| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.5003174603174603, | |
| "eval_steps": 500, | |
| "global_step": 394, | |
| "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 | |
| } | |
| ], | |
| "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": 8.734007407002255e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |