153,
+      "step": 1528
+    },
+    {
+      "epoch": 1.9396825396825397,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.8421624545452815e-06,
+      "loss": 1.1135,
+      "step": 1529
+    },
+    {
+      "epoch": 1.940952380952381,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.838572081522191e-06,
+      "loss": 1.029,
+      "step": 1530
+    },
+    {
+      "epoch": 1.942222222222222,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.834981348815262e-06,
+      "loss": 1.0614,
+      "step": 1531
+    },
+    {
+      "epoch": 1.9434920634920636,
+      "grad_norm": 1.140625,
+      "learning_rate": 3.831390260239128e-06,
+      "loss": 1.1094,
+      "step": 1532
+    },
+    {
+      "epoch": 1.9447619047619047,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.8277988196088e-06,
+      "loss": 1.0785,
+      "step": 1533
+    },
+    {
+      "epoch": 1.946031746031746,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.824207030739665e-06,
+      "loss": 1.0686,
+      "step": 1534
+    },
+    {
+      "epoch": 1.9473015873015873,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.820614897447478e-06,
+      "loss": 1.065,
+      "step": 1535
+    },
+    {
+      "epoch": 1.9485714285714286,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.81702242354836e-06,
+      "loss": 1.0723,
+      "step": 1536
+    },
+    {
+      "epoch": 1.94984126984127,
+      "grad_norm": 1.0,
+      "learning_rate": 3.813429612858798e-06,
+      "loss": 1.1189,
+      "step": 1537
+    },
+    {
+      "epoch": 1.951111111111111,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.809836469195631e-06,
+      "loss": 1.0674,
+      "step": 1538
+    },
+    {
+      "epoch": 1.9523809523809523,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.806242996376052e-06,
+      "loss": 1.0869,
+      "step": 1539
+    },
+    {
+      "epoch": 1.9536507936507936,
+      "grad_norm": 1.0,
+      "learning_rate": 3.8026491982176097e-06,
+      "loss": 1.1377,
+      "step": 1540
+    },
+    {
+      "epoch": 1.954920634920635,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.799055078538194e-06,
+      "loss": 1.1337,
+      "step": 1541
+    },
+    {
+      "epoch": 1.9561904761904763,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.7954606411560347e-06,
+      "loss": 1.1662,
+      "step": 1542
+    },
+    {
+      "epoch": 1.9574603174603173,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.791865889889702e-06,
+      "loss": 1.0977,
+      "step": 1543
+    },
+    {
+      "epoch": 1.9587301587301589,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.788270828558101e-06,
+      "loss": 1.0802,
+      "step": 1544
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.784675460980462e-06,
+      "loss": 1.0882,
+      "step": 1545
+    },
+    {
+      "epoch": 1.9612698412698413,
+      "grad_norm": 1.0,
+      "learning_rate": 3.781079790976344e-06,
+      "loss": 1.0936,
+      "step": 1546
+    },
+    {
+      "epoch": 1.9625396825396826,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.777483822365624e-06,
+      "loss": 1.0968,
+      "step": 1547
+    },
+    {
+      "epoch": 1.9638095238095237,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.7738875589685004e-06,
+      "loss": 1.1422,
+      "step": 1548
+    },
+    {
+      "epoch": 1.9650793650793652,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.7702910046054817e-06,
+      "loss": 1.0776,
+      "step": 1549
+    },
+    {
+      "epoch": 1.9663492063492063,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.766694163097386e-06,
+      "loss": 1.1129,
+      "step": 1550
+    },
+    {
+      "epoch": 1.9676190476190476,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.7630970382653382e-06,
+      "loss": 1.1185,
+      "step": 1551
+    },
+    {
+      "epoch": 1.968888888888889,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.7594996339307617e-06,
+      "loss": 1.1754,
+      "step": 1552
+    },
+    {
+      "epoch": 1.97015873015873,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.7559019539153776e-06,
+      "loss": 1.1038,
+      "step": 1553
+    },
+    {
+      "epoch": 1.9714285714285715,
+      "grad_norm": 1.0,
+      "learning_rate": 3.752304002041203e-06,
+      "loss": 1.103,
+      "step": 1554
+    },
+    {
+      "epoch": 1.9726984126984126,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.748705782130538e-06,
+      "loss": 1.0537,
+      "step": 1555
+    },
+    {
+      "epoch": 1.973968253968254,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.7451072980059715e-06,
+      "loss": 1.1092,
+      "step": 1556
+    },
+    {
+      "epoch": 1.9752380952380952,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.7415085534903725e-06,
+      "loss": 1.0675,
+      "step": 1557
+    },
+    {
+      "epoch": 1.9765079365079365,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.737909552406886e-06,
+      "loss": 1.0737,
+      "step": 1558
+    },
+    {
+      "epoch": 1.9777777777777779,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.73431029857893e-06,
+      "loss": 1.1477,
+      "step": 1559
+    },
+    {
+      "epoch": 1.979047619047619,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.7307107958301905e-06,
+      "loss": 1.1316,
+      "step": 1560
+    },
+    {
+      "epoch": 1.9803174603174605,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.727111047984618e-06,
+      "loss": 1.1262,
+      "step": 1561
+    },
+    {
+      "epoch": 1.9815873015873016,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.723511058866424e-06,
+      "loss": 1.14,
+      "step": 1562
+    },
+    {
+      "epoch": 1.9828571428571429,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.7199108323000767e-06,
+      "loss": 1.1392,
+      "step": 1563
+    },
+    {
+      "epoch": 1.9841269841269842,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.716310372110295e-06,
+      "loss": 1.096,
+      "step": 1564
+    },
+    {
+      "epoch": 1.9853968253968253,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.7127096821220467e-06,
+      "loss": 1.0792,
+      "step": 1565
+    },
+    {
+      "epoch": 1.9866666666666668,
+      "grad_norm": 1.0,
+      "learning_rate": 3.7091087661605437e-06,
+      "loss": 1.1489,
+      "step": 1566
+    },
+    {
+      "epoch": 1.987936507936508,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.7055076280512384e-06,
+      "loss": 1.14,
+      "step": 1567
+    },
+    {
+      "epoch": 1.9892063492063492,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.701906271619819e-06,
+      "loss": 1.1288,
+      "step": 1568
+    },
+    {
+      "epoch": 1.9904761904761905,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.6983047006922063e-06,
+      "loss": 1.0907,
+      "step": 1569
+    },
+    {
+      "epoch": 1.9917460317460316,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.694702919094546e-06,
+      "loss": 1.1328,
+      "step": 1570
+    },
+    {
+      "epoch": 1.9930158730158731,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.691100930653212e-06,
+      "loss": 1.1367,
+      "step": 1571
+    },
+    {
+      "epoch": 1.9942857142857142,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.6874987391947947e-06,
+      "loss": 1.043,
+      "step": 1572
+    },
+    {
+      "epoch": 1.9955555555555555,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.683896348546102e-06,
+      "loss": 1.1243,
+      "step": 1573
+    },
+    {
+      "epoch": 1.9968253968253968,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.6802937625341506e-06,
+      "loss": 1.1258,
+      "step": 1574
+    },
+    {
+      "epoch": 1.9980952380952381,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.676690984986169e-06,
+      "loss": 1.164,
+      "step": 1575
+    },
+    {
+      "epoch": 1.9993650793650795,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.673088019729587e-06,
+      "loss": 1.1386,
+      "step": 1576
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.669484870592032e-06,
+      "loss": 0.5283,
+      "step": 1577
+    },
+    {
+      "epoch": 2.0006349206349205,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.665881541401329e-06,
+      "loss": 0.5619,
+      "step": 1578
+    },
+    {
+      "epoch": 2.0006349206349205,
+      "eval_loss": 1.1575638055801392,
+      "eval_runtime": 89.0193,
+      "eval_samples_per_second": 48.697,
+      "eval_steps_per_second": 6.089,
+      "step": 1578
+    },
+    {
+      "epoch": 2.001269841269841,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.6622780359854946e-06,
+      "loss": 1.1188,
+      "step": 1579
+    },
+    {
+      "epoch": 2.0025396825396826,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.6586743581727326e-06,
+      "loss": 1.1121,
+      "step": 1580
+    },
+    {
+      "epoch": 2.0038095238095237,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.6550705117914276e-06,
+      "loss": 1.0873,
+      "step": 1581
+    },
+    {
+      "epoch": 2.0050793650793652,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.6514665006701447e-06,
+      "loss": 1.1009,
+      "step": 1582
+    },
+    {
+      "epoch": 2.0063492063492063,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.6478623286376248e-06,
+      "loss": 1.0493,
+      "step": 1583
+    },
+    {
+      "epoch": 2.0076190476190474,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.6442579995227805e-06,
+      "loss": 1.1485,
+      "step": 1584
+    },
+    {
+      "epoch": 2.008888888888889,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.6406535171546864e-06,
+      "loss": 1.1112,
+      "step": 1585
+    },
+    {
+      "epoch": 2.01015873015873,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.637048885362587e-06,
+      "loss": 1.1392,
+      "step": 1586
+    },
+    {
+      "epoch": 2.0114285714285716,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.6334441079758788e-06,
+      "loss": 1.1524,
+      "step": 1587
+    },
+    {
+      "epoch": 2.0126984126984127,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.629839188824118e-06,
+      "loss": 1.0846,
+      "step": 1588
+    },
+    {
+      "epoch": 2.013968253968254,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.62623413173701e-06,
+      "loss": 1.0632,
+      "step": 1589
+    },
+    {
+      "epoch": 2.0152380952380953,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.6226289405444057e-06,
+      "loss": 1.0533,
+      "step": 1590
+    },
+    {
+      "epoch": 2.0165079365079364,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.619023619076297e-06,
+      "loss": 1.1303,
+      "step": 1591
+    },
+    {
+      "epoch": 2.017777777777778,
+      "grad_norm": 1.0,
+      "learning_rate": 3.6154181711628185e-06,
+      "loss": 1.1772,
+      "step": 1592
+    },
+    {
+      "epoch": 2.019047619047619,
+      "grad_norm": 1.0,
+      "learning_rate": 3.611812600634237e-06,
+      "loss": 1.0997,
+      "step": 1593
+    },
+    {
+      "epoch": 2.0203174603174605,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.6082069113209474e-06,
+      "loss": 1.1378,
+      "step": 1594
+    },
+    {
+      "epoch": 2.0215873015873016,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.6046011070534754e-06,
+      "loss": 1.0801,
+      "step": 1595
+    },
+    {
+      "epoch": 2.0228571428571427,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.6009951916624636e-06,
+      "loss": 1.0544,
+      "step": 1596
+    },
+    {
+      "epoch": 2.024126984126984,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.5973891689786764e-06,
+      "loss": 1.1424,
+      "step": 1597
+    },
+    {
+      "epoch": 2.0253968253968253,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.593783042832992e-06,
+      "loss": 1.1791,
+      "step": 1598
+    },
+    {
+      "epoch": 2.026666666666667,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.5901768170563963e-06,
+      "loss": 1.1071,
+      "step": 1599
+    },
+    {
+      "epoch": 2.027936507936508,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.5865704954799826e-06,
+      "loss": 1.1094,
+      "step": 1600
+    },
+    {
+      "epoch": 2.029206349206349,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.582964081934946e-06,
+      "loss": 1.0959,
+      "step": 1601
+    },
+    {
+      "epoch": 2.0304761904761905,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.5793575802525783e-06,
+      "loss": 1.1588,
+      "step": 1602
+    },
+    {
+      "epoch": 2.0317460317460316,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.5757509942642663e-06,
+      "loss": 1.1089,
+      "step": 1603
+    },
+    {
+      "epoch": 2.033015873015873,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.5721443278014855e-06,
+      "loss": 1.1385,
+      "step": 1604
+    },
+    {
+      "epoch": 2.0342857142857143,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.5685375846957967e-06,
+      "loss": 1.1326,
+      "step": 1605
+    },
+    {
+      "epoch": 2.0355555555555553,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.564930768778842e-06,
+      "loss": 1.156,
+      "step": 1606
+    },
+    {
+      "epoch": 2.036825396825397,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.561323883882344e-06,
+      "loss": 1.0762,
+      "step": 1607
+    },
+    {
+      "epoch": 2.038095238095238,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.557716933838093e-06,
+      "loss": 1.1483,
+      "step": 1608
+    },
+    {
+      "epoch": 2.0393650793650795,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.5541099224779517e-06,
+      "loss": 1.1206,
+      "step": 1609
+    },
+    {
+      "epoch": 2.0406349206349206,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.55050285363385e-06,
+      "loss": 1.0709,
+      "step": 1610
+    },
+    {
+      "epoch": 2.041904761904762,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.5468957311377735e-06,
+      "loss": 1.0941,
+      "step": 1611
+    },
+    {
+      "epoch": 2.043174603174603,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.54328855882177e-06,
+      "loss": 1.051,
+      "step": 1612
+    },
+    {
+      "epoch": 2.0444444444444443,
+      "grad_norm": 0.875,
+      "learning_rate": 3.5396813405179376e-06,
+      "loss": 1.0019,
+      "step": 1613
+    },
+    {
+      "epoch": 2.045714285714286,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.536074080058423e-06,
+      "loss": 1.1298,
+      "step": 1614
+    },
+    {
+      "epoch": 2.046984126984127,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.5324667812754193e-06,
+      "loss": 1.1687,
+      "step": 1615
+    },
+    {
+      "epoch": 2.0482539682539684,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.5288594480011592e-06,
+      "loss": 1.0941,
+      "step": 1616
+    },
+    {
+      "epoch": 2.0495238095238095,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.525252084067912e-06,
+      "loss": 1.1288,
+      "step": 1617
+    },
+    {
+      "epoch": 2.0507936507936506,
+      "grad_norm": 1.1015625,
+      "learning_rate": 3.5216446933079787e-06,
+      "loss": 1.0944,
+      "step": 1618
+    },
+    {
+      "epoch": 2.052063492063492,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.518037279553693e-06,
+      "loss": 1.1013,
+      "step": 1619
+    },
+    {
+      "epoch": 2.0533333333333332,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.5144298466374074e-06,
+      "loss": 1.1078,
+      "step": 1620
+    },
+    {
+      "epoch": 2.0546031746031748,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.5108223983914974e-06,
+      "loss": 1.1235,
+      "step": 1621
+    },
+    {
+      "epoch": 2.055873015873016,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.5072149386483557e-06,
+      "loss": 1.1266,
+      "step": 1622
+    },
+    {
+      "epoch": 2.057142857142857,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.5036074712403857e-06,
+      "loss": 1.1639,
+      "step": 1623
+    },
+    {
+      "epoch": 2.0584126984126985,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.5e-06,
+      "loss": 1.1113,
+      "step": 1624
+    },
+    {
+      "epoch": 2.0596825396825396,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.4963925287596137e-06,
+      "loss": 1.1378,
+      "step": 1625
+    },
+    {
+      "epoch": 2.060952380952381,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.492785061351644e-06,
+      "loss": 1.0979,
+      "step": 1626
+    },
+    {
+      "epoch": 2.062222222222222,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.4891776016085025e-06,
+      "loss": 1.066,
+      "step": 1627
+    },
+    {
+      "epoch": 2.0634920634920633,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.4855701533625925e-06,
+      "loss": 1.1784,
+      "step": 1628
+    },
+    {
+      "epoch": 2.064761904761905,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.4819627204463077e-06,
+      "loss": 1.0794,
+      "step": 1629
+    },
+    {
+      "epoch": 2.066031746031746,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.478355306692021e-06,
+      "loss": 1.0571,
+      "step": 1630
+    },
+    {
+      "epoch": 2.0673015873015874,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.4747479159320884e-06,
+      "loss": 1.0963,
+      "step": 1631
+    },
+    {
+      "epoch": 2.0685714285714285,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.4711405519988415e-06,
+      "loss": 1.1355,
+      "step": 1632
+    },
+    {
+      "epoch": 2.06984126984127,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.467533218724581e-06,
+      "loss": 1.193,
+      "step": 1633
+    },
+    {
+      "epoch": 2.071111111111111,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.4639259199415767e-06,
+      "loss": 1.1166,
+      "step": 1634
+    },
+    {
+      "epoch": 2.072380952380952,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.4603186594820627e-06,
+      "loss": 1.0351,
+      "step": 1635
+    },
+    {
+      "epoch": 2.0736507936507937,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.45671144117823e-06,
+      "loss": 1.1471,
+      "step": 1636
+    },
+    {
+      "epoch": 2.074920634920635,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.453104268862226e-06,
+      "loss": 1.0593,
+      "step": 1637
+    },
+    {
+      "epoch": 2.0761904761904764,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.4494971463661514e-06,
+      "loss": 1.1042,
+      "step": 1638
+    },
+    {
+      "epoch": 2.0774603174603175,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.4458900775220487e-06,
+      "loss": 1.0667,
+      "step": 1639
+    },
+    {
+      "epoch": 2.0787301587301585,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.4422830661619072e-06,
+      "loss": 1.0711,
+      "step": 1640
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.438676116117657e-06,
+      "loss": 1.1181,
+      "step": 1641
+    },
+    {
+      "epoch": 2.081269841269841,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.435069231221158e-06,
+      "loss": 1.0383,
+      "step": 1642
+    },
+    {
+      "epoch": 2.0825396825396827,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.431462415304204e-06,
+      "loss": 1.0495,
+      "step": 1643
+    },
+    {
+      "epoch": 2.083809523809524,
+      "grad_norm": 1.0,
+      "learning_rate": 3.4278556721985157e-06,
+      "loss": 1.0865,
+      "step": 1644
+    },
+    {
+      "epoch": 2.085079365079365,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.424249005735734e-06,
+      "loss": 1.0705,
+      "step": 1645
+    },
+    {
+      "epoch": 2.0863492063492064,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.420642419747422e-06,
+      "loss": 1.0912,
+      "step": 1646
+    },
+    {
+      "epoch": 2.0876190476190475,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.4170359180650556e-06,
+      "loss": 1.0855,
+      "step": 1647
+    },
+    {
+      "epoch": 2.088888888888889,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.4134295045200173e-06,
+      "loss": 1.1909,
+      "step": 1648
+    },
+    {
+      "epoch": 2.09015873015873,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.4098231829436036e-06,
+      "loss": 1.1589,
+      "step": 1649
+    },
+    {
+      "epoch": 2.0914285714285716,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.4062169571670087e-06,
+      "loss": 1.1057,
+      "step": 1650
+    },
+    {
+      "epoch": 2.0926984126984127,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.402610831021324e-06,
+      "loss": 1.1359,
+      "step": 1651
+    },
+    {
+      "epoch": 2.093968253968254,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.3990048083375368e-06,
+      "loss": 1.145,
+      "step": 1652
+    },
+    {
+      "epoch": 2.0952380952380953,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.395398892946526e-06,
+      "loss": 1.1437,
+      "step": 1653
+    },
+    {
+      "epoch": 2.0965079365079364,
+      "grad_norm": 1.0,
+      "learning_rate": 3.391793088679053e-06,
+      "loss": 1.1066,
+      "step": 1654
+    },
+    {
+      "epoch": 2.097777777777778,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.3881873993657634e-06,
+      "loss": 1.1463,
+      "step": 1655
+    },
+    {
+      "epoch": 2.099047619047619,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.384581828837181e-06,
+      "loss": 1.0738,
+      "step": 1656
+    },
+    {
+      "epoch": 2.10031746031746,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.380976380923703e-06,
+      "loss": 1.1684,
+      "step": 1657
+    },
+    {
+      "epoch": 2.1015873015873017,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.3773710594555947e-06,
+      "loss": 1.1076,
+      "step": 1658
+    },
+    {
+      "epoch": 2.1028571428571428,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.373765868262989e-06,
+      "loss": 1.1151,
+      "step": 1659
+    },
+    {
+      "epoch": 2.1041269841269843,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.370160811175882e-06,
+      "loss": 1.0252,
+      "step": 1660
+    },
+    {
+      "epoch": 2.1053968253968254,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.3665558920241216e-06,
+      "loss": 1.0858,
+      "step": 1661
+    },
+    {
+      "epoch": 2.1066666666666665,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.362951114637413e-06,
+      "loss": 1.0822,
+      "step": 1662
+    },
+    {
+      "epoch": 2.107936507936508,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.359346482845314e-06,
+      "loss": 1.0881,
+      "step": 1663
+    },
+    {
+      "epoch": 2.109206349206349,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.3557420004772203e-06,
+      "loss": 1.0764,
+      "step": 1664
+    },
+    {
+      "epoch": 2.1104761904761906,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.3521376713623743e-06,
+      "loss": 1.0295,
+      "step": 1665
+    },
+    {
+      "epoch": 2.1117460317460317,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.3485334993298552e-06,
+      "loss": 1.0734,
+      "step": 1666
+    },
+    {
+      "epoch": 2.113015873015873,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.3449294882085723e-06,
+      "loss": 1.0289,
+      "step": 1667
+    },
+    {
+      "epoch": 2.1142857142857143,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.341325641827267e-06,
+      "loss": 1.1468,
+      "step": 1668
+    },
+    {
+      "epoch": 2.1155555555555554,
+      "grad_norm": 1.0,
+      "learning_rate": 3.3377219640145053e-06,
+      "loss": 1.1006,
+      "step": 1669
+    },
+    {
+      "epoch": 2.116825396825397,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.334118458598671e-06,
+      "loss": 1.1155,
+      "step": 1670
+    },
+    {
+      "epoch": 2.118095238095238,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.3305151294079673e-06,
+      "loss": 1.1421,
+      "step": 1671
+    },
+    {
+      "epoch": 2.1193650793650796,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.326911980270414e-06,
+      "loss": 1.1595,
+      "step": 1672
+    },
+    {
+      "epoch": 2.1206349206349207,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.323309015013831e-06,
+      "loss": 1.1544,
+      "step": 1673
+    },
+    {
+      "epoch": 2.1219047619047617,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.319706237465849e-06,
+      "loss": 1.0962,
+      "step": 1674
+    },
+    {
+      "epoch": 2.1231746031746033,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.3161036514538992e-06,
+      "loss": 1.1188,
+      "step": 1675
+    },
+    {
+      "epoch": 2.1244444444444444,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.3125012608052056e-06,
+      "loss": 1.1019,
+      "step": 1676
+    },
+    {
+      "epoch": 2.125714285714286,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.308899069346788e-06,
+      "loss": 1.0536,
+      "step": 1677
+    },
+    {
+      "epoch": 2.126984126984127,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.305297080905455e-06,
+      "loss": 1.1097,
+      "step": 1678
+    },
+    {
+      "epoch": 2.128253968253968,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.301695299307794e-06,
+      "loss": 1.1417,
+      "step": 1679
+    },
+    {
+      "epoch": 2.1295238095238096,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.2980937283801805e-06,
+      "loss": 1.0972,
+      "step": 1680
+    },
+    {
+      "epoch": 2.1307936507936507,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.2944923719487624e-06,
+      "loss": 1.0071,
+      "step": 1681
+    },
+    {
+      "epoch": 2.132063492063492,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.290891233839457e-06,
+      "loss": 1.0581,
+      "step": 1682
+    },
+    {
+      "epoch": 2.1333333333333333,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.287290317877954e-06,
+      "loss": 1.0501,
+      "step": 1683
+    },
+    {
+      "epoch": 2.134603174603175,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.2836896278897062e-06,
+      "loss": 1.0731,
+      "step": 1684
+    },
+    {
+      "epoch": 2.135873015873016,
+      "grad_norm": 1.0,
+      "learning_rate": 3.280089167699924e-06,
+      "loss": 1.127,
+      "step": 1685
+    },
+    {
+      "epoch": 2.137142857142857,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.2764889411335757e-06,
+      "loss": 1.0615,
+      "step": 1686
+    },
+    {
+      "epoch": 2.1384126984126985,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.2728889520153823e-06,
+      "loss": 1.1027,
+      "step": 1687
+    },
+    {
+      "epoch": 2.1396825396825396,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.26928920416981e-06,
+      "loss": 1.1408,
+      "step": 1688
+    },
+    {
+      "epoch": 2.140952380952381,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.26568970142107e-06,
+      "loss": 1.1034,
+      "step": 1689
+    },
+    {
+      "epoch": 2.1422222222222222,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.262090447593115e-06,
+      "loss": 1.0953,
+      "step": 1690
+    },
+    {
+      "epoch": 2.1434920634920633,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.2584914465096283e-06,
+      "loss": 1.0978,
+      "step": 1691
+    },
+    {
+      "epoch": 2.144761904761905,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.254892701994029e-06,
+      "loss": 1.1599,
+      "step": 1692
+    },
+    {
+      "epoch": 2.146031746031746,
+      "grad_norm": 1.09375,
+      "learning_rate": 3.2512942178694626e-06,
+      "loss": 1.1698,
+      "step": 1693
+    },
+    {
