| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 504, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.003968253968253968, | |
| "grad_norm": 648.0144653320312, | |
| "learning_rate": 3.846153846153847e-07, | |
| "loss": 1.4175, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.007936507936507936, | |
| "grad_norm": 694.8490600585938, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 1.4409, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.011904761904761904, | |
| "grad_norm": 644.5820922851562, | |
| "learning_rate": 1.153846153846154e-06, | |
| "loss": 1.4856, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.015873015873015872, | |
| "grad_norm": 533.597900390625, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 1.5221, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01984126984126984, | |
| "grad_norm": 95.04138946533203, | |
| "learning_rate": 1.9230769230769234e-06, | |
| "loss": 1.618, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.023809523809523808, | |
| "grad_norm": 302.99554443359375, | |
| "learning_rate": 2.307692307692308e-06, | |
| "loss": 1.4584, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.027777777777777776, | |
| "grad_norm": 309.1397399902344, | |
| "learning_rate": 2.6923076923076923e-06, | |
| "loss": 1.3838, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.031746031746031744, | |
| "grad_norm": 203.5092315673828, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.3827, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03571428571428571, | |
| "grad_norm": 65.24649047851562, | |
| "learning_rate": 3.4615384615384617e-06, | |
| "loss": 1.262, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03968253968253968, | |
| "grad_norm": 130.4703826904297, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 1.3211, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04365079365079365, | |
| "grad_norm": 68.12091827392578, | |
| "learning_rate": 4.230769230769231e-06, | |
| "loss": 1.3073, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.047619047619047616, | |
| "grad_norm": 48.97660446166992, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 1.2135, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.051587301587301584, | |
| "grad_norm": 83.72801971435547, | |
| "learning_rate": 5e-06, | |
| "loss": 1.1344, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.05555555555555555, | |
| "grad_norm": 36.81949234008789, | |
| "learning_rate": 5.384615384615385e-06, | |
| "loss": 1.1234, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.05952380952380952, | |
| "grad_norm": 20.137876510620117, | |
| "learning_rate": 5.769230769230769e-06, | |
| "loss": 1.0783, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06349206349206349, | |
| "grad_norm": 46.69076919555664, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 1.043, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.06746031746031746, | |
| "grad_norm": 86.63626861572266, | |
| "learning_rate": 6.538461538461539e-06, | |
| "loss": 1.0176, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 35.93476104736328, | |
| "learning_rate": 6.923076923076923e-06, | |
| "loss": 1.012, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.07539682539682539, | |
| "grad_norm": 58.39170455932617, | |
| "learning_rate": 7.307692307692308e-06, | |
| "loss": 0.9421, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.07936507936507936, | |
| "grad_norm": 73.09225463867188, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.9197, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08333333333333333, | |
| "grad_norm": 17.957416534423828, | |
| "learning_rate": 8.076923076923077e-06, | |
| "loss": 0.8901, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0873015873015873, | |
| "grad_norm": 18.88216209411621, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 1.0445, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.09126984126984126, | |
| "grad_norm": 17.8280086517334, | |
| "learning_rate": 8.846153846153847e-06, | |
| "loss": 1.1157, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.09523809523809523, | |
| "grad_norm": 20.809980392456055, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 1.0306, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0992063492063492, | |
| "grad_norm": 82.48164367675781, | |
| "learning_rate": 9.615384615384616e-06, | |
| "loss": 0.9471, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.10317460317460317, | |
| "grad_norm": 9.758552551269531, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6123, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.10714285714285714, | |
| "grad_norm": 3.8075623512268066, | |
| "learning_rate": 9.999892010284378e-06, | |
| "loss": 0.6156, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 12.073590278625488, | |
| "learning_rate": 9.999568045802216e-06, | |
| "loss": 0.6354, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.11507936507936507, | |
| "grad_norm": 8.46188735961914, | |
| "learning_rate": 9.999028120547456e-06, | |
| "loss": 0.5497, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.11904761904761904, | |
| "grad_norm": 7.2497239112854, | |
| "learning_rate": 9.99827225784264e-06, | |
| "loss": 0.6059, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12301587301587301, | |
| "grad_norm": 27.255783081054688, | |
| "learning_rate": 9.99730049033793e-06, | |
| "loss": 0.5631, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.12698412698412698, | |
| "grad_norm": 2.5536746978759766, | |
| "learning_rate": 9.996112860009689e-06, | |
| "loss": 0.5835, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.13095238095238096, | |
| "grad_norm": 3.371126413345337, | |
| "learning_rate": 9.994709418158652e-06, | |
| "loss": 0.5346, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.1349206349206349, | |
| "grad_norm": 2.407099485397339, | |
| "learning_rate": 9.993090225407743e-06, | |
| "loss": 0.6248, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1388888888888889, | |
| "grad_norm": 2.6606991291046143, | |
| "learning_rate": 9.991255351699422e-06, | |
| "loss": 0.5521, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 2.3420910835266113, | |
| "learning_rate": 9.98920487629269e-06, | |
| "loss": 0.6289, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.14682539682539683, | |
| "grad_norm": 2.7614691257476807, | |
| "learning_rate": 9.986938887759643e-06, | |
| "loss": 0.6279, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.15079365079365079, | |
| "grad_norm": 2.5013375282287598, | |
| "learning_rate": 9.98445748398167e-06, | |
| "loss": 0.615, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.15476190476190477, | |
| "grad_norm": 2.933556079864502, | |
| "learning_rate": 9.981760772145201e-06, | |
| "loss": 0.6374, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.15873015873015872, | |
| "grad_norm": 2.4485886096954346, | |
| "learning_rate": 9.978848868737099e-06, | |
| "loss": 0.6512, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1626984126984127, | |
| "grad_norm": 2.976875066757202, | |
| "learning_rate": 9.975721899539607e-06, | |
| "loss": 0.6453, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.16666666666666666, | |
| "grad_norm": 2.8370003700256348, | |
| "learning_rate": 9.972379999624935e-06, | |
| "loss": 0.6895, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.17063492063492064, | |
| "grad_norm": 2.5592763423919678, | |
| "learning_rate": 9.968823313349412e-06, | |
| "loss": 0.6005, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.1746031746031746, | |
| "grad_norm": 2.805567502975464, | |
| "learning_rate": 9.96505199434725e-06, | |
| "loss": 0.6138, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.17857142857142858, | |
| "grad_norm": 2.789598226547241, | |
| "learning_rate": 9.961066205523917e-06, | |
| "loss": 0.6533, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.18253968253968253, | |
| "grad_norm": 3.0848469734191895, | |
| "learning_rate": 9.956866119049095e-06, | |
| "loss": 0.6645, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1865079365079365, | |
| "grad_norm": 3.980682611465454, | |
| "learning_rate": 9.952451916349242e-06, | |
| "loss": 0.6034, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.19047619047619047, | |
| "grad_norm": 2.5931661128997803, | |
| "learning_rate": 9.947823788099754e-06, | |
| "loss": 0.5567, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.19444444444444445, | |
| "grad_norm": 2.911203384399414, | |
| "learning_rate": 9.942981934216731e-06, | |
| "loss": 0.5614, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.1984126984126984, | |
| "grad_norm": 3.0429320335388184, | |
| "learning_rate": 9.937926563848345e-06, | |
| "loss": 0.6008, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.20238095238095238, | |
| "grad_norm": 2.6976630687713623, | |
| "learning_rate": 9.9326578953658e-06, | |
| "loss": 0.5025, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.20634920634920634, | |
| "grad_norm": 2.5022969245910645, | |
| "learning_rate": 9.9271761563539e-06, | |
| "loss": 0.4798, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.21031746031746032, | |
| "grad_norm": 2.5305087566375732, | |
| "learning_rate": 9.921481583601218e-06, | |
| "loss": 0.5471, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "grad_norm": 2.8594768047332764, | |
| "learning_rate": 9.915574423089872e-06, | |
| "loss": 0.4727, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.21825396825396826, | |
| "grad_norm": 5.546922206878662, | |
| "learning_rate": 9.909454929984894e-06, | |
| "loss": 0.4817, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 3.0452191829681396, | |
| "learning_rate": 9.903123368623216e-06, | |
| "loss": 0.4687, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.2261904761904762, | |
| "grad_norm": 2.8640449047088623, | |
| "learning_rate": 9.896580012502238e-06, | |
| "loss": 0.5576, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.23015873015873015, | |
| "grad_norm": 3.2537431716918945, | |
| "learning_rate": 9.889825144268029e-06, | |
| "loss": 0.4854, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.23412698412698413, | |
| "grad_norm": 2.716897487640381, | |
| "learning_rate": 9.882859055703109e-06, | |
| "loss": 0.4641, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.23809523809523808, | |
| "grad_norm": 2.7615535259246826, | |
| "learning_rate": 9.875682047713847e-06, | |
| "loss": 0.4824, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.24206349206349206, | |
| "grad_norm": 3.1990036964416504, | |
| "learning_rate": 9.868294430317464e-06, | |
| "loss": 0.5235, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.24603174603174602, | |
| "grad_norm": 3.1204071044921875, | |
| "learning_rate": 9.860696522628638e-06, | |
| "loss": 0.5236, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 3.176624298095703, | |
| "learning_rate": 9.852888652845729e-06, | |
| "loss": 0.5321, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.25396825396825395, | |
| "grad_norm": 3.3741488456726074, | |
| "learning_rate": 9.84487115823659e-06, | |
| "loss": 0.5117, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.25793650793650796, | |
| "grad_norm": 3.497760772705078, | |
| "learning_rate": 9.836644385124006e-06, | |
| "loss": 0.4979, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.2619047619047619, | |
| "grad_norm": 3.474196434020996, | |
| "learning_rate": 9.828208688870736e-06, | |
| "loss": 0.5353, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.26587301587301587, | |
| "grad_norm": 3.723007917404175, | |
| "learning_rate": 9.81956443386415e-06, | |
| "loss": 0.5654, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.2698412698412698, | |
| "grad_norm": 4.143252372741699, | |
| "learning_rate": 9.810711993500506e-06, | |
| "loss": 0.4795, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.27380952380952384, | |
| "grad_norm": 4.070223808288574, | |
| "learning_rate": 9.801651750168815e-06, | |
| "loss": 0.4903, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.2777777777777778, | |
| "grad_norm": 4.025697231292725, | |
| "learning_rate": 9.792384095234312e-06, | |
| "loss": 0.4735, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.28174603174603174, | |
| "grad_norm": 4.2407965660095215, | |
| "learning_rate": 9.782909429021568e-06, | |
| "loss": 0.4705, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 4.298872947692871, | |
| "learning_rate": 9.773228160797187e-06, | |
| "loss": 0.486, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.2896825396825397, | |
| "grad_norm": 3.741116523742676, | |
| "learning_rate": 9.76334070875213e-06, | |
| "loss": 0.479, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.29365079365079366, | |
| "grad_norm": 3.841461658477783, | |
| "learning_rate": 9.753247499983649e-06, | |
| "loss": 0.4248, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.2976190476190476, | |
| "grad_norm": 4.278059959411621, | |
| "learning_rate": 9.742948970476845e-06, | |
| "loss": 0.4267, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.30158730158730157, | |
| "grad_norm": 3.1821653842926025, | |
| "learning_rate": 9.732445565085823e-06, | |
| "loss": 0.3589, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.3055555555555556, | |
| "grad_norm": 3.5140159130096436, | |
| "learning_rate": 9.721737737514492e-06, | |
| "loss": 0.337, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.30952380952380953, | |
| "grad_norm": 3.5480399131774902, | |
| "learning_rate": 9.71082595029695e-06, | |
| "loss": 0.3834, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.3134920634920635, | |
| "grad_norm": 3.601264238357544, | |
| "learning_rate": 9.699710674777519e-06, | |
| "loss": 0.3401, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.31746031746031744, | |
| "grad_norm": 3.6704108715057373, | |
| "learning_rate": 9.688392391090374e-06, | |
| "loss": 0.3911, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.32142857142857145, | |
| "grad_norm": 3.646101236343384, | |
| "learning_rate": 9.676871588138812e-06, | |
| "loss": 0.3487, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.3253968253968254, | |
| "grad_norm": 3.7638654708862305, | |
| "learning_rate": 9.665148763574123e-06, | |
| "loss": 0.3537, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.32936507936507936, | |
| "grad_norm": 3.739696502685547, | |
| "learning_rate": 9.653224423774107e-06, | |
| "loss": 0.3341, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 3.6597673892974854, | |
| "learning_rate": 9.64109908382119e-06, | |
| "loss": 0.3197, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.3373015873015873, | |
| "grad_norm": 3.643167734146118, | |
| "learning_rate": 9.628773267480177e-06, | |
| "loss": 0.3852, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.3412698412698413, | |
| "grad_norm": 3.7483153343200684, | |
| "learning_rate": 9.616247507175624e-06, | |
| "loss": 0.3564, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.34523809523809523, | |
| "grad_norm": 3.868358850479126, | |
| "learning_rate": 9.603522343968852e-06, | |
| "loss": 0.379, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.3492063492063492, | |
| "grad_norm": 3.770395040512085, | |
| "learning_rate": 9.590598327534563e-06, | |
| "loss": 0.3483, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.3531746031746032, | |
| "grad_norm": 3.850552558898926, | |
| "learning_rate": 9.577476016137105e-06, | |
| "loss": 0.33, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 3.7598023414611816, | |
| "learning_rate": 9.56415597660634e-06, | |
| "loss": 0.355, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3611111111111111, | |
| "grad_norm": 4.041435718536377, | |
| "learning_rate": 9.550638784313187e-06, | |
| "loss": 0.3417, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.36507936507936506, | |
| "grad_norm": 4.028756141662598, | |
| "learning_rate": 9.536925023144742e-06, | |
| "loss": 0.3107, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.36904761904761907, | |
| "grad_norm": 3.9944610595703125, | |
| "learning_rate": 9.523015285479076e-06, | |
| "loss": 0.3532, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.373015873015873, | |
| "grad_norm": 3.9798991680145264, | |
| "learning_rate": 9.508910172159635e-06, | |
| "loss": 0.3187, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.376984126984127, | |
| "grad_norm": 3.8477606773376465, | |
| "learning_rate": 9.494610292469287e-06, | |
| "loss": 0.3011, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.38095238095238093, | |
| "grad_norm": 3.9732916355133057, | |
| "learning_rate": 9.48011626410401e-06, | |
| "loss": 0.331, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.38492063492063494, | |
| "grad_norm": 3.4221317768096924, | |
| "learning_rate": 9.465428713146206e-06, | |
| "loss": 0.3296, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.3888888888888889, | |
| "grad_norm": 3.0487005710601807, | |
| "learning_rate": 9.450548274037652e-06, | |
| "loss": 0.3098, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.39285714285714285, | |
| "grad_norm": 2.9913902282714844, | |
| "learning_rate": 9.435475589552107e-06, | |
| "loss": 0.2884, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.3968253968253968, | |
| "grad_norm": 3.7084836959838867, | |
| "learning_rate": 9.420211310767534e-06, | |
| "loss": 0.3116, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4007936507936508, | |
| "grad_norm": 2.4643120765686035, | |
| "learning_rate": 9.40475609703798e-06, | |
| "loss": 0.2482, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.40476190476190477, | |
| "grad_norm": 2.7057974338531494, | |
| "learning_rate": 9.389110615965102e-06, | |
| "loss": 0.2594, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.4087301587301587, | |
| "grad_norm": 2.6204943656921387, | |
| "learning_rate": 9.37327554336932e-06, | |
| "loss": 0.1999, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.4126984126984127, | |
| "grad_norm": 2.613316059112549, | |
| "learning_rate": 9.35725156326063e-06, | |
| "loss": 0.2377, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.4166666666666667, | |
| "grad_norm": 2.509354829788208, | |
| "learning_rate": 9.341039367809056e-06, | |
| "loss": 0.2559, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.42063492063492064, | |
| "grad_norm": 2.508958101272583, | |
| "learning_rate": 9.324639657314742e-06, | |
| "loss": 0.261, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.4246031746031746, | |
| "grad_norm": 2.3309414386749268, | |
| "learning_rate": 9.308053140177722e-06, | |
| "loss": 0.2201, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 2.375749349594116, | |
| "learning_rate": 9.291280532867301e-06, | |
| "loss": 0.215, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.43253968253968256, | |
| "grad_norm": 2.2472996711730957, | |
| "learning_rate": 9.27432255989112e-06, | |
| "loss": 0.25, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.4365079365079365, | |
| "grad_norm": 2.2145283222198486, | |
| "learning_rate": 9.257179953763846e-06, | |
| "loss": 0.2424, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.44047619047619047, | |
| "grad_norm": 2.066765069961548, | |
| "learning_rate": 9.239853454975548e-06, | |
| "loss": 0.2298, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 1.9276115894317627, | |
| "learning_rate": 9.222343811959694e-06, | |
| "loss": 0.2463, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.44841269841269843, | |
| "grad_norm": 1.9066914319992065, | |
| "learning_rate": 9.204651781060832e-06, | |
| "loss": 0.1999, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.4523809523809524, | |
| "grad_norm": 1.8612842559814453, | |
| "learning_rate": 9.186778126501916e-06, | |
| "loss": 0.2194, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.45634920634920634, | |
| "grad_norm": 1.8626900911331177, | |
| "learning_rate": 9.168723620351298e-06, | |
| "loss": 0.2281, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.4603174603174603, | |
| "grad_norm": 1.7394243478775024, | |
| "learning_rate": 9.150489042489368e-06, | |
| "loss": 0.2006, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.4642857142857143, | |
| "grad_norm": 1.784781575202942, | |
| "learning_rate": 9.13207518057488e-06, | |
| "loss": 0.2339, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.46825396825396826, | |
| "grad_norm": 1.5163494348526, | |
| "learning_rate": 9.113482830010918e-06, | |
| "loss": 0.1893, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.4722222222222222, | |
| "grad_norm": 1.6937190294265747, | |
| "learning_rate": 9.094712793910541e-06, | |
| "loss": 0.2179, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.47619047619047616, | |
| "grad_norm": 1.551163911819458, | |
| "learning_rate": 9.075765883062093e-06, | |
| "loss": 0.2127, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4801587301587302, | |
| "grad_norm": 1.5405539274215698, | |
| "learning_rate": 9.056642915894182e-06, | |
| "loss": 0.2084, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.48412698412698413, | |
| "grad_norm": 1.432281255722046, | |
| "learning_rate": 9.037344718440321e-06, | |
| "loss": 0.2308, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.4880952380952381, | |
| "grad_norm": 1.4656060934066772, | |
| "learning_rate": 9.017872124303255e-06, | |
| "loss": 0.2294, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.49206349206349204, | |
| "grad_norm": 1.4060044288635254, | |
| "learning_rate": 8.99822597461894e-06, | |
| "loss": 0.1836, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.49603174603174605, | |
| "grad_norm": 1.4943199157714844, | |
| "learning_rate": 8.978407118020226e-06, | |
| "loss": 0.1944, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.338510513305664, | |
| "learning_rate": 8.958416410600188e-06, | |
| "loss": 0.2127, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.503968253968254, | |
| "grad_norm": 1.1287288665771484, | |
| "learning_rate": 8.938254715875152e-06, | |
| "loss": 0.1771, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.5079365079365079, | |
| "grad_norm": 1.3611695766448975, | |
| "learning_rate": 8.917922904747385e-06, | |
| "loss": 0.168, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.5119047619047619, | |
| "grad_norm": 1.3361289501190186, | |
| "learning_rate": 8.897421855467491e-06, | |
| "loss": 0.2142, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.5158730158730159, | |
| "grad_norm": 1.4268479347229004, | |
| "learning_rate": 8.876752453596462e-06, | |
| "loss": 0.2552, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5198412698412699, | |
| "grad_norm": 1.3106672763824463, | |
| "learning_rate": 8.85591559196743e-06, | |
| "loss": 0.2317, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.5238095238095238, | |
| "grad_norm": 1.1432324647903442, | |
| "learning_rate": 8.834912170647102e-06, | |
| "loss": 0.2043, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.5277777777777778, | |
| "grad_norm": 1.1050715446472168, | |
| "learning_rate": 8.813743096896872e-06, | |
| "loss": 0.189, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.5317460317460317, | |
| "grad_norm": 1.2125810384750366, | |
| "learning_rate": 8.792409285133644e-06, | |
| "loss": 0.1922, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.5357142857142857, | |
| "grad_norm": 1.149783968925476, | |
| "learning_rate": 8.770911656890325e-06, | |
| "loss": 0.2148, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.5396825396825397, | |
| "grad_norm": 1.3137670755386353, | |
| "learning_rate": 8.749251140776016e-06, | |
| "loss": 0.2036, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.5436507936507936, | |
| "grad_norm": 1.276073694229126, | |
| "learning_rate": 8.727428672435911e-06, | |
| "loss": 0.2031, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.5476190476190477, | |
| "grad_norm": 1.5819066762924194, | |
| "learning_rate": 8.705445194510868e-06, | |
| "loss": 0.1966, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.5515873015873016, | |
| "grad_norm": 1.5322155952453613, | |
| "learning_rate": 8.6833016565967e-06, | |
| "loss": 0.2281, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 1.4338527917861938, | |
| "learning_rate": 8.660999015203152e-06, | |
| "loss": 0.1856, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5595238095238095, | |
| "grad_norm": 1.5081082582473755, | |
| "learning_rate": 8.638538233712581e-06, | |
| "loss": 0.1981, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.5634920634920635, | |
| "grad_norm": 1.409002423286438, | |
| "learning_rate": 8.615920282338355e-06, | |
| "loss": 0.2, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5674603174603174, | |
| "grad_norm": 1.612339973449707, | |
| "learning_rate": 8.593146138082925e-06, | |
| "loss": 0.2534, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 1.5412747859954834, | |
| "learning_rate": 8.570216784695637e-06, | |
| "loss": 0.193, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.5753968253968254, | |
| "grad_norm": 1.7418545484542847, | |
| "learning_rate": 8.54713321263023e-06, | |
| "loss": 0.1858, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5793650793650794, | |
| "grad_norm": 1.6208196878433228, | |
| "learning_rate": 8.52389641900206e-06, | |
| "loss": 0.1878, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.5833333333333334, | |
| "grad_norm": 2.062012195587158, | |
| "learning_rate": 8.50050740754502e-06, | |
| "loss": 0.2345, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5873015873015873, | |
| "grad_norm": 1.5701236724853516, | |
| "learning_rate": 8.476967188568187e-06, | |
| "loss": 0.1876, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.5912698412698413, | |
| "grad_norm": 1.480595350265503, | |
| "learning_rate": 8.453276778912186e-06, | |
| "loss": 0.2026, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.5952380952380952, | |
| "grad_norm": 1.4031518697738647, | |
| "learning_rate": 8.429437201905254e-06, | |
| "loss": 0.168, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5992063492063492, | |
| "grad_norm": 1.276462197303772, | |
| "learning_rate": 8.405449487319049e-06, | |
| "loss": 0.1616, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.6031746031746031, | |
| "grad_norm": 1.3217096328735352, | |
| "learning_rate": 8.38131467132416e-06, | |
| "loss": 0.151, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.6071428571428571, | |
| "grad_norm": 1.5035640001296997, | |
| "learning_rate": 8.357033796445356e-06, | |
| "loss": 0.1768, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.6111111111111112, | |
| "grad_norm": 1.3976829051971436, | |
| "learning_rate": 8.332607911516545e-06, | |
| "loss": 0.1487, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.6150793650793651, | |
| "grad_norm": 1.5825048685073853, | |
| "learning_rate": 8.308038071635475e-06, | |
| "loss": 0.2662, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.6190476190476191, | |
| "grad_norm": 1.4428508281707764, | |
| "learning_rate": 8.283325338118154e-06, | |
| "loss": 0.1698, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.623015873015873, | |
| "grad_norm": 1.5432616472244263, | |
| "learning_rate": 8.258470778453005e-06, | |
| "loss": 0.2058, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.626984126984127, | |
| "grad_norm": 1.4593292474746704, | |
| "learning_rate": 8.233475466254766e-06, | |
| "loss": 0.1697, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.6309523809523809, | |
| "grad_norm": 1.5151939392089844, | |
| "learning_rate": 8.208340481218094e-06, | |
| "loss": 0.1782, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.6349206349206349, | |
| "grad_norm": 1.4148085117340088, | |
| "learning_rate": 8.183066909070946e-06, | |
| "loss": 0.165, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.6388888888888888, | |
| "grad_norm": 1.4057435989379883, | |
| "learning_rate": 8.15765584152767e-06, | |
| "loss": 0.1692, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.6428571428571429, | |
| "grad_norm": 1.428735375404358, | |
| "learning_rate": 8.132108376241849e-06, | |
| "loss": 0.1469, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.6468253968253969, | |
| "grad_norm": 1.4047014713287354, | |
| "learning_rate": 8.106425616758886e-06, | |
| "loss": 0.1916, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.6507936507936508, | |
| "grad_norm": 1.518492579460144, | |
| "learning_rate": 8.08060867246834e-06, | |
| "loss": 0.2098, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.6547619047619048, | |
| "grad_norm": 2.7570230960845947, | |
| "learning_rate": 8.054658658555998e-06, | |
| "loss": 0.2004, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.6587301587301587, | |
| "grad_norm": 1.6120089292526245, | |
| "learning_rate": 8.028576695955711e-06, | |
| "loss": 0.1967, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.6626984126984127, | |
| "grad_norm": 1.460198163986206, | |
| "learning_rate": 8.002363911300966e-06, | |
| "loss": 0.1747, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 1.5127792358398438, | |
| "learning_rate": 7.976021436876232e-06, | |
| "loss": 0.1613, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.6706349206349206, | |
| "grad_norm": 1.6299463510513306, | |
| "learning_rate": 7.949550410568033e-06, | |
| "loss": 0.1714, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.6746031746031746, | |
| "grad_norm": 1.5469965934753418, | |
| "learning_rate": 7.92295197581581e-06, | |
| "loss": 0.1473, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6785714285714286, | |
| "grad_norm": 1.7051180601119995, | |
| "learning_rate": 7.89622728156253e-06, | |
| "loss": 0.1655, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.6825396825396826, | |
| "grad_norm": 1.5857340097427368, | |
| "learning_rate": 7.869377482205042e-06, | |
| "loss": 0.2132, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.6865079365079365, | |
| "grad_norm": 1.3160123825073242, | |
| "learning_rate": 7.842403737544226e-06, | |
| "loss": 0.1654, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.6904761904761905, | |
| "grad_norm": 1.4576656818389893, | |
| "learning_rate": 7.815307212734888e-06, | |
| "loss": 0.1394, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.6944444444444444, | |
| "grad_norm": 1.421958565711975, | |
| "learning_rate": 7.788089078235432e-06, | |
| "loss": 0.1757, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6984126984126984, | |
| "grad_norm": 1.3313640356063843, | |
| "learning_rate": 7.7607505097573e-06, | |
| "loss": 0.1826, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.7023809523809523, | |
| "grad_norm": 1.1764626502990723, | |
| "learning_rate": 7.733292688214182e-06, | |
| "loss": 0.1427, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.7063492063492064, | |
| "grad_norm": 1.2739901542663574, | |
| "learning_rate": 7.705716799671019e-06, | |
| "loss": 0.1741, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.7103174603174603, | |
| "grad_norm": 1.3538535833358765, | |
| "learning_rate": 7.678024035292757e-06, | |
| "loss": 0.1465, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 1.2715352773666382, | |
| "learning_rate": 7.650215591292888e-06, | |
| "loss": 0.1371, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7182539682539683, | |
| "grad_norm": 1.2927345037460327, | |
| "learning_rate": 7.622292668881805e-06, | |
| "loss": 0.1712, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.7222222222222222, | |
| "grad_norm": 1.2348829507827759, | |
| "learning_rate": 7.594256474214883e-06, | |
| "loss": 0.146, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.7261904761904762, | |
| "grad_norm": 1.2945867776870728, | |
| "learning_rate": 7.566108218340399e-06, | |
| "loss": 0.1303, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.7301587301587301, | |
| "grad_norm": 1.3601559400558472, | |
| "learning_rate": 7.537849117147212e-06, | |
| "loss": 0.172, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.7341269841269841, | |
| "grad_norm": 1.309390664100647, | |
| "learning_rate": 7.509480391312243e-06, | |
| "loss": 0.1186, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.7380952380952381, | |
| "grad_norm": 1.5198652744293213, | |
| "learning_rate": 7.481003266247745e-06, | |
| "loss": 0.1882, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.7420634920634921, | |
| "grad_norm": 1.530924916267395, | |
| "learning_rate": 7.452418972048372e-06, | |
| "loss": 0.1742, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.746031746031746, | |
| "grad_norm": 1.3118865489959717, | |
| "learning_rate": 7.4237287434380485e-06, | |
| "loss": 0.1736, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 1.4929744005203247, | |
| "learning_rate": 7.394933819716625e-06, | |
| "loss": 0.1883, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.753968253968254, | |
| "grad_norm": 1.2493473291397095, | |
| "learning_rate": 7.366035444706346e-06, | |
| "loss": 0.1367, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.7579365079365079, | |
| "grad_norm": 1.4079400300979614, | |
| "learning_rate": 7.337034866698138e-06, | |
| "loss": 0.155, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.7619047619047619, | |
| "grad_norm": 1.4902135133743286, | |
| "learning_rate": 7.307933338397667e-06, | |
| "loss": 0.2089, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.7658730158730159, | |
| "grad_norm": 1.3614805936813354, | |
| "learning_rate": 7.278732116871239e-06, | |
| "loss": 0.1658, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.7698412698412699, | |
| "grad_norm": 1.376658320426941, | |
| "learning_rate": 7.249432463491498e-06, | |
| "loss": 0.1363, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.7738095238095238, | |
| "grad_norm": 1.5702760219573975, | |
| "learning_rate": 7.220035643882938e-06, | |
| "loss": 0.1511, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 1.440657377243042, | |
| "learning_rate": 7.190542927867234e-06, | |
| "loss": 0.1343, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.7817460317460317, | |
| "grad_norm": 3.7832703590393066, | |
| "learning_rate": 7.160955589408395e-06, | |
| "loss": 0.2226, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.7857142857142857, | |
| "grad_norm": 2.7817816734313965, | |
| "learning_rate": 7.131274906557725e-06, | |
| "loss": 0.1822, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.7896825396825397, | |
| "grad_norm": 1.5844206809997559, | |
| "learning_rate": 7.101502161398626e-06, | |
| "loss": 0.1486, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.7936507936507936, | |
| "grad_norm": 1.188583254814148, | |
| "learning_rate": 7.0716386399912075e-06, | |
| "loss": 0.1397, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7976190476190477, | |
| "grad_norm": 1.0894182920455933, | |
| "learning_rate": 7.041685632316748e-06, | |
| "loss": 0.1455, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.8015873015873016, | |
| "grad_norm": 0.9938972592353821, | |
| "learning_rate": 7.0116444322219575e-06, | |
| "loss": 0.1156, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.8055555555555556, | |
| "grad_norm": 1.0027174949645996, | |
| "learning_rate": 6.981516337363099e-06, | |
| "loss": 0.1089, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.8095238095238095, | |
| "grad_norm": 1.0513209104537964, | |
| "learning_rate": 6.95130264914993e-06, | |
| "loss": 0.1236, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.8134920634920635, | |
| "grad_norm": 1.0481796264648438, | |
| "learning_rate": 6.9210046726894885e-06, | |
| "loss": 0.1115, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.8174603174603174, | |
| "grad_norm": 1.1058838367462158, | |
| "learning_rate": 6.890623716729724e-06, | |
| "loss": 0.1475, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.8214285714285714, | |
| "grad_norm": 1.4549169540405273, | |
| "learning_rate": 6.860161093602949e-06, | |
| "loss": 0.1455, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.8253968253968254, | |
| "grad_norm": 1.0771286487579346, | |
| "learning_rate": 6.829618119169169e-06, | |
| "loss": 0.1467, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.8293650793650794, | |
| "grad_norm": 1.1434047222137451, | |
| "learning_rate": 6.798996112759233e-06, | |
| "loss": 0.1497, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 1.1114635467529297, | |
| "learning_rate": 6.768296397117848e-06, | |
| "loss": 0.1664, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.8373015873015873, | |
| "grad_norm": 1.13973069190979, | |
| "learning_rate": 6.737520298346438e-06, | |
| "loss": 0.1261, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.8412698412698413, | |
| "grad_norm": 1.0329458713531494, | |
| "learning_rate": 6.706669145845863e-06, | |
| "loss": 0.1117, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.8452380952380952, | |
| "grad_norm": 1.2193344831466675, | |
| "learning_rate": 6.6757442722590015e-06, | |
| "loss": 0.1784, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.8492063492063492, | |
| "grad_norm": 1.1052571535110474, | |
| "learning_rate": 6.6447470134131685e-06, | |
| "loss": 0.1241, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.8531746031746031, | |
| "grad_norm": 1.015051245689392, | |
| "learning_rate": 6.613678708262439e-06, | |
| "loss": 0.1057, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 1.20777428150177, | |
| "learning_rate": 6.5825406988297815e-06, | |
| "loss": 0.1465, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.8611111111111112, | |
| "grad_norm": 1.1608431339263916, | |
| "learning_rate": 6.551334330149114e-06, | |
| "loss": 0.1493, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.8650793650793651, | |
| "grad_norm": 1.062906265258789, | |
| "learning_rate": 6.520060950207186e-06, | |
| "loss": 0.1456, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.8690476190476191, | |
| "grad_norm": 1.1560730934143066, | |
| "learning_rate": 6.488721909885359e-06, | |
| "loss": 0.1772, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.873015873015873, | |
| "grad_norm": 0.9284862279891968, | |
| "learning_rate": 6.457318562901257e-06, | |
| "loss": 0.0989, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.876984126984127, | |
| "grad_norm": 1.1579439640045166, | |
| "learning_rate": 6.425852265750282e-06, | |
| "loss": 0.1563, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.8809523809523809, | |
| "grad_norm": 1.201245903968811, | |
| "learning_rate": 6.394324377647028e-06, | |
| "loss": 0.1564, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.8849206349206349, | |
| "grad_norm": 1.1407997608184814, | |
| "learning_rate": 6.362736260466561e-06, | |
| "loss": 0.1306, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 1.1680482625961304, | |
| "learning_rate": 6.331089278685599e-06, | |
| "loss": 0.1834, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.8928571428571429, | |
| "grad_norm": 1.1824140548706055, | |
| "learning_rate": 6.299384799323568e-06, | |
| "loss": 0.1118, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.8968253968253969, | |
| "grad_norm": 1.023852825164795, | |
| "learning_rate": 6.267624191883551e-06, | |
| "loss": 0.1337, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.9007936507936508, | |
| "grad_norm": 0.9579774737358093, | |
| "learning_rate": 6.235808828293135e-06, | |
| "loss": 0.1268, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.9047619047619048, | |
| "grad_norm": 0.9918968677520752, | |
| "learning_rate": 6.203940082845144e-06, | |
| "loss": 0.1637, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.9087301587301587, | |
| "grad_norm": 0.7822407484054565, | |
| "learning_rate": 6.172019332138285e-06, | |
| "loss": 0.1063, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.9126984126984127, | |
| "grad_norm": 0.9548757076263428, | |
| "learning_rate": 6.140047955017672e-06, | |
| "loss": 0.1543, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.9166666666666666, | |
| "grad_norm": 0.8214371204376221, | |
| "learning_rate": 6.108027332515276e-06, | |
| "loss": 0.0956, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.9206349206349206, | |
| "grad_norm": 1.0052070617675781, | |
| "learning_rate": 6.075958847790262e-06, | |
| "loss": 0.1316, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.9246031746031746, | |
| "grad_norm": 0.8131003975868225, | |
| "learning_rate": 6.043843886069251e-06, | |
| "loss": 0.1167, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.9285714285714286, | |
| "grad_norm": 0.9374865293502808, | |
| "learning_rate": 6.011683834586474e-06, | |
| "loss": 0.1267, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.9325396825396826, | |
| "grad_norm": 1.0154731273651123, | |
| "learning_rate": 5.979480082523858e-06, | |
| "loss": 0.1606, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.9365079365079365, | |
| "grad_norm": 0.9171246290206909, | |
| "learning_rate": 5.947234020951015e-06, | |
| "loss": 0.1327, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.9404761904761905, | |
| "grad_norm": 0.9893870949745178, | |
| "learning_rate": 5.914947042765149e-06, | |
| "loss": 0.1545, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.9444444444444444, | |
| "grad_norm": 0.9642376899719238, | |
| "learning_rate": 5.882620542630901e-06, | |
| "loss": 0.1438, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.9484126984126984, | |
| "grad_norm": 0.9926443099975586, | |
| "learning_rate": 5.850255916920093e-06, | |
| "loss": 0.1348, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.9523809523809523, | |
| "grad_norm": 0.9060367345809937, | |
| "learning_rate": 5.817854563651415e-06, | |
| "loss": 0.1437, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.9563492063492064, | |
| "grad_norm": 0.9371214509010315, | |
| "learning_rate": 5.785417882430035e-06, | |
| "loss": 0.1142, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.9603174603174603, | |
| "grad_norm": 0.9900795817375183, | |
| "learning_rate": 5.752947274387147e-06, | |
| "loss": 0.1292, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.9642857142857143, | |
| "grad_norm": 1.0389565229415894, | |
| "learning_rate": 5.720444142119445e-06, | |
| "loss": 0.1506, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.9682539682539683, | |
| "grad_norm": 0.8546874523162842, | |
| "learning_rate": 5.687909889628529e-06, | |
| "loss": 0.1165, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.9722222222222222, | |
| "grad_norm": 0.949036180973053, | |
| "learning_rate": 5.6553459222602714e-06, | |
| "loss": 0.1499, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.9761904761904762, | |
| "grad_norm": 0.8515863418579102, | |
| "learning_rate": 5.622753646644102e-06, | |
| "loss": 0.0989, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.9801587301587301, | |
| "grad_norm": 0.9659066200256348, | |
| "learning_rate": 5.59013447063225e-06, | |
| "loss": 0.1601, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.9841269841269841, | |
| "grad_norm": 0.977457582950592, | |
| "learning_rate": 5.557489803238934e-06, | |
| "loss": 0.1228, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.9880952380952381, | |
| "grad_norm": 0.804831326007843, | |
| "learning_rate": 5.524821054579491e-06, | |
| "loss": 0.1309, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.9920634920634921, | |
| "grad_norm": 0.8933465480804443, | |
| "learning_rate": 5.492129635809473e-06, | |
| "loss": 0.1132, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.996031746031746, | |
| "grad_norm": 0.9376035928726196, | |
| "learning_rate": 5.459416959063688e-06, | |
| "loss": 0.1322, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.9543865323066711, | |
| "learning_rate": 5.426684437395196e-06, | |
| "loss": 0.1712, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.003968253968254, | |
| "grad_norm": 0.8264181017875671, | |
| "learning_rate": 5.393933484714284e-06, | |
| "loss": 0.1055, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.007936507936508, | |
| "grad_norm": 0.822293758392334, | |
| "learning_rate": 5.361165515727374e-06, | |
| "loss": 0.1113, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.0119047619047619, | |
| "grad_norm": 0.8232570886611938, | |
| "learning_rate": 5.328381945875933e-06, | |
| "loss": 0.1183, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.0158730158730158, | |
| "grad_norm": 0.9493715167045593, | |
| "learning_rate": 5.295584191275308e-06, | |
| "loss": 0.1222, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.0198412698412698, | |
| "grad_norm": 1.0286954641342163, | |
| "learning_rate": 5.26277366865358e-06, | |
| "loss": 0.1521, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.0238095238095237, | |
| "grad_norm": 0.8530657887458801, | |
| "learning_rate": 5.229951795290353e-06, | |
| "loss": 0.1163, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.0277777777777777, | |
| "grad_norm": 0.7479215860366821, | |
| "learning_rate": 5.197119988955534e-06, | |
| "loss": 0.0753, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.0317460317460316, | |
| "grad_norm": 0.7635036706924438, | |
| "learning_rate": 5.164279667848094e-06, | |
| "loss": 0.0935, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.0357142857142858, | |
| "grad_norm": 0.7508280873298645, | |
| "learning_rate": 5.131432250534809e-06, | |
| "loss": 0.0713, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.0396825396825398, | |
| "grad_norm": 0.9552590250968933, | |
| "learning_rate": 5.0985791558889785e-06, | |
| "loss": 0.1145, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.0436507936507937, | |
| "grad_norm": 0.9875309467315674, | |
| "learning_rate": 5.065721803029146e-06, | |
| "loss": 0.128, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.0476190476190477, | |
| "grad_norm": 0.9727612137794495, | |
| "learning_rate": 5.032861611257783e-06, | |
| "loss": 0.095, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.0515873015873016, | |
| "grad_norm": 0.9066519141197205, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0982, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.0555555555555556, | |
| "grad_norm": 0.9620884656906128, | |
| "learning_rate": 4.967138388742218e-06, | |
| "loss": 0.1131, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.0595238095238095, | |
| "grad_norm": 0.8276383876800537, | |
| "learning_rate": 4.934278196970857e-06, | |
| "loss": 0.0992, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.0634920634920635, | |
| "grad_norm": 1.015833854675293, | |
| "learning_rate": 4.9014208441110215e-06, | |
| "loss": 0.1297, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.0674603174603174, | |
| "grad_norm": 0.7764245271682739, | |
| "learning_rate": 4.868567749465192e-06, | |
| "loss": 0.085, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.0714285714285714, | |
| "grad_norm": 0.8442825675010681, | |
| "learning_rate": 4.835720332151907e-06, | |
| "loss": 0.0893, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.0753968253968254, | |
| "grad_norm": 0.823527991771698, | |
| "learning_rate": 4.802880011044467e-06, | |
| "loss": 0.0937, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.0793650793650793, | |
| "grad_norm": 0.8294535279273987, | |
| "learning_rate": 4.770048204709648e-06, | |
| "loss": 0.0872, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.0833333333333333, | |
| "grad_norm": 0.8440665602684021, | |
| "learning_rate": 4.73722633134642e-06, | |
| "loss": 0.0848, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.0873015873015872, | |
| "grad_norm": 0.9344722628593445, | |
| "learning_rate": 4.7044158087246926e-06, | |
| "loss": 0.1102, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.0912698412698412, | |
| "grad_norm": 0.898384690284729, | |
| "learning_rate": 4.67161805412407e-06, | |
| "loss": 0.0996, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.0952380952380953, | |
| "grad_norm": 0.8309195637702942, | |
| "learning_rate": 4.6388344842726266e-06, | |
| "loss": 0.0899, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.0992063492063493, | |
| "grad_norm": 0.7668273448944092, | |
| "learning_rate": 4.606066515285719e-06, | |
| "loss": 0.1004, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.1031746031746033, | |
| "grad_norm": 0.7124539017677307, | |
| "learning_rate": 4.573315562604804e-06, | |
| "loss": 0.0837, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.1071428571428572, | |
| "grad_norm": 0.8069830536842346, | |
| "learning_rate": 4.540583040936313e-06, | |
| "loss": 0.09, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 0.7966949343681335, | |
| "learning_rate": 4.5078703641905275e-06, | |
| "loss": 0.0795, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.1150793650793651, | |
| "grad_norm": 0.7558461427688599, | |
| "learning_rate": 4.4751789454205105e-06, | |
| "loss": 0.0854, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.119047619047619, | |
| "grad_norm": 0.8736066818237305, | |
| "learning_rate": 4.442510196761068e-06, | |
| "loss": 0.0925, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.123015873015873, | |
| "grad_norm": 0.8083801865577698, | |
| "learning_rate": 4.409865529367751e-06, | |
| "loss": 0.0882, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.126984126984127, | |
| "grad_norm": 0.818231999874115, | |
| "learning_rate": 4.377246353355899e-06, | |
| "loss": 0.083, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.130952380952381, | |
| "grad_norm": 0.8110062479972839, | |
| "learning_rate": 4.34465407773973e-06, | |
| "loss": 0.0832, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.1349206349206349, | |
| "grad_norm": 0.9061889052391052, | |
| "learning_rate": 4.312090110371473e-06, | |
| "loss": 0.1133, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.1388888888888888, | |
| "grad_norm": 0.8491413593292236, | |
| "learning_rate": 4.279555857880558e-06, | |
| "loss": 0.0803, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 0.8797329664230347, | |
| "learning_rate": 4.247052725612853e-06, | |
| "loss": 0.0903, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.1468253968253967, | |
| "grad_norm": 0.8875324726104736, | |
| "learning_rate": 4.214582117569966e-06, | |
| "loss": 0.0985, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.1507936507936507, | |
| "grad_norm": 0.8876000642776489, | |
| "learning_rate": 4.182145436348587e-06, | |
| "loss": 0.0997, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.1547619047619047, | |
| "grad_norm": 0.8417052626609802, | |
| "learning_rate": 4.1497440830799084e-06, | |
| "loss": 0.086, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.1587301587301586, | |
| "grad_norm": 0.8717465996742249, | |
| "learning_rate": 4.1173794573691e-06, | |
| "loss": 0.1065, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.1626984126984128, | |
| "grad_norm": 0.893736720085144, | |
| "learning_rate": 4.0850529572348505e-06, | |
| "loss": 0.0857, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.1666666666666667, | |
| "grad_norm": 0.8343749642372131, | |
| "learning_rate": 4.052765979048986e-06, | |
| "loss": 0.0894, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.1706349206349207, | |
| "grad_norm": 0.8286570310592651, | |
| "learning_rate": 4.0205199174761435e-06, | |
| "loss": 0.0871, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.1746031746031746, | |
| "grad_norm": 0.8456138372421265, | |
| "learning_rate": 3.988316165413528e-06, | |
| "loss": 0.0871, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.1785714285714286, | |
| "grad_norm": 0.9013364911079407, | |
| "learning_rate": 3.956156113930752e-06, | |
| "loss": 0.082, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.1825396825396826, | |
| "grad_norm": 0.859642744064331, | |
| "learning_rate": 3.924041152209739e-06, | |
| "loss": 0.1053, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.1865079365079365, | |
| "grad_norm": 0.8950112462043762, | |
| "learning_rate": 3.891972667484726e-06, | |
| "loss": 0.0966, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.1904761904761905, | |
| "grad_norm": 0.8393763899803162, | |
| "learning_rate": 3.859952044982329e-06, | |
| "loss": 0.1012, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.1944444444444444, | |
| "grad_norm": 0.8542735576629639, | |
| "learning_rate": 3.827980667861716e-06, | |
| "loss": 0.0998, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.1984126984126984, | |
| "grad_norm": 0.8967165946960449, | |
| "learning_rate": 3.7960599171548572e-06, | |
| "loss": 0.0796, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.2023809523809523, | |
| "grad_norm": 0.8004995584487915, | |
| "learning_rate": 3.764191171706867e-06, | |
| "loss": 0.0991, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.2063492063492063, | |
| "grad_norm": 0.7223465442657471, | |
| "learning_rate": 3.732375808116451e-06, | |
| "loss": 0.0776, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.2103174603174602, | |
| "grad_norm": 0.8522040843963623, | |
| "learning_rate": 3.7006152006764336e-06, | |
| "loss": 0.0997, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.2142857142857142, | |
| "grad_norm": 0.8484842777252197, | |
| "learning_rate": 3.6689107213144025e-06, | |
| "loss": 0.0819, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.2182539682539684, | |
| "grad_norm": 0.9578449130058289, | |
| "learning_rate": 3.6372637395334416e-06, | |
| "loss": 0.1186, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.2222222222222223, | |
| "grad_norm": 0.8413398265838623, | |
| "learning_rate": 3.6056756223529734e-06, | |
| "loss": 0.0922, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.2261904761904763, | |
| "grad_norm": 0.6764865517616272, | |
| "learning_rate": 3.574147734249719e-06, | |
| "loss": 0.0748, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.2301587301587302, | |
| "grad_norm": 0.8983719944953918, | |
| "learning_rate": 3.542681437098745e-06, | |
| "loss": 0.1026, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.2341269841269842, | |
| "grad_norm": 0.9965245723724365, | |
| "learning_rate": 3.5112780901146426e-06, | |
| "loss": 0.103, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.2380952380952381, | |
| "grad_norm": 0.8493394255638123, | |
| "learning_rate": 3.479939049792817e-06, | |
| "loss": 0.0972, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.242063492063492, | |
| "grad_norm": 0.7720935940742493, | |
| "learning_rate": 3.448665669850888e-06, | |
| "loss": 0.0849, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.246031746031746, | |
| "grad_norm": 0.9442537426948547, | |
| "learning_rate": 3.4174593011702197e-06, | |
| "loss": 0.1175, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.8026649951934814, | |
| "learning_rate": 3.386321291737563e-06, | |
| "loss": 0.0762, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.253968253968254, | |
| "grad_norm": 0.9502577185630798, | |
| "learning_rate": 3.3552529865868323e-06, | |
| "loss": 0.0999, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.257936507936508, | |
| "grad_norm": 0.7329561114311218, | |
| "learning_rate": 3.3242557277410015e-06, | |
| "loss": 0.0741, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.2619047619047619, | |
| "grad_norm": 0.8264954686164856, | |
| "learning_rate": 3.2933308541541365e-06, | |
| "loss": 0.0998, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.2658730158730158, | |
| "grad_norm": 0.7596467137336731, | |
| "learning_rate": 3.2624797016535626e-06, | |
| "loss": 0.0797, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.2698412698412698, | |
| "grad_norm": 0.9133650064468384, | |
| "learning_rate": 3.2317036028821523e-06, | |
| "loss": 0.1184, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.2738095238095237, | |
| "grad_norm": 0.8063773512840271, | |
| "learning_rate": 3.201003887240768e-06, | |
| "loss": 0.0885, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.2777777777777777, | |
| "grad_norm": 0.8149939775466919, | |
| "learning_rate": 3.1703818808308327e-06, | |
| "loss": 0.0733, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.2817460317460316, | |
| "grad_norm": 1.0018041133880615, | |
| "learning_rate": 3.1398389063970512e-06, | |
| "loss": 0.0965, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "grad_norm": 0.8329199552536011, | |
| "learning_rate": 3.1093762832702775e-06, | |
| "loss": 0.1019, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.2896825396825398, | |
| "grad_norm": 0.8745952248573303, | |
| "learning_rate": 3.0789953273105123e-06, | |
| "loss": 0.1007, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.2936507936507937, | |
| "grad_norm": 1.0435259342193604, | |
| "learning_rate": 3.048697350850073e-06, | |
| "loss": 0.1308, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.2976190476190477, | |
| "grad_norm": 0.9064452648162842, | |
| "learning_rate": 3.0184836626369034e-06, | |
| "loss": 0.1006, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.3015873015873016, | |
| "grad_norm": 0.7629454135894775, | |
| "learning_rate": 2.988355567778043e-06, | |
| "loss": 0.084, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.3055555555555556, | |
| "grad_norm": 0.8011732697486877, | |
| "learning_rate": 2.9583143676832526e-06, | |
| "loss": 0.0921, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.3095238095238095, | |
| "grad_norm": 0.7690629959106445, | |
| "learning_rate": 2.9283613600087933e-06, | |
| "loss": 0.0665, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.3134920634920635, | |
| "grad_norm": 0.7121254205703735, | |
| "learning_rate": 2.8984978386013767e-06, | |
| "loss": 0.0864, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.3174603174603174, | |
| "grad_norm": 0.7874686121940613, | |
| "learning_rate": 2.8687250934422774e-06, | |
| "loss": 0.0838, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.3214285714285714, | |
| "grad_norm": 0.9707236289978027, | |
| "learning_rate": 2.839044410591606e-06, | |
| "loss": 0.1283, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.3253968253968254, | |
| "grad_norm": 0.834000289440155, | |
| "learning_rate": 2.809457072132766e-06, | |
| "loss": 0.1012, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.3293650793650793, | |
| "grad_norm": 0.7457403540611267, | |
| "learning_rate": 2.779964356117063e-06, | |
| "loss": 0.0788, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.8331637382507324, | |
| "learning_rate": 2.750567536508504e-06, | |
| "loss": 0.0905, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.3373015873015874, | |
| "grad_norm": 0.8110689520835876, | |
| "learning_rate": 2.7212678831287627e-06, | |
| "loss": 0.0928, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.3412698412698414, | |
| "grad_norm": 0.8174662590026855, | |
| "learning_rate": 2.692066661602333e-06, | |
| "loss": 0.084, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.3452380952380953, | |
| "grad_norm": 0.8551895618438721, | |
| "learning_rate": 2.662965133301862e-06, | |
| "loss": 0.0957, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.3492063492063493, | |
| "grad_norm": 0.839144766330719, | |
| "learning_rate": 2.633964555293654e-06, | |
| "loss": 0.0912, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.3531746031746033, | |
| "grad_norm": 0.7665494680404663, | |
| "learning_rate": 2.605066180283378e-06, | |
| "loss": 0.0785, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.3571428571428572, | |
| "grad_norm": 0.880729079246521, | |
| "learning_rate": 2.576271256561953e-06, | |
| "loss": 0.087, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.3611111111111112, | |
| "grad_norm": 0.8345591425895691, | |
| "learning_rate": 2.5475810279516287e-06, | |
| "loss": 0.0961, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.3650793650793651, | |
| "grad_norm": 0.8339279294013977, | |
| "learning_rate": 2.5189967337522574e-06, | |
| "loss": 0.0863, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.369047619047619, | |
| "grad_norm": 0.9567959904670715, | |
| "learning_rate": 2.49051960868776e-06, | |
| "loss": 0.1068, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.373015873015873, | |
| "grad_norm": 0.7667821645736694, | |
| "learning_rate": 2.46215088285279e-06, | |
| "loss": 0.0833, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.376984126984127, | |
| "grad_norm": 0.8011244535446167, | |
| "learning_rate": 2.433891781659603e-06, | |
| "loss": 0.072, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.380952380952381, | |
| "grad_norm": 0.8509224653244019, | |
| "learning_rate": 2.4057435257851173e-06, | |
| "loss": 0.1056, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.3849206349206349, | |
| "grad_norm": 0.7882251143455505, | |
| "learning_rate": 2.377707331118196e-06, | |
| "loss": 0.0941, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.3888888888888888, | |
| "grad_norm": 0.7005813121795654, | |
| "learning_rate": 2.349784408707112e-06, | |
| "loss": 0.0848, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.3928571428571428, | |
| "grad_norm": 0.7593621015548706, | |
| "learning_rate": 2.3219759647072467e-06, | |
| "loss": 0.0758, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.3968253968253967, | |
| "grad_norm": 0.86487877368927, | |
| "learning_rate": 2.2942832003289823e-06, | |
| "loss": 0.0895, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.4007936507936507, | |
| "grad_norm": 0.8020721673965454, | |
| "learning_rate": 2.2667073117858185e-06, | |
| "loss": 0.1081, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.4047619047619047, | |
| "grad_norm": 0.6789891123771667, | |
| "learning_rate": 2.2392494902427027e-06, | |
| "loss": 0.0715, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.4087301587301586, | |
| "grad_norm": 0.6987098455429077, | |
| "learning_rate": 2.2119109217645697e-06, | |
| "loss": 0.0854, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.4126984126984126, | |
| "grad_norm": 0.8296271562576294, | |
| "learning_rate": 2.1846927872651135e-06, | |
| "loss": 0.0812, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.4166666666666667, | |
| "grad_norm": 0.7869781851768494, | |
| "learning_rate": 2.1575962624557754e-06, | |
| "loss": 0.0922, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.4206349206349207, | |
| "grad_norm": 0.8419052362442017, | |
| "learning_rate": 2.1306225177949584e-06, | |
| "loss": 0.1003, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.4246031746031746, | |
| "grad_norm": 0.7135988473892212, | |
| "learning_rate": 2.1037727184374705e-06, | |
| "loss": 0.0724, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 0.7993552684783936, | |
| "learning_rate": 2.07704802418419e-06, | |
| "loss": 0.0928, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.4325396825396826, | |
| "grad_norm": 0.911687970161438, | |
| "learning_rate": 2.050449589431969e-06, | |
| "loss": 0.1044, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.4365079365079365, | |
| "grad_norm": 0.7924899458885193, | |
| "learning_rate": 2.023978563123771e-06, | |
| "loss": 0.0795, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.4404761904761905, | |
| "grad_norm": 0.8037704229354858, | |
| "learning_rate": 1.997636088699035e-06, | |
| "loss": 0.0884, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.4444444444444444, | |
| "grad_norm": 0.9572436213493347, | |
| "learning_rate": 1.9714233040442915e-06, | |
| "loss": 0.0924, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.4484126984126984, | |
| "grad_norm": 0.7599306106567383, | |
| "learning_rate": 1.9453413414440043e-06, | |
| "loss": 0.0883, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.4523809523809523, | |
| "grad_norm": 0.936923623085022, | |
| "learning_rate": 1.919391327531663e-06, | |
| "loss": 0.1162, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.4563492063492063, | |
| "grad_norm": 0.7141274213790894, | |
| "learning_rate": 1.8935743832411163e-06, | |
| "loss": 0.0745, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.4603174603174602, | |
| "grad_norm": 0.9221301078796387, | |
| "learning_rate": 1.8678916237581524e-06, | |
| "loss": 0.0947, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.4642857142857144, | |
| "grad_norm": 0.926031768321991, | |
| "learning_rate": 1.8423441584723312e-06, | |
| "loss": 0.0945, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.4682539682539684, | |
| "grad_norm": 0.8945465087890625, | |
| "learning_rate": 1.816933090929055e-06, | |
| "loss": 0.0934, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.4722222222222223, | |
| "grad_norm": 0.8531501293182373, | |
| "learning_rate": 1.791659518781908e-06, | |
| "loss": 0.0761, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.4761904761904763, | |
| "grad_norm": 0.9297811985015869, | |
| "learning_rate": 1.7665245337452368e-06, | |
| "loss": 0.0991, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.4801587301587302, | |
| "grad_norm": 0.8548992276191711, | |
| "learning_rate": 1.7415292215469948e-06, | |
| "loss": 0.0818, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.4841269841269842, | |
| "grad_norm": 0.8387070298194885, | |
| "learning_rate": 1.716674661881848e-06, | |
| "loss": 0.1002, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.4880952380952381, | |
| "grad_norm": 1.011826753616333, | |
| "learning_rate": 1.6919619283645262e-06, | |
| "loss": 0.1029, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.492063492063492, | |
| "grad_norm": 0.7841417193412781, | |
| "learning_rate": 1.667392088483456e-06, | |
| "loss": 0.0861, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.496031746031746, | |
| "grad_norm": 0.8473195433616638, | |
| "learning_rate": 1.6429662035546451e-06, | |
| "loss": 0.0818, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.636673629283905, | |
| "learning_rate": 1.6186853286758397e-06, | |
| "loss": 0.0693, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.503968253968254, | |
| "grad_norm": 0.6218969821929932, | |
| "learning_rate": 1.5945505126809524e-06, | |
| "loss": 0.0574, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.507936507936508, | |
| "grad_norm": 0.6572891473770142, | |
| "learning_rate": 1.570562798094747e-06, | |
| "loss": 0.064, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.5119047619047619, | |