+      "epoch": 2.1473015873015875,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.2476959979587977e-06,
+      "loss": 1.0336,
+      "step": 1694
+    },
+    {
+      "epoch": 2.1485714285714286,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.2440980460846214e-06,
+      "loss": 1.0691,
+      "step": 1695
+    },
+    {
+      "epoch": 2.1498412698412697,
+      "grad_norm": 1.0,
+      "learning_rate": 3.240500366069239e-06,
+      "loss": 1.1295,
+      "step": 1696
+    },
+    {
+      "epoch": 2.151111111111111,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.2369029617346617e-06,
+      "loss": 1.1279,
+      "step": 1697
+    },
+    {
+      "epoch": 2.1523809523809523,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.2333058369026136e-06,
+      "loss": 1.1438,
+      "step": 1698
+    },
+    {
+      "epoch": 2.153650793650794,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.2297089953945195e-06,
+      "loss": 1.0495,
+      "step": 1699
+    },
+    {
+      "epoch": 2.154920634920635,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.2261124410315003e-06,
+      "loss": 1.1117,
+      "step": 1700
+    },
+    {
+      "epoch": 2.156190476190476,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.222516177634376e-06,
+      "loss": 1.0946,
+      "step": 1701
+    },
+    {
+      "epoch": 2.1574603174603175,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.2189202090236574e-06,
+      "loss": 1.0813,
+      "step": 1702
+    },
+    {
+      "epoch": 2.1587301587301586,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.2153245390195386e-06,
+      "loss": 1.1138,
+      "step": 1703
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.211729171441899e-06,
+      "loss": 1.0996,
+      "step": 1704
+    },
+    {
+      "epoch": 2.1612698412698412,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.2081341101102978e-06,
+      "loss": 1.1335,
+      "step": 1705
+    },
+    {
+      "epoch": 2.1625396825396823,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.2045393588439657e-06,
+      "loss": 1.1013,
+      "step": 1706
+    },
+    {
+      "epoch": 2.163809523809524,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.2009449214618065e-06,
+      "loss": 1.0708,
+      "step": 1707
+    },
+    {
+      "epoch": 2.165079365079365,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.197350801782391e-06,
+      "loss": 1.0554,
+      "step": 1708
+    },
+    {
+      "epoch": 2.1663492063492065,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.1937570036239484e-06,
+      "loss": 1.1047,
+      "step": 1709
+    },
+    {
+      "epoch": 2.1676190476190476,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.1901635308043696e-06,
+      "loss": 1.1375,
+      "step": 1710
+    },
+    {
+      "epoch": 2.168888888888889,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.1865703871412033e-06,
+      "loss": 1.1244,
+      "step": 1711
+    },
+    {
+      "epoch": 2.17015873015873,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.18297757645164e-06,
+      "loss": 1.0371,
+      "step": 1712
+    },
+    {
+      "epoch": 2.1714285714285713,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.179385102552523e-06,
+      "loss": 1.1099,
+      "step": 1713
+    },
+    {
+      "epoch": 2.172698412698413,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.175792969260336e-06,
+      "loss": 1.1233,
+      "step": 1714
+    },
+    {
+      "epoch": 2.173968253968254,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.1722011803912004e-06,
+      "loss": 1.1149,
+      "step": 1715
+    },
+    {
+      "epoch": 2.1752380952380954,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.1686097397608727e-06,
+      "loss": 1.1355,
+      "step": 1716
+    },
+    {
+      "epoch": 2.1765079365079365,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.1650186511847393e-06,
+      "loss": 1.0709,
+      "step": 1717
+    },
+    {
+      "epoch": 2.1777777777777776,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.1614279184778094e-06,
+      "loss": 1.1508,
+      "step": 1718
+    },
+    {
+      "epoch": 2.179047619047619,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.1578375454547193e-06,
+      "loss": 1.1545,
+      "step": 1719
+    },
+    {
+      "epoch": 2.18031746031746,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.15424753592972e-06,
+      "loss": 1.0869,
+      "step": 1720
+    },
+    {
+      "epoch": 2.1815873015873017,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.1506578937166786e-06,
+      "loss": 1.131,
+      "step": 1721
+    },
+    {
+      "epoch": 2.182857142857143,
+      "grad_norm": 1.0,
+      "learning_rate": 3.147068622629068e-06,
+      "loss": 1.0826,
+      "step": 1722
+    },
+    {
+      "epoch": 2.1841269841269844,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.1434797264799704e-06,
+      "loss": 1.0663,
+      "step": 1723
+    },
+    {
+      "epoch": 2.1853968253968254,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.1398912090820686e-06,
+      "loss": 1.0954,
+      "step": 1724
+    },
+    {
+      "epoch": 2.1866666666666665,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.136303074247643e-06,
+      "loss": 1.1323,
+      "step": 1725
+    },
+    {
+      "epoch": 2.187936507936508,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.1327153257885674e-06,
+      "loss": 1.0114,
+      "step": 1726
+    },
+    {
+      "epoch": 2.189206349206349,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.129127967516306e-06,
+      "loss": 1.0659,
+      "step": 1727
+    },
+    {
+      "epoch": 2.1904761904761907,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.125541003241907e-06,
+      "loss": 1.041,
+      "step": 1728
+    },
+    {
+      "epoch": 2.1917460317460318,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.1219544367760006e-06,
+      "loss": 1.0414,
+      "step": 1729
+    },
+    {
+      "epoch": 2.193015873015873,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.1183682719287973e-06,
+      "loss": 1.0896,
+      "step": 1730
+    },
+    {
+      "epoch": 2.1942857142857144,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.1147825125100742e-06,
+      "loss": 1.1423,
+      "step": 1731
+    },
+    {
+      "epoch": 2.1955555555555555,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.111197162329184e-06,
+      "loss": 1.1852,
+      "step": 1732
+    },
+    {
+      "epoch": 2.196825396825397,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.1076122251950428e-06,
+      "loss": 1.1249,
+      "step": 1733
+    },
+    {
+      "epoch": 2.198095238095238,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.104027704916126e-06,
+      "loss": 1.202,
+      "step": 1734
+    },
+    {
+      "epoch": 2.199365079365079,
+      "grad_norm": 1.1328125,
+      "learning_rate": 3.1004436053004685e-06,
+      "loss": 1.1197,
+      "step": 1735
+    },
+    {
+      "epoch": 2.2006349206349207,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.096859930155658e-06,
+      "loss": 1.0592,
+      "step": 1736
+    },
+    {
+      "epoch": 2.201904761904762,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.0932766832888297e-06,
+      "loss": 1.0943,
+      "step": 1737
+    },
+    {
+      "epoch": 2.2031746031746033,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.0896938685066655e-06,
+      "loss": 1.0431,
+      "step": 1738
+    },
+    {
+      "epoch": 2.2044444444444444,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.0861114896153885e-06,
+      "loss": 1.1072,
+      "step": 1739
+    },
+    {
+      "epoch": 2.2057142857142855,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.082529550420755e-06,
+      "loss": 1.0871,
+      "step": 1740
+    },
+    {
+      "epoch": 2.206984126984127,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.0789480547280586e-06,
+      "loss": 1.0951,
+      "step": 1741
+    },
+    {
+      "epoch": 2.208253968253968,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.0753670063421223e-06,
+      "loss": 1.1225,
+      "step": 1742
+    },
+    {
+      "epoch": 2.2095238095238097,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.0717864090672884e-06,
+      "loss": 1.0767,
+      "step": 1743
+    },
+    {
+      "epoch": 2.2107936507936508,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.0682062667074243e-06,
+      "loss": 1.0996,
+      "step": 1744
+    },
+    {
+      "epoch": 2.212063492063492,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.0646265830659147e-06,
+      "loss": 1.1124,
+      "step": 1745
+    },
+    {
+      "epoch": 2.2133333333333334,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.0610473619456536e-06,
+      "loss": 1.0755,
+      "step": 1746
+    },
+    {
+      "epoch": 2.2146031746031745,
+      "grad_norm": 1.1015625,
+      "learning_rate": 3.057468607149046e-06,
+      "loss": 1.1035,
+      "step": 1747
+    },
+    {
+      "epoch": 2.215873015873016,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.0538903224780027e-06,
+      "loss": 1.0679,
+      "step": 1748
+    },
+    {
+      "epoch": 2.217142857142857,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.0503125117339295e-06,
+      "loss": 1.119,
+      "step": 1749
+    },
+    {
+      "epoch": 2.2184126984126986,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.0467351787177354e-06,
+      "loss": 1.1171,
+      "step": 1750
+    },
+    {
+      "epoch": 2.2196825396825397,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.0431583272298204e-06,
+      "loss": 1.028,
+      "step": 1751
+    },
+    {
+      "epoch": 2.220952380952381,
+      "grad_norm": 1.0,
+      "learning_rate": 3.0395819610700676e-06,
+      "loss": 1.0899,
+      "step": 1752
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.0360060840378504e-06,
+      "loss": 1.2035,
+      "step": 1753
+    },
+    {
+      "epoch": 2.2234920634920634,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.0324306999320217e-06,
+      "loss": 1.0518,
+      "step": 1754
+    },
+    {
+      "epoch": 2.224761904761905,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.028855812550908e-06,
+      "loss": 1.0406,
+      "step": 1755
+    },
+    {
+      "epoch": 2.226031746031746,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.02528142569231e-06,
+      "loss": 1.0926,
+      "step": 1756
+    },
+    {
+      "epoch": 2.227301587301587,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.021707543153498e-06,
+      "loss": 1.1265,
+      "step": 1757
+    },
+    {
+      "epoch": 2.2285714285714286,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.0181341687312035e-06,
+      "loss": 1.1466,
+      "step": 1758
+    },
+    {
+      "epoch": 2.2298412698412697,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.0145613062216198e-06,
+      "loss": 1.1103,
+      "step": 1759
+    },
+    {
+      "epoch": 2.2311111111111113,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.010988959420398e-06,
+      "loss": 1.1407,
+      "step": 1760
+    },
+    {
+      "epoch": 2.2323809523809524,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.0074171321226373e-06,
+      "loss": 1.0754,
+      "step": 1761
+    },
+    {
+      "epoch": 2.233650793650794,
+      "grad_norm": 1.140625,
+      "learning_rate": 3.0038458281228877e-06,
+      "loss": 1.1313,
+      "step": 1762
+    },
+    {
+      "epoch": 2.234920634920635,
+      "grad_norm": 1.0,
+      "learning_rate": 3.000275051215144e-06,
+      "loss": 1.057,
+      "step": 1763
+    },
+    {
+      "epoch": 2.236190476190476,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.996704805192839e-06,
+      "loss": 1.1695,
+      "step": 1764
+    },
+    {
+      "epoch": 2.2374603174603176,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.9931350938488425e-06,
+      "loss": 1.1953,
+      "step": 1765
+    },
+    {
+      "epoch": 2.2387301587301587,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.989565920975457e-06,
+      "loss": 1.0479,
+      "step": 1766
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.985997290364411e-06,
+      "loss": 1.0766,
+      "step": 1767
+    },
+    {
+      "epoch": 2.2412698412698413,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.9824292058068586e-06,
+      "loss": 1.1509,
+      "step": 1768
+    },
+    {
+      "epoch": 2.2425396825396824,
+      "grad_norm": 1.140625,
+      "learning_rate": 2.9788616710933747e-06,
+      "loss": 1.1135,
+      "step": 1769
+    },
+    {
+      "epoch": 2.243809523809524,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.9752946900139457e-06,
+      "loss": 1.0459,
+      "step": 1770
+    },
+    {
+      "epoch": 2.245079365079365,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.971728266357974e-06,
+      "loss": 1.1093,
+      "step": 1771
+    },
+    {
+      "epoch": 2.2463492063492065,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.9681624039142707e-06,
+      "loss": 1.0816,
+      "step": 1772
+    },
+    {
+      "epoch": 2.2476190476190476,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.9645971064710454e-06,
+      "loss": 1.1727,
+      "step": 1773
+    },
+    {
+      "epoch": 2.2488888888888887,
+      "grad_norm": 1.1328125,
+      "learning_rate": 2.9610323778159122e-06,
+      "loss": 1.1136,
+      "step": 1774
+    },
+    {
+      "epoch": 2.2501587301587302,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.95746822173588e-06,
+      "loss": 1.1094,
+      "step": 1775
+    },
+    {
+      "epoch": 2.2514285714285713,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.9539046420173476e-06,
+      "loss": 1.1155,
+      "step": 1776
+    },
+    {
+      "epoch": 2.252698412698413,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.9503416424461035e-06,
+      "loss": 1.1436,
+      "step": 1777
+    },
+    {
+      "epoch": 2.253968253968254,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.94677922680732e-06,
+      "loss": 1.0944,
+      "step": 1778
+    },
+    {
+      "epoch": 2.255238095238095,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.9432173988855457e-06,
+      "loss": 1.1316,
+      "step": 1779
+    },
+    {
+      "epoch": 2.2565079365079366,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.9396561624647085e-06,
+      "loss": 1.0358,
+      "step": 1780
+    },
+    {
+      "epoch": 2.2577777777777777,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.9360955213281087e-06,
+      "loss": 1.1072,
+      "step": 1781
+    },
+    {
+      "epoch": 2.259047619047619,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.9325354792584094e-06,
+      "loss": 1.0899,
+      "step": 1782
+    },
+    {
+      "epoch": 2.2603174603174603,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.9289760400376406e-06,
+      "loss": 1.0689,
+      "step": 1783
+    },
+    {
+      "epoch": 2.2615873015873014,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.925417207447192e-06,
+      "loss": 1.047,
+      "step": 1784
+    },
+    {
+      "epoch": 2.262857142857143,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.921858985267809e-06,
+      "loss": 1.1,
+      "step": 1785
+    },
+    {
+      "epoch": 2.264126984126984,
+      "grad_norm": 1.109375,
+      "learning_rate": 2.918301377279586e-06,
+      "loss": 1.1428,
+      "step": 1786
+    },
+    {
+      "epoch": 2.2653968253968255,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.9147443872619674e-06,
+      "loss": 1.152,
+      "step": 1787
+    },
+    {
+      "epoch": 2.2666666666666666,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.9111880189937417e-06,
+      "loss": 1.0685,
+      "step": 1788
+    },
+    {
+      "epoch": 2.2679365079365077,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.907632276253033e-06,
+      "loss": 1.0719,
+      "step": 1789
+    },
+    {
+      "epoch": 2.2692063492063492,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.9040771628173044e-06,
+      "loss": 1.1062,
+      "step": 1790
+    },
+    {
+      "epoch": 2.2704761904761903,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.9005226824633515e-06,
+      "loss": 1.0439,
+      "step": 1791
+    },
+    {
+      "epoch": 2.271746031746032,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.896968838967292e-06,
+      "loss": 1.1011,
+      "step": 1792
+    },
+    {
+      "epoch": 2.273015873015873,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.8934156361045714e-06,
+      "loss": 1.077,
+      "step": 1793
+    },
+    {
+      "epoch": 2.2742857142857145,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.889863077649955e-06,
+      "loss": 1.1206,
+      "step": 1794
+    },
+    {
+      "epoch": 2.2755555555555556,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.88631116737752e-06,
+      "loss": 1.0656,
+      "step": 1795
+    },
+    {
+      "epoch": 2.2768253968253966,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.8827599090606576e-06,
+      "loss": 1.0776,
+      "step": 1796
+    },
+    {
+      "epoch": 2.278095238095238,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.8792093064720675e-06,
+      "loss": 1.0576,
+      "step": 1797
+    },
+    {
+      "epoch": 2.2793650793650793,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.8756593633837486e-06,
+      "loss": 1.0584,
+      "step": 1798
+    },
+    {
+      "epoch": 2.280634920634921,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.872110083567003e-06,
+      "loss": 1.0951,
+      "step": 1799
+    },
+    {
+      "epoch": 2.281904761904762,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.8685614707924283e-06,
+      "loss": 1.1586,
+      "step": 1800
+    },
+    {
+      "epoch": 2.2831746031746034,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.865013528829909e-06,
+      "loss": 1.116,
+      "step": 1801
+    },
+    {
+      "epoch": 2.2844444444444445,
+      "grad_norm": 1.0,
+      "learning_rate": 2.861466261448621e-06,
+      "loss": 1.0048,
+      "step": 1802
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.8579196724170244e-06,
+      "loss": 1.1141,
+      "step": 1803
+    },
+    {
+      "epoch": 2.286984126984127,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.8543737655028544e-06,
+      "loss": 1.1467,
+      "step": 1804
+    },
+    {
+      "epoch": 2.288253968253968,
+      "grad_norm": 1.1171875,
+      "learning_rate": 2.850828544473125e-06,
+      "loss": 1.1634,
+      "step": 1805
+    },
+    {
+      "epoch": 2.2895238095238097,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.847284013094121e-06,
+      "loss": 1.1187,
+      "step": 1806
+    },
+    {
+      "epoch": 2.290793650793651,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.8437401751313927e-06,
+      "loss": 1.1661,
+      "step": 1807
+    },
+    {
+      "epoch": 2.292063492063492,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.840197034349756e-06,
+      "loss": 1.1115,
+      "step": 1808
+    },
+    {
+      "epoch": 2.2933333333333334,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.8366545945132864e-06,
+      "loss": 1.0922,
+      "step": 1809
+    },
+    {
+      "epoch": 2.2946031746031745,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.833112859385311e-06,
+      "loss": 1.0849,
+      "step": 1810
+    },
+    {
+      "epoch": 2.295873015873016,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.829571832728411e-06,
+      "loss": 1.117,
+      "step": 1811
+    },
+    {
+      "epoch": 2.297142857142857,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.826031518304418e-06,
+      "loss": 1.0563,
+      "step": 1812
+    },
+    {
+      "epoch": 2.2984126984126982,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.8224919198744003e-06,
+      "loss": 1.0758,
+      "step": 1813
+    },
+    {
+      "epoch": 2.2996825396825398,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.8189530411986697e-06,
+      "loss": 1.0432,
+      "step": 1814
+    },
+    {
+      "epoch": 2.300952380952381,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.8154148860367745e-06,
+      "loss": 1.0024,
+      "step": 1815
+    },
+    {
+      "epoch": 2.3022222222222224,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.8118774581474903e-06,
+      "loss": 1.0775,
+      "step": 1816
+    },
+    {
+      "epoch": 2.3034920634920635,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.8083407612888232e-06,
+      "loss": 1.0708,
+      "step": 1817
+    },
+    {
+      "epoch": 2.3047619047619046,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.8048047992180034e-06,
+      "loss": 1.1002,
+      "step": 1818
+    },
+    {
+      "epoch": 2.306031746031746,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.8012695756914745e-06,
+      "loss": 1.0636,
+      "step": 1819
+    },
+    {
+      "epoch": 2.307301587301587,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.797735094464904e-06,
+      "loss": 1.1096,
+      "step": 1820
+    },
+    {
+      "epoch": 2.3085714285714287,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.794201359293167e-06,
+      "loss": 1.1431,
+      "step": 1821
+    },
+    {
+      "epoch": 2.30984126984127,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.790668373930342e-06,
+      "loss": 1.0675,
+      "step": 1822
+    },
+    {
+      "epoch": 2.311111111111111,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.7871361421297176e-06,
+      "loss": 1.1974,
+      "step": 1823
+    },
+    {
+      "epoch": 2.3123809523809524,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.783604667643778e-06,
+      "loss": 1.1248,
+      "step": 1824
+    },
+    {
+      "epoch": 2.3136507936507935,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.7800739542242037e-06,
+      "loss": 1.1497,
+      "step": 1825
+    },
+    {
+      "epoch": 2.314920634920635,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.776544005621866e-06,
+      "loss": 1.0875,
+      "step": 1826
+    },
+    {
+      "epoch": 2.316190476190476,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.7730148255868272e-06,
+      "loss": 1.0432,
+      "step": 1827
+    },
+    {
+      "epoch": 2.317460317460317,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.7694864178683274e-06,
+      "loss": 1.2387,
+      "step": 1828
+    },
+    {
+      "epoch": 2.3187301587301588,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.76595878621479e-06,
+      "loss": 1.0651,
+      "step": 1829
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.762431934373815e-06,
+      "loss": 0.9725,
+      "step": 1830
+    },
+    {
+      "epoch": 2.3212698412698414,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.758905866092169e-06,
+      "loss": 1.0772,
+      "step": 1831
+    },
+    {
+      "epoch": 2.3225396825396825,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.7553805851157908e-06,
+      "loss": 1.1864,
+      "step": 1832
+    },
+    {
+      "epoch": 2.323809523809524,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.751856095189782e-06,
+      "loss": 1.0956,
+      "step": 1833
+    },
+    {
+      "epoch": 2.325079365079365,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.748332400058401e-06,
+      "loss": 1.0685,
+      "step": 1834
+    },
+    {
+      "epoch": 2.326349206349206,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.744809503465066e-06,
+      "loss": 1.1514,
+      "step": 1835
+    },
+    {
+      "epoch": 2.3276190476190477,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.7412874091523443e-06,
+      "loss": 1.1161,
+      "step": 1836
+    },
+    {
+      "epoch": 2.328888888888889,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.7377661208619507e-06,
+      "loss": 1.1107,
+      "step": 1837
+    },
+    {
+      "epoch": 2.3301587301587303,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.7342456423347447e-06,
+      "loss": 1.1623,
+      "step": 1838
+    },
+    {
+      "epoch": 2.3314285714285714,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.730725977310727e-06,
+      "loss": 1.0283,
+      "step": 1839
+    },
+    {
+      "epoch": 2.332698412698413,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.7272071295290287e-06,
+      "loss": 1.0891,
+      "step": 1840
+    },
+    {
+      "epoch": 2.333968253968254,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.7236891027279193e-06,
+      "loss": 1.1403,
+      "step": 1841
+    },
+    {
+      "epoch": 2.335238095238095,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.720171900644792e-06,
+      "loss": 1.1341,
+      "step": 1842
+    },
+    {
+      "epoch": 2.3365079365079366,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.716655527016165e-06,
+      "loss": 1.1241,
+      "step": 1843
+    },
+    {
+      "epoch": 2.3377777777777777,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.713139985577677e-06,
+      "loss": 1.1266,
+      "step": 1844
+    },
+    {
+      "epoch": 2.3390476190476193,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.7096252800640815e-06,
+      "loss": 1.1383,
+      "step": 1845
+    },
+    {
+      "epoch": 2.3403174603174604,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.7061114142092446e-06,
+      "loss": 1.1058,
+      "step": 1846
+    },
+    {
+      "epoch": 2.3415873015873014,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.7025983917461403e-06,
+      "loss": 1.0996,
+      "step": 1847
+    },
+    {
+      "epoch": 2.342857142857143,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.6990862164068457e-06,
+      "loss": 1.0692,
+      "step": 1848
+    },
+    {
+      "epoch": 2.344126984126984,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.6955748919225423e-06,
+      "loss": 1.0931,
+      "step": 1849
+    },
+    {
+      "epoch": 2.3453968253968256,
+      "grad_norm": 1.0,
+      "learning_rate": 2.6920644220234987e-06,
+      "loss": 1.1498,
+      "step": 1850
+    },
+    {
+      "epoch": 2.3466666666666667,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.6885548104390847e-06,
+      "loss": 1.1273,
+      "step": 1851
+    },
+    {
+      "epoch": 2.3479365079365078,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.6850460608977553e-06,
+      "loss": 1.0613,
+      "step": 1852
+    },
+    {
+      "epoch": 2.3492063492063493,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.681538177127047e-06,
+      "loss": 1.053,
+      "step": 1853
+    },
+    {
+      "epoch": 2.3504761904761904,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.6780311628535786e-06,
+      "loss": 1.0882,
+      "step": 1854
+    },
+    {
+      "epoch": 2.351746031746032,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.674525021803048e-06,
+      "loss": 1.0766,
+      "step": 1855
+    },
+    {
+      "epoch": 2.353015873015873,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.671019757700219e-06,
+      "loss": 1.1448,
+      "step": 1856
+    },
+    {
+      "epoch": 2.354285714285714,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.66751537426893e-06,
+      "loss": 1.1185,
+      "step": 1857
+    },
+    {
+      "epoch": 2.3555555555555556,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.6640118752320805e-06,
+      "loss": 1.1182,
+      "step": 1858
+    },
+    {
+      "epoch": 2.3568253968253967,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.660509264311631e-06,
+      "loss": 1.1698,
+      "step": 1859
+    },
+    {
+      "epoch": 2.3580952380952382,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.657007545228599e-06,
+      "loss": 1.1668,
+      "step": 1860
+    },
+    {
+      "epoch": 2.3593650793650793,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.653506721703055e-06,
+      "loss": 1.1547,
+      "step": 1861
+    },
+    {
+      "epoch": 2.3606349206349204,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.650006797454115e-06,
+      "loss": 1.1429,
+      "step": 1862
+    },
+    {
+      "epoch": 2.361904761904762,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.6465077761999445e-06,
+      "loss": 1.1267,
+      "step": 1863
+    },
+    {
+      "epoch": 2.363174603174603,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.6430096616577462e-06,
+      "loss": 1.1491,
+      "step": 1864
+    },
+    {
+      "epoch": 2.3644444444444446,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.63951245754376e-06,
+      "loss": 1.0985,
+      "step": 1865
+    },
+    {
+      "epoch": 2.3657142857142857,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.63601616757326e-06,
+      "loss": 1.1629,
+      "step": 1866
+    },
+    {
+      "epoch": 2.3669841269841267,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.6325207954605487e-06,
+      "loss": 1.1326,
+      "step": 1867
+    },
+    {
+      "epoch": 2.3682539682539683,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.629026344918951e-06,
+      "loss": 1.0767,
+      "step": 1868
+    },
+    {
+      "epoch": 2.3695238095238094,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.6255328196608177e-06,
+      "loss": 1.0746,
+      "step": 1869
+    },
+    {
+      "epoch": 2.370793650793651,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.6220402233975134e-06,
+      "loss": 1.0798,
+      "step": 1870
+    },
+    {
+      "epoch": 2.372063492063492,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.618548559839414e-06,
+      "loss": 1.0402,
+      "step": 1871
+    },
+    {
+      "epoch": 2.3733333333333335,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.6150578326959075e-06,
+      "loss": 1.1098,
+      "step": 1872
+    },
+    {
+      "epoch": 2.3746031746031746,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.611568045675388e-06,
+      "loss": 1.1225,
+      "step": 1873
+    },
+    {
+      "epoch": 2.3758730158730157,
+      "grad_norm": 1.09375,
+      "learning_rate": 2.6080792024852476e-06,
+      "loss": 1.1165,
+      "step": 1874
+    },
+    {
+      "epoch": 2.3771428571428572,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.604591306831877e-06,
+      "loss": 1.1329,
+      "step": 1875
+    },
+    {
+      "epoch": 2.3784126984126983,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.6011043624206627e-06,
+      "loss": 1.0614,
+      "step": 1876
+    },
+    {
+      "epoch": 2.37968253968254,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.597618372955976e-06,
+      "loss": 0.9886,
+      "step": 1877
+    },
+    {
+      "epoch": 2.380952380952381,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.594133342141177e-06,
+      "loss": 1.0381,
+      "step": 1878
+    },
+    {
+      "epoch": 2.3822222222222225,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.5906492736786086e-06,
+      "loss": 1.0849,
+      "step": 1879
+    },
+    {
+      "epoch": 2.3834920634920636,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.5871661712695866e-06,
+      "loss": 1.0896,
+      "step": 1880
+    },
+    {
+      "epoch": 2.3847619047619046,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.5836840386144032e-06,
+      "loss": 1.1288,
+      "step": 1881
+    },
+    {
+      "epoch": 2.386031746031746,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.5802028794123236e-06,
+      "loss": 1.1489,
+      "step": 1882
+    },
+    {
+      "epoch": 2.3873015873015873,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.5767226973615724e-06,
+      "loss": 1.0566,
+      "step": 1883
+    },
+    {
+      "epoch": 2.388571428571429,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.5732434961593405e-06,
+      "loss": 1.1293,
+      "step": 1884
+    },
+    {
+      "epoch": 2.38984126984127,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.5697652795017767e-06,
+      "loss": 1.0864,
+      "step": 1885
+    },
+    {
+      "epoch": 2.391111111111111,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.5662880510839806e-06,
+      "loss": 1.0624,
+      "step": 1886
+    },
+    {
+      "epoch": 2.3923809523809525,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.5628118146000057e-06,
+      "loss": 1.1134,
+      "step": 1887
+    },
+    {
+      "epoch": 2.3936507936507936,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.5593365737428515e-06,
+      "loss": 1.098,
+      "step": 1888
+    },
+    {
+      "epoch": 2.394920634920635,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.555862332204455e-06,
+      "loss": 1.0941,
+      "step": 1889
+    },
+    {
+      "epoch": 2.396190476190476,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.5523890936756973e-06,
+      "loss": 1.0794,
+      "step": 1890
+    },
+    {
+      "epoch": 2.3974603174603173,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.5489168618463925e-06,
+      "loss": 1.0828,
+      "step": 1891
+    },
+    {
+      "epoch": 2.398730158730159,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.545445640405282e-06,
+      "loss": 1.0805,
+      "step": 1892
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.541975433040038e-06,
+      "loss": 1.0684,
+      "step": 1893
+    },
+    {
+      "epoch": 2.4012698412698414,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.538506243437254e-06,
+      "loss": 1.0645,
+      "step": 1894
+    },
+    {
+      "epoch": 2.4025396825396825,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.5350380752824398e-06,
+      "loss": 1.1507,
+      "step": 1895
+    },
+    {
+      "epoch": 2.4038095238095236,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.5315709322600235e-06,
+      "loss": 1.1797,
+      "step": 1896
+    },
+    {
+      "epoch": 2.405079365079365,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.5281048180533435e-06,
+      "loss": 1.0862,
+      "step": 1897
+    },
+    {
+      "epoch": 2.4063492063492062,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.5246397363446434e-06,
+      "loss": 1.0854,
+      "step": 1898
+    },
+    {
+      "epoch": 2.4076190476190478,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.521175690815071e-06,
+      "loss": 1.1069,
+      "step": 1899
+    },
+    {
+      "epoch": 2.408888888888889,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.517712685144675e-06,
+      "loss": 1.1071,
+      "step": 1900
+    },
+    {
+      "epoch": 2.41015873015873,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.5142507230123958e-06,
+      "loss": 1.1429,
+      "step": 1901
+    },
+    {
+      "epoch": 2.4114285714285715,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.510789808096067e-06,
+      "loss": 1.1749,
+      "step": 1902
+    },
+    {
+      "epoch": 2.4126984126984126,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.5073299440724123e-06,
+      "loss": 1.1322,
+      "step": 1903
+    },
+    {
+      "epoch": 2.413968253968254,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.5038711346170343e-06,
+      "loss": 1.1336,
+      "step": 1904
+    },
+    {
+      "epoch": 2.415238095238095,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.5004133834044173e-06,
+      "loss": 1.0746,
+      "step": 1905
+    },
+    {
+      "epoch": 2.4165079365079363,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.4969566941079237e-06,
+      "loss": 1.0861,
+      "step": 1906
+    },
+    {
+      "epoch": 2.417777777777778,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.4935010703997837e-06,
+      "loss": 1.0976,
+      "step": 1907
+    },
+    {
+      "epoch": 2.419047619047619,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.490046515951098e-06,
+      "loss": 1.1116,
+      "step": 1908
+    },
+    {
+      "epoch": 2.4203174603174604,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.4865930344318318e-06,
+      "loss": 1.1143,
+      "step": 1909
+    },
+    {
+      "epoch": 2.4215873015873015,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.4831406295108073e-06,
+      "loss": 1.1464,
+      "step": 1910
+    },
+    {
+      "epoch": 2.422857142857143,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.4796893048557065e-06,
+      "loss": 1.0728,
+      "step": 1911
+    },
+    {
+      "epoch": 2.424126984126984,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.476239064133061e-06,
+      "loss": 1.056,
+      "step": 1912
+    },
+    {
+      "epoch": 2.425396825396825,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.472789911008256e-06,
+      "loss": 1.0837,
+      "step": 1913
+    },
+    {
+      "epoch": 2.4266666666666667,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.4693418491455134e-06,
+      "loss": 1.1147,
+      "step": 1914
+    },
+    {
+      "epoch": 2.427936507936508,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.4658948822079015e-06,
+      "loss": 1.1255,
+      "step": 1915
+    },
+    {
+      "epoch": 2.4292063492063494,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.462449013857324e-06,
+      "loss": 1.061,
+      "step": 1916
+    },
+    {
+      "epoch": 2.4304761904761905,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.4590042477545174e-06,
+      "loss": 1.0575,
+      "step": 1917
+    },
+    {
+      "epoch": 2.431746031746032,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.455560587559046e-06,
+      "loss": 1.0712,
+      "step": 1918
+    },
+    {
+      "epoch": 2.433015873015873,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.452118036929303e-06,
+      "loss": 1.1059,
+      "step": 1919
+    },
+    {
+      "epoch": 2.434285714285714,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.448676599522497e-06,
+      "loss": 1.0617,
+      "step": 1920
+    },
+    {
+      "epoch": 2.4355555555555557,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.4452362789946585e-06,
+      "loss": 1.1436,
+      "step": 1921
+    },
+    {
+      "epoch": 2.436825396825397,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.441797079000633e-06,
+      "loss": 1.1023,
+      "step": 1922
+    },
+    {
+      "epoch": 2.4380952380952383,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.4383590031940674e-06,
+      "loss": 1.1191,
+      "step": 1923
+    },
+    {
+      "epoch": 2.4393650793650794,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.4349220552274235e-06,
+      "loss": 1.1114,
+      "step": 1924
+    },
+    {
+      "epoch": 2.4406349206349205,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.4314862387519605e-06,
+      "loss": 1.1117,
+      "step": 1925
+    },
+    {
+      "epoch": 2.441904761904762,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.4280515574177355e-06,
+      "loss": 1.1458,
+      "step": 1926
+    },
+    {
+      "epoch": 2.443174603174603,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.4246180148736005e-06,
+      "loss": 1.066,
+      "step": 1927
+    },
+    {
+      "epoch": 2.4444444444444446,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.4211856147671983e-06,
+      "loss": 1.0545,
+      "step": 1928
+    },
+    {
+      "epoch": 2.4457142857142857,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.417754360744956e-06,
+      "loss": 1.1227,
+      "step": 1929
+    },
+    {
+      "epoch": 2.446984126984127,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.4143242564520854e-06,
+      "loss": 1.0712,
+      "step": 1930
+    },
+    {
+      "epoch": 2.4482539682539683,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.410895305532577e-06,
+      "loss": 1.1392,
+      "step": 1931
+    },
+    {
+      "epoch": 2.4495238095238094,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.407467511629192e-06,
+      "loss": 1.1064,
+      "step": 1932
+    },
+    {
+      "epoch": 2.450793650793651,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.404040878383467e-06,
+      "loss": 1.0607,
+      "step": 1933
+    },
+    {
+      "epoch": 2.452063492063492,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.400615409435705e-06,
+      "loss": 1.0725,
+      "step": 1934
+    },
+    {
+      "epoch": 2.453333333333333,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.3971911084249687e-06,
+      "loss": 1.0903,
+      "step": 1935
+    },
+    {
+      "epoch": 2.4546031746031747,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.3937679789890837e-06,
+      "loss": 1.0725,
+      "step": 1936
+    },
+    {
+      "epoch": 2.4558730158730158,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.3903460247646305e-06,
+      "loss": 1.0601,
+      "step": 1937
+    },
+    {
+      "epoch": 2.4571428571428573,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.3869252493869386e-06,
+      "loss": 1.1456,
+      "step": 1938
+    },
+    {
+      "epoch": 2.4584126984126984,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.3835056564900872e-06,
+      "loss": 1.1521,
+      "step": 1939
+    },
+    {
+      "epoch": 2.4596825396825395,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.3800872497069003e-06,
+      "loss": 1.0804,
+      "step": 1940
+    },
+    {
+      "epoch": 2.460952380952381,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.3766700326689385e-06,
+      "loss": 1.0186,
+      "step": 1941
+    },
+    {
+      "epoch": 2.462222222222222,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.3732540090065007e-06,
+      "loss": 1.0648,
+      "step": 1942
+    },
+    {
+      "epoch": 2.4634920634920636,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.369839182348619e-06,
+      "loss": 1.0986,
+      "step": 1943
+    },
+    {
+      "epoch": 2.4647619047619047,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.3664255563230506e-06,
+      "loss": 1.051,
+      "step": 1944
+    },
+    {
+      "epoch": 2.466031746031746,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.36301313455628e-06,
+      "loss": 1.1296,
+      "step": 1945
+    },
+    {
+      "epoch": 2.4673015873015873,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.359601920673513e-06,
+      "loss": 1.0191,
+      "step": 1946
+    },
+    {
+      "epoch": 2.4685714285714284,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.356191918298668e-06,
+      "loss": 1.0894,
+      "step": 1947
+    },
+    {
+      "epoch": 2.46984126984127,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.3527831310543813e-06,
+      "loss": 1.0448,
+      "step": 1948
+    },
+    {
+      "epoch": 2.471111111111111,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.3493755625619958e-06,
+      "loss": 1.131,
+      "step": 1949
+    },
+    {
+      "epoch": 2.4723809523809526,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.3459692164415576e-06,
+      "loss": 1.0821,
+      "step": 1950
+    },
+    {
+      "epoch": 2.4736507936507937,
+      "grad_norm": 1.0,
+      "learning_rate": 2.342564096311818e-06,
+      "loss": 1.08,
+      "step": 1951
+    },
+    {
+      "epoch": 2.4749206349206347,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.339160205790226e-06,
+      "loss": 1.1783,
+      "step": 1952
+    },
+    {
+      "epoch": 2.4761904761904763,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.33575754849292e-06,
+      "loss": 1.2117,
+      "step": 1953
+    },
+    {
+      "epoch": 2.4774603174603174,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.3323561280347326e-06,
+      "loss": 1.093,
+      "step": 1954
+    },
+    {
+      "epoch": 2.478730158730159,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.3289559480291806e-06,
+      "loss": 1.11,
+      "step": 1955
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.0,
+      "learning_rate": 2.3255570120884617e-06,
+      "loss": 1.0685,
+      "step": 1956
+    },
+    {
+      "epoch": 2.4812698412698415,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.3221593238234553e-06,
+      "loss": 1.0753,
+      "step": 1957
+    },
+    {
+      "epoch": 2.4825396825396826,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.3187628868437134e-06,
+      "loss": 1.2114,
+      "step": 1958
+    },
+    {
+      "epoch": 2.4838095238095237,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.3153677047574567e-06,
+      "loss": 1.1823,
+      "step": 1959
+    },
+    {
+      "epoch": 2.485079365079365,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.311973781171577e-06,
+      "loss": 1.0855,
+      "step": 1960
+    },
+    {
+      "epoch": 2.4863492063492063,
+      "grad_norm": 1.0,
+      "learning_rate": 2.3085811196916274e-06,
+      "loss": 1.1072,
+      "step": 1961
+    },
+    {
+      "epoch": 2.487619047619048,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.305189723921817e-06,
+      "loss": 1.1988,
+      "step": 1962
+    },
+    {
+      "epoch": 2.488888888888889,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.3017995974650146e-06,
+      "loss": 1.1193,
+      "step": 1963
+    },
+    {
+      "epoch": 2.49015873015873,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.2984107439227395e-06,
+      "loss": 1.0405,
+      "step": 1964
+    },
+    {
+      "epoch": 2.4914285714285715,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.2950231668951576e-06,
+      "loss": 1.1319,
+      "step": 1965
+    },
+    {
+      "epoch": 2.4926984126984126,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.291636869981079e-06,
+      "loss": 1.0125,
+      "step": 1966
+    },
+    {
+      "epoch": 2.493968253968254,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.2882518567779545e-06,
+      "loss": 1.1485,
+      "step": 1967
+    },
+    {
+      "epoch": 2.4952380952380953,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.2848681308818707e-06,
+      "loss": 1.0248,
+      "step": 1968
+    },
+    {
+      "epoch": 2.4965079365079363,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.2814856958875457e-06,
+      "loss": 1.1444,
+      "step": 1969
+    },
+    {
+      "epoch": 2.497777777777778,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.27810455538833e-06,
+      "loss": 1.0855,
+      "step": 1970
+    }
+  ],
+  "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": 4.362570197203157e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}