| "grad_norm": 0.8124316334724426, | |
| "learning_rate": 1.5467232210878153e-06, | |
| "loss": 0.1065, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.5158730158730158, | |
| "grad_norm": 0.8636795878410339, | |
| "learning_rate": 1.5230328114318127e-06, | |
| "loss": 0.0923, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.5198412698412698, | |
| "grad_norm": 0.6614148020744324, | |
| "learning_rate": 1.4994925924549797e-06, | |
| "loss": 0.0655, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.5238095238095237, | |
| "grad_norm": 0.7655928730964661, | |
| "learning_rate": 1.4761035809979395e-06, | |
| "loss": 0.0665, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.5277777777777777, | |
| "grad_norm": 0.7094256281852722, | |
| "learning_rate": 1.452866787369771e-06, | |
| "loss": 0.0872, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.5317460317460316, | |
| "grad_norm": 0.761876106262207, | |
| "learning_rate": 1.4297832153043657e-06, | |
| "loss": 0.083, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.5357142857142856, | |
| "grad_norm": 0.7579266428947449, | |
| "learning_rate": 1.4068538619170763e-06, | |
| "loss": 0.0706, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.5396825396825395, | |
| "grad_norm": 0.7755815386772156, | |
| "learning_rate": 1.3840797176616467e-06, | |
| "loss": 0.0756, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.5436507936507935, | |
| "grad_norm": 0.8676632642745972, | |
| "learning_rate": 1.3614617662874197e-06, | |
| "loss": 0.0971, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.5476190476190477, | |
| "grad_norm": 0.818500280380249, | |
| "learning_rate": 1.3390009847968505e-06, | |
| "loss": 0.0866, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.5515873015873016, | |
| "grad_norm": 0.7812132835388184, | |
| "learning_rate": 1.316698343403302e-06, | |
| "loss": 0.0992, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.5555555555555556, | |
| "grad_norm": 0.9807851314544678, | |
| "learning_rate": 1.2945548054891322e-06, | |
| "loss": 0.1187, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.5595238095238095, | |
| "grad_norm": 0.8131131529808044, | |
| "learning_rate": 1.27257132756409e-06, | |
| "loss": 0.0795, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.5634920634920635, | |
| "grad_norm": 0.8454893827438354, | |
| "learning_rate": 1.2507488592239848e-06, | |
| "loss": 0.0901, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.5674603174603174, | |
| "grad_norm": 0.8869048357009888, | |
| "learning_rate": 1.2290883431096778e-06, | |
| "loss": 0.0841, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 0.7952851057052612, | |
| "learning_rate": 1.2075907148663579e-06, | |
| "loss": 0.0808, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.5753968253968254, | |
| "grad_norm": 0.7806260585784912, | |
| "learning_rate": 1.186256903103129e-06, | |
| "loss": 0.073, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.5793650793650795, | |
| "grad_norm": 0.945780336856842, | |
| "learning_rate": 1.1650878293528994e-06, | |
| "loss": 0.1026, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.5833333333333335, | |
| "grad_norm": 0.8170139789581299, | |
| "learning_rate": 1.1440844080325703e-06, | |
| "loss": 0.0902, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.5873015873015874, | |
| "grad_norm": 0.7791316509246826, | |
| "learning_rate": 1.1232475464035386e-06, | |
| "loss": 0.0874, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.5912698412698414, | |
| "grad_norm": 0.6370510458946228, | |
| "learning_rate": 1.10257814453251e-06, | |
| "loss": 0.0623, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.5952380952380953, | |
| "grad_norm": 0.8416541814804077, | |
| "learning_rate": 1.0820770952526155e-06, | |
| "loss": 0.0873, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.5992063492063493, | |
| "grad_norm": 0.6123059988021851, | |
| "learning_rate": 1.0617452841248494e-06, | |
| "loss": 0.0813, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.6031746031746033, | |
| "grad_norm": 0.7546087503433228, | |
| "learning_rate": 1.0415835893998116e-06, | |
| "loss": 0.0899, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.6071428571428572, | |
| "grad_norm": 0.7951581478118896, | |
| "learning_rate": 1.0215928819797744e-06, | |
| "loss": 0.0974, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.6111111111111112, | |
| "grad_norm": 0.657884418964386, | |
| "learning_rate": 1.0017740253810608e-06, | |
| "loss": 0.0573, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.6150793650793651, | |
| "grad_norm": 0.6941938400268555, | |
| "learning_rate": 9.821278756967467e-07, | |
| "loss": 0.0771, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.619047619047619, | |
| "grad_norm": 0.9812749028205872, | |
| "learning_rate": 9.62655281559679e-07, | |
| "loss": 0.1157, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.623015873015873, | |
| "grad_norm": 0.7745327949523926, | |
| "learning_rate": 9.433570841058187e-07, | |
| "loss": 0.0801, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.626984126984127, | |
| "grad_norm": 0.7946932911872864, | |
| "learning_rate": 9.242341169379077e-07, | |
| "loss": 0.0858, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.630952380952381, | |
| "grad_norm": 0.7916086912155151, | |
| "learning_rate": 9.052872060894613e-07, | |
| "loss": 0.0848, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.6349206349206349, | |
| "grad_norm": 0.8336475491523743, | |
| "learning_rate": 8.865171699890835e-07, | |
| "loss": 0.0788, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.6388888888888888, | |
| "grad_norm": 0.8196171522140503, | |
| "learning_rate": 8.679248194251211e-07, | |
| "loss": 0.0944, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.6428571428571428, | |
| "grad_norm": 0.7034598588943481, | |
| "learning_rate": 8.495109575106331e-07, | |
| "loss": 0.0756, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.6468253968253967, | |
| "grad_norm": 0.7516794204711914, | |
| "learning_rate": 8.312763796487038e-07, | |
| "loss": 0.0745, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.6507936507936507, | |
| "grad_norm": 0.7654904723167419, | |
| "learning_rate": 8.132218734980852e-07, | |
| "loss": 0.0795, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.6547619047619047, | |
| "grad_norm": 0.7922310829162598, | |
| "learning_rate": 7.953482189391687e-07, | |
| "loss": 0.0877, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.6587301587301586, | |
| "grad_norm": 0.8557181358337402, | |
| "learning_rate": 7.776561880403072e-07, | |
| "loss": 0.0873, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.6626984126984126, | |
| "grad_norm": 1.0167126655578613, | |
| "learning_rate": 7.601465450244528e-07, | |
| "loss": 0.1491, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.7713876962661743, | |
| "learning_rate": 7.42820046236154e-07, | |
| "loss": 0.0758, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.6706349206349205, | |
| "grad_norm": 0.8583239912986755, | |
| "learning_rate": 7.256774401088817e-07, | |
| "loss": 0.0933, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.6746031746031746, | |
| "grad_norm": 0.8612584471702576, | |
| "learning_rate": 7.087194671326986e-07, | |
| "loss": 0.0775, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.6785714285714286, | |
| "grad_norm": 0.9081663489341736, | |
| "learning_rate": 6.91946859822279e-07, | |
| "loss": 0.1025, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.6825396825396826, | |
| "grad_norm": 0.8462613224983215, | |
| "learning_rate": 6.753603426852589e-07, | |
| "loss": 0.1076, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.6865079365079365, | |
| "grad_norm": 0.6927803754806519, | |
| "learning_rate": 6.589606321909464e-07, | |
| "loss": 0.0801, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.6904761904761905, | |
| "grad_norm": 0.8397090435028076, | |
| "learning_rate": 6.427484367393699e-07, | |
| "loss": 0.1019, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.6944444444444444, | |
| "grad_norm": 0.7412496209144592, | |
| "learning_rate": 6.267244566306801e-07, | |
| "loss": 0.0779, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.6984126984126984, | |
| "grad_norm": 0.7467244863510132, | |
| "learning_rate": 6.108893840348995e-07, | |
| "loss": 0.1018, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.7023809523809523, | |
| "grad_norm": 0.7235032320022583, | |
| "learning_rate": 5.952439029620222e-07, | |
| "loss": 0.0737, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.7063492063492065, | |
| "grad_norm": 0.6594804525375366, | |
| "learning_rate": 5.797886892324695e-07, | |
| "loss": 0.0657, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.7103174603174605, | |
| "grad_norm": 0.6685808897018433, | |
| "learning_rate": 5.645244104478947e-07, | |
| "loss": 0.0755, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 0.8202279210090637, | |
| "learning_rate": 5.494517259623478e-07, | |
| "loss": 0.1106, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.7182539682539684, | |
| "grad_norm": 0.7880169153213501, | |
| "learning_rate": 5.34571286853795e-07, | |
| "loss": 0.1068, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.7222222222222223, | |
| "grad_norm": 0.8468673229217529, | |
| "learning_rate": 5.198837358959901e-07, | |
| "loss": 0.1055, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.7261904761904763, | |
| "grad_norm": 0.6636783480644226, | |
| "learning_rate": 5.05389707530714e-07, | |
| "loss": 0.0618, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.7301587301587302, | |
| "grad_norm": 0.6822192072868347, | |
| "learning_rate": 4.91089827840367e-07, | |
| "loss": 0.0759, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.7341269841269842, | |
| "grad_norm": 0.8493044376373291, | |
| "learning_rate": 4.769847145209244e-07, | |
| "loss": 0.1057, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.7380952380952381, | |
| "grad_norm": 0.7789743542671204, | |
| "learning_rate": 4.6307497685525894e-07, | |
| "loss": 0.0842, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.742063492063492, | |
| "grad_norm": 0.7724023461341858, | |
| "learning_rate": 4.4936121568681546e-07, | |
| "loss": 0.0817, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.746031746031746, | |
| "grad_norm": 0.7365525364875793, | |
| "learning_rate": 4.3584402339366174e-07, | |
| "loss": 0.0714, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.7248952388763428, | |
| "learning_rate": 4.225239838628981e-07, | |
| "loss": 0.0727, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.753968253968254, | |
| "grad_norm": 0.7953075170516968, | |
| "learning_rate": 4.0940167246543595e-07, | |
| "loss": 0.0842, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.757936507936508, | |
| "grad_norm": 0.8194500803947449, | |
| "learning_rate": 3.964776560311484e-07, | |
| "loss": 0.0915, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.7619047619047619, | |
| "grad_norm": 0.7962750196456909, | |
| "learning_rate": 3.8375249282437743e-07, | |
| "loss": 0.0882, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.7658730158730158, | |
| "grad_norm": 0.8540671467781067, | |
| "learning_rate": 3.71226732519826e-07, | |
| "loss": 0.0976, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.7698412698412698, | |
| "grad_norm": 0.8313424587249756, | |
| "learning_rate": 3.589009161788104e-07, | |
| "loss": 0.0807, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.7738095238095237, | |
| "grad_norm": 0.749427855014801, | |
| "learning_rate": 3.4677557622589175e-07, | |
| "loss": 0.0748, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.8093726634979248, | |
| "learning_rate": 3.3485123642587657e-07, | |
| "loss": 0.081, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.7817460317460316, | |
| "grad_norm": 0.7896418571472168, | |
| "learning_rate": 3.2312841186118937e-07, | |
| "loss": 0.0789, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.7857142857142856, | |
| "grad_norm": 0.8458564281463623, | |
| "learning_rate": 3.116076089096265e-07, | |
| "loss": 0.0841, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.7896825396825395, | |
| "grad_norm": 0.759649395942688, | |
| "learning_rate": 3.0028932522248256e-07, | |
| "loss": 0.0793, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.7936507936507935, | |
| "grad_norm": 0.7106239795684814, | |
| "learning_rate": 2.8917404970305096e-07, | |
| "loss": 0.0705, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.7976190476190477, | |
| "grad_norm": 0.7886518836021423, | |
| "learning_rate": 2.782622624855097e-07, | |
| "loss": 0.0826, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.8015873015873016, | |
| "grad_norm": 0.711067795753479, | |
| "learning_rate": 2.6755443491417786e-07, | |
| "loss": 0.0799, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.8055555555555556, | |
| "grad_norm": 0.6752684712409973, | |
| "learning_rate": 2.570510295231571e-07, | |
| "loss": 0.0665, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.8095238095238095, | |
| "grad_norm": 0.8066158890724182, | |
| "learning_rate": 2.467525000163523e-07, | |
| "loss": 0.0855, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.8134920634920635, | |
| "grad_norm": 0.8242791891098022, | |
| "learning_rate": 2.36659291247871e-07, | |
| "loss": 0.0976, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.8174603174603174, | |
| "grad_norm": 0.6588331460952759, | |
| "learning_rate": 2.2677183920281342e-07, | |
| "loss": 0.0543, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.8214285714285714, | |
| "grad_norm": 0.7499187588691711, | |
| "learning_rate": 2.1709057097843266e-07, | |
| "loss": 0.0726, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.8253968253968254, | |
| "grad_norm": 0.7212550640106201, | |
| "learning_rate": 2.0761590476568893e-07, | |
| "loss": 0.0658, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.8293650793650795, | |
| "grad_norm": 0.8763760328292847, | |
| "learning_rate": 1.9834824983118673e-07, | |
| "loss": 0.1061, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.8333333333333335, | |
| "grad_norm": 0.6587537527084351, | |
| "learning_rate": 1.892880064994934e-07, | |
| "loss": 0.061, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.8373015873015874, | |
| "grad_norm": 0.8592242002487183, | |
| "learning_rate": 1.8043556613585143e-07, | |
| "loss": 0.0956, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.8412698412698414, | |
| "grad_norm": 0.885543942451477, | |
| "learning_rate": 1.7179131112926628e-07, | |
| "loss": 0.0961, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.8452380952380953, | |
| "grad_norm": 0.7622782588005066, | |
| "learning_rate": 1.6335561487599406e-07, | |
| "loss": 0.0757, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.8492063492063493, | |
| "grad_norm": 1.004665493965149, | |
| "learning_rate": 1.551288417634106e-07, | |
| "loss": 0.1354, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.8531746031746033, | |
| "grad_norm": 0.8439547419548035, | |
| "learning_rate": 1.471113471542712e-07, | |
| "loss": 0.089, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 0.8595339059829712, | |
| "learning_rate": 1.3930347737136195e-07, | |
| "loss": 0.1045, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.8611111111111112, | |
| "grad_norm": 0.9769722819328308, | |
| "learning_rate": 1.3170556968253756e-07, | |
| "loss": 0.1181, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.8650793650793651, | |
| "grad_norm": 0.8348788619041443, | |
| "learning_rate": 1.2431795228615372e-07, | |
| "loss": 0.0841, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.869047619047619, | |
| "grad_norm": 0.8991684317588806, | |
| "learning_rate": 1.1714094429689127e-07, | |
| "loss": 0.0836, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.873015873015873, | |
| "grad_norm": 0.8475848436355591, | |
| "learning_rate": 1.1017485573197151e-07, | |
| "loss": 0.0913, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.876984126984127, | |
| "grad_norm": 0.7161443829536438, | |
| "learning_rate": 1.0341998749776316e-07, | |
| "loss": 0.0636, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.880952380952381, | |
| "grad_norm": 0.8380317687988281, | |
| "learning_rate": 9.687663137678605e-08, | |
| "loss": 0.0989, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.8849206349206349, | |
| "grad_norm": 0.9291573762893677, | |
| "learning_rate": 9.054507001510727e-08, | |
| "loss": 0.1095, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.8888888888888888, | |
| "grad_norm": 0.655913233757019, | |
| "learning_rate": 8.442557691013042e-08, | |
| "loss": 0.075, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.8928571428571428, | |
| "grad_norm": 0.8449409008026123, | |
| "learning_rate": 7.851841639878399e-08, | |
| "loss": 0.0803, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.8968253968253967, | |
| "grad_norm": 0.7550128698348999, | |
| "learning_rate": 7.282384364610207e-08, | |
| "loss": 0.0971, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.9007936507936507, | |
| "grad_norm": 0.6306801438331604, | |
| "learning_rate": 6.734210463420099e-08, | |
| "loss": 0.0603, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.9047619047619047, | |
| "grad_norm": 0.6543687582015991, | |
| "learning_rate": 6.207343615165562e-08, | |
| "loss": 0.0695, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.9087301587301586, | |
| "grad_norm": 0.7484958171844482, | |
| "learning_rate": 5.701806578327029e-08, | |
| "loss": 0.0658, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.9126984126984126, | |
| "grad_norm": 0.7087326049804688, | |
| "learning_rate": 5.21762119002478e-08, | |
| "loss": 0.0733, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.9166666666666665, | |
| "grad_norm": 0.6761496663093567, | |
| "learning_rate": 4.7548083650759134e-08, | |
| "loss": 0.066, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.9206349206349205, | |
| "grad_norm": 0.5993028283119202, | |
| "learning_rate": 4.31338809509052e-08, | |
| "loss": 0.0547, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.9246031746031746, | |
| "grad_norm": 0.8561880588531494, | |
| "learning_rate": 3.8933794476083143e-08, | |
| "loss": 0.07, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.9285714285714286, | |
| "grad_norm": 0.7461079955101013, | |
| "learning_rate": 3.494800565275125e-08, | |
| "loss": 0.0708, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.9325396825396826, | |
| "grad_norm": 0.7032601237297058, | |
| "learning_rate": 3.1176686650589147e-08, | |
| "loss": 0.074, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.9365079365079365, | |
| "grad_norm": 0.7916510105133057, | |
| "learning_rate": 2.7620000375064848e-08, | |
| "loss": 0.0924, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.9404761904761905, | |
| "grad_norm": 0.7835144400596619, | |
| "learning_rate": 2.4278100460393138e-08, | |
| "loss": 0.0689, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.9444444444444444, | |
| "grad_norm": 0.9007583260536194, | |
| "learning_rate": 2.115113126290258e-08, | |
| "loss": 0.1015, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.9484126984126984, | |
| "grad_norm": 0.8017187714576721, | |
| "learning_rate": 1.8239227854799368e-08, | |
| "loss": 0.0728, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.9523809523809523, | |
| "grad_norm": 0.8309543132781982, | |
| "learning_rate": 1.554251601833201e-08, | |
| "loss": 0.0911, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.9563492063492065, | |
| "grad_norm": 0.7955175042152405, | |
| "learning_rate": 1.3061112240357887e-08, | |
| "loss": 0.0696, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 1.9603174603174605, | |
| "grad_norm": 0.7506862282752991, | |
| "learning_rate": 1.0795123707312283e-08, | |
| "loss": 0.0723, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 1.9642857142857144, | |
| "grad_norm": 0.8498867154121399, | |
| "learning_rate": 8.744648300578196e-09, | |
| "loss": 0.0876, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.9682539682539684, | |
| "grad_norm": 0.8584848046302795, | |
| "learning_rate": 6.9097745922580564e-09, | |
| "loss": 0.0809, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 1.9722222222222223, | |
| "grad_norm": 0.7752880454063416, | |
| "learning_rate": 5.2905818413478975e-09, | |
| "loss": 0.0683, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 1.9761904761904763, | |
| "grad_norm": 0.7957850694656372, | |
| "learning_rate": 3.887139990313427e-09, | |
| "loss": 0.0838, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 1.9801587301587302, | |
| "grad_norm": 0.8344635367393494, | |
| "learning_rate": 2.699509662069666e-09, | |
| "loss": 0.1012, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 1.9841269841269842, | |
| "grad_norm": 0.6328668594360352, | |
| "learning_rate": 1.7277421573608234e-09, | |
| "loss": 0.0706, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.9880952380952381, | |
| "grad_norm": 0.7299294471740723, | |
| "learning_rate": 9.71879452545399e-10, | |
| "loss": 0.0803, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 1.992063492063492, | |
| "grad_norm": 0.9233019948005676, | |
| "learning_rate": 4.3195419778319095e-10, | |
| "loss": 0.1024, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 1.996031746031746, | |
| "grad_norm": 0.8229629993438721, | |
| "learning_rate": 1.0798971562364647e-10, | |
| "loss": 0.0778, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.6106996536254883, | |
| "learning_rate": 0.0, | |
| "loss": 0.0529, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 504, | |
| "total_flos": 54880760659968.0, | |
| "train_loss": 0.22782912893989493, | |
| "train_runtime": 3942.2537, | |
| "train_samples_per_second": 2.041, | |
| "train_steps_per_second": 0.128 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 504, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 54880760659968.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |