| { | |
| "best_metric": 0.5934335589408875, | |
| "best_model_checkpoint": "/ephemeral/models/qwen-describe_tasks_uniform_1slice/checkpoint-171", | |
| "epoch": 6.0, | |
| "eval_steps": 57, | |
| "global_step": 1344, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004464285714285714, | |
| "grad_norm": 11.030957221984863, | |
| "learning_rate": 0.0, | |
| "loss": 1.3544, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.008928571428571428, | |
| "grad_norm": 7.5021185874938965, | |
| "learning_rate": 2.626495350371936e-06, | |
| "loss": 1.081, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.013392857142857142, | |
| "grad_norm": 7.396653652191162, | |
| "learning_rate": 4.162896638657993e-06, | |
| "loss": 1.0844, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.017857142857142856, | |
| "grad_norm": 6.913914203643799, | |
| "learning_rate": 5.252990700743872e-06, | |
| "loss": 0.9979, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.022321428571428572, | |
| "grad_norm": 6.602043628692627, | |
| "learning_rate": 6.098533345119624e-06, | |
| "loss": 0.7692, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.026785714285714284, | |
| "grad_norm": 4.961360454559326, | |
| "learning_rate": 6.7893919890299284e-06, | |
| "loss": 0.7355, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.03125, | |
| "grad_norm": 3.8967952728271484, | |
| "learning_rate": 7.373504649628066e-06, | |
| "loss": 0.6837, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.03571428571428571, | |
| "grad_norm": 5.37336540222168, | |
| "learning_rate": 7.879486051115807e-06, | |
| "loss": 0.7389, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.04017857142857143, | |
| "grad_norm": 4.335018157958984, | |
| "learning_rate": 8.325793277315987e-06, | |
| "loss": 0.6734, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.044642857142857144, | |
| "grad_norm": 3.5685887336730957, | |
| "learning_rate": 8.72502869549156e-06, | |
| "loss": 0.6469, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.049107142857142856, | |
| "grad_norm": 3.8976097106933594, | |
| "learning_rate": 9.086181061280522e-06, | |
| "loss": 0.5893, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.05357142857142857, | |
| "grad_norm": 3.3721907138824463, | |
| "learning_rate": 9.415887339401865e-06, | |
| "loss": 0.5792, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.05803571428571429, | |
| "grad_norm": 3.204407215118408, | |
| "learning_rate": 9.719187714029216e-06, | |
| "loss": 0.6511, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0625, | |
| "grad_norm": 2.965876817703247, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5734, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.06696428571428571, | |
| "grad_norm": 3.5149614810943604, | |
| "learning_rate": 9.999986052613417e-06, | |
| "loss": 0.5893, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 3.0840070247650146, | |
| "learning_rate": 9.99994421053148e-06, | |
| "loss": 0.6035, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.07589285714285714, | |
| "grad_norm": 2.63883113861084, | |
| "learning_rate": 9.999874473987653e-06, | |
| "loss": 0.5674, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.08035714285714286, | |
| "grad_norm": 2.7957351207733154, | |
| "learning_rate": 9.999776843371027e-06, | |
| "loss": 0.6093, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.08482142857142858, | |
| "grad_norm": 2.5564727783203125, | |
| "learning_rate": 9.99965131922634e-06, | |
| "loss": 0.5237, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.08928571428571429, | |
| "grad_norm": 2.794530153274536, | |
| "learning_rate": 9.999497902253949e-06, | |
| "loss": 0.5938, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09375, | |
| "grad_norm": 2.564220428466797, | |
| "learning_rate": 9.999316593309849e-06, | |
| "loss": 0.605, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.09821428571428571, | |
| "grad_norm": 2.927748918533325, | |
| "learning_rate": 9.999107393405655e-06, | |
| "loss": 0.5935, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.10267857142857142, | |
| "grad_norm": 3.0531842708587646, | |
| "learning_rate": 9.998870303708601e-06, | |
| "loss": 0.4989, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.10714285714285714, | |
| "grad_norm": 2.589479923248291, | |
| "learning_rate": 9.998605325541531e-06, | |
| "loss": 0.5839, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.11160714285714286, | |
| "grad_norm": 2.739987850189209, | |
| "learning_rate": 9.998312460382895e-06, | |
| "loss": 0.5932, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.11607142857142858, | |
| "grad_norm": 2.6279497146606445, | |
| "learning_rate": 9.997991709866738e-06, | |
| "loss": 0.5965, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.12053571428571429, | |
| "grad_norm": 2.1124532222747803, | |
| "learning_rate": 9.997643075782691e-06, | |
| "loss": 0.475, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 2.6372640132904053, | |
| "learning_rate": 9.997266560075961e-06, | |
| "loss": 0.5592, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.12946428571428573, | |
| "grad_norm": 3.1374351978302, | |
| "learning_rate": 9.996862164847323e-06, | |
| "loss": 0.5208, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.13392857142857142, | |
| "grad_norm": 2.652562379837036, | |
| "learning_rate": 9.996429892353107e-06, | |
| "loss": 0.4983, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.13839285714285715, | |
| "grad_norm": 2.6861250400543213, | |
| "learning_rate": 9.99596974500518e-06, | |
| "loss": 0.5353, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 2.821516990661621, | |
| "learning_rate": 9.995481725370941e-06, | |
| "loss": 0.5777, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.14732142857142858, | |
| "grad_norm": 2.599327325820923, | |
| "learning_rate": 9.994965836173303e-06, | |
| "loss": 0.5224, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.15178571428571427, | |
| "grad_norm": 3.059565544128418, | |
| "learning_rate": 9.994422080290675e-06, | |
| "loss": 0.5117, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.15625, | |
| "grad_norm": 2.7172787189483643, | |
| "learning_rate": 9.99385046075695e-06, | |
| "loss": 0.5436, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.16071428571428573, | |
| "grad_norm": 2.469592571258545, | |
| "learning_rate": 9.993250980761487e-06, | |
| "loss": 0.5107, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.16517857142857142, | |
| "grad_norm": 2.5674359798431396, | |
| "learning_rate": 9.99262364364909e-06, | |
| "loss": 0.5125, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.16964285714285715, | |
| "grad_norm": 2.548276901245117, | |
| "learning_rate": 9.991968452919999e-06, | |
| "loss": 0.4619, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.17410714285714285, | |
| "grad_norm": 2.671201229095459, | |
| "learning_rate": 9.991285412229854e-06, | |
| "loss": 0.5114, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.17857142857142858, | |
| "grad_norm": 2.5654966831207275, | |
| "learning_rate": 9.99057452538969e-06, | |
| "loss": 0.5276, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.18303571428571427, | |
| "grad_norm": 2.3860409259796143, | |
| "learning_rate": 9.989835796365911e-06, | |
| "loss": 0.4979, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.1875, | |
| "grad_norm": 2.5768401622772217, | |
| "learning_rate": 9.989069229280264e-06, | |
| "loss": 0.5313, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.19196428571428573, | |
| "grad_norm": 2.3968393802642822, | |
| "learning_rate": 9.988274828409821e-06, | |
| "loss": 0.5388, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.19642857142857142, | |
| "grad_norm": 2.1060421466827393, | |
| "learning_rate": 9.987452598186947e-06, | |
| "loss": 0.487, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.20089285714285715, | |
| "grad_norm": 2.427931070327759, | |
| "learning_rate": 9.986602543199292e-06, | |
| "loss": 0.5106, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.20535714285714285, | |
| "grad_norm": 2.4082748889923096, | |
| "learning_rate": 9.985724668189744e-06, | |
| "loss": 0.5157, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.20982142857142858, | |
| "grad_norm": 2.3398637771606445, | |
| "learning_rate": 9.98481897805642e-06, | |
| "loss": 0.5287, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "grad_norm": 2.396620750427246, | |
| "learning_rate": 9.983885477852628e-06, | |
| "loss": 0.5999, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.21875, | |
| "grad_norm": 2.5863988399505615, | |
| "learning_rate": 9.982924172786847e-06, | |
| "loss": 0.5622, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.22321428571428573, | |
| "grad_norm": 2.56740665435791, | |
| "learning_rate": 9.981935068222687e-06, | |
| "loss": 0.5167, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.22767857142857142, | |
| "grad_norm": 2.1008498668670654, | |
| "learning_rate": 9.980918169678872e-06, | |
| "loss": 0.4426, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.23214285714285715, | |
| "grad_norm": 3.540738821029663, | |
| "learning_rate": 9.979873482829199e-06, | |
| "loss": 0.5136, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.23660714285714285, | |
| "grad_norm": 2.5687336921691895, | |
| "learning_rate": 9.978801013502511e-06, | |
| "loss": 0.5229, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.24107142857142858, | |
| "grad_norm": 2.21964168548584, | |
| "learning_rate": 9.977700767682665e-06, | |
| "loss": 0.4715, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.24553571428571427, | |
| "grad_norm": 2.621365547180176, | |
| "learning_rate": 9.976572751508497e-06, | |
| "loss": 0.5466, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 2.3530430793762207, | |
| "learning_rate": 9.975416971273787e-06, | |
| "loss": 0.5023, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.2544642857142857, | |
| "grad_norm": 2.344916820526123, | |
| "learning_rate": 9.974233433427222e-06, | |
| "loss": 0.4531, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.2544642857142857, | |
| "eval_loss": 0.6133683323860168, | |
| "eval_runtime": 3.5046, | |
| "eval_samples_per_second": 17.12, | |
| "eval_steps_per_second": 1.141, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.25892857142857145, | |
| "grad_norm": 1.9012365341186523, | |
| "learning_rate": 9.97302214457237e-06, | |
| "loss": 0.4561, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.26339285714285715, | |
| "grad_norm": 2.1153564453125, | |
| "learning_rate": 9.971783111467635e-06, | |
| "loss": 0.4818, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.26785714285714285, | |
| "grad_norm": 2.0768792629241943, | |
| "learning_rate": 9.970516341026211e-06, | |
| "loss": 0.4822, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.27232142857142855, | |
| "grad_norm": 2.338013172149658, | |
| "learning_rate": 9.969221840316066e-06, | |
| "loss": 0.5536, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.2767857142857143, | |
| "grad_norm": 2.0366456508636475, | |
| "learning_rate": 9.967899616559879e-06, | |
| "loss": 0.4722, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.28125, | |
| "grad_norm": 2.3978164196014404, | |
| "learning_rate": 9.966549677135015e-06, | |
| "loss": 0.4918, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 2.4098730087280273, | |
| "learning_rate": 9.965172029573479e-06, | |
| "loss": 0.4849, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.29017857142857145, | |
| "grad_norm": 2.1472318172454834, | |
| "learning_rate": 9.96376668156187e-06, | |
| "loss": 0.4636, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.29464285714285715, | |
| "grad_norm": 2.1821885108947754, | |
| "learning_rate": 9.962333640941349e-06, | |
| "loss": 0.5001, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.29910714285714285, | |
| "grad_norm": 2.1246650218963623, | |
| "learning_rate": 9.960872915707582e-06, | |
| "loss": 0.5134, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.30357142857142855, | |
| "grad_norm": 1.918825626373291, | |
| "learning_rate": 9.959384514010703e-06, | |
| "loss": 0.4598, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.3080357142857143, | |
| "grad_norm": 2.18090558052063, | |
| "learning_rate": 9.95786844415527e-06, | |
| "loss": 0.535, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.3125, | |
| "grad_norm": 2.012615203857422, | |
| "learning_rate": 9.956324714600212e-06, | |
| "loss": 0.5072, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.3169642857142857, | |
| "grad_norm": 2.2373433113098145, | |
| "learning_rate": 9.95475333395879e-06, | |
| "loss": 0.4815, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.32142857142857145, | |
| "grad_norm": 2.2920775413513184, | |
| "learning_rate": 9.95315431099854e-06, | |
| "loss": 0.4943, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.32589285714285715, | |
| "grad_norm": 2.103111743927002, | |
| "learning_rate": 9.951527654641231e-06, | |
| "loss": 0.4899, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.33035714285714285, | |
| "grad_norm": 2.1315195560455322, | |
| "learning_rate": 9.949873373962814e-06, | |
| "loss": 0.5516, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.33482142857142855, | |
| "grad_norm": 2.0358729362487793, | |
| "learning_rate": 9.948191478193365e-06, | |
| "loss": 0.4557, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.3392857142857143, | |
| "grad_norm": 1.9133963584899902, | |
| "learning_rate": 9.946481976717046e-06, | |
| "loss": 0.4406, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.34375, | |
| "grad_norm": 1.8832437992095947, | |
| "learning_rate": 9.944744879072043e-06, | |
| "loss": 0.4518, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.3482142857142857, | |
| "grad_norm": 2.119741439819336, | |
| "learning_rate": 9.942980194950511e-06, | |
| "loss": 0.5114, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.35267857142857145, | |
| "grad_norm": 2.543752670288086, | |
| "learning_rate": 9.941187934198528e-06, | |
| "loss": 0.546, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 2.132417917251587, | |
| "learning_rate": 9.939368106816038e-06, | |
| "loss": 0.4919, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.36160714285714285, | |
| "grad_norm": 2.2283120155334473, | |
| "learning_rate": 9.937520722956789e-06, | |
| "loss": 0.5794, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.36607142857142855, | |
| "grad_norm": 2.0629560947418213, | |
| "learning_rate": 9.93564579292828e-06, | |
| "loss": 0.4665, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.3705357142857143, | |
| "grad_norm": 3.117424249649048, | |
| "learning_rate": 9.933743327191711e-06, | |
| "loss": 0.4809, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 2.185500144958496, | |
| "learning_rate": 9.93181333636191e-06, | |
| "loss": 0.5341, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.3794642857142857, | |
| "grad_norm": 2.1382153034210205, | |
| "learning_rate": 9.929855831207288e-06, | |
| "loss": 0.4604, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.38392857142857145, | |
| "grad_norm": 2.0280189514160156, | |
| "learning_rate": 9.92787082264977e-06, | |
| "loss": 0.4251, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.38839285714285715, | |
| "grad_norm": 1.953497290611267, | |
| "learning_rate": 9.925858321764733e-06, | |
| "loss": 0.5032, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.39285714285714285, | |
| "grad_norm": 2.049941062927246, | |
| "learning_rate": 9.923818339780954e-06, | |
| "loss": 0.5318, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.39732142857142855, | |
| "grad_norm": 2.191439390182495, | |
| "learning_rate": 9.921750888080534e-06, | |
| "loss": 0.4664, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.4017857142857143, | |
| "grad_norm": 2.2432761192321777, | |
| "learning_rate": 9.91965597819885e-06, | |
| "loss": 0.4706, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.40625, | |
| "grad_norm": 1.9182389974594116, | |
| "learning_rate": 9.917533621824476e-06, | |
| "loss": 0.4089, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.4107142857142857, | |
| "grad_norm": 1.8381813764572144, | |
| "learning_rate": 9.915383830799129e-06, | |
| "loss": 0.4085, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.41517857142857145, | |
| "grad_norm": 2.1741788387298584, | |
| "learning_rate": 9.91320661711759e-06, | |
| "loss": 0.5138, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.41964285714285715, | |
| "grad_norm": 2.1558420658111572, | |
| "learning_rate": 9.911001992927655e-06, | |
| "loss": 0.4869, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.42410714285714285, | |
| "grad_norm": 2.0384535789489746, | |
| "learning_rate": 9.908769970530049e-06, | |
| "loss": 0.5034, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 2.0241775512695312, | |
| "learning_rate": 9.90651056237837e-06, | |
| "loss": 0.5012, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.4330357142857143, | |
| "grad_norm": 2.040360689163208, | |
| "learning_rate": 9.904223781079017e-06, | |
| "loss": 0.5085, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.4375, | |
| "grad_norm": 2.2121329307556152, | |
| "learning_rate": 9.901909639391111e-06, | |
| "loss": 0.5127, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.4419642857142857, | |
| "grad_norm": 2.294875383377075, | |
| "learning_rate": 9.899568150226435e-06, | |
| "loss": 0.4317, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.44642857142857145, | |
| "grad_norm": 2.159607410430908, | |
| "learning_rate": 9.897199326649362e-06, | |
| "loss": 0.4729, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.45089285714285715, | |
| "grad_norm": 1.8338195085525513, | |
| "learning_rate": 9.894803181876765e-06, | |
| "loss": 0.4525, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.45535714285714285, | |
| "grad_norm": 1.8389171361923218, | |
| "learning_rate": 9.892379729277972e-06, | |
| "loss": 0.4579, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.45982142857142855, | |
| "grad_norm": 1.8244115114212036, | |
| "learning_rate": 9.889928982374663e-06, | |
| "loss": 0.37, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.4642857142857143, | |
| "grad_norm": 2.351200819015503, | |
| "learning_rate": 9.887450954840812e-06, | |
| "loss": 0.4758, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.46875, | |
| "grad_norm": 2.3291563987731934, | |
| "learning_rate": 9.884945660502607e-06, | |
| "loss": 0.4573, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.4732142857142857, | |
| "grad_norm": 1.9490771293640137, | |
| "learning_rate": 9.882413113338364e-06, | |
| "loss": 0.487, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.47767857142857145, | |
| "grad_norm": 2.136775255203247, | |
| "learning_rate": 9.879853327478465e-06, | |
| "loss": 0.4859, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.48214285714285715, | |
| "grad_norm": 2.3490424156188965, | |
| "learning_rate": 9.877266317205268e-06, | |
| "loss": 0.4642, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.48660714285714285, | |
| "grad_norm": 2.140052556991577, | |
| "learning_rate": 9.874652096953028e-06, | |
| "loss": 0.4657, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.49107142857142855, | |
| "grad_norm": 2.1019790172576904, | |
| "learning_rate": 9.872010681307821e-06, | |
| "loss": 0.5293, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.4955357142857143, | |
| "grad_norm": 2.066462755203247, | |
| "learning_rate": 9.869342085007458e-06, | |
| "loss": 0.4226, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.9048563241958618, | |
| "learning_rate": 9.866646322941405e-06, | |
| "loss": 0.4657, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.5044642857142857, | |
| "grad_norm": 1.835811734199524, | |
| "learning_rate": 9.863923410150704e-06, | |
| "loss": 0.4676, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.5089285714285714, | |
| "grad_norm": 2.1999454498291016, | |
| "learning_rate": 9.861173361827876e-06, | |
| "loss": 0.4419, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.5089285714285714, | |
| "eval_loss": 0.6013900637626648, | |
| "eval_runtime": 3.4939, | |
| "eval_samples_per_second": 17.173, | |
| "eval_steps_per_second": 1.145, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.5133928571428571, | |
| "grad_norm": 1.9584894180297852, | |
| "learning_rate": 9.858396193316853e-06, | |
| "loss": 0.4626, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.5178571428571429, | |
| "grad_norm": 1.9644426107406616, | |
| "learning_rate": 9.855591920112883e-06, | |
| "loss": 0.4588, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.5223214285714286, | |
| "grad_norm": 1.712542176246643, | |
| "learning_rate": 9.85276055786244e-06, | |
| "loss": 0.4963, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.5267857142857143, | |
| "grad_norm": 2.043294906616211, | |
| "learning_rate": 9.849902122363148e-06, | |
| "loss": 0.4623, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.53125, | |
| "grad_norm": 1.8057119846343994, | |
| "learning_rate": 9.847016629563683e-06, | |
| "loss": 0.4633, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.5357142857142857, | |
| "grad_norm": 2.072108030319214, | |
| "learning_rate": 9.844104095563689e-06, | |
| "loss": 0.5328, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.5401785714285714, | |
| "grad_norm": 2.03574800491333, | |
| "learning_rate": 9.841164536613685e-06, | |
| "loss": 0.4889, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.5446428571428571, | |
| "grad_norm": 1.9355207681655884, | |
| "learning_rate": 9.83819796911498e-06, | |
| "loss": 0.53, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.5491071428571429, | |
| "grad_norm": 1.9122416973114014, | |
| "learning_rate": 9.83520440961957e-06, | |
| "loss": 0.3849, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.5535714285714286, | |
| "grad_norm": 1.7899999618530273, | |
| "learning_rate": 9.83218387483006e-06, | |
| "loss": 0.4963, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.5580357142857143, | |
| "grad_norm": 1.8941091299057007, | |
| "learning_rate": 9.829136381599563e-06, | |
| "loss": 0.4781, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.5625, | |
| "grad_norm": 1.7798947095870972, | |
| "learning_rate": 9.826061946931605e-06, | |
| "loss": 0.4165, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.5669642857142857, | |
| "grad_norm": 1.9236619472503662, | |
| "learning_rate": 9.822960587980034e-06, | |
| "loss": 0.4841, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 1.7892321348190308, | |
| "learning_rate": 9.81983232204892e-06, | |
| "loss": 0.4859, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.5758928571428571, | |
| "grad_norm": 1.9513256549835205, | |
| "learning_rate": 9.816677166592462e-06, | |
| "loss": 0.4075, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.5803571428571429, | |
| "grad_norm": 1.8629826307296753, | |
| "learning_rate": 9.81349513921489e-06, | |
| "loss": 0.4481, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5848214285714286, | |
| "grad_norm": 1.646044135093689, | |
| "learning_rate": 9.810286257670365e-06, | |
| "loss": 0.4173, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.5892857142857143, | |
| "grad_norm": 2.023775339126587, | |
| "learning_rate": 9.807050539862884e-06, | |
| "loss": 0.4307, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.59375, | |
| "grad_norm": 1.763723611831665, | |
| "learning_rate": 9.803788003846175e-06, | |
| "loss": 0.4549, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.5982142857142857, | |
| "grad_norm": 1.967829942703247, | |
| "learning_rate": 9.800498667823595e-06, | |
| "loss": 0.4401, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.6026785714285714, | |
| "grad_norm": 1.759050726890564, | |
| "learning_rate": 9.797182550148039e-06, | |
| "loss": 0.4588, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.6071428571428571, | |
| "grad_norm": 1.9382320642471313, | |
| "learning_rate": 9.793839669321828e-06, | |
| "loss": 0.4162, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.6116071428571429, | |
| "grad_norm": 1.9155679941177368, | |
| "learning_rate": 9.790470043996604e-06, | |
| "loss": 0.4854, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.6160714285714286, | |
| "grad_norm": 1.8347245454788208, | |
| "learning_rate": 9.78707369297324e-06, | |
| "loss": 0.4797, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.6205357142857143, | |
| "grad_norm": 1.8676778078079224, | |
| "learning_rate": 9.783650635201714e-06, | |
| "loss": 0.5195, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 1.8418855667114258, | |
| "learning_rate": 9.780200889781021e-06, | |
| "loss": 0.4262, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.6294642857142857, | |
| "grad_norm": 1.934958815574646, | |
| "learning_rate": 9.776724475959061e-06, | |
| "loss": 0.5029, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.6339285714285714, | |
| "grad_norm": 1.6102712154388428, | |
| "learning_rate": 9.773221413132525e-06, | |
| "loss": 0.4372, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.6383928571428571, | |
| "grad_norm": 1.7224174737930298, | |
| "learning_rate": 9.769691720846801e-06, | |
| "loss": 0.4196, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.6428571428571429, | |
| "grad_norm": 1.9889607429504395, | |
| "learning_rate": 9.766135418795848e-06, | |
| "loss": 0.4587, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.6473214285714286, | |
| "grad_norm": 1.9140293598175049, | |
| "learning_rate": 9.762552526822098e-06, | |
| "loss": 0.4267, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.6517857142857143, | |
| "grad_norm": 1.9280264377593994, | |
| "learning_rate": 9.758943064916342e-06, | |
| "loss": 0.4194, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.65625, | |
| "grad_norm": 1.6826374530792236, | |
| "learning_rate": 9.755307053217622e-06, | |
| "loss": 0.4084, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.6607142857142857, | |
| "grad_norm": 2.0488829612731934, | |
| "learning_rate": 9.751644512013106e-06, | |
| "loss": 0.4627, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.6651785714285714, | |
| "grad_norm": 1.9179582595825195, | |
| "learning_rate": 9.74795546173799e-06, | |
| "loss": 0.4595, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.6696428571428571, | |
| "grad_norm": 2.0788657665252686, | |
| "learning_rate": 9.744239922975377e-06, | |
| "loss": 0.4916, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.6741071428571429, | |
| "grad_norm": 1.7633529901504517, | |
| "learning_rate": 9.740497916456163e-06, | |
| "loss": 0.4813, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.6785714285714286, | |
| "grad_norm": 1.8860759735107422, | |
| "learning_rate": 9.736729463058921e-06, | |
| "loss": 0.5132, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.6830357142857143, | |
| "grad_norm": 1.725066065788269, | |
| "learning_rate": 9.732934583809782e-06, | |
| "loss": 0.4973, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.6875, | |
| "grad_norm": 1.7949278354644775, | |
| "learning_rate": 9.729113299882324e-06, | |
| "loss": 0.4693, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.6919642857142857, | |
| "grad_norm": 1.7584055662155151, | |
| "learning_rate": 9.725265632597448e-06, | |
| "loss": 0.4737, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.6964285714285714, | |
| "grad_norm": 1.766491174697876, | |
| "learning_rate": 9.721391603423263e-06, | |
| "loss": 0.4587, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.7008928571428571, | |
| "grad_norm": 1.7748725414276123, | |
| "learning_rate": 9.717491233974962e-06, | |
| "loss": 0.441, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.7053571428571429, | |
| "grad_norm": 1.8221688270568848, | |
| "learning_rate": 9.713564546014707e-06, | |
| "loss": 0.4942, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.7098214285714286, | |
| "grad_norm": 1.745335578918457, | |
| "learning_rate": 9.7096115614515e-06, | |
| "loss": 0.4398, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 1.9421252012252808, | |
| "learning_rate": 9.705632302341073e-06, | |
| "loss": 0.4656, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.71875, | |
| "grad_norm": 1.720116138458252, | |
| "learning_rate": 9.701626790885749e-06, | |
| "loss": 0.4361, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.7232142857142857, | |
| "grad_norm": 1.8272347450256348, | |
| "learning_rate": 9.69759504943433e-06, | |
| "loss": 0.4062, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.7276785714285714, | |
| "grad_norm": 1.6778669357299805, | |
| "learning_rate": 9.69353710048197e-06, | |
| "loss": 0.4743, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.7321428571428571, | |
| "grad_norm": 1.8853040933609009, | |
| "learning_rate": 9.68945296667004e-06, | |
| "loss": 0.4481, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.7366071428571429, | |
| "grad_norm": 1.826070785522461, | |
| "learning_rate": 9.685342670786025e-06, | |
| "loss": 0.5064, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.7410714285714286, | |
| "grad_norm": 2.058189868927002, | |
| "learning_rate": 9.681206235763367e-06, | |
| "loss": 0.488, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.7455357142857143, | |
| "grad_norm": 1.6986044645309448, | |
| "learning_rate": 9.677043684681358e-06, | |
| "loss": 0.4543, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 1.7179213762283325, | |
| "learning_rate": 9.672855040765006e-06, | |
| "loss": 0.4026, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.7544642857142857, | |
| "grad_norm": 1.7653132677078247, | |
| "learning_rate": 9.668640327384899e-06, | |
| "loss": 0.4711, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.7589285714285714, | |
| "grad_norm": 1.9445300102233887, | |
| "learning_rate": 9.664399568057087e-06, | |
| "loss": 0.4854, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.7633928571428571, | |
| "grad_norm": 1.8008228540420532, | |
| "learning_rate": 9.660132786442937e-06, | |
| "loss": 0.4672, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.7633928571428571, | |
| "eval_loss": 0.5934335589408875, | |
| "eval_runtime": 3.4095, | |
| "eval_samples_per_second": 17.598, | |
| "eval_steps_per_second": 1.173, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.7678571428571429, | |
| "grad_norm": 1.8834118843078613, | |
| "learning_rate": 9.655840006349014e-06, | |
| "loss": 0.4698, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.7723214285714286, | |
| "grad_norm": 1.642304539680481, | |
| "learning_rate": 9.651521251726936e-06, | |
| "loss": 0.4687, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.7767857142857143, | |
| "grad_norm": 1.7180356979370117, | |
| "learning_rate": 9.64717654667325e-06, | |
| "loss": 0.5118, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.78125, | |
| "grad_norm": 1.6180371046066284, | |
| "learning_rate": 9.642805915429291e-06, | |
| "loss": 0.4241, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.7857142857142857, | |
| "grad_norm": 1.7322157621383667, | |
| "learning_rate": 9.638409382381052e-06, | |
| "loss": 0.4465, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.7901785714285714, | |
| "grad_norm": 1.9031792879104614, | |
| "learning_rate": 9.633986972059047e-06, | |
| "loss": 0.4716, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.7946428571428571, | |
| "grad_norm": 1.7738388776779175, | |
| "learning_rate": 9.629538709138166e-06, | |
| "loss": 0.4896, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.7991071428571429, | |
| "grad_norm": 1.725033164024353, | |
| "learning_rate": 9.625064618437549e-06, | |
| "loss": 0.4508, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.8035714285714286, | |
| "grad_norm": 1.687565565109253, | |
| "learning_rate": 9.620564724920443e-06, | |
| "loss": 0.4277, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.8080357142857143, | |
| "grad_norm": 1.5750185251235962, | |
| "learning_rate": 9.616039053694058e-06, | |
| "loss": 0.4168, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.8125, | |
| "grad_norm": 1.7201429605484009, | |
| "learning_rate": 9.611487630009436e-06, | |
| "loss": 0.4097, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.8169642857142857, | |
| "grad_norm": 1.8075188398361206, | |
| "learning_rate": 9.606910479261301e-06, | |
| "loss": 0.4857, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.8214285714285714, | |
| "grad_norm": 1.4948920011520386, | |
| "learning_rate": 9.602307626987925e-06, | |
| "loss": 0.4004, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.8258928571428571, | |
| "grad_norm": 1.8233128786087036, | |
| "learning_rate": 9.597679098870978e-06, | |
| "loss": 0.4808, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.8303571428571429, | |
| "grad_norm": 1.698188304901123, | |
| "learning_rate": 9.593024920735393e-06, | |
| "loss": 0.4741, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.8348214285714286, | |
| "grad_norm": 1.7174502611160278, | |
| "learning_rate": 9.588345118549214e-06, | |
| "loss": 0.4313, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.8392857142857143, | |
| "grad_norm": 1.6591637134552002, | |
| "learning_rate": 9.583639718423457e-06, | |
| "loss": 0.453, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.84375, | |
| "grad_norm": 1.4759143590927124, | |
| "learning_rate": 9.57890874661196e-06, | |
| "loss": 0.3768, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.8482142857142857, | |
| "grad_norm": 1.6473032236099243, | |
| "learning_rate": 9.57415222951124e-06, | |
| "loss": 0.4448, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.8526785714285714, | |
| "grad_norm": 1.6935791969299316, | |
| "learning_rate": 9.569370193660348e-06, | |
| "loss": 0.4112, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 1.8300622701644897, | |
| "learning_rate": 9.564562665740708e-06, | |
| "loss": 0.5206, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.8616071428571429, | |
| "grad_norm": 1.512624979019165, | |
| "learning_rate": 9.559729672575985e-06, | |
| "loss": 0.4014, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.8660714285714286, | |
| "grad_norm": 1.5721509456634521, | |
| "learning_rate": 9.554871241131923e-06, | |
| "loss": 0.4845, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.8705357142857143, | |
| "grad_norm": 1.6285744905471802, | |
| "learning_rate": 9.549987398516206e-06, | |
| "loss": 0.486, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.875, | |
| "grad_norm": 1.7633181810379028, | |
| "learning_rate": 9.54507817197829e-06, | |
| "loss": 0.429, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.8794642857142857, | |
| "grad_norm": 1.6904608011245728, | |
| "learning_rate": 9.540143588909268e-06, | |
| "loss": 0.4263, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.8839285714285714, | |
| "grad_norm": 1.7905316352844238, | |
| "learning_rate": 9.535183676841709e-06, | |
| "loss": 0.457, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.8883928571428571, | |
| "grad_norm": 1.888141393661499, | |
| "learning_rate": 9.530198463449507e-06, | |
| "loss": 0.4649, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.8928571428571429, | |
| "grad_norm": 1.853598713874817, | |
| "learning_rate": 9.525187976547718e-06, | |
| "loss": 0.46, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.8973214285714286, | |
| "grad_norm": 1.5846953392028809, | |
| "learning_rate": 9.520152244092421e-06, | |
| "loss": 0.362, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.9017857142857143, | |
| "grad_norm": 1.608396053314209, | |
| "learning_rate": 9.515091294180546e-06, | |
| "loss": 0.4001, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.90625, | |
| "grad_norm": 1.8720006942749023, | |
| "learning_rate": 9.510005155049729e-06, | |
| "loss": 0.48, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.9107142857142857, | |
| "grad_norm": 1.741791009902954, | |
| "learning_rate": 9.504893855078144e-06, | |
| "loss": 0.381, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.9151785714285714, | |
| "grad_norm": 1.5903561115264893, | |
| "learning_rate": 9.499757422784358e-06, | |
| "loss": 0.4422, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.9196428571428571, | |
| "grad_norm": 1.6735451221466064, | |
| "learning_rate": 9.494595886827157e-06, | |
| "loss": 0.4973, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.9241071428571429, | |
| "grad_norm": 1.6444984674453735, | |
| "learning_rate": 9.489409276005393e-06, | |
| "loss": 0.4464, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.9285714285714286, | |
| "grad_norm": 1.6434756517410278, | |
| "learning_rate": 9.48419761925783e-06, | |
| "loss": 0.5094, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.9330357142857143, | |
| "grad_norm": 1.8670734167099, | |
| "learning_rate": 9.478960945662974e-06, | |
| "loss": 0.3986, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.9375, | |
| "grad_norm": 1.7340041399002075, | |
| "learning_rate": 9.473699284438908e-06, | |
| "loss": 0.4024, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.9419642857142857, | |
| "grad_norm": 1.8192938566207886, | |
| "learning_rate": 9.468412664943137e-06, | |
| "loss": 0.515, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.9464285714285714, | |
| "grad_norm": 1.5921608209609985, | |
| "learning_rate": 9.463101116672423e-06, | |
| "loss": 0.4257, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.9508928571428571, | |
| "grad_norm": 2.0308141708374023, | |
| "learning_rate": 9.457764669262615e-06, | |
| "loss": 0.4918, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.9553571428571429, | |
| "grad_norm": 1.6878437995910645, | |
| "learning_rate": 9.452403352488488e-06, | |
| "loss": 0.4433, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.9598214285714286, | |
| "grad_norm": 1.9883629083633423, | |
| "learning_rate": 9.447017196263578e-06, | |
| "loss": 0.4836, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.9642857142857143, | |
| "grad_norm": 1.67111337184906, | |
| "learning_rate": 9.441606230640012e-06, | |
| "loss": 0.448, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.96875, | |
| "grad_norm": 1.7054740190505981, | |
| "learning_rate": 9.436170485808338e-06, | |
| "loss": 0.449, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.9732142857142857, | |
| "grad_norm": 1.690224289894104, | |
| "learning_rate": 9.430709992097364e-06, | |
| "loss": 0.4924, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.9776785714285714, | |
| "grad_norm": 1.5849170684814453, | |
| "learning_rate": 9.425224779973986e-06, | |
| "loss": 0.4583, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.9821428571428571, | |
| "grad_norm": 1.726426124572754, | |
| "learning_rate": 9.41971488004301e-06, | |
| "loss": 0.4982, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.9866071428571429, | |
| "grad_norm": 1.7829738855361938, | |
| "learning_rate": 9.414180323046991e-06, | |
| "loss": 0.4642, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.9910714285714286, | |
| "grad_norm": 1.6020630598068237, | |
| "learning_rate": 9.408621139866067e-06, | |
| "loss": 0.3982, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.9955357142857143, | |
| "grad_norm": 1.7414110898971558, | |
| "learning_rate": 9.403037361517762e-06, | |
| "loss": 0.4485, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.4676933288574219, | |
| "learning_rate": 9.397429019156841e-06, | |
| "loss": 0.4268, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.0044642857142858, | |
| "grad_norm": 1.4603513479232788, | |
| "learning_rate": 9.391796144075123e-06, | |
| "loss": 0.341, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.0089285714285714, | |
| "grad_norm": 1.742700457572937, | |
| "learning_rate": 9.386138767701306e-06, | |
| "loss": 0.3227, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.0133928571428572, | |
| "grad_norm": 1.5671008825302124, | |
| "learning_rate": 9.380456921600785e-06, | |
| "loss": 0.3646, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.0178571428571428, | |
| "grad_norm": 1.6107045412063599, | |
| "learning_rate": 9.374750637475499e-06, | |
| "loss": 0.3366, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.0178571428571428, | |
| "eval_loss": 0.5993052124977112, | |
| "eval_runtime": 3.4011, | |
| "eval_samples_per_second": 17.641, | |
| "eval_steps_per_second": 1.176, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.0223214285714286, | |
| "grad_norm": 1.5372443199157715, | |
| "learning_rate": 9.36901994716373e-06, | |
| "loss": 0.3582, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.0267857142857142, | |
| "grad_norm": 1.5077764987945557, | |
| "learning_rate": 9.363264882639936e-06, | |
| "loss": 0.3136, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.03125, | |
| "grad_norm": 1.6883026361465454, | |
| "learning_rate": 9.357485476014573e-06, | |
| "loss": 0.3196, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.0357142857142858, | |
| "grad_norm": 1.380987524986267, | |
| "learning_rate": 9.351681759533914e-06, | |
| "loss": 0.2765, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.0401785714285714, | |
| "grad_norm": 1.6997498273849487, | |
| "learning_rate": 9.345853765579865e-06, | |
| "loss": 0.3005, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.0446428571428572, | |
| "grad_norm": 1.550504446029663, | |
| "learning_rate": 9.340001526669794e-06, | |
| "loss": 0.2848, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.0491071428571428, | |
| "grad_norm": 1.7140612602233887, | |
| "learning_rate": 9.33412507545634e-06, | |
| "loss": 0.3395, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.0535714285714286, | |
| "grad_norm": 1.783673644065857, | |
| "learning_rate": 9.32822444472724e-06, | |
| "loss": 0.3454, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.0580357142857142, | |
| "grad_norm": 1.7617640495300293, | |
| "learning_rate": 9.322299667405134e-06, | |
| "loss": 0.3444, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.0625, | |
| "grad_norm": 1.7787531614303589, | |
| "learning_rate": 9.31635077654739e-06, | |
| "loss": 0.3595, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.0669642857142858, | |
| "grad_norm": 1.4490877389907837, | |
| "learning_rate": 9.310377805345926e-06, | |
| "loss": 0.2953, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.0714285714285714, | |
| "grad_norm": 1.6861026287078857, | |
| "learning_rate": 9.304380787127003e-06, | |
| "loss": 0.3182, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.0758928571428572, | |
| "grad_norm": 1.58506441116333, | |
| "learning_rate": 9.298359755351065e-06, | |
| "loss": 0.3151, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.0803571428571428, | |
| "grad_norm": 1.7218842506408691, | |
| "learning_rate": 9.29231474361253e-06, | |
| "loss": 0.3803, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.0848214285714286, | |
| "grad_norm": 1.6209449768066406, | |
| "learning_rate": 9.28624578563962e-06, | |
| "loss": 0.3161, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.0892857142857142, | |
| "grad_norm": 1.5501459836959839, | |
| "learning_rate": 9.280152915294162e-06, | |
| "loss": 0.3917, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.09375, | |
| "grad_norm": 1.5111147165298462, | |
| "learning_rate": 9.274036166571402e-06, | |
| "loss": 0.3649, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.0982142857142858, | |
| "grad_norm": 1.4189096689224243, | |
| "learning_rate": 9.267895573599819e-06, | |
| "loss": 0.3162, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.1026785714285714, | |
| "grad_norm": 1.57675039768219, | |
| "learning_rate": 9.261731170640923e-06, | |
| "loss": 0.3331, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.1071428571428572, | |
| "grad_norm": 1.4706840515136719, | |
| "learning_rate": 9.255542992089086e-06, | |
| "loss": 0.3054, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.1116071428571428, | |
| "grad_norm": 1.4259796142578125, | |
| "learning_rate": 9.24933107247132e-06, | |
| "loss": 0.3209, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.1160714285714286, | |
| "grad_norm": 1.526444435119629, | |
| "learning_rate": 9.243095446447113e-06, | |
| "loss": 0.3066, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.1205357142857142, | |
| "grad_norm": 1.4234224557876587, | |
| "learning_rate": 9.23683614880822e-06, | |
| "loss": 0.3107, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.125, | |
| "grad_norm": 1.6062707901000977, | |
| "learning_rate": 9.230553214478469e-06, | |
| "loss": 0.3621, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.1294642857142858, | |
| "grad_norm": 1.5600666999816895, | |
| "learning_rate": 9.224246678513569e-06, | |
| "loss": 0.3225, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.1339285714285714, | |
| "grad_norm": 1.6989763975143433, | |
| "learning_rate": 9.217916576100922e-06, | |
| "loss": 0.3107, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.1383928571428572, | |
| "grad_norm": 1.603786587715149, | |
| "learning_rate": 9.211562942559408e-06, | |
| "loss": 0.3434, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 1.5290991067886353, | |
| "learning_rate": 9.20518581333921e-06, | |
| "loss": 0.3242, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.1473214285714286, | |
| "grad_norm": 1.4983583688735962, | |
| "learning_rate": 9.1987852240216e-06, | |
| "loss": 0.3315, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.1517857142857142, | |
| "grad_norm": 1.5081202983856201, | |
| "learning_rate": 9.192361210318745e-06, | |
| "loss": 0.3156, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.15625, | |
| "grad_norm": 1.575416088104248, | |
| "learning_rate": 9.185913808073513e-06, | |
| "loss": 0.3083, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.1607142857142858, | |
| "grad_norm": 1.6067839860916138, | |
| "learning_rate": 9.179443053259263e-06, | |
| "loss": 0.3597, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.1651785714285714, | |
| "grad_norm": 1.5789872407913208, | |
| "learning_rate": 9.172948981979654e-06, | |
| "loss": 0.3639, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.1696428571428572, | |
| "grad_norm": 1.6132181882858276, | |
| "learning_rate": 9.166431630468438e-06, | |
| "loss": 0.3473, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.1741071428571428, | |
| "grad_norm": 1.4658143520355225, | |
| "learning_rate": 9.159891035089262e-06, | |
| "loss": 0.275, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.1785714285714286, | |
| "grad_norm": 1.6115272045135498, | |
| "learning_rate": 9.153327232335455e-06, | |
| "loss": 0.337, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.1830357142857142, | |
| "grad_norm": 1.3804121017456055, | |
| "learning_rate": 9.146740258829844e-06, | |
| "loss": 0.2805, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.1875, | |
| "grad_norm": 1.6690940856933594, | |
| "learning_rate": 9.140130151324526e-06, | |
| "loss": 0.3812, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.1919642857142858, | |
| "grad_norm": 1.4038200378417969, | |
| "learning_rate": 9.13349694670068e-06, | |
| "loss": 0.3075, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.1964285714285714, | |
| "grad_norm": 1.4030662775039673, | |
| "learning_rate": 9.126840681968357e-06, | |
| "loss": 0.3192, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.2008928571428572, | |
| "grad_norm": 1.6158102750778198, | |
| "learning_rate": 9.120161394266266e-06, | |
| "loss": 0.3123, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.2053571428571428, | |
| "grad_norm": 1.463722586631775, | |
| "learning_rate": 9.113459120861579e-06, | |
| "loss": 0.2831, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.2098214285714286, | |
| "grad_norm": 1.7193071842193604, | |
| "learning_rate": 9.106733899149715e-06, | |
| "loss": 0.3422, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.2142857142857142, | |
| "grad_norm": 1.5576410293579102, | |
| "learning_rate": 9.099985766654132e-06, | |
| "loss": 0.3379, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.21875, | |
| "grad_norm": 1.4065855741500854, | |
| "learning_rate": 9.093214761026121e-06, | |
| "loss": 0.2972, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.2232142857142858, | |
| "grad_norm": 1.2878752946853638, | |
| "learning_rate": 9.08642092004459e-06, | |
| "loss": 0.301, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.2276785714285714, | |
| "grad_norm": 1.5659253597259521, | |
| "learning_rate": 9.079604281615868e-06, | |
| "loss": 0.3259, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.2321428571428572, | |
| "grad_norm": 1.4330716133117676, | |
| "learning_rate": 9.072764883773464e-06, | |
| "loss": 0.3241, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.2366071428571428, | |
| "grad_norm": 1.465423583984375, | |
| "learning_rate": 9.065902764677897e-06, | |
| "loss": 0.2896, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.2410714285714286, | |
| "grad_norm": 1.60066819190979, | |
| "learning_rate": 9.059017962616435e-06, | |
| "loss": 0.3486, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.2455357142857142, | |
| "grad_norm": 1.688224196434021, | |
| "learning_rate": 9.052110516002925e-06, | |
| "loss": 0.3625, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.7048122882843018, | |
| "learning_rate": 9.04518046337755e-06, | |
| "loss": 0.3627, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.2544642857142856, | |
| "grad_norm": 1.699864149093628, | |
| "learning_rate": 9.038227843406628e-06, | |
| "loss": 0.3429, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.2589285714285714, | |
| "grad_norm": 1.503157377243042, | |
| "learning_rate": 9.031252694882386e-06, | |
| "loss": 0.3415, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.2633928571428572, | |
| "grad_norm": 1.475395917892456, | |
| "learning_rate": 9.024255056722753e-06, | |
| "loss": 0.3211, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.2678571428571428, | |
| "grad_norm": 1.4830482006072998, | |
| "learning_rate": 9.017234967971143e-06, | |
| "loss": 0.3258, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.2723214285714286, | |
| "grad_norm": 1.499875783920288, | |
| "learning_rate": 9.010192467796228e-06, | |
| "loss": 0.3257, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.2723214285714286, | |
| "eval_loss": 0.6085591912269592, | |
| "eval_runtime": 3.4657, | |
| "eval_samples_per_second": 17.313, | |
| "eval_steps_per_second": 1.154, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.2767857142857144, | |
| "grad_norm": 1.5229018926620483, | |
| "learning_rate": 9.003127595491723e-06, | |
| "loss": 0.3377, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.28125, | |
| "grad_norm": 1.5909078121185303, | |
| "learning_rate": 8.996040390476177e-06, | |
| "loss": 0.3675, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "grad_norm": 1.553013801574707, | |
| "learning_rate": 8.988930892292738e-06, | |
| "loss": 0.3638, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.2901785714285714, | |
| "grad_norm": 1.5105496644973755, | |
| "learning_rate": 8.981799140608938e-06, | |
| "loss": 0.3437, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.2946428571428572, | |
| "grad_norm": 1.563785195350647, | |
| "learning_rate": 8.974645175216478e-06, | |
| "loss": 0.3299, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.2991071428571428, | |
| "grad_norm": 1.5746451616287231, | |
| "learning_rate": 8.967469036030996e-06, | |
| "loss": 0.3344, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.3035714285714286, | |
| "grad_norm": 1.735245943069458, | |
| "learning_rate": 8.960270763091853e-06, | |
| "loss": 0.3105, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.3080357142857144, | |
| "grad_norm": 1.3750625848770142, | |
| "learning_rate": 8.953050396561904e-06, | |
| "loss": 0.3074, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.3125, | |
| "grad_norm": 1.5357894897460938, | |
| "learning_rate": 8.94580797672727e-06, | |
| "loss": 0.3292, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.3169642857142856, | |
| "grad_norm": 1.5508729219436646, | |
| "learning_rate": 8.938543543997129e-06, | |
| "loss": 0.2959, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.3214285714285714, | |
| "grad_norm": 1.7151294946670532, | |
| "learning_rate": 8.931257138903474e-06, | |
| "loss": 0.3023, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.3258928571428572, | |
| "grad_norm": 1.6977832317352295, | |
| "learning_rate": 8.923948802100891e-06, | |
| "loss": 0.352, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.3303571428571428, | |
| "grad_norm": 1.724831461906433, | |
| "learning_rate": 8.916618574366338e-06, | |
| "loss": 0.3564, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.3348214285714286, | |
| "grad_norm": 1.5105293989181519, | |
| "learning_rate": 8.909266496598917e-06, | |
| "loss": 0.315, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.3392857142857144, | |
| "grad_norm": 1.3590584993362427, | |
| "learning_rate": 8.901892609819632e-06, | |
| "loss": 0.2753, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.34375, | |
| "grad_norm": 1.5936496257781982, | |
| "learning_rate": 8.894496955171182e-06, | |
| "loss": 0.357, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.3482142857142856, | |
| "grad_norm": 1.6498358249664307, | |
| "learning_rate": 8.887079573917713e-06, | |
| "loss": 0.3318, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.3526785714285714, | |
| "grad_norm": 1.6604431867599487, | |
| "learning_rate": 8.879640507444598e-06, | |
| "loss": 0.3314, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.3571428571428572, | |
| "grad_norm": 1.5546607971191406, | |
| "learning_rate": 8.872179797258202e-06, | |
| "loss": 0.3534, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.3616071428571428, | |
| "grad_norm": 1.6600899696350098, | |
| "learning_rate": 8.86469748498565e-06, | |
| "loss": 0.3231, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.3660714285714286, | |
| "grad_norm": 1.4908391237258911, | |
| "learning_rate": 8.8571936123746e-06, | |
| "loss": 0.308, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.3705357142857144, | |
| "grad_norm": 1.4231539964675903, | |
| "learning_rate": 8.849668221293e-06, | |
| "loss": 0.3309, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.375, | |
| "grad_norm": 1.4389381408691406, | |
| "learning_rate": 8.842121353728867e-06, | |
| "loss": 0.3231, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.3794642857142856, | |
| "grad_norm": 1.5428330898284912, | |
| "learning_rate": 8.834553051790044e-06, | |
| "loss": 0.3523, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.3839285714285714, | |
| "grad_norm": 1.3884952068328857, | |
| "learning_rate": 8.826963357703964e-06, | |
| "loss": 0.2901, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.3883928571428572, | |
| "grad_norm": 1.549255609512329, | |
| "learning_rate": 8.819352313817424e-06, | |
| "loss": 0.3387, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.3928571428571428, | |
| "grad_norm": 1.483252763748169, | |
| "learning_rate": 8.811719962596338e-06, | |
| "loss": 0.3238, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.3973214285714286, | |
| "grad_norm": 1.4185503721237183, | |
| "learning_rate": 8.804066346625506e-06, | |
| "loss": 0.314, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.4017857142857144, | |
| "grad_norm": 1.6147483587265015, | |
| "learning_rate": 8.796391508608372e-06, | |
| "loss": 0.3628, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.40625, | |
| "grad_norm": 1.513182282447815, | |
| "learning_rate": 8.788695491366795e-06, | |
| "loss": 0.3209, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.4107142857142856, | |
| "grad_norm": 1.4564861059188843, | |
| "learning_rate": 8.780978337840796e-06, | |
| "loss": 0.3037, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.4151785714285714, | |
| "grad_norm": 1.4925004243850708, | |
| "learning_rate": 8.773240091088335e-06, | |
| "loss": 0.3171, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.4196428571428572, | |
| "grad_norm": 1.5218322277069092, | |
| "learning_rate": 8.765480794285054e-06, | |
| "loss": 0.3329, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.4241071428571428, | |
| "grad_norm": 1.4326412677764893, | |
| "learning_rate": 8.757700490724046e-06, | |
| "loss": 0.3119, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 1.6294825077056885, | |
| "learning_rate": 8.749899223815618e-06, | |
| "loss": 0.3568, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.4330357142857144, | |
| "grad_norm": 1.4534196853637695, | |
| "learning_rate": 8.742077037087032e-06, | |
| "loss": 0.3247, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.4375, | |
| "grad_norm": 1.4942129850387573, | |
| "learning_rate": 8.734233974182276e-06, | |
| "loss": 0.3189, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.4419642857142856, | |
| "grad_norm": 1.6860888004302979, | |
| "learning_rate": 8.726370078861825e-06, | |
| "loss": 0.3466, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.4464285714285714, | |
| "grad_norm": 1.5800039768218994, | |
| "learning_rate": 8.718485395002377e-06, | |
| "loss": 0.3636, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.4508928571428572, | |
| "grad_norm": 1.4802786111831665, | |
| "learning_rate": 8.710579966596625e-06, | |
| "loss": 0.3442, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.4553571428571428, | |
| "grad_norm": 1.4833781719207764, | |
| "learning_rate": 8.702653837753005e-06, | |
| "loss": 0.3162, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.4598214285714286, | |
| "grad_norm": 1.6002053022384644, | |
| "learning_rate": 8.694707052695459e-06, | |
| "loss": 0.3312, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.4642857142857144, | |
| "grad_norm": 1.5287590026855469, | |
| "learning_rate": 8.686739655763166e-06, | |
| "loss": 0.3525, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.46875, | |
| "grad_norm": 1.4307093620300293, | |
| "learning_rate": 8.678751691410323e-06, | |
| "loss": 0.3121, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.4732142857142856, | |
| "grad_norm": 1.4360722303390503, | |
| "learning_rate": 8.670743204205875e-06, | |
| "loss": 0.3461, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.4776785714285714, | |
| "grad_norm": 1.3801894187927246, | |
| "learning_rate": 8.662714238833278e-06, | |
| "loss": 0.2942, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.4821428571428572, | |
| "grad_norm": 1.4250258207321167, | |
| "learning_rate": 8.654664840090247e-06, | |
| "loss": 0.3665, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.4866071428571428, | |
| "grad_norm": 1.5609809160232544, | |
| "learning_rate": 8.6465950528885e-06, | |
| "loss": 0.3197, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.4910714285714286, | |
| "grad_norm": 1.4296934604644775, | |
| "learning_rate": 8.638504922253518e-06, | |
| "loss": 0.3132, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.4955357142857144, | |
| "grad_norm": 1.4732247591018677, | |
| "learning_rate": 8.63039449332429e-06, | |
| "loss": 0.3335, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.5229887962341309, | |
| "learning_rate": 8.62226381135305e-06, | |
| "loss": 0.3197, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.5044642857142856, | |
| "grad_norm": 1.5183864831924438, | |
| "learning_rate": 8.614112921705045e-06, | |
| "loss": 0.3109, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.5089285714285714, | |
| "grad_norm": 1.5982593297958374, | |
| "learning_rate": 8.605941869858265e-06, | |
| "loss": 0.3363, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.5133928571428572, | |
| "grad_norm": 1.5128353834152222, | |
| "learning_rate": 8.597750701403197e-06, | |
| "loss": 0.3177, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.5178571428571428, | |
| "grad_norm": 1.383318305015564, | |
| "learning_rate": 8.589539462042566e-06, | |
| "loss": 0.3099, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.5223214285714286, | |
| "grad_norm": 1.4567229747772217, | |
| "learning_rate": 8.581308197591088e-06, | |
| "loss": 0.3105, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.5267857142857144, | |
| "grad_norm": 1.5694351196289062, | |
| "learning_rate": 8.573056953975208e-06, | |
| "loss": 0.3521, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.5267857142857144, | |
| "eval_loss": 0.6054974794387817, | |
| "eval_runtime": 3.4752, | |
| "eval_samples_per_second": 17.265, | |
| "eval_steps_per_second": 1.151, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.53125, | |
| "grad_norm": 1.499888300895691, | |
| "learning_rate": 8.56478577723284e-06, | |
| "loss": 0.3472, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.5357142857142856, | |
| "grad_norm": 1.5075103044509888, | |
| "learning_rate": 8.556494713513123e-06, | |
| "loss": 0.3353, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.5401785714285714, | |
| "grad_norm": 1.4308085441589355, | |
| "learning_rate": 8.548183809076146e-06, | |
| "loss": 0.31, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.5446428571428572, | |
| "grad_norm": 1.4208667278289795, | |
| "learning_rate": 8.539853110292708e-06, | |
| "loss": 0.3331, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.5491071428571428, | |
| "grad_norm": 1.6008033752441406, | |
| "learning_rate": 8.531502663644046e-06, | |
| "loss": 0.3568, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.5535714285714286, | |
| "grad_norm": 1.471472978591919, | |
| "learning_rate": 8.523132515721586e-06, | |
| "loss": 0.3379, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.5580357142857144, | |
| "grad_norm": 1.4036483764648438, | |
| "learning_rate": 8.51474271322667e-06, | |
| "loss": 0.3488, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.5625, | |
| "grad_norm": 1.4578651189804077, | |
| "learning_rate": 8.506333302970306e-06, | |
| "loss": 0.3206, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.5669642857142856, | |
| "grad_norm": 1.3492405414581299, | |
| "learning_rate": 8.497904331872909e-06, | |
| "loss": 0.2995, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 1.5430017709732056, | |
| "learning_rate": 8.489455846964027e-06, | |
| "loss": 0.3503, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.5758928571428572, | |
| "grad_norm": 1.4253814220428467, | |
| "learning_rate": 8.480987895382086e-06, | |
| "loss": 0.3621, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.5803571428571428, | |
| "grad_norm": 1.615607500076294, | |
| "learning_rate": 8.472500524374129e-06, | |
| "loss": 0.3578, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.5848214285714286, | |
| "grad_norm": 1.4999765157699585, | |
| "learning_rate": 8.463993781295552e-06, | |
| "loss": 0.3241, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.5892857142857144, | |
| "grad_norm": 1.4261177778244019, | |
| "learning_rate": 8.45546771360983e-06, | |
| "loss": 0.3021, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.59375, | |
| "grad_norm": 1.4563229084014893, | |
| "learning_rate": 8.44692236888827e-06, | |
| "loss": 0.3194, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.5982142857142856, | |
| "grad_norm": 1.3366395235061646, | |
| "learning_rate": 8.43835779480973e-06, | |
| "loss": 0.3295, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.6026785714285714, | |
| "grad_norm": 1.6574403047561646, | |
| "learning_rate": 8.429774039160355e-06, | |
| "loss": 0.3375, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.6071428571428572, | |
| "grad_norm": 1.7103718519210815, | |
| "learning_rate": 8.421171149833322e-06, | |
| "loss": 0.3367, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.6116071428571428, | |
| "grad_norm": 1.6518172025680542, | |
| "learning_rate": 8.412549174828558e-06, | |
| "loss": 0.3802, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.6160714285714286, | |
| "grad_norm": 1.59016752243042, | |
| "learning_rate": 8.403908162252481e-06, | |
| "loss": 0.3342, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.6205357142857144, | |
| "grad_norm": 1.6090682744979858, | |
| "learning_rate": 8.395248160317728e-06, | |
| "loss": 0.3196, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.625, | |
| "grad_norm": 1.4735345840454102, | |
| "learning_rate": 8.386569217342893e-06, | |
| "loss": 0.3211, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.6294642857142856, | |
| "grad_norm": 1.554945707321167, | |
| "learning_rate": 8.377871381752246e-06, | |
| "loss": 0.3309, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.6339285714285714, | |
| "grad_norm": 1.4980627298355103, | |
| "learning_rate": 8.369154702075466e-06, | |
| "loss": 0.3394, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.6383928571428572, | |
| "grad_norm": 1.5292226076126099, | |
| "learning_rate": 8.360419226947383e-06, | |
| "loss": 0.319, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.6428571428571428, | |
| "grad_norm": 1.5913363695144653, | |
| "learning_rate": 8.351665005107686e-06, | |
| "loss": 0.3279, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.6473214285714286, | |
| "grad_norm": 1.526674509048462, | |
| "learning_rate": 8.34289208540067e-06, | |
| "loss": 0.3564, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.6517857142857144, | |
| "grad_norm": 1.5326498746871948, | |
| "learning_rate": 8.334100516774946e-06, | |
| "loss": 0.3462, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.65625, | |
| "grad_norm": 1.512019395828247, | |
| "learning_rate": 8.325290348283186e-06, | |
| "loss": 0.3624, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.6607142857142856, | |
| "grad_norm": 1.5395188331604004, | |
| "learning_rate": 8.316461629081833e-06, | |
| "loss": 0.3028, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.6651785714285714, | |
| "grad_norm": 1.4373003244400024, | |
| "learning_rate": 8.307614408430839e-06, | |
| "loss": 0.3623, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.6696428571428572, | |
| "grad_norm": 1.5330880880355835, | |
| "learning_rate": 8.298748735693382e-06, | |
| "loss": 0.3412, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.6741071428571428, | |
| "grad_norm": 1.408527135848999, | |
| "learning_rate": 8.289864660335595e-06, | |
| "loss": 0.3248, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.6785714285714286, | |
| "grad_norm": 1.5303860902786255, | |
| "learning_rate": 8.280962231926288e-06, | |
| "loss": 0.3394, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.6830357142857144, | |
| "grad_norm": 1.5478315353393555, | |
| "learning_rate": 8.27204150013667e-06, | |
| "loss": 0.334, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.6875, | |
| "grad_norm": 1.4640541076660156, | |
| "learning_rate": 8.263102514740082e-06, | |
| "loss": 0.3116, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.6919642857142856, | |
| "grad_norm": 1.5886211395263672, | |
| "learning_rate": 8.2541453256117e-06, | |
| "loss": 0.3146, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.6964285714285714, | |
| "grad_norm": 1.450217366218567, | |
| "learning_rate": 8.245169982728276e-06, | |
| "loss": 0.3425, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.7008928571428572, | |
| "grad_norm": 1.4083800315856934, | |
| "learning_rate": 8.23617653616785e-06, | |
| "loss": 0.3078, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.7053571428571428, | |
| "grad_norm": 1.4849356412887573, | |
| "learning_rate": 8.227165036109468e-06, | |
| "loss": 0.328, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.7098214285714286, | |
| "grad_norm": 1.6867343187332153, | |
| "learning_rate": 8.218135532832909e-06, | |
| "loss": 0.3688, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 1.5400327444076538, | |
| "learning_rate": 8.209088076718398e-06, | |
| "loss": 0.3378, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.71875, | |
| "grad_norm": 1.5973129272460938, | |
| "learning_rate": 8.20002271824633e-06, | |
| "loss": 0.3734, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.7232142857142856, | |
| "grad_norm": 1.532235860824585, | |
| "learning_rate": 8.190939507996992e-06, | |
| "loss": 0.3582, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.7276785714285714, | |
| "grad_norm": 1.5533599853515625, | |
| "learning_rate": 8.181838496650266e-06, | |
| "loss": 0.3479, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.7321428571428572, | |
| "grad_norm": 1.6157996654510498, | |
| "learning_rate": 8.17271973498536e-06, | |
| "loss": 0.3596, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.7366071428571428, | |
| "grad_norm": 1.5097051858901978, | |
| "learning_rate": 8.163583273880519e-06, | |
| "loss": 0.3298, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.7410714285714286, | |
| "grad_norm": 1.4383480548858643, | |
| "learning_rate": 8.154429164312742e-06, | |
| "loss": 0.3054, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.7455357142857144, | |
| "grad_norm": 1.43813157081604, | |
| "learning_rate": 8.145257457357502e-06, | |
| "loss": 0.3249, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 1.5752897262573242, | |
| "learning_rate": 8.136068204188448e-06, | |
| "loss": 0.3342, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.7544642857142856, | |
| "grad_norm": 1.3768292665481567, | |
| "learning_rate": 8.12686145607714e-06, | |
| "loss": 0.3069, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.7589285714285714, | |
| "grad_norm": 1.6397638320922852, | |
| "learning_rate": 8.11763726439274e-06, | |
| "loss": 0.3843, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.7633928571428572, | |
| "grad_norm": 1.5874100923538208, | |
| "learning_rate": 8.108395680601742e-06, | |
| "loss": 0.3226, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.7678571428571428, | |
| "grad_norm": 1.5031986236572266, | |
| "learning_rate": 8.099136756267682e-06, | |
| "loss": 0.3176, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.7723214285714286, | |
| "grad_norm": 1.420452356338501, | |
| "learning_rate": 8.089860543050843e-06, | |
| "loss": 0.3259, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.7767857142857144, | |
| "grad_norm": 1.6253381967544556, | |
| "learning_rate": 8.080567092707973e-06, | |
| "loss": 0.3457, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.78125, | |
| "grad_norm": 1.4841629266738892, | |
| "learning_rate": 8.071256457091994e-06, | |
| "loss": 0.3201, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.78125, | |
| "eval_loss": 0.6055915951728821, | |
| "eval_runtime": 3.4629, | |
| "eval_samples_per_second": 17.327, | |
| "eval_steps_per_second": 1.155, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.7857142857142856, | |
| "grad_norm": 1.4217731952667236, | |
| "learning_rate": 8.06192868815172e-06, | |
| "loss": 0.3229, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.7901785714285714, | |
| "grad_norm": 1.5509140491485596, | |
| "learning_rate": 8.05258383793155e-06, | |
| "loss": 0.3586, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.7946428571428572, | |
| "grad_norm": 1.4150621891021729, | |
| "learning_rate": 8.043221958571193e-06, | |
| "loss": 0.3282, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.7991071428571428, | |
| "grad_norm": 1.4854953289031982, | |
| "learning_rate": 8.033843102305376e-06, | |
| "loss": 0.3426, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.8035714285714286, | |
| "grad_norm": 1.3903532028198242, | |
| "learning_rate": 8.024447321463545e-06, | |
| "loss": 0.3215, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.8080357142857144, | |
| "grad_norm": 1.4859530925750732, | |
| "learning_rate": 8.015034668469576e-06, | |
| "loss": 0.3505, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.8125, | |
| "grad_norm": 1.5304354429244995, | |
| "learning_rate": 8.005605195841485e-06, | |
| "loss": 0.323, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.8169642857142856, | |
| "grad_norm": 1.44026517868042, | |
| "learning_rate": 7.996158956191135e-06, | |
| "loss": 0.2791, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.8214285714285714, | |
| "grad_norm": 1.3094749450683594, | |
| "learning_rate": 7.986696002223936e-06, | |
| "loss": 0.2975, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.8258928571428572, | |
| "grad_norm": 1.4733068943023682, | |
| "learning_rate": 7.97721638673856e-06, | |
| "loss": 0.3331, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.8303571428571428, | |
| "grad_norm": 1.403172254562378, | |
| "learning_rate": 7.967720162626643e-06, | |
| "loss": 0.3336, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.8348214285714286, | |
| "grad_norm": 1.4719949960708618, | |
| "learning_rate": 7.958207382872486e-06, | |
| "loss": 0.3417, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.8392857142857144, | |
| "grad_norm": 1.4690333604812622, | |
| "learning_rate": 7.94867810055276e-06, | |
| "loss": 0.307, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.84375, | |
| "grad_norm": 1.5231949090957642, | |
| "learning_rate": 7.93913236883622e-06, | |
| "loss": 0.3421, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.8482142857142856, | |
| "grad_norm": 1.6096547842025757, | |
| "learning_rate": 7.929570240983393e-06, | |
| "loss": 0.3273, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.8526785714285714, | |
| "grad_norm": 1.3807638883590698, | |
| "learning_rate": 7.919991770346295e-06, | |
| "loss": 0.3431, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 1.413569688796997, | |
| "learning_rate": 7.910397010368122e-06, | |
| "loss": 0.33, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.8616071428571428, | |
| "grad_norm": 1.435351014137268, | |
| "learning_rate": 7.900786014582957e-06, | |
| "loss": 0.3581, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.8660714285714286, | |
| "grad_norm": 1.4194928407669067, | |
| "learning_rate": 7.891158836615472e-06, | |
| "loss": 0.2989, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.8705357142857144, | |
| "grad_norm": 1.4297808408737183, | |
| "learning_rate": 7.881515530180629e-06, | |
| "loss": 0.3542, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.875, | |
| "grad_norm": 1.4238284826278687, | |
| "learning_rate": 7.871856149083377e-06, | |
| "loss": 0.3257, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.8794642857142856, | |
| "grad_norm": 1.3757729530334473, | |
| "learning_rate": 7.862180747218354e-06, | |
| "loss": 0.3359, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.8839285714285714, | |
| "grad_norm": 1.5045543909072876, | |
| "learning_rate": 7.852489378569588e-06, | |
| "loss": 0.3401, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.8883928571428572, | |
| "grad_norm": 1.4873552322387695, | |
| "learning_rate": 7.84278209721019e-06, | |
| "loss": 0.3424, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.8928571428571428, | |
| "grad_norm": 1.2040387392044067, | |
| "learning_rate": 7.83305895730206e-06, | |
| "loss": 0.2713, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.8973214285714286, | |
| "grad_norm": 1.522424340248108, | |
| "learning_rate": 7.823320013095578e-06, | |
| "loss": 0.3308, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.9017857142857144, | |
| "grad_norm": 1.4239548444747925, | |
| "learning_rate": 7.81356531892931e-06, | |
| "loss": 0.3225, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.90625, | |
| "grad_norm": 1.491533875465393, | |
| "learning_rate": 7.803794929229689e-06, | |
| "loss": 0.3185, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.9107142857142856, | |
| "grad_norm": 1.5129714012145996, | |
| "learning_rate": 7.794008898510731e-06, | |
| "loss": 0.3497, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.9151785714285714, | |
| "grad_norm": 1.5560288429260254, | |
| "learning_rate": 7.784207281373716e-06, | |
| "loss": 0.36, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.9196428571428572, | |
| "grad_norm": 1.469099998474121, | |
| "learning_rate": 7.774390132506892e-06, | |
| "loss": 0.3236, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.9241071428571428, | |
| "grad_norm": 1.5299158096313477, | |
| "learning_rate": 7.764557506685162e-06, | |
| "loss": 0.3857, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.9285714285714286, | |
| "grad_norm": 1.5339336395263672, | |
| "learning_rate": 7.754709458769787e-06, | |
| "loss": 0.3451, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.9330357142857144, | |
| "grad_norm": 1.4419758319854736, | |
| "learning_rate": 7.744846043708076e-06, | |
| "loss": 0.3194, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.9375, | |
| "grad_norm": 1.580003023147583, | |
| "learning_rate": 7.734967316533074e-06, | |
| "loss": 0.3381, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.9419642857142856, | |
| "grad_norm": 1.5472431182861328, | |
| "learning_rate": 7.725073332363265e-06, | |
| "loss": 0.3528, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.9464285714285714, | |
| "grad_norm": 1.468518853187561, | |
| "learning_rate": 7.715164146402259e-06, | |
| "loss": 0.3329, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.9508928571428572, | |
| "grad_norm": 1.5422027111053467, | |
| "learning_rate": 7.705239813938486e-06, | |
| "loss": 0.3392, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.9553571428571428, | |
| "grad_norm": 1.3538919687271118, | |
| "learning_rate": 7.69530039034488e-06, | |
| "loss": 0.3179, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.9598214285714286, | |
| "grad_norm": 1.3986053466796875, | |
| "learning_rate": 7.685345931078579e-06, | |
| "loss": 0.3533, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.9642857142857144, | |
| "grad_norm": 1.484214186668396, | |
| "learning_rate": 7.675376491680617e-06, | |
| "loss": 0.3466, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.96875, | |
| "grad_norm": 1.5053229331970215, | |
| "learning_rate": 7.665392127775605e-06, | |
| "loss": 0.3464, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.9732142857142856, | |
| "grad_norm": 1.4051772356033325, | |
| "learning_rate": 7.65539289507143e-06, | |
| "loss": 0.3184, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.9776785714285714, | |
| "grad_norm": 1.3952518701553345, | |
| "learning_rate": 7.645378849358931e-06, | |
| "loss": 0.2796, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.9821428571428572, | |
| "grad_norm": 1.409744381904602, | |
| "learning_rate": 7.635350046511609e-06, | |
| "loss": 0.3369, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.9866071428571428, | |
| "grad_norm": 1.5328154563903809, | |
| "learning_rate": 7.625306542485289e-06, | |
| "loss": 0.3415, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.9910714285714286, | |
| "grad_norm": 1.4076818227767944, | |
| "learning_rate": 7.615248393317833e-06, | |
| "loss": 0.3431, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.9955357142857144, | |
| "grad_norm": 1.6852420568466187, | |
| "learning_rate": 7.605175655128809e-06, | |
| "loss": 0.3177, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.4953469038009644, | |
| "learning_rate": 7.595088384119186e-06, | |
| "loss": 0.3471, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 2.0044642857142856, | |
| "grad_norm": 1.3344584703445435, | |
| "learning_rate": 7.58498663657102e-06, | |
| "loss": 0.2074, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 2.0089285714285716, | |
| "grad_norm": 1.3186531066894531, | |
| "learning_rate": 7.57487046884714e-06, | |
| "loss": 0.2191, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.013392857142857, | |
| "grad_norm": 1.155795931816101, | |
| "learning_rate": 7.5647399373908296e-06, | |
| "loss": 0.1733, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 2.017857142857143, | |
| "grad_norm": 1.2673778533935547, | |
| "learning_rate": 7.554595098725515e-06, | |
| "loss": 0.2102, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 2.0223214285714284, | |
| "grad_norm": 1.1908435821533203, | |
| "learning_rate": 7.544436009454454e-06, | |
| "loss": 0.1796, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 2.0267857142857144, | |
| "grad_norm": 1.380186676979065, | |
| "learning_rate": 7.534262726260413e-06, | |
| "loss": 0.2077, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 2.03125, | |
| "grad_norm": 1.4421330690383911, | |
| "learning_rate": 7.524075305905351e-06, | |
| "loss": 0.1972, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 2.0357142857142856, | |
| "grad_norm": 1.6806164979934692, | |
| "learning_rate": 7.513873805230111e-06, | |
| "loss": 0.2034, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 2.0357142857142856, | |
| "eval_loss": 0.6833150386810303, | |
| "eval_runtime": 3.4113, | |
| "eval_samples_per_second": 17.588, | |
| "eval_steps_per_second": 1.173, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 2.0401785714285716, | |
| "grad_norm": 1.8982454538345337, | |
| "learning_rate": 7.5036582811540935e-06, | |
| "loss": 0.2251, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 2.044642857142857, | |
| "grad_norm": 1.90556001663208, | |
| "learning_rate": 7.493428790674943e-06, | |
| "loss": 0.1763, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 2.049107142857143, | |
| "grad_norm": 1.8609174489974976, | |
| "learning_rate": 7.483185390868232e-06, | |
| "loss": 0.1971, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 2.0535714285714284, | |
| "grad_norm": 1.6848630905151367, | |
| "learning_rate": 7.472928138887134e-06, | |
| "loss": 0.1822, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.0580357142857144, | |
| "grad_norm": 1.8364639282226562, | |
| "learning_rate": 7.462657091962122e-06, | |
| "loss": 0.2003, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 2.0625, | |
| "grad_norm": 1.6772575378417969, | |
| "learning_rate": 7.452372307400626e-06, | |
| "loss": 0.1848, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 2.0669642857142856, | |
| "grad_norm": 1.6733424663543701, | |
| "learning_rate": 7.442073842586733e-06, | |
| "loss": 0.1877, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 2.0714285714285716, | |
| "grad_norm": 1.5740326642990112, | |
| "learning_rate": 7.43176175498086e-06, | |
| "loss": 0.1693, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 2.075892857142857, | |
| "grad_norm": 1.621090054512024, | |
| "learning_rate": 7.421436102119427e-06, | |
| "loss": 0.2113, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 2.080357142857143, | |
| "grad_norm": 1.4219175577163696, | |
| "learning_rate": 7.411096941614543e-06, | |
| "loss": 0.1687, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 2.0848214285714284, | |
| "grad_norm": 1.4048285484313965, | |
| "learning_rate": 7.400744331153684e-06, | |
| "loss": 0.1832, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 2.0892857142857144, | |
| "grad_norm": 1.2322577238082886, | |
| "learning_rate": 7.390378328499372e-06, | |
| "loss": 0.1558, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 2.09375, | |
| "grad_norm": 1.5150905847549438, | |
| "learning_rate": 7.3799989914888506e-06, | |
| "loss": 0.2065, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 2.0982142857142856, | |
| "grad_norm": 1.4450438022613525, | |
| "learning_rate": 7.3696063780337566e-06, | |
| "loss": 0.1938, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 2.1026785714285716, | |
| "grad_norm": 1.6060644388198853, | |
| "learning_rate": 7.359200546119813e-06, | |
| "loss": 0.1865, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 2.107142857142857, | |
| "grad_norm": 1.4423115253448486, | |
| "learning_rate": 7.3487815538064865e-06, | |
| "loss": 0.1867, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 2.111607142857143, | |
| "grad_norm": 1.3800395727157593, | |
| "learning_rate": 7.338349459226678e-06, | |
| "loss": 0.1674, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 2.1160714285714284, | |
| "grad_norm": 1.5749934911727905, | |
| "learning_rate": 7.327904320586387e-06, | |
| "loss": 0.2004, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 2.1205357142857144, | |
| "grad_norm": 1.5107316970825195, | |
| "learning_rate": 7.3174461961644e-06, | |
| "loss": 0.1827, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 2.125, | |
| "grad_norm": 1.3470244407653809, | |
| "learning_rate": 7.3069751443119505e-06, | |
| "loss": 0.1696, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 2.1294642857142856, | |
| "grad_norm": 1.3709880113601685, | |
| "learning_rate": 7.296491223452407e-06, | |
| "loss": 0.1768, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 2.1339285714285716, | |
| "grad_norm": 1.4795838594436646, | |
| "learning_rate": 7.285994492080934e-06, | |
| "loss": 0.1915, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 2.138392857142857, | |
| "grad_norm": 1.5192097425460815, | |
| "learning_rate": 7.275485008764183e-06, | |
| "loss": 0.192, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 1.7044153213500977, | |
| "learning_rate": 7.2649628321399415e-06, | |
| "loss": 0.1945, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.1473214285714284, | |
| "grad_norm": 1.4719630479812622, | |
| "learning_rate": 7.254428020916829e-06, | |
| "loss": 0.1842, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 2.1517857142857144, | |
| "grad_norm": 1.4197455644607544, | |
| "learning_rate": 7.243880633873957e-06, | |
| "loss": 0.1977, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 2.15625, | |
| "grad_norm": 1.5410860776901245, | |
| "learning_rate": 7.2333207298606075e-06, | |
| "loss": 0.1899, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 2.1607142857142856, | |
| "grad_norm": 1.4274650812149048, | |
| "learning_rate": 7.222748367795892e-06, | |
| "loss": 0.1977, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 2.1651785714285716, | |
| "grad_norm": 1.4208344221115112, | |
| "learning_rate": 7.212163606668442e-06, | |
| "loss": 0.1799, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 2.169642857142857, | |
| "grad_norm": 1.4087142944335938, | |
| "learning_rate": 7.201566505536065e-06, | |
| "loss": 0.1876, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 2.174107142857143, | |
| "grad_norm": 1.5957927703857422, | |
| "learning_rate": 7.190957123525417e-06, | |
| "loss": 0.1926, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 2.1785714285714284, | |
| "grad_norm": 1.4640475511550903, | |
| "learning_rate": 7.180335519831685e-06, | |
| "loss": 0.1962, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 2.1830357142857144, | |
| "grad_norm": 1.6668205261230469, | |
| "learning_rate": 7.169701753718232e-06, | |
| "loss": 0.1923, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 2.1875, | |
| "grad_norm": 1.4836277961730957, | |
| "learning_rate": 7.159055884516297e-06, | |
| "loss": 0.1698, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.1919642857142856, | |
| "grad_norm": 1.4313628673553467, | |
| "learning_rate": 7.148397971624636e-06, | |
| "loss": 0.1823, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 2.1964285714285716, | |
| "grad_norm": 1.4819891452789307, | |
| "learning_rate": 7.137728074509211e-06, | |
| "loss": 0.2001, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 2.200892857142857, | |
| "grad_norm": 1.4184925556182861, | |
| "learning_rate": 7.127046252702847e-06, | |
| "loss": 0.1985, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 2.205357142857143, | |
| "grad_norm": 1.4677350521087646, | |
| "learning_rate": 7.116352565804904e-06, | |
| "loss": 0.1803, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 2.2098214285714284, | |
| "grad_norm": 1.4225105047225952, | |
| "learning_rate": 7.105647073480939e-06, | |
| "loss": 0.1934, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 2.2142857142857144, | |
| "grad_norm": 1.5292863845825195, | |
| "learning_rate": 7.0949298354623855e-06, | |
| "loss": 0.1983, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 2.21875, | |
| "grad_norm": 1.6830313205718994, | |
| "learning_rate": 7.084200911546205e-06, | |
| "loss": 0.2033, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 2.2232142857142856, | |
| "grad_norm": 1.462639331817627, | |
| "learning_rate": 7.073460361594565e-06, | |
| "loss": 0.1775, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 2.2276785714285716, | |
| "grad_norm": 1.3703210353851318, | |
| "learning_rate": 7.0627082455344984e-06, | |
| "loss": 0.167, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 2.232142857142857, | |
| "grad_norm": 1.4754226207733154, | |
| "learning_rate": 7.0519446233575715e-06, | |
| "loss": 0.1945, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.236607142857143, | |
| "grad_norm": 1.5474300384521484, | |
| "learning_rate": 7.041169555119552e-06, | |
| "loss": 0.1792, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 2.2410714285714284, | |
| "grad_norm": 1.745498776435852, | |
| "learning_rate": 7.030383100940068e-06, | |
| "loss": 0.2106, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 2.2455357142857144, | |
| "grad_norm": 1.510685920715332, | |
| "learning_rate": 7.019585321002276e-06, | |
| "loss": 0.1952, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 1.4274405241012573, | |
| "learning_rate": 7.008776275552522e-06, | |
| "loss": 0.1823, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 2.2544642857142856, | |
| "grad_norm": 1.5458096265792847, | |
| "learning_rate": 6.997956024900014e-06, | |
| "loss": 0.1768, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 2.2589285714285716, | |
| "grad_norm": 1.645890474319458, | |
| "learning_rate": 6.9871246294164775e-06, | |
| "loss": 0.1986, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 2.263392857142857, | |
| "grad_norm": 1.677795171737671, | |
| "learning_rate": 6.9762821495358194e-06, | |
| "loss": 0.2074, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 2.267857142857143, | |
| "grad_norm": 1.4883826971054077, | |
| "learning_rate": 6.965428645753792e-06, | |
| "loss": 0.1853, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 2.2723214285714284, | |
| "grad_norm": 1.3959269523620605, | |
| "learning_rate": 6.954564178627655e-06, | |
| "loss": 0.1731, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 2.2767857142857144, | |
| "grad_norm": 1.4190350770950317, | |
| "learning_rate": 6.943688808775843e-06, | |
| "loss": 0.1996, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.28125, | |
| "grad_norm": 1.4735363721847534, | |
| "learning_rate": 6.9328025968776155e-06, | |
| "loss": 0.1716, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 2.2857142857142856, | |
| "grad_norm": 1.458999752998352, | |
| "learning_rate": 6.921905603672733e-06, | |
| "loss": 0.1894, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 2.2901785714285716, | |
| "grad_norm": 1.522794246673584, | |
| "learning_rate": 6.910997889961098e-06, | |
| "loss": 0.2079, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 2.2901785714285716, | |
| "eval_loss": 0.6830385327339172, | |
| "eval_runtime": 3.1216, | |
| "eval_samples_per_second": 19.221, | |
| "eval_steps_per_second": 1.281, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 2.294642857142857, | |
| "grad_norm": 1.3990702629089355, | |
| "learning_rate": 6.900079516602445e-06, | |
| "loss": 0.1609, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 2.299107142857143, | |
| "grad_norm": 1.5663005113601685, | |
| "learning_rate": 6.889150544515972e-06, | |
| "loss": 0.1873, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 2.3035714285714284, | |
| "grad_norm": 1.6061094999313354, | |
| "learning_rate": 6.8782110346800155e-06, | |
| "loss": 0.1913, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 2.3080357142857144, | |
| "grad_norm": 1.5188199281692505, | |
| "learning_rate": 6.867261048131712e-06, | |
| "loss": 0.1748, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 2.3125, | |
| "grad_norm": 1.574725866317749, | |
| "learning_rate": 6.856300645966645e-06, | |
| "loss": 0.1963, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 2.3169642857142856, | |
| "grad_norm": 1.462173581123352, | |
| "learning_rate": 6.845329889338519e-06, | |
| "loss": 0.1942, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 2.3214285714285716, | |
| "grad_norm": 1.5870957374572754, | |
| "learning_rate": 6.834348839458806e-06, | |
| "loss": 0.1953, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.325892857142857, | |
| "grad_norm": 1.4638605117797852, | |
| "learning_rate": 6.823357557596416e-06, | |
| "loss": 0.173, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 2.330357142857143, | |
| "grad_norm": 1.6311603784561157, | |
| "learning_rate": 6.81235610507734e-06, | |
| "loss": 0.1966, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 2.3348214285714284, | |
| "grad_norm": 1.397397756576538, | |
| "learning_rate": 6.801344543284324e-06, | |
| "loss": 0.1798, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 2.3392857142857144, | |
| "grad_norm": 1.4249341487884521, | |
| "learning_rate": 6.790322933656515e-06, | |
| "loss": 0.1752, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 2.34375, | |
| "grad_norm": 1.5181248188018799, | |
| "learning_rate": 6.779291337689122e-06, | |
| "loss": 0.1919, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.3482142857142856, | |
| "grad_norm": 1.5693392753601074, | |
| "learning_rate": 6.768249816933074e-06, | |
| "loss": 0.2068, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 2.3526785714285716, | |
| "grad_norm": 1.3944882154464722, | |
| "learning_rate": 6.757198432994674e-06, | |
| "loss": 0.1921, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.357142857142857, | |
| "grad_norm": 1.467804193496704, | |
| "learning_rate": 6.7461372475352585e-06, | |
| "loss": 0.1736, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 2.361607142857143, | |
| "grad_norm": 1.5602641105651855, | |
| "learning_rate": 6.73506632227085e-06, | |
| "loss": 0.2016, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 2.3660714285714284, | |
| "grad_norm": 1.4512310028076172, | |
| "learning_rate": 6.723985718971818e-06, | |
| "loss": 0.1734, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.3705357142857144, | |
| "grad_norm": 1.5196360349655151, | |
| "learning_rate": 6.712895499462527e-06, | |
| "loss": 0.1815, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 2.375, | |
| "grad_norm": 1.38957679271698, | |
| "learning_rate": 6.701795725620995e-06, | |
| "loss": 0.1815, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 2.3794642857142856, | |
| "grad_norm": 1.3005454540252686, | |
| "learning_rate": 6.69068645937855e-06, | |
| "loss": 0.1751, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 2.3839285714285716, | |
| "grad_norm": 1.3817120790481567, | |
| "learning_rate": 6.6795677627194835e-06, | |
| "loss": 0.1728, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 2.388392857142857, | |
| "grad_norm": 1.559007167816162, | |
| "learning_rate": 6.668439697680704e-06, | |
| "loss": 0.1765, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 2.392857142857143, | |
| "grad_norm": 1.5712296962738037, | |
| "learning_rate": 6.65730232635139e-06, | |
| "loss": 0.2081, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 2.3973214285714284, | |
| "grad_norm": 1.3464829921722412, | |
| "learning_rate": 6.6461557108726435e-06, | |
| "loss": 0.1874, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 2.4017857142857144, | |
| "grad_norm": 1.4296271800994873, | |
| "learning_rate": 6.634999913437148e-06, | |
| "loss": 0.1625, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 2.40625, | |
| "grad_norm": 1.5054280757904053, | |
| "learning_rate": 6.623834996288816e-06, | |
| "loss": 0.185, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 2.4107142857142856, | |
| "grad_norm": 1.50546133518219, | |
| "learning_rate": 6.6126610217224405e-06, | |
| "loss": 0.2131, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.4151785714285716, | |
| "grad_norm": 1.6412122249603271, | |
| "learning_rate": 6.601478052083356e-06, | |
| "loss": 0.2126, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 2.419642857142857, | |
| "grad_norm": 1.5263912677764893, | |
| "learning_rate": 6.59028614976708e-06, | |
| "loss": 0.2023, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 2.424107142857143, | |
| "grad_norm": 1.551417589187622, | |
| "learning_rate": 6.579085377218973e-06, | |
| "loss": 0.2003, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 2.4285714285714284, | |
| "grad_norm": 1.6211479902267456, | |
| "learning_rate": 6.567875796933888e-06, | |
| "loss": 0.2003, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 2.4330357142857144, | |
| "grad_norm": 1.471747875213623, | |
| "learning_rate": 6.556657471455817e-06, | |
| "loss": 0.1965, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 2.4375, | |
| "grad_norm": 1.4076069593429565, | |
| "learning_rate": 6.54543046337755e-06, | |
| "loss": 0.2109, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 2.4419642857142856, | |
| "grad_norm": 1.5845184326171875, | |
| "learning_rate": 6.534194835340321e-06, | |
| "loss": 0.2135, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 2.4464285714285716, | |
| "grad_norm": 1.4862483739852905, | |
| "learning_rate": 6.522950650033454e-06, | |
| "loss": 0.1866, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 2.450892857142857, | |
| "grad_norm": 1.5533431768417358, | |
| "learning_rate": 6.511697970194024e-06, | |
| "loss": 0.2, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 2.455357142857143, | |
| "grad_norm": 1.3386541604995728, | |
| "learning_rate": 6.500436858606501e-06, | |
| "loss": 0.1801, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.4598214285714284, | |
| "grad_norm": 1.4485145807266235, | |
| "learning_rate": 6.4891673781023975e-06, | |
| "loss": 0.1829, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 2.4642857142857144, | |
| "grad_norm": 1.59889554977417, | |
| "learning_rate": 6.477889591559926e-06, | |
| "loss": 0.2162, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 2.46875, | |
| "grad_norm": 1.4210249185562134, | |
| "learning_rate": 6.466603561903633e-06, | |
| "loss": 0.1716, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 2.4732142857142856, | |
| "grad_norm": 1.406024694442749, | |
| "learning_rate": 6.455309352104065e-06, | |
| "loss": 0.1924, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 2.4776785714285716, | |
| "grad_norm": 1.4636434316635132, | |
| "learning_rate": 6.444007025177407e-06, | |
| "loss": 0.1902, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 2.482142857142857, | |
| "grad_norm": 1.5032724142074585, | |
| "learning_rate": 6.4326966441851355e-06, | |
| "loss": 0.1878, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 2.486607142857143, | |
| "grad_norm": 1.3488855361938477, | |
| "learning_rate": 6.4213782722336625e-06, | |
| "loss": 0.1592, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 2.4910714285714284, | |
| "grad_norm": 1.7310540676116943, | |
| "learning_rate": 6.4100519724739875e-06, | |
| "loss": 0.2106, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 2.4955357142857144, | |
| "grad_norm": 1.4981244802474976, | |
| "learning_rate": 6.3987178081013446e-06, | |
| "loss": 0.1887, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 1.3576077222824097, | |
| "learning_rate": 6.387375842354843e-06, | |
| "loss": 0.1877, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.5044642857142856, | |
| "grad_norm": 1.4860981702804565, | |
| "learning_rate": 6.376026138517125e-06, | |
| "loss": 0.1788, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 2.508928571428571, | |
| "grad_norm": 1.5492916107177734, | |
| "learning_rate": 6.364668759914005e-06, | |
| "loss": 0.1964, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 2.513392857142857, | |
| "grad_norm": 1.440119743347168, | |
| "learning_rate": 6.353303769914121e-06, | |
| "loss": 0.2057, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 2.517857142857143, | |
| "grad_norm": 1.3989627361297607, | |
| "learning_rate": 6.341931231928577e-06, | |
| "loss": 0.1675, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 2.522321428571429, | |
| "grad_norm": 1.6142938137054443, | |
| "learning_rate": 6.330551209410593e-06, | |
| "loss": 0.1856, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 2.5267857142857144, | |
| "grad_norm": 1.6988365650177002, | |
| "learning_rate": 6.319163765855146e-06, | |
| "loss": 0.2122, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 2.53125, | |
| "grad_norm": 1.641639232635498, | |
| "learning_rate": 6.307768964798623e-06, | |
| "loss": 0.1836, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 2.5357142857142856, | |
| "grad_norm": 1.6539298295974731, | |
| "learning_rate": 6.296366869818458e-06, | |
| "loss": 0.203, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 2.540178571428571, | |
| "grad_norm": 1.6499872207641602, | |
| "learning_rate": 6.284957544532783e-06, | |
| "loss": 0.2115, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 2.544642857142857, | |
| "grad_norm": 1.6138975620269775, | |
| "learning_rate": 6.273541052600074e-06, | |
| "loss": 0.2228, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.544642857142857, | |
| "eval_loss": 0.6881652474403381, | |
| "eval_runtime": 3.0998, | |
| "eval_samples_per_second": 19.356, | |
| "eval_steps_per_second": 1.29, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.549107142857143, | |
| "grad_norm": 1.3579055070877075, | |
| "learning_rate": 6.2621174577187895e-06, | |
| "loss": 0.1622, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 2.553571428571429, | |
| "grad_norm": 1.519002079963684, | |
| "learning_rate": 6.250686823627022e-06, | |
| "loss": 0.1875, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 2.5580357142857144, | |
| "grad_norm": 1.2916154861450195, | |
| "learning_rate": 6.239249214102139e-06, | |
| "loss": 0.1616, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 2.5625, | |
| "grad_norm": 1.4468685388565063, | |
| "learning_rate": 6.2278046929604265e-06, | |
| "loss": 0.2001, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 2.5669642857142856, | |
| "grad_norm": 1.4290978908538818, | |
| "learning_rate": 6.216353324056732e-06, | |
| "loss": 0.1818, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.571428571428571, | |
| "grad_norm": 1.3633805513381958, | |
| "learning_rate": 6.204895171284115e-06, | |
| "loss": 0.1724, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 2.575892857142857, | |
| "grad_norm": 1.4299393892288208, | |
| "learning_rate": 6.193430298573481e-06, | |
| "loss": 0.2044, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 2.580357142857143, | |
| "grad_norm": 1.428096890449524, | |
| "learning_rate": 6.181958769893234e-06, | |
| "loss": 0.2038, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 2.584821428571429, | |
| "grad_norm": 1.4805070161819458, | |
| "learning_rate": 6.17048064924891e-06, | |
| "loss": 0.2023, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 2.5892857142857144, | |
| "grad_norm": 1.5858566761016846, | |
| "learning_rate": 6.158996000682829e-06, | |
| "loss": 0.1874, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.59375, | |
| "grad_norm": 1.547018051147461, | |
| "learning_rate": 6.147504888273736e-06, | |
| "loss": 0.1938, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 2.5982142857142856, | |
| "grad_norm": 1.5140594244003296, | |
| "learning_rate": 6.136007376136429e-06, | |
| "loss": 0.1989, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 2.602678571428571, | |
| "grad_norm": 1.2419382333755493, | |
| "learning_rate": 6.124503528421429e-06, | |
| "loss": 0.1524, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 2.607142857142857, | |
| "grad_norm": 1.5366337299346924, | |
| "learning_rate": 6.112993409314594e-06, | |
| "loss": 0.1853, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 2.611607142857143, | |
| "grad_norm": 1.4070125818252563, | |
| "learning_rate": 6.101477083036783e-06, | |
| "loss": 0.1767, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 2.616071428571429, | |
| "grad_norm": 1.5408363342285156, | |
| "learning_rate": 6.0899546138434785e-06, | |
| "loss": 0.2016, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 2.6205357142857144, | |
| "grad_norm": 1.5800507068634033, | |
| "learning_rate": 6.0784260660244475e-06, | |
| "loss": 0.1665, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 2.625, | |
| "grad_norm": 1.4633119106292725, | |
| "learning_rate": 6.066891503903363e-06, | |
| "loss": 0.1846, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 2.6294642857142856, | |
| "grad_norm": 1.5278315544128418, | |
| "learning_rate": 6.0553509918374635e-06, | |
| "loss": 0.1761, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 2.633928571428571, | |
| "grad_norm": 1.4787242412567139, | |
| "learning_rate": 6.0438045942171775e-06, | |
| "loss": 0.1836, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.638392857142857, | |
| "grad_norm": 1.4805656671524048, | |
| "learning_rate": 6.032252375465778e-06, | |
| "loss": 0.1823, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 2.642857142857143, | |
| "grad_norm": 1.4597275257110596, | |
| "learning_rate": 6.020694400039017e-06, | |
| "loss": 0.1751, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 2.647321428571429, | |
| "grad_norm": 1.5701682567596436, | |
| "learning_rate": 6.009130732424758e-06, | |
| "loss": 0.1856, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 2.6517857142857144, | |
| "grad_norm": 1.5376973152160645, | |
| "learning_rate": 5.997561437142636e-06, | |
| "loss": 0.2012, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 2.65625, | |
| "grad_norm": 1.3718751668930054, | |
| "learning_rate": 5.985986578743676e-06, | |
| "loss": 0.1746, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.6607142857142856, | |
| "grad_norm": 1.4416507482528687, | |
| "learning_rate": 5.974406221809949e-06, | |
| "loss": 0.1865, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 2.665178571428571, | |
| "grad_norm": 1.628366470336914, | |
| "learning_rate": 5.962820430954198e-06, | |
| "loss": 0.2131, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 2.669642857142857, | |
| "grad_norm": 1.4710111618041992, | |
| "learning_rate": 5.951229270819494e-06, | |
| "loss": 0.197, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 2.674107142857143, | |
| "grad_norm": 1.3397061824798584, | |
| "learning_rate": 5.9396328060788576e-06, | |
| "loss": 0.1876, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 2.678571428571429, | |
| "grad_norm": 1.5016379356384277, | |
| "learning_rate": 5.928031101434908e-06, | |
| "loss": 0.1825, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.6830357142857144, | |
| "grad_norm": 1.597392201423645, | |
| "learning_rate": 5.916424221619507e-06, | |
| "loss": 0.2096, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 2.6875, | |
| "grad_norm": 1.4211905002593994, | |
| "learning_rate": 5.904812231393383e-06, | |
| "loss": 0.1961, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 2.6919642857142856, | |
| "grad_norm": 1.3317689895629883, | |
| "learning_rate": 5.893195195545784e-06, | |
| "loss": 0.1727, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 2.696428571428571, | |
| "grad_norm": 1.5371907949447632, | |
| "learning_rate": 5.8815731788941064e-06, | |
| "loss": 0.1877, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 2.700892857142857, | |
| "grad_norm": 1.4972670078277588, | |
| "learning_rate": 5.86994624628354e-06, | |
| "loss": 0.1853, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.705357142857143, | |
| "grad_norm": 1.468634843826294, | |
| "learning_rate": 5.858314462586697e-06, | |
| "loss": 0.1787, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 2.709821428571429, | |
| "grad_norm": 1.6285043954849243, | |
| "learning_rate": 5.846677892703268e-06, | |
| "loss": 0.2082, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 2.7142857142857144, | |
| "grad_norm": 1.4436808824539185, | |
| "learning_rate": 5.835036601559634e-06, | |
| "loss": 0.1811, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 2.71875, | |
| "grad_norm": 1.5260608196258545, | |
| "learning_rate": 5.82339065410853e-06, | |
| "loss": 0.1872, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 2.7232142857142856, | |
| "grad_norm": 1.5837960243225098, | |
| "learning_rate": 5.811740115328665e-06, | |
| "loss": 0.1817, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.727678571428571, | |
| "grad_norm": 1.5583548545837402, | |
| "learning_rate": 5.800085050224367e-06, | |
| "loss": 0.1993, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 2.732142857142857, | |
| "grad_norm": 1.5718779563903809, | |
| "learning_rate": 5.7884255238252175e-06, | |
| "loss": 0.1955, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 2.736607142857143, | |
| "grad_norm": 1.5729767084121704, | |
| "learning_rate": 5.776761601185692e-06, | |
| "loss": 0.1868, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 2.741071428571429, | |
| "grad_norm": 1.5369676351547241, | |
| "learning_rate": 5.765093347384793e-06, | |
| "loss": 0.1974, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 2.7455357142857144, | |
| "grad_norm": 1.5321385860443115, | |
| "learning_rate": 5.753420827525691e-06, | |
| "loss": 0.1991, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 1.6197004318237305, | |
| "learning_rate": 5.741744106735354e-06, | |
| "loss": 0.1967, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 2.7544642857142856, | |
| "grad_norm": 1.4618470668792725, | |
| "learning_rate": 5.730063250164196e-06, | |
| "loss": 0.1778, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 2.758928571428571, | |
| "grad_norm": 1.6011563539505005, | |
| "learning_rate": 5.718378322985702e-06, | |
| "loss": 0.1986, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 2.763392857142857, | |
| "grad_norm": 1.3535871505737305, | |
| "learning_rate": 5.70668939039607e-06, | |
| "loss": 0.179, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 2.767857142857143, | |
| "grad_norm": 1.5589332580566406, | |
| "learning_rate": 5.694996517613847e-06, | |
| "loss": 0.2123, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.772321428571429, | |
| "grad_norm": 1.483581304550171, | |
| "learning_rate": 5.683299769879562e-06, | |
| "loss": 0.189, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 2.7767857142857144, | |
| "grad_norm": 1.5656222105026245, | |
| "learning_rate": 5.6715992124553685e-06, | |
| "loss": 0.1915, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 2.78125, | |
| "grad_norm": 1.5247652530670166, | |
| "learning_rate": 5.6598949106246734e-06, | |
| "loss": 0.2024, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 2.7857142857142856, | |
| "grad_norm": 1.7072973251342773, | |
| "learning_rate": 5.648186929691776e-06, | |
| "loss": 0.2113, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 2.790178571428571, | |
| "grad_norm": 1.5222718715667725, | |
| "learning_rate": 5.6364753349815035e-06, | |
| "loss": 0.2002, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.794642857142857, | |
| "grad_norm": 1.42404305934906, | |
| "learning_rate": 5.624760191838845e-06, | |
| "loss": 0.1804, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 2.799107142857143, | |
| "grad_norm": 1.5567388534545898, | |
| "learning_rate": 5.61304156562859e-06, | |
| "loss": 0.2147, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 2.799107142857143, | |
| "eval_loss": 0.6883992552757263, | |
| "eval_runtime": 3.1165, | |
| "eval_samples_per_second": 19.252, | |
| "eval_steps_per_second": 1.283, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 2.803571428571429, | |
| "grad_norm": 1.4494566917419434, | |
| "learning_rate": 5.60131952173496e-06, | |
| "loss": 0.1921, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 2.8080357142857144, | |
| "grad_norm": 1.530175805091858, | |
| "learning_rate": 5.589594125561246e-06, | |
| "loss": 0.1979, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 2.8125, | |
| "grad_norm": 1.4815398454666138, | |
| "learning_rate": 5.577865442529447e-06, | |
| "loss": 0.1999, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.8169642857142856, | |
| "grad_norm": 1.4520913362503052, | |
| "learning_rate": 5.566133538079893e-06, | |
| "loss": 0.1783, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 2.821428571428571, | |
| "grad_norm": 1.5236479043960571, | |
| "learning_rate": 5.554398477670895e-06, | |
| "loss": 0.1875, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 2.825892857142857, | |
| "grad_norm": 1.3968005180358887, | |
| "learning_rate": 5.54266032677837e-06, | |
| "loss": 0.1919, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 2.830357142857143, | |
| "grad_norm": 1.4452468156814575, | |
| "learning_rate": 5.53091915089548e-06, | |
| "loss": 0.1802, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.834821428571429, | |
| "grad_norm": 1.512412190437317, | |
| "learning_rate": 5.5191750155322595e-06, | |
| "loss": 0.195, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 2.8392857142857144, | |
| "grad_norm": 1.4154279232025146, | |
| "learning_rate": 5.507427986215265e-06, | |
| "loss": 0.1846, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.84375, | |
| "grad_norm": 1.6457233428955078, | |
| "learning_rate": 5.49567812848719e-06, | |
| "loss": 0.1981, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 2.8482142857142856, | |
| "grad_norm": 1.6645605564117432, | |
| "learning_rate": 5.483925507906514e-06, | |
| "loss": 0.1992, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 2.852678571428571, | |
| "grad_norm": 1.491758942604065, | |
| "learning_rate": 5.4721701900471335e-06, | |
| "loss": 0.188, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 1.5240974426269531, | |
| "learning_rate": 5.460412240497993e-06, | |
| "loss": 0.1966, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.861607142857143, | |
| "grad_norm": 1.5371315479278564, | |
| "learning_rate": 5.448651724862716e-06, | |
| "loss": 0.1846, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 2.866071428571429, | |
| "grad_norm": 1.4583534002304077, | |
| "learning_rate": 5.436888708759253e-06, | |
| "loss": 0.1906, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.8705357142857144, | |
| "grad_norm": 1.4962140321731567, | |
| "learning_rate": 5.425123257819494e-06, | |
| "loss": 0.1965, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 2.875, | |
| "grad_norm": 1.5177336931228638, | |
| "learning_rate": 5.413355437688926e-06, | |
| "loss": 0.1751, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 2.8794642857142856, | |
| "grad_norm": 1.5701833963394165, | |
| "learning_rate": 5.401585314026248e-06, | |
| "loss": 0.1948, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.883928571428571, | |
| "grad_norm": 1.6026036739349365, | |
| "learning_rate": 5.3898129525030105e-06, | |
| "loss": 0.2127, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 2.888392857142857, | |
| "grad_norm": 1.6437232494354248, | |
| "learning_rate": 5.378038418803256e-06, | |
| "loss": 0.1991, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 2.892857142857143, | |
| "grad_norm": 1.5540730953216553, | |
| "learning_rate": 5.366261778623143e-06, | |
| "loss": 0.2023, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 2.897321428571429, | |
| "grad_norm": 1.5963166952133179, | |
| "learning_rate": 5.354483097670584e-06, | |
| "loss": 0.1991, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 2.9017857142857144, | |
| "grad_norm": 1.7462104558944702, | |
| "learning_rate": 5.342702441664875e-06, | |
| "loss": 0.2079, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.90625, | |
| "grad_norm": 1.6072155237197876, | |
| "learning_rate": 5.3309198763363345e-06, | |
| "loss": 0.207, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 2.9107142857142856, | |
| "grad_norm": 1.5229613780975342, | |
| "learning_rate": 5.319135467425937e-06, | |
| "loss": 0.1961, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 2.915178571428571, | |
| "grad_norm": 1.3907989263534546, | |
| "learning_rate": 5.3073492806849405e-06, | |
| "loss": 0.1754, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 2.919642857142857, | |
| "grad_norm": 1.3996022939682007, | |
| "learning_rate": 5.295561381874518e-06, | |
| "loss": 0.1684, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 2.924107142857143, | |
| "grad_norm": 1.448462724685669, | |
| "learning_rate": 5.2837718367654036e-06, | |
| "loss": 0.1932, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 2.928571428571429, | |
| "grad_norm": 1.5644668340682983, | |
| "learning_rate": 5.2719807111375096e-06, | |
| "loss": 0.1985, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 2.9330357142857144, | |
| "grad_norm": 1.6308317184448242, | |
| "learning_rate": 5.260188070779573e-06, | |
| "loss": 0.1931, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 2.9375, | |
| "grad_norm": 1.47319757938385, | |
| "learning_rate": 5.248393981488777e-06, | |
| "loss": 0.1859, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 2.9419642857142856, | |
| "grad_norm": 1.4166940450668335, | |
| "learning_rate": 5.236598509070389e-06, | |
| "loss": 0.175, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 2.946428571428571, | |
| "grad_norm": 1.5478429794311523, | |
| "learning_rate": 5.2248017193374e-06, | |
| "loss": 0.2117, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.950892857142857, | |
| "grad_norm": 1.5836637020111084, | |
| "learning_rate": 5.2130036781101455e-06, | |
| "loss": 0.1859, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 2.955357142857143, | |
| "grad_norm": 1.584566593170166, | |
| "learning_rate": 5.201204451215945e-06, | |
| "loss": 0.1871, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 2.959821428571429, | |
| "grad_norm": 1.43748939037323, | |
| "learning_rate": 5.18940410448873e-06, | |
| "loss": 0.1822, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 2.9642857142857144, | |
| "grad_norm": 1.4014517068862915, | |
| "learning_rate": 5.1776027037686895e-06, | |
| "loss": 0.1731, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 2.96875, | |
| "grad_norm": 1.5529756546020508, | |
| "learning_rate": 5.165800314901883e-06, | |
| "loss": 0.1923, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 2.9732142857142856, | |
| "grad_norm": 1.4991168975830078, | |
| "learning_rate": 5.15399700373989e-06, | |
| "loss": 0.1984, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 2.977678571428571, | |
| "grad_norm": 1.5158687829971313, | |
| "learning_rate": 5.142192836139432e-06, | |
| "loss": 0.1903, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 2.982142857142857, | |
| "grad_norm": 1.5199565887451172, | |
| "learning_rate": 5.130387877962012e-06, | |
| "loss": 0.1709, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 2.986607142857143, | |
| "grad_norm": 1.5208170413970947, | |
| "learning_rate": 5.118582195073542e-06, | |
| "loss": 0.1964, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 2.991071428571429, | |
| "grad_norm": 1.45652437210083, | |
| "learning_rate": 5.1067758533439804e-06, | |
| "loss": 0.1957, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.9955357142857144, | |
| "grad_norm": 1.3445968627929688, | |
| "learning_rate": 5.094968918646954e-06, | |
| "loss": 0.1752, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.3375955820083618, | |
| "learning_rate": 5.0831614568594105e-06, | |
| "loss": 0.1758, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 3.0044642857142856, | |
| "grad_norm": 1.089407205581665, | |
| "learning_rate": 5.071353533861225e-06, | |
| "loss": 0.1099, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 3.0089285714285716, | |
| "grad_norm": 1.0388957262039185, | |
| "learning_rate": 5.059545215534859e-06, | |
| "loss": 0.1054, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 3.013392857142857, | |
| "grad_norm": 1.2963316440582275, | |
| "learning_rate": 5.047736567764967e-06, | |
| "loss": 0.114, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 3.017857142857143, | |
| "grad_norm": 1.0460807085037231, | |
| "learning_rate": 5.0359276564380514e-06, | |
| "loss": 0.0957, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 3.0223214285714284, | |
| "grad_norm": 1.1701343059539795, | |
| "learning_rate": 5.024118547442083e-06, | |
| "loss": 0.0955, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 3.0267857142857144, | |
| "grad_norm": 1.2869516611099243, | |
| "learning_rate": 5.012309306666129e-06, | |
| "loss": 0.1006, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 3.03125, | |
| "grad_norm": 1.0782946348190308, | |
| "learning_rate": 5.000500000000001e-06, | |
| "loss": 0.0765, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 3.0357142857142856, | |
| "grad_norm": 1.4111815690994263, | |
| "learning_rate": 4.988690693333873e-06, | |
| "loss": 0.0861, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 3.0401785714285716, | |
| "grad_norm": 1.5609370470046997, | |
| "learning_rate": 4.97688145255792e-06, | |
| "loss": 0.1035, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 3.044642857142857, | |
| "grad_norm": 1.3300373554229736, | |
| "learning_rate": 4.965072343561948e-06, | |
| "loss": 0.0937, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 3.049107142857143, | |
| "grad_norm": 1.6226136684417725, | |
| "learning_rate": 4.953263432235034e-06, | |
| "loss": 0.0875, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 3.0535714285714284, | |
| "grad_norm": 1.5775179862976074, | |
| "learning_rate": 4.941454784465144e-06, | |
| "loss": 0.0839, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 3.0535714285714284, | |
| "eval_loss": 0.8500283360481262, | |
| "eval_runtime": 3.0932, | |
| "eval_samples_per_second": 19.397, | |
| "eval_steps_per_second": 1.293, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 3.0580357142857144, | |
| "grad_norm": 1.7753103971481323, | |
| "learning_rate": 4.929646466138777e-06, | |
| "loss": 0.0929, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 3.0625, | |
| "grad_norm": 1.6143715381622314, | |
| "learning_rate": 4.917838543140591e-06, | |
| "loss": 0.0753, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 3.0669642857142856, | |
| "grad_norm": 1.5157322883605957, | |
| "learning_rate": 4.906031081353047e-06, | |
| "loss": 0.0869, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 3.0714285714285716, | |
| "grad_norm": 1.7490545511245728, | |
| "learning_rate": 4.8942241466560226e-06, | |
| "loss": 0.0908, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 3.075892857142857, | |
| "grad_norm": 1.8048044443130493, | |
| "learning_rate": 4.882417804926457e-06, | |
| "loss": 0.0893, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 3.080357142857143, | |
| "grad_norm": 1.587409496307373, | |
| "learning_rate": 4.870612122037989e-06, | |
| "loss": 0.0918, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 3.0848214285714284, | |
| "grad_norm": 1.6927273273468018, | |
| "learning_rate": 4.858807163860569e-06, | |
| "loss": 0.0969, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 3.0892857142857144, | |
| "grad_norm": 1.8378304243087769, | |
| "learning_rate": 4.847002996260113e-06, | |
| "loss": 0.0923, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 3.09375, | |
| "grad_norm": 1.6607656478881836, | |
| "learning_rate": 4.835199685098117e-06, | |
| "loss": 0.1028, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 3.0982142857142856, | |
| "grad_norm": 1.640812635421753, | |
| "learning_rate": 4.823397296231313e-06, | |
| "loss": 0.0849, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 3.1026785714285716, | |
| "grad_norm": 1.707801103591919, | |
| "learning_rate": 4.8115958955112715e-06, | |
| "loss": 0.0944, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 3.107142857142857, | |
| "grad_norm": 1.6389654874801636, | |
| "learning_rate": 4.799795548784058e-06, | |
| "loss": 0.0872, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 3.111607142857143, | |
| "grad_norm": 1.723669409751892, | |
| "learning_rate": 4.787996321889856e-06, | |
| "loss": 0.0956, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 3.1160714285714284, | |
| "grad_norm": 1.560467004776001, | |
| "learning_rate": 4.7761982806626015e-06, | |
| "loss": 0.1029, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 3.1205357142857144, | |
| "grad_norm": 1.4397633075714111, | |
| "learning_rate": 4.764401490929613e-06, | |
| "loss": 0.087, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 3.125, | |
| "grad_norm": 1.480446457862854, | |
| "learning_rate": 4.752606018511225e-06, | |
| "loss": 0.0883, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.1294642857142856, | |
| "grad_norm": 1.4872808456420898, | |
| "learning_rate": 4.740811929220429e-06, | |
| "loss": 0.1094, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 3.1339285714285716, | |
| "grad_norm": 1.4935928583145142, | |
| "learning_rate": 4.729019288862492e-06, | |
| "loss": 0.0976, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 3.138392857142857, | |
| "grad_norm": 1.3575384616851807, | |
| "learning_rate": 4.717228163234599e-06, | |
| "loss": 0.0837, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 3.142857142857143, | |
| "grad_norm": 1.112143874168396, | |
| "learning_rate": 4.705438618125482e-06, | |
| "loss": 0.0696, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 3.1473214285714284, | |
| "grad_norm": 1.4292476177215576, | |
| "learning_rate": 4.693650719315062e-06, | |
| "loss": 0.1014, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 3.1517857142857144, | |
| "grad_norm": 1.388718605041504, | |
| "learning_rate": 4.681864532574064e-06, | |
| "loss": 0.1, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 3.15625, | |
| "grad_norm": 1.3625158071517944, | |
| "learning_rate": 4.670080123663668e-06, | |
| "loss": 0.1026, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 3.1607142857142856, | |
| "grad_norm": 1.1807421445846558, | |
| "learning_rate": 4.658297558335127e-06, | |
| "loss": 0.0884, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 3.1651785714285716, | |
| "grad_norm": 1.36719810962677, | |
| "learning_rate": 4.64651690232942e-06, | |
| "loss": 0.0948, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 3.169642857142857, | |
| "grad_norm": 1.239861011505127, | |
| "learning_rate": 4.634738221376858e-06, | |
| "loss": 0.0783, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 3.174107142857143, | |
| "grad_norm": 1.1686639785766602, | |
| "learning_rate": 4.622961581196743e-06, | |
| "loss": 0.0785, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 3.1785714285714284, | |
| "grad_norm": 1.2871395349502563, | |
| "learning_rate": 4.611187047496989e-06, | |
| "loss": 0.0838, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 3.1830357142857144, | |
| "grad_norm": 1.3209705352783203, | |
| "learning_rate": 4.599414685973754e-06, | |
| "loss": 0.0802, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 3.1875, | |
| "grad_norm": 1.2725917100906372, | |
| "learning_rate": 4.587644562311077e-06, | |
| "loss": 0.0856, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 3.1919642857142856, | |
| "grad_norm": 1.506381630897522, | |
| "learning_rate": 4.575876742180506e-06, | |
| "loss": 0.0974, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 3.1964285714285716, | |
| "grad_norm": 1.2153488397598267, | |
| "learning_rate": 4.56411129124075e-06, | |
| "loss": 0.085, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 3.200892857142857, | |
| "grad_norm": 1.4479278326034546, | |
| "learning_rate": 4.552348275137285e-06, | |
| "loss": 0.1026, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 3.205357142857143, | |
| "grad_norm": 1.657431960105896, | |
| "learning_rate": 4.54058775950201e-06, | |
| "loss": 0.1017, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 3.2098214285714284, | |
| "grad_norm": 1.4893465042114258, | |
| "learning_rate": 4.528829809952867e-06, | |
| "loss": 0.0917, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 3.2142857142857144, | |
| "grad_norm": 1.6063541173934937, | |
| "learning_rate": 4.517074492093487e-06, | |
| "loss": 0.1133, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 3.21875, | |
| "grad_norm": 1.5053054094314575, | |
| "learning_rate": 4.505321871512813e-06, | |
| "loss": 0.1028, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 3.2232142857142856, | |
| "grad_norm": 1.6949694156646729, | |
| "learning_rate": 4.493572013784737e-06, | |
| "loss": 0.1063, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 3.2276785714285716, | |
| "grad_norm": 1.3530110120773315, | |
| "learning_rate": 4.481824984467742e-06, | |
| "loss": 0.0957, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 3.232142857142857, | |
| "grad_norm": 1.4270166158676147, | |
| "learning_rate": 4.470080849104521e-06, | |
| "loss": 0.0959, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 3.236607142857143, | |
| "grad_norm": 1.7574278116226196, | |
| "learning_rate": 4.458339673221631e-06, | |
| "loss": 0.1066, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 3.2410714285714284, | |
| "grad_norm": 1.531517744064331, | |
| "learning_rate": 4.446601522329105e-06, | |
| "loss": 0.086, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 3.2455357142857144, | |
| "grad_norm": 1.428818702697754, | |
| "learning_rate": 4.434866461920108e-06, | |
| "loss": 0.0859, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 1.7895373106002808, | |
| "learning_rate": 4.4231345574705555e-06, | |
| "loss": 0.1118, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 3.2544642857142856, | |
| "grad_norm": 1.335732102394104, | |
| "learning_rate": 4.4114058744387535e-06, | |
| "loss": 0.0875, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 3.2589285714285716, | |
| "grad_norm": 1.547389268875122, | |
| "learning_rate": 4.399680478265042e-06, | |
| "loss": 0.1004, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 3.263392857142857, | |
| "grad_norm": 1.5613607168197632, | |
| "learning_rate": 4.387958434371413e-06, | |
| "loss": 0.0989, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 3.267857142857143, | |
| "grad_norm": 1.3108863830566406, | |
| "learning_rate": 4.376239808161157e-06, | |
| "loss": 0.0791, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 3.2723214285714284, | |
| "grad_norm": 1.4128496646881104, | |
| "learning_rate": 4.364524665018496e-06, | |
| "loss": 0.088, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 3.2767857142857144, | |
| "grad_norm": 1.280126690864563, | |
| "learning_rate": 4.3528130703082245e-06, | |
| "loss": 0.0838, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 3.28125, | |
| "grad_norm": 1.370133399963379, | |
| "learning_rate": 4.341105089375328e-06, | |
| "loss": 0.0833, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 3.2857142857142856, | |
| "grad_norm": 1.5750563144683838, | |
| "learning_rate": 4.329400787544633e-06, | |
| "loss": 0.0924, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 3.2901785714285716, | |
| "grad_norm": 1.218309760093689, | |
| "learning_rate": 4.317700230120438e-06, | |
| "loss": 0.0807, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 3.294642857142857, | |
| "grad_norm": 1.4388004541397095, | |
| "learning_rate": 4.306003482386156e-06, | |
| "loss": 0.1089, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 3.299107142857143, | |
| "grad_norm": 1.3800266981124878, | |
| "learning_rate": 4.2943106096039315e-06, | |
| "loss": 0.0916, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 3.3035714285714284, | |
| "grad_norm": 1.3160336017608643, | |
| "learning_rate": 4.282621677014299e-06, | |
| "loss": 0.0877, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 3.3080357142857144, | |
| "grad_norm": 1.3781185150146484, | |
| "learning_rate": 4.270936749835805e-06, | |
| "loss": 0.0984, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 3.3080357142857144, | |
| "eval_loss": 0.8194601535797119, | |
| "eval_runtime": 3.0964, | |
| "eval_samples_per_second": 19.378, | |
| "eval_steps_per_second": 1.292, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 3.3125, | |
| "grad_norm": 1.310025691986084, | |
| "learning_rate": 4.259255893264647e-06, | |
| "loss": 0.0893, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 3.3169642857142856, | |
| "grad_norm": 1.4620391130447388, | |
| "learning_rate": 4.247579172474312e-06, | |
| "loss": 0.1009, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 3.3214285714285716, | |
| "grad_norm": 1.4170101881027222, | |
| "learning_rate": 4.235906652615207e-06, | |
| "loss": 0.1037, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 3.325892857142857, | |
| "grad_norm": 1.8959122896194458, | |
| "learning_rate": 4.224238398814309e-06, | |
| "loss": 0.0969, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 3.330357142857143, | |
| "grad_norm": 1.463104248046875, | |
| "learning_rate": 4.212574476174784e-06, | |
| "loss": 0.0902, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 3.3348214285714284, | |
| "grad_norm": 1.4386403560638428, | |
| "learning_rate": 4.2009149497756355e-06, | |
| "loss": 0.0984, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 3.3392857142857144, | |
| "grad_norm": 1.3516470193862915, | |
| "learning_rate": 4.189259884671336e-06, | |
| "loss": 0.0875, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 3.34375, | |
| "grad_norm": 1.3817871809005737, | |
| "learning_rate": 4.177609345891472e-06, | |
| "loss": 0.0881, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 3.3482142857142856, | |
| "grad_norm": 1.302051305770874, | |
| "learning_rate": 4.165963398440368e-06, | |
| "loss": 0.0976, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.3526785714285716, | |
| "grad_norm": 1.4859888553619385, | |
| "learning_rate": 4.1543221072967334e-06, | |
| "loss": 0.0977, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 3.357142857142857, | |
| "grad_norm": 1.3460721969604492, | |
| "learning_rate": 4.142685537413303e-06, | |
| "loss": 0.086, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 3.361607142857143, | |
| "grad_norm": 1.467089056968689, | |
| "learning_rate": 4.1310537537164615e-06, | |
| "loss": 0.0999, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 3.3660714285714284, | |
| "grad_norm": 1.6154288053512573, | |
| "learning_rate": 4.119426821105895e-06, | |
| "loss": 0.0937, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 3.3705357142857144, | |
| "grad_norm": 1.610485315322876, | |
| "learning_rate": 4.107804804454215e-06, | |
| "loss": 0.0977, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 3.375, | |
| "grad_norm": 1.5039902925491333, | |
| "learning_rate": 4.096187768606617e-06, | |
| "loss": 0.0934, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 3.3794642857142856, | |
| "grad_norm": 1.3222577571868896, | |
| "learning_rate": 4.084575778380495e-06, | |
| "loss": 0.0777, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 3.3839285714285716, | |
| "grad_norm": 1.5685445070266724, | |
| "learning_rate": 4.072968898565094e-06, | |
| "loss": 0.0969, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 3.388392857142857, | |
| "grad_norm": 1.5284277200698853, | |
| "learning_rate": 4.061367193921145e-06, | |
| "loss": 0.1104, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 3.392857142857143, | |
| "grad_norm": 1.5759763717651367, | |
| "learning_rate": 4.049770729180508e-06, | |
| "loss": 0.0991, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 3.3973214285714284, | |
| "grad_norm": 1.490787386894226, | |
| "learning_rate": 4.038179569045803e-06, | |
| "loss": 0.0934, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 3.4017857142857144, | |
| "grad_norm": 1.490193486213684, | |
| "learning_rate": 4.026593778190052e-06, | |
| "loss": 0.1003, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 3.40625, | |
| "grad_norm": 1.3332754373550415, | |
| "learning_rate": 4.015013421256324e-06, | |
| "loss": 0.0855, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 3.4107142857142856, | |
| "grad_norm": 1.195845127105713, | |
| "learning_rate": 4.0034385628573655e-06, | |
| "loss": 0.0811, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 3.4151785714285716, | |
| "grad_norm": 1.3731515407562256, | |
| "learning_rate": 3.991869267575243e-06, | |
| "loss": 0.0848, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 3.419642857142857, | |
| "grad_norm": 1.6035957336425781, | |
| "learning_rate": 3.9803055999609855e-06, | |
| "loss": 0.0925, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 3.424107142857143, | |
| "grad_norm": 1.4509934186935425, | |
| "learning_rate": 3.9687476245342234e-06, | |
| "loss": 0.0963, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 3.4285714285714284, | |
| "grad_norm": 1.679079532623291, | |
| "learning_rate": 3.957195405782824e-06, | |
| "loss": 0.0976, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 3.4330357142857144, | |
| "grad_norm": 1.6150476932525635, | |
| "learning_rate": 3.9456490081625396e-06, | |
| "loss": 0.101, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 3.4375, | |
| "grad_norm": 1.3960528373718262, | |
| "learning_rate": 3.934108496096638e-06, | |
| "loss": 0.0918, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 3.4419642857142856, | |
| "grad_norm": 1.4752354621887207, | |
| "learning_rate": 3.922573933975555e-06, | |
| "loss": 0.0957, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 3.4464285714285716, | |
| "grad_norm": 1.4538674354553223, | |
| "learning_rate": 3.911045386156523e-06, | |
| "loss": 0.0991, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 3.450892857142857, | |
| "grad_norm": 1.5427879095077515, | |
| "learning_rate": 3.899522916963219e-06, | |
| "loss": 0.0964, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 3.455357142857143, | |
| "grad_norm": 1.506783366203308, | |
| "learning_rate": 3.888006590685407e-06, | |
| "loss": 0.0907, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 3.4598214285714284, | |
| "grad_norm": 1.3656387329101562, | |
| "learning_rate": 3.876496471578572e-06, | |
| "loss": 0.0939, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 3.4642857142857144, | |
| "grad_norm": 1.509194254875183, | |
| "learning_rate": 3.864992623863572e-06, | |
| "loss": 0.0962, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 3.46875, | |
| "grad_norm": 1.249863862991333, | |
| "learning_rate": 3.853495111726265e-06, | |
| "loss": 0.0884, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 3.4732142857142856, | |
| "grad_norm": 1.534104585647583, | |
| "learning_rate": 3.84200399931717e-06, | |
| "loss": 0.0983, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 3.4776785714285716, | |
| "grad_norm": 1.4144586324691772, | |
| "learning_rate": 3.8305193507510905e-06, | |
| "loss": 0.0954, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 3.482142857142857, | |
| "grad_norm": 1.5498822927474976, | |
| "learning_rate": 3.819041230106768e-06, | |
| "loss": 0.1033, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 3.486607142857143, | |
| "grad_norm": 1.2330667972564697, | |
| "learning_rate": 3.807569701426519e-06, | |
| "loss": 0.0825, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 3.4910714285714284, | |
| "grad_norm": 1.3459640741348267, | |
| "learning_rate": 3.7961048287158865e-06, | |
| "loss": 0.0847, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 3.4955357142857144, | |
| "grad_norm": 1.4080768823623657, | |
| "learning_rate": 3.784646675943269e-06, | |
| "loss": 0.0901, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 1.5240070819854736, | |
| "learning_rate": 3.773195307039574e-06, | |
| "loss": 0.0994, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 3.5044642857142856, | |
| "grad_norm": 1.3279098272323608, | |
| "learning_rate": 3.7617507858978615e-06, | |
| "loss": 0.089, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 3.508928571428571, | |
| "grad_norm": 1.2906996011734009, | |
| "learning_rate": 3.7503131763729785e-06, | |
| "loss": 0.0892, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 3.513392857142857, | |
| "grad_norm": 1.472264051437378, | |
| "learning_rate": 3.738882542281212e-06, | |
| "loss": 0.0889, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 3.517857142857143, | |
| "grad_norm": 1.4592229127883911, | |
| "learning_rate": 3.727458947399927e-06, | |
| "loss": 0.0923, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 3.522321428571429, | |
| "grad_norm": 1.5087579488754272, | |
| "learning_rate": 3.7160424554672187e-06, | |
| "loss": 0.0954, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 3.5267857142857144, | |
| "grad_norm": 1.4825618267059326, | |
| "learning_rate": 3.7046331301815435e-06, | |
| "loss": 0.0937, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 3.53125, | |
| "grad_norm": 1.3035829067230225, | |
| "learning_rate": 3.6932310352013796e-06, | |
| "loss": 0.0827, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 3.5357142857142856, | |
| "grad_norm": 1.418879508972168, | |
| "learning_rate": 3.6818362341448545e-06, | |
| "loss": 0.0957, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 3.540178571428571, | |
| "grad_norm": 1.6021920442581177, | |
| "learning_rate": 3.670448790589408e-06, | |
| "loss": 0.0964, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 3.544642857142857, | |
| "grad_norm": 1.3729946613311768, | |
| "learning_rate": 3.659068768071425e-06, | |
| "loss": 0.085, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 3.549107142857143, | |
| "grad_norm": 1.3656620979309082, | |
| "learning_rate": 3.6476962300858793e-06, | |
| "loss": 0.0946, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 3.553571428571429, | |
| "grad_norm": 1.2834999561309814, | |
| "learning_rate": 3.6363312400859963e-06, | |
| "loss": 0.084, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 3.5580357142857144, | |
| "grad_norm": 1.388552188873291, | |
| "learning_rate": 3.6249738614828765e-06, | |
| "loss": 0.0903, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 3.5625, | |
| "grad_norm": 1.4588837623596191, | |
| "learning_rate": 3.613624157645159e-06, | |
| "loss": 0.0961, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 3.5625, | |
| "eval_loss": 0.8479318618774414, | |
| "eval_runtime": 3.1035, | |
| "eval_samples_per_second": 19.333, | |
| "eval_steps_per_second": 1.289, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 3.5669642857142856, | |
| "grad_norm": 1.5527163743972778, | |
| "learning_rate": 3.6022821918986563e-06, | |
| "loss": 0.0891, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 1.3039300441741943, | |
| "learning_rate": 3.590948027526012e-06, | |
| "loss": 0.0961, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.575892857142857, | |
| "grad_norm": 1.2164970636367798, | |
| "learning_rate": 3.579621727766339e-06, | |
| "loss": 0.0763, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 3.580357142857143, | |
| "grad_norm": 1.3291783332824707, | |
| "learning_rate": 3.568303355814867e-06, | |
| "loss": 0.0914, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 3.584821428571429, | |
| "grad_norm": 1.267117977142334, | |
| "learning_rate": 3.5569929748225945e-06, | |
| "loss": 0.086, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 3.5892857142857144, | |
| "grad_norm": 1.430953860282898, | |
| "learning_rate": 3.5456906478959367e-06, | |
| "loss": 0.1026, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 3.59375, | |
| "grad_norm": 1.5599972009658813, | |
| "learning_rate": 3.534396438096369e-06, | |
| "loss": 0.0978, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 3.5982142857142856, | |
| "grad_norm": 1.3565053939819336, | |
| "learning_rate": 3.5231104084400745e-06, | |
| "loss": 0.0893, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 3.602678571428571, | |
| "grad_norm": 1.3777987957000732, | |
| "learning_rate": 3.5118326218976013e-06, | |
| "loss": 0.091, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 3.607142857142857, | |
| "grad_norm": 1.642444133758545, | |
| "learning_rate": 3.5005631413935006e-06, | |
| "loss": 0.0943, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 3.611607142857143, | |
| "grad_norm": 1.4887057542800903, | |
| "learning_rate": 3.4893020298059784e-06, | |
| "loss": 0.0829, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 3.616071428571429, | |
| "grad_norm": 1.5320430994033813, | |
| "learning_rate": 3.4780493499665478e-06, | |
| "loss": 0.1012, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 3.6205357142857144, | |
| "grad_norm": 1.5760090351104736, | |
| "learning_rate": 3.4668051646596825e-06, | |
| "loss": 0.101, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 3.625, | |
| "grad_norm": 1.5013108253479004, | |
| "learning_rate": 3.455569536622451e-06, | |
| "loss": 0.0926, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 3.6294642857142856, | |
| "grad_norm": 1.4752159118652344, | |
| "learning_rate": 3.4443425285441847e-06, | |
| "loss": 0.0981, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 3.633928571428571, | |
| "grad_norm": 1.3213896751403809, | |
| "learning_rate": 3.433124203066113e-06, | |
| "loss": 0.092, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 3.638392857142857, | |
| "grad_norm": 1.4325315952301025, | |
| "learning_rate": 3.421914622781028e-06, | |
| "loss": 0.0994, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 3.642857142857143, | |
| "grad_norm": 1.538521647453308, | |
| "learning_rate": 3.4107138502329225e-06, | |
| "loss": 0.0989, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 3.647321428571429, | |
| "grad_norm": 1.4363764524459839, | |
| "learning_rate": 3.399521947916646e-06, | |
| "loss": 0.1054, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 3.6517857142857144, | |
| "grad_norm": 1.3556541204452515, | |
| "learning_rate": 3.3883389782775604e-06, | |
| "loss": 0.0883, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 3.65625, | |
| "grad_norm": 1.4465206861495972, | |
| "learning_rate": 3.377165003711185e-06, | |
| "loss": 0.0826, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 3.6607142857142856, | |
| "grad_norm": 1.4532499313354492, | |
| "learning_rate": 3.3660000865628523e-06, | |
| "loss": 0.0808, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.665178571428571, | |
| "grad_norm": 1.5245791673660278, | |
| "learning_rate": 3.3548442891273553e-06, | |
| "loss": 0.0914, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 3.669642857142857, | |
| "grad_norm": 1.1719828844070435, | |
| "learning_rate": 3.343697673648611e-06, | |
| "loss": 0.0849, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 3.674107142857143, | |
| "grad_norm": 1.5081385374069214, | |
| "learning_rate": 3.332560302319297e-06, | |
| "loss": 0.1125, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 3.678571428571429, | |
| "grad_norm": 1.331972599029541, | |
| "learning_rate": 3.321432237280518e-06, | |
| "loss": 0.0891, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 3.6830357142857144, | |
| "grad_norm": 1.2731318473815918, | |
| "learning_rate": 3.3103135406214506e-06, | |
| "loss": 0.0855, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 3.6875, | |
| "grad_norm": 1.3402756452560425, | |
| "learning_rate": 3.2992042743790055e-06, | |
| "loss": 0.0852, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 3.6919642857142856, | |
| "grad_norm": 1.49905264377594, | |
| "learning_rate": 3.2881045005374747e-06, | |
| "loss": 0.0901, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 3.696428571428571, | |
| "grad_norm": 1.3129043579101562, | |
| "learning_rate": 3.277014281028181e-06, | |
| "loss": 0.0823, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 3.700892857142857, | |
| "grad_norm": 1.5265233516693115, | |
| "learning_rate": 3.2659336777291497e-06, | |
| "loss": 0.1079, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 3.705357142857143, | |
| "grad_norm": 1.3291493654251099, | |
| "learning_rate": 3.254862752464743e-06, | |
| "loss": 0.0854, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 3.709821428571429, | |
| "grad_norm": 1.509037733078003, | |
| "learning_rate": 3.243801567005329e-06, | |
| "loss": 0.1013, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 3.7142857142857144, | |
| "grad_norm": 1.7585151195526123, | |
| "learning_rate": 3.232750183066928e-06, | |
| "loss": 0.0994, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 3.71875, | |
| "grad_norm": 1.5366458892822266, | |
| "learning_rate": 3.2217086623108796e-06, | |
| "loss": 0.1006, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 3.7232142857142856, | |
| "grad_norm": 1.3573994636535645, | |
| "learning_rate": 3.2106770663434867e-06, | |
| "loss": 0.0855, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 3.727678571428571, | |
| "grad_norm": 1.3388910293579102, | |
| "learning_rate": 3.1996554567156774e-06, | |
| "loss": 0.0859, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 3.732142857142857, | |
| "grad_norm": 1.2037252187728882, | |
| "learning_rate": 3.18864389492266e-06, | |
| "loss": 0.0857, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 3.736607142857143, | |
| "grad_norm": 1.4536150693893433, | |
| "learning_rate": 3.1776424424035857e-06, | |
| "loss": 0.0983, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 3.741071428571429, | |
| "grad_norm": 1.406284213066101, | |
| "learning_rate": 3.1666511605411947e-06, | |
| "loss": 0.0899, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 3.7455357142857144, | |
| "grad_norm": 1.4425653219223022, | |
| "learning_rate": 3.155670110661482e-06, | |
| "loss": 0.0889, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 1.3152062892913818, | |
| "learning_rate": 3.144699354033356e-06, | |
| "loss": 0.0859, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 3.7544642857142856, | |
| "grad_norm": 1.456082820892334, | |
| "learning_rate": 3.1337389518682894e-06, | |
| "loss": 0.0937, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 3.758928571428571, | |
| "grad_norm": 1.3939170837402344, | |
| "learning_rate": 3.122788965319985e-06, | |
| "loss": 0.0802, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 3.763392857142857, | |
| "grad_norm": 1.2668402194976807, | |
| "learning_rate": 3.1118494554840284e-06, | |
| "loss": 0.0849, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 3.767857142857143, | |
| "grad_norm": 1.369589924812317, | |
| "learning_rate": 3.100920483397556e-06, | |
| "loss": 0.076, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 3.772321428571429, | |
| "grad_norm": 1.4516035318374634, | |
| "learning_rate": 3.090002110038903e-06, | |
| "loss": 0.0892, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 3.7767857142857144, | |
| "grad_norm": 1.5191524028778076, | |
| "learning_rate": 3.0790943963272697e-06, | |
| "loss": 0.0962, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 3.78125, | |
| "grad_norm": 1.4074844121932983, | |
| "learning_rate": 3.0681974031223854e-06, | |
| "loss": 0.0942, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 3.7857142857142856, | |
| "grad_norm": 1.3058712482452393, | |
| "learning_rate": 3.0573111912241575e-06, | |
| "loss": 0.085, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 3.790178571428571, | |
| "grad_norm": 1.3393832445144653, | |
| "learning_rate": 3.0464358213723455e-06, | |
| "loss": 0.0894, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 3.794642857142857, | |
| "grad_norm": 1.4606190919876099, | |
| "learning_rate": 3.0355713542462086e-06, | |
| "loss": 0.1035, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.799107142857143, | |
| "grad_norm": 1.4510114192962646, | |
| "learning_rate": 3.024717850464181e-06, | |
| "loss": 0.0965, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 3.803571428571429, | |
| "grad_norm": 1.7563194036483765, | |
| "learning_rate": 3.0138753705835234e-06, | |
| "loss": 0.0993, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 3.8080357142857144, | |
| "grad_norm": 1.5832990407943726, | |
| "learning_rate": 3.003043975099988e-06, | |
| "loss": 0.094, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 3.8125, | |
| "grad_norm": 1.3764246702194214, | |
| "learning_rate": 2.99222372444748e-06, | |
| "loss": 0.0797, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 3.8169642857142856, | |
| "grad_norm": 1.3575708866119385, | |
| "learning_rate": 2.9814146789977278e-06, | |
| "loss": 0.0884, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 3.8169642857142856, | |
| "eval_loss": 0.8522133231163025, | |
| "eval_runtime": 3.118, | |
| "eval_samples_per_second": 19.243, | |
| "eval_steps_per_second": 1.283, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 3.821428571428571, | |
| "grad_norm": 1.6227210760116577, | |
| "learning_rate": 2.970616899059934e-06, | |
| "loss": 0.0896, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 3.825892857142857, | |
| "grad_norm": 1.3933652639389038, | |
| "learning_rate": 2.959830444880447e-06, | |
| "loss": 0.089, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 3.830357142857143, | |
| "grad_norm": 1.4572827816009521, | |
| "learning_rate": 2.949055376642428e-06, | |
| "loss": 0.0906, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 3.834821428571429, | |
| "grad_norm": 1.3022706508636475, | |
| "learning_rate": 2.9382917544655025e-06, | |
| "loss": 0.0836, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 3.8392857142857144, | |
| "grad_norm": 1.3461490869522095, | |
| "learning_rate": 2.9275396384054373e-06, | |
| "loss": 0.0843, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 3.84375, | |
| "grad_norm": 1.421582579612732, | |
| "learning_rate": 2.9167990884537943e-06, | |
| "loss": 0.0972, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 3.8482142857142856, | |
| "grad_norm": 1.499457597732544, | |
| "learning_rate": 2.906070164537616e-06, | |
| "loss": 0.1037, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 3.852678571428571, | |
| "grad_norm": 1.422713279724121, | |
| "learning_rate": 2.8953529265190618e-06, | |
| "loss": 0.0898, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 3.857142857142857, | |
| "grad_norm": 1.5362988710403442, | |
| "learning_rate": 2.884647434195099e-06, | |
| "loss": 0.0992, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 3.861607142857143, | |
| "grad_norm": 1.393685221672058, | |
| "learning_rate": 2.873953747297153e-06, | |
| "loss": 0.0786, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 3.866071428571429, | |
| "grad_norm": 1.4596489667892456, | |
| "learning_rate": 2.863271925490791e-06, | |
| "loss": 0.1016, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 3.8705357142857144, | |
| "grad_norm": 1.3821113109588623, | |
| "learning_rate": 2.8526020283753658e-06, | |
| "loss": 0.0918, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 3.875, | |
| "grad_norm": 1.502488613128662, | |
| "learning_rate": 2.841944115483703e-06, | |
| "loss": 0.0933, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 3.8794642857142856, | |
| "grad_norm": 1.3570479154586792, | |
| "learning_rate": 2.8312982462817686e-06, | |
| "loss": 0.0929, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 3.883928571428571, | |
| "grad_norm": 1.5335980653762817, | |
| "learning_rate": 2.820664480168317e-06, | |
| "loss": 0.1003, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 3.888392857142857, | |
| "grad_norm": 1.459232211112976, | |
| "learning_rate": 2.810042876474584e-06, | |
| "loss": 0.1046, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 3.892857142857143, | |
| "grad_norm": 1.4594147205352783, | |
| "learning_rate": 2.799433494463935e-06, | |
| "loss": 0.1012, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 3.897321428571429, | |
| "grad_norm": 1.5626919269561768, | |
| "learning_rate": 2.7888363933315593e-06, | |
| "loss": 0.1023, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 3.9017857142857144, | |
| "grad_norm": 1.3069941997528076, | |
| "learning_rate": 2.7782516322041087e-06, | |
| "loss": 0.0817, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 3.90625, | |
| "grad_norm": 1.4590051174163818, | |
| "learning_rate": 2.767679270139394e-06, | |
| "loss": 0.0885, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 3.9107142857142856, | |
| "grad_norm": 1.47640860080719, | |
| "learning_rate": 2.7571193661260427e-06, | |
| "loss": 0.092, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 3.915178571428571, | |
| "grad_norm": 1.5121464729309082, | |
| "learning_rate": 2.746571979083172e-06, | |
| "loss": 0.0954, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 3.919642857142857, | |
| "grad_norm": 1.3360768556594849, | |
| "learning_rate": 2.736037167860061e-06, | |
| "loss": 0.0983, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 3.924107142857143, | |
| "grad_norm": 1.395384669303894, | |
| "learning_rate": 2.725514991235818e-06, | |
| "loss": 0.0953, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 3.928571428571429, | |
| "grad_norm": 1.3785486221313477, | |
| "learning_rate": 2.7150055079190663e-06, | |
| "loss": 0.0825, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.9330357142857144, | |
| "grad_norm": 1.4748543500900269, | |
| "learning_rate": 2.704508776547595e-06, | |
| "loss": 0.0903, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 3.9375, | |
| "grad_norm": 1.3787004947662354, | |
| "learning_rate": 2.6940248556880512e-06, | |
| "loss": 0.0912, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 3.9419642857142856, | |
| "grad_norm": 1.3161265850067139, | |
| "learning_rate": 2.6835538038356017e-06, | |
| "loss": 0.08, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 3.946428571428571, | |
| "grad_norm": 1.4215610027313232, | |
| "learning_rate": 2.6730956794136138e-06, | |
| "loss": 0.088, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 3.950892857142857, | |
| "grad_norm": 1.396683692932129, | |
| "learning_rate": 2.6626505407733255e-06, | |
| "loss": 0.0925, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 3.955357142857143, | |
| "grad_norm": 1.4125168323516846, | |
| "learning_rate": 2.6522184461935153e-06, | |
| "loss": 0.0888, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 3.959821428571429, | |
| "grad_norm": 1.6444463729858398, | |
| "learning_rate": 2.6417994538801882e-06, | |
| "loss": 0.1116, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 3.9642857142857144, | |
| "grad_norm": 1.562724232673645, | |
| "learning_rate": 2.6313936219662435e-06, | |
| "loss": 0.0936, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 3.96875, | |
| "grad_norm": 1.3996388912200928, | |
| "learning_rate": 2.6210010085111507e-06, | |
| "loss": 0.0949, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 3.9732142857142856, | |
| "grad_norm": 1.3662428855895996, | |
| "learning_rate": 2.6106216715006282e-06, | |
| "loss": 0.0883, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.977678571428571, | |
| "grad_norm": 1.5796185731887817, | |
| "learning_rate": 2.600255668846316e-06, | |
| "loss": 0.0824, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 3.982142857142857, | |
| "grad_norm": 1.5451363325119019, | |
| "learning_rate": 2.58990305838546e-06, | |
| "loss": 0.1012, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 3.986607142857143, | |
| "grad_norm": 1.4925421476364136, | |
| "learning_rate": 2.5795638978805755e-06, | |
| "loss": 0.0859, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 3.991071428571429, | |
| "grad_norm": 1.4481643438339233, | |
| "learning_rate": 2.5692382450191404e-06, | |
| "loss": 0.1046, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 3.9955357142857144, | |
| "grad_norm": 1.4745389223098755, | |
| "learning_rate": 2.558926157413266e-06, | |
| "loss": 0.1084, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.4540438652038574, | |
| "learning_rate": 2.5486276925993746e-06, | |
| "loss": 0.0899, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 4.004464285714286, | |
| "grad_norm": 0.8183535933494568, | |
| "learning_rate": 2.5383429080378807e-06, | |
| "loss": 0.0459, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 4.008928571428571, | |
| "grad_norm": 0.8469629287719727, | |
| "learning_rate": 2.5280718611128657e-06, | |
| "loss": 0.0473, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 4.013392857142857, | |
| "grad_norm": 0.7638592720031738, | |
| "learning_rate": 2.517814609131772e-06, | |
| "loss": 0.0451, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 4.017857142857143, | |
| "grad_norm": 0.8340522646903992, | |
| "learning_rate": 2.507571209325058e-06, | |
| "loss": 0.0452, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.022321428571429, | |
| "grad_norm": 0.9095687866210938, | |
| "learning_rate": 2.4973417188459074e-06, | |
| "loss": 0.0472, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 4.026785714285714, | |
| "grad_norm": 0.8626049160957336, | |
| "learning_rate": 2.4871261947698892e-06, | |
| "loss": 0.0485, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 4.03125, | |
| "grad_norm": 1.161464810371399, | |
| "learning_rate": 2.4769246940946487e-06, | |
| "loss": 0.0539, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 4.035714285714286, | |
| "grad_norm": 0.7767446637153625, | |
| "learning_rate": 2.4667372737395894e-06, | |
| "loss": 0.0431, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 4.040178571428571, | |
| "grad_norm": 0.8548541069030762, | |
| "learning_rate": 2.4565639905455455e-06, | |
| "loss": 0.0429, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 4.044642857142857, | |
| "grad_norm": 1.0133311748504639, | |
| "learning_rate": 2.446404901274486e-06, | |
| "loss": 0.0457, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 4.049107142857143, | |
| "grad_norm": 0.8490503430366516, | |
| "learning_rate": 2.436260062609173e-06, | |
| "loss": 0.0424, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 4.053571428571429, | |
| "grad_norm": 0.9454727172851562, | |
| "learning_rate": 2.4261295311528632e-06, | |
| "loss": 0.038, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 4.058035714285714, | |
| "grad_norm": 0.9559707641601562, | |
| "learning_rate": 2.4160133634289804e-06, | |
| "loss": 0.0424, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 4.0625, | |
| "grad_norm": 1.1564035415649414, | |
| "learning_rate": 2.4059116158808147e-06, | |
| "loss": 0.0504, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 4.066964285714286, | |
| "grad_norm": 1.1515920162200928, | |
| "learning_rate": 2.395824344871193e-06, | |
| "loss": 0.0468, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 4.071428571428571, | |
| "grad_norm": 1.1842188835144043, | |
| "learning_rate": 2.385751606682167e-06, | |
| "loss": 0.0474, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 4.071428571428571, | |
| "eval_loss": 1.0089935064315796, | |
| "eval_runtime": 3.4591, | |
| "eval_samples_per_second": 17.346, | |
| "eval_steps_per_second": 1.156, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 4.075892857142857, | |
| "grad_norm": 0.9655628800392151, | |
| "learning_rate": 2.3756934575147117e-06, | |
| "loss": 0.0399, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 4.080357142857143, | |
| "grad_norm": 1.0869725942611694, | |
| "learning_rate": 2.365649953488393e-06, | |
| "loss": 0.0383, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 4.084821428571429, | |
| "grad_norm": 1.3358078002929688, | |
| "learning_rate": 2.3556211506410708e-06, | |
| "loss": 0.0382, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 4.089285714285714, | |
| "grad_norm": 1.2573343515396118, | |
| "learning_rate": 2.3456071049285717e-06, | |
| "loss": 0.0459, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 4.09375, | |
| "grad_norm": 1.1886146068572998, | |
| "learning_rate": 2.3356078722243963e-06, | |
| "loss": 0.0407, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 4.098214285714286, | |
| "grad_norm": 1.2139469385147095, | |
| "learning_rate": 2.325623508319385e-06, | |
| "loss": 0.047, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 4.102678571428571, | |
| "grad_norm": 0.9977007508277893, | |
| "learning_rate": 2.3156540689214227e-06, | |
| "loss": 0.0339, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 4.107142857142857, | |
| "grad_norm": 1.0386406183242798, | |
| "learning_rate": 2.3056996096551228e-06, | |
| "loss": 0.0383, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 4.111607142857143, | |
| "grad_norm": 1.2945815324783325, | |
| "learning_rate": 2.2957601860615152e-06, | |
| "loss": 0.0442, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 4.116071428571429, | |
| "grad_norm": 1.2519358396530151, | |
| "learning_rate": 2.285835853597742e-06, | |
| "loss": 0.0429, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 4.120535714285714, | |
| "grad_norm": 1.4103801250457764, | |
| "learning_rate": 2.2759266676367345e-06, | |
| "loss": 0.0538, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 4.125, | |
| "grad_norm": 1.0328574180603027, | |
| "learning_rate": 2.266032683466928e-06, | |
| "loss": 0.0382, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 4.129464285714286, | |
| "grad_norm": 1.3296153545379639, | |
| "learning_rate": 2.2561539562919265e-06, | |
| "loss": 0.0408, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 4.133928571428571, | |
| "grad_norm": 1.1139670610427856, | |
| "learning_rate": 2.246290541230214e-06, | |
| "loss": 0.0433, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 4.138392857142857, | |
| "grad_norm": 1.2616791725158691, | |
| "learning_rate": 2.236442493314839e-06, | |
| "loss": 0.0412, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 4.142857142857143, | |
| "grad_norm": 1.1220238208770752, | |
| "learning_rate": 2.2266098674931094e-06, | |
| "loss": 0.0414, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 4.147321428571429, | |
| "grad_norm": 1.056321620941162, | |
| "learning_rate": 2.216792718626286e-06, | |
| "loss": 0.0457, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 4.151785714285714, | |
| "grad_norm": 1.0423365831375122, | |
| "learning_rate": 2.2069911014892712e-06, | |
| "loss": 0.0386, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 4.15625, | |
| "grad_norm": 0.9723503589630127, | |
| "learning_rate": 2.197205070770313e-06, | |
| "loss": 0.0316, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 4.160714285714286, | |
| "grad_norm": 1.3372763395309448, | |
| "learning_rate": 2.1874346810706925e-06, | |
| "loss": 0.043, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 4.165178571428571, | |
| "grad_norm": 0.9659892916679382, | |
| "learning_rate": 2.177679986904422e-06, | |
| "loss": 0.0361, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 4.169642857142857, | |
| "grad_norm": 1.2238346338272095, | |
| "learning_rate": 2.1679410426979412e-06, | |
| "loss": 0.0546, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 4.174107142857143, | |
| "grad_norm": 1.1462628841400146, | |
| "learning_rate": 2.1582179027898102e-06, | |
| "loss": 0.0485, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 4.178571428571429, | |
| "grad_norm": 1.2392868995666504, | |
| "learning_rate": 2.148510621430414e-06, | |
| "loss": 0.0453, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 4.183035714285714, | |
| "grad_norm": 0.8607609272003174, | |
| "learning_rate": 2.1388192527816472e-06, | |
| "loss": 0.032, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 4.1875, | |
| "grad_norm": 1.1780694723129272, | |
| "learning_rate": 2.1291438509166236e-06, | |
| "loss": 0.0465, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 4.191964285714286, | |
| "grad_norm": 1.0053000450134277, | |
| "learning_rate": 2.119484469819371e-06, | |
| "loss": 0.0456, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 4.196428571428571, | |
| "grad_norm": 0.9943235516548157, | |
| "learning_rate": 2.109841163384528e-06, | |
| "loss": 0.0408, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 4.200892857142857, | |
| "grad_norm": 1.1572273969650269, | |
| "learning_rate": 2.100213985417045e-06, | |
| "loss": 0.041, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 4.205357142857143, | |
| "grad_norm": 0.9679527878761292, | |
| "learning_rate": 2.090602989631878e-06, | |
| "loss": 0.0394, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 4.209821428571429, | |
| "grad_norm": 0.9852113723754883, | |
| "learning_rate": 2.081008229653706e-06, | |
| "loss": 0.041, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 4.214285714285714, | |
| "grad_norm": 1.0761950016021729, | |
| "learning_rate": 2.071429759016607e-06, | |
| "loss": 0.0426, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 4.21875, | |
| "grad_norm": 1.483879804611206, | |
| "learning_rate": 2.061867631163781e-06, | |
| "loss": 0.0506, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 4.223214285714286, | |
| "grad_norm": 1.103798270225525, | |
| "learning_rate": 2.0523218994472408e-06, | |
| "loss": 0.0352, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 4.227678571428571, | |
| "grad_norm": 1.1449997425079346, | |
| "learning_rate": 2.0427926171275157e-06, | |
| "loss": 0.0416, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 4.232142857142857, | |
| "grad_norm": 1.024032711982727, | |
| "learning_rate": 2.033279837373359e-06, | |
| "loss": 0.0421, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 4.236607142857143, | |
| "grad_norm": 0.9040204882621765, | |
| "learning_rate": 2.023783613261439e-06, | |
| "loss": 0.0357, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 4.241071428571429, | |
| "grad_norm": 1.1004376411437988, | |
| "learning_rate": 2.0143039977760663e-06, | |
| "loss": 0.0468, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 4.245535714285714, | |
| "grad_norm": 1.3727236986160278, | |
| "learning_rate": 2.0048410438088675e-06, | |
| "loss": 0.058, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "grad_norm": 1.0734888315200806, | |
| "learning_rate": 1.995394804158516e-06, | |
| "loss": 0.045, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 4.254464285714286, | |
| "grad_norm": 1.0533781051635742, | |
| "learning_rate": 1.9859653315304254e-06, | |
| "loss": 0.0432, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 4.258928571428571, | |
| "grad_norm": 0.9554777145385742, | |
| "learning_rate": 1.976552678536456e-06, | |
| "loss": 0.0429, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 4.263392857142857, | |
| "grad_norm": 1.208379864692688, | |
| "learning_rate": 1.9671568976946257e-06, | |
| "loss": 0.0548, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 4.267857142857143, | |
| "grad_norm": 1.1880334615707397, | |
| "learning_rate": 1.9577780414288066e-06, | |
| "loss": 0.0449, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 4.272321428571429, | |
| "grad_norm": 1.1660935878753662, | |
| "learning_rate": 1.9484161620684524e-06, | |
| "loss": 0.0474, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 4.276785714285714, | |
| "grad_norm": 1.281361699104309, | |
| "learning_rate": 1.939071311848282e-06, | |
| "loss": 0.0468, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 4.28125, | |
| "grad_norm": 1.1080694198608398, | |
| "learning_rate": 1.9297435429080076e-06, | |
| "loss": 0.0491, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 0.9572265148162842, | |
| "learning_rate": 1.9204329072920285e-06, | |
| "loss": 0.035, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 4.290178571428571, | |
| "grad_norm": 1.2457669973373413, | |
| "learning_rate": 1.911139456949158e-06, | |
| "loss": 0.049, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 4.294642857142857, | |
| "grad_norm": 1.1637389659881592, | |
| "learning_rate": 1.9018632437323198e-06, | |
| "loss": 0.0481, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 4.299107142857143, | |
| "grad_norm": 1.0944490432739258, | |
| "learning_rate": 1.892604319398259e-06, | |
| "loss": 0.0516, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 4.303571428571429, | |
| "grad_norm": 1.129134178161621, | |
| "learning_rate": 1.883362735607262e-06, | |
| "loss": 0.0405, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 4.308035714285714, | |
| "grad_norm": 1.2214771509170532, | |
| "learning_rate": 1.8741385439228616e-06, | |
| "loss": 0.0455, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 4.3125, | |
| "grad_norm": 1.0837950706481934, | |
| "learning_rate": 1.8649317958115534e-06, | |
| "loss": 0.0445, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 4.316964285714286, | |
| "grad_norm": 1.048578143119812, | |
| "learning_rate": 1.8557425426424989e-06, | |
| "loss": 0.0488, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 4.321428571428571, | |
| "grad_norm": 0.9222402572631836, | |
| "learning_rate": 1.8465708356872592e-06, | |
| "loss": 0.0336, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 4.325892857142857, | |
| "grad_norm": 1.0440882444381714, | |
| "learning_rate": 1.8374167261194826e-06, | |
| "loss": 0.046, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 4.325892857142857, | |
| "eval_loss": 1.0081344842910767, | |
| "eval_runtime": 3.1161, | |
| "eval_samples_per_second": 19.255, | |
| "eval_steps_per_second": 1.284, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 4.330357142857143, | |
| "grad_norm": 1.0037401914596558, | |
| "learning_rate": 1.8282802650146408e-06, | |
| "loss": 0.0423, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 4.334821428571429, | |
| "grad_norm": 1.0888147354125977, | |
| "learning_rate": 1.8191615033497345e-06, | |
| "loss": 0.0426, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 4.339285714285714, | |
| "grad_norm": 1.0715805292129517, | |
| "learning_rate": 1.810060492003008e-06, | |
| "loss": 0.0444, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 4.34375, | |
| "grad_norm": 1.0277068614959717, | |
| "learning_rate": 1.800977281753671e-06, | |
| "loss": 0.0428, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 4.348214285714286, | |
| "grad_norm": 0.999724268913269, | |
| "learning_rate": 1.7919119232816049e-06, | |
| "loss": 0.0387, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 4.352678571428571, | |
| "grad_norm": 0.9766676425933838, | |
| "learning_rate": 1.7828644671670943e-06, | |
| "loss": 0.0393, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 4.357142857142857, | |
| "grad_norm": 1.0714170932769775, | |
| "learning_rate": 1.773834963890534e-06, | |
| "loss": 0.0458, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 4.361607142857143, | |
| "grad_norm": 0.8471158742904663, | |
| "learning_rate": 1.764823463832151e-06, | |
| "loss": 0.0364, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 4.366071428571429, | |
| "grad_norm": 1.0079231262207031, | |
| "learning_rate": 1.7558300172717234e-06, | |
| "loss": 0.0364, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 4.370535714285714, | |
| "grad_norm": 1.1584757566452026, | |
| "learning_rate": 1.7468546743882997e-06, | |
| "loss": 0.0442, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 4.375, | |
| "grad_norm": 1.0798187255859375, | |
| "learning_rate": 1.7378974852599203e-06, | |
| "loss": 0.0366, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 4.379464285714286, | |
| "grad_norm": 1.2951469421386719, | |
| "learning_rate": 1.7289584998633307e-06, | |
| "loss": 0.0501, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 4.383928571428571, | |
| "grad_norm": 1.370492696762085, | |
| "learning_rate": 1.7200377680737148e-06, | |
| "loss": 0.0511, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 4.388392857142857, | |
| "grad_norm": 1.24025297164917, | |
| "learning_rate": 1.7111353396644071e-06, | |
| "loss": 0.0433, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 4.392857142857143, | |
| "grad_norm": 1.1378741264343262, | |
| "learning_rate": 1.7022512643066196e-06, | |
| "loss": 0.0417, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 4.397321428571429, | |
| "grad_norm": 1.17164146900177, | |
| "learning_rate": 1.6933855915691622e-06, | |
| "loss": 0.0464, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 4.401785714285714, | |
| "grad_norm": 1.1803674697875977, | |
| "learning_rate": 1.6845383709181676e-06, | |
| "loss": 0.0404, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 4.40625, | |
| "grad_norm": 1.1585767269134521, | |
| "learning_rate": 1.675709651716817e-06, | |
| "loss": 0.0417, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 4.410714285714286, | |
| "grad_norm": 1.1010210514068604, | |
| "learning_rate": 1.6668994832250556e-06, | |
| "loss": 0.0429, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 4.415178571428571, | |
| "grad_norm": 1.0791648626327515, | |
| "learning_rate": 1.6581079145993323e-06, | |
| "loss": 0.0457, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 4.419642857142857, | |
| "grad_norm": 1.114737868309021, | |
| "learning_rate": 1.649334994892314e-06, | |
| "loss": 0.0502, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 4.424107142857143, | |
| "grad_norm": 1.1722322702407837, | |
| "learning_rate": 1.640580773052618e-06, | |
| "loss": 0.0444, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 4.428571428571429, | |
| "grad_norm": 1.1338437795639038, | |
| "learning_rate": 1.6318452979245355e-06, | |
| "loss": 0.0462, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 4.433035714285714, | |
| "grad_norm": 1.1225347518920898, | |
| "learning_rate": 1.6231286182477555e-06, | |
| "loss": 0.049, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 4.4375, | |
| "grad_norm": 1.1152468919754028, | |
| "learning_rate": 1.6144307826571085e-06, | |
| "loss": 0.0474, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 4.441964285714286, | |
| "grad_norm": 0.970764696598053, | |
| "learning_rate": 1.6057518396822724e-06, | |
| "loss": 0.0474, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 4.446428571428571, | |
| "grad_norm": 1.2339911460876465, | |
| "learning_rate": 1.5970918377475208e-06, | |
| "loss": 0.0497, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 4.450892857142857, | |
| "grad_norm": 1.1019359827041626, | |
| "learning_rate": 1.5884508251714436e-06, | |
| "loss": 0.0419, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 4.455357142857143, | |
| "grad_norm": 1.013564109802246, | |
| "learning_rate": 1.5798288501666793e-06, | |
| "loss": 0.0405, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 4.459821428571429, | |
| "grad_norm": 1.2404049634933472, | |
| "learning_rate": 1.5712259608396462e-06, | |
| "loss": 0.0437, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 4.464285714285714, | |
| "grad_norm": 0.8601024746894836, | |
| "learning_rate": 1.5626422051902709e-06, | |
| "loss": 0.0374, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.46875, | |
| "grad_norm": 1.117348074913025, | |
| "learning_rate": 1.5540776311117304e-06, | |
| "loss": 0.042, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 4.473214285714286, | |
| "grad_norm": 1.1788359880447388, | |
| "learning_rate": 1.5455322863901704e-06, | |
| "loss": 0.0436, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 4.477678571428571, | |
| "grad_norm": 1.2720320224761963, | |
| "learning_rate": 1.5370062187044502e-06, | |
| "loss": 0.0491, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 4.482142857142857, | |
| "grad_norm": 1.1377935409545898, | |
| "learning_rate": 1.5284994756258718e-06, | |
| "loss": 0.0447, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 4.486607142857143, | |
| "grad_norm": 1.1015208959579468, | |
| "learning_rate": 1.5200121046179151e-06, | |
| "loss": 0.0441, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 4.491071428571429, | |
| "grad_norm": 1.0011322498321533, | |
| "learning_rate": 1.511544153035975e-06, | |
| "loss": 0.0407, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 4.495535714285714, | |
| "grad_norm": 1.0262796878814697, | |
| "learning_rate": 1.5030956681270903e-06, | |
| "loss": 0.0368, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 0.9674072861671448, | |
| "learning_rate": 1.4946666970296932e-06, | |
| "loss": 0.0366, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 4.504464285714286, | |
| "grad_norm": 1.2644139528274536, | |
| "learning_rate": 1.486257286773331e-06, | |
| "loss": 0.0472, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 4.508928571428571, | |
| "grad_norm": 1.3127970695495605, | |
| "learning_rate": 1.4778674842784168e-06, | |
| "loss": 0.0517, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 4.513392857142857, | |
| "grad_norm": 0.9656462669372559, | |
| "learning_rate": 1.4694973363559539e-06, | |
| "loss": 0.0347, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 4.517857142857143, | |
| "grad_norm": 1.1587086915969849, | |
| "learning_rate": 1.4611468897072933e-06, | |
| "loss": 0.0403, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 4.522321428571429, | |
| "grad_norm": 1.2970781326293945, | |
| "learning_rate": 1.4528161909238569e-06, | |
| "loss": 0.0361, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 4.526785714285714, | |
| "grad_norm": 1.1997908353805542, | |
| "learning_rate": 1.44450528648688e-06, | |
| "loss": 0.0469, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 4.53125, | |
| "grad_norm": 1.2376347780227661, | |
| "learning_rate": 1.4362142227671607e-06, | |
| "loss": 0.0453, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 4.535714285714286, | |
| "grad_norm": 1.1609876155853271, | |
| "learning_rate": 1.427943046024793e-06, | |
| "loss": 0.0404, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 4.540178571428571, | |
| "grad_norm": 1.313492774963379, | |
| "learning_rate": 1.4196918024089133e-06, | |
| "loss": 0.0503, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 4.544642857142857, | |
| "grad_norm": 1.1651771068572998, | |
| "learning_rate": 1.4114605379574342e-06, | |
| "loss": 0.0426, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 4.549107142857143, | |
| "grad_norm": 1.1554261445999146, | |
| "learning_rate": 1.4032492985968057e-06, | |
| "loss": 0.0516, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 4.553571428571429, | |
| "grad_norm": 1.2204504013061523, | |
| "learning_rate": 1.3950581301417365e-06, | |
| "loss": 0.0426, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 4.558035714285714, | |
| "grad_norm": 1.2661011219024658, | |
| "learning_rate": 1.3868870782949565e-06, | |
| "loss": 0.0505, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 4.5625, | |
| "grad_norm": 0.979673445224762, | |
| "learning_rate": 1.3787361886469509e-06, | |
| "loss": 0.0422, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 4.566964285714286, | |
| "grad_norm": 1.139962911605835, | |
| "learning_rate": 1.3706055066757116e-06, | |
| "loss": 0.0474, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 4.571428571428571, | |
| "grad_norm": 1.2725175619125366, | |
| "learning_rate": 1.3624950777464828e-06, | |
| "loss": 0.0467, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 4.575892857142857, | |
| "grad_norm": 1.189316987991333, | |
| "learning_rate": 1.3544049471115017e-06, | |
| "loss": 0.0438, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 4.580357142857143, | |
| "grad_norm": 0.9822104573249817, | |
| "learning_rate": 1.3463351599097552e-06, | |
| "loss": 0.0386, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 4.580357142857143, | |
| "eval_loss": 1.028843641281128, | |
| "eval_runtime": 3.1044, | |
| "eval_samples_per_second": 19.327, | |
| "eval_steps_per_second": 1.288, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 4.584821428571429, | |
| "grad_norm": 1.140833854675293, | |
| "learning_rate": 1.3382857611667233e-06, | |
| "loss": 0.0433, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 4.589285714285714, | |
| "grad_norm": 0.9858164191246033, | |
| "learning_rate": 1.3302567957941265e-06, | |
| "loss": 0.0394, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 4.59375, | |
| "grad_norm": 1.0154980421066284, | |
| "learning_rate": 1.3222483085896786e-06, | |
| "loss": 0.0412, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 4.598214285714286, | |
| "grad_norm": 1.2253535985946655, | |
| "learning_rate": 1.3142603442368346e-06, | |
| "loss": 0.0508, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 4.602678571428571, | |
| "grad_norm": 1.0019500255584717, | |
| "learning_rate": 1.3062929473045442e-06, | |
| "loss": 0.041, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 4.607142857142857, | |
| "grad_norm": 1.0669821500778198, | |
| "learning_rate": 1.2983461622469953e-06, | |
| "loss": 0.0456, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 4.611607142857143, | |
| "grad_norm": 1.2003577947616577, | |
| "learning_rate": 1.290420033403377e-06, | |
| "loss": 0.0423, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 4.616071428571429, | |
| "grad_norm": 1.1979588270187378, | |
| "learning_rate": 1.2825146049976244e-06, | |
| "loss": 0.0426, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 4.620535714285714, | |
| "grad_norm": 1.0805418491363525, | |
| "learning_rate": 1.2746299211381755e-06, | |
| "loss": 0.0453, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 4.625, | |
| "grad_norm": 0.9814077615737915, | |
| "learning_rate": 1.2667660258177241e-06, | |
| "loss": 0.0398, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 4.629464285714286, | |
| "grad_norm": 0.8897956013679504, | |
| "learning_rate": 1.258922962912969e-06, | |
| "loss": 0.0334, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 4.633928571428571, | |
| "grad_norm": 1.0321533679962158, | |
| "learning_rate": 1.2511007761843839e-06, | |
| "loss": 0.0467, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 4.638392857142857, | |
| "grad_norm": 1.1807042360305786, | |
| "learning_rate": 1.2432995092759538e-06, | |
| "loss": 0.0459, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 4.642857142857143, | |
| "grad_norm": 1.181643009185791, | |
| "learning_rate": 1.2355192057149477e-06, | |
| "loss": 0.043, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 4.647321428571429, | |
| "grad_norm": 1.0699180364608765, | |
| "learning_rate": 1.2277599089116662e-06, | |
| "loss": 0.043, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 4.651785714285714, | |
| "grad_norm": 1.0686554908752441, | |
| "learning_rate": 1.220021662159204e-06, | |
| "loss": 0.0533, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 4.65625, | |
| "grad_norm": 1.014110803604126, | |
| "learning_rate": 1.2123045086332076e-06, | |
| "loss": 0.0417, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 4.660714285714286, | |
| "grad_norm": 1.0047358274459839, | |
| "learning_rate": 1.204608491391628e-06, | |
| "loss": 0.0355, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 4.665178571428571, | |
| "grad_norm": 0.8971364498138428, | |
| "learning_rate": 1.1969336533744965e-06, | |
| "loss": 0.0361, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 4.669642857142857, | |
| "grad_norm": 1.230128288269043, | |
| "learning_rate": 1.1892800374036632e-06, | |
| "loss": 0.0462, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 4.674107142857143, | |
| "grad_norm": 1.1491485834121704, | |
| "learning_rate": 1.181647686182576e-06, | |
| "loss": 0.0449, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 4.678571428571429, | |
| "grad_norm": 1.171077847480774, | |
| "learning_rate": 1.174036642296036e-06, | |
| "loss": 0.0382, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 4.683035714285714, | |
| "grad_norm": 1.138710856437683, | |
| "learning_rate": 1.166446948209957e-06, | |
| "loss": 0.0472, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 4.6875, | |
| "grad_norm": 0.9967400431632996, | |
| "learning_rate": 1.1588786462711347e-06, | |
| "loss": 0.0438, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 4.691964285714286, | |
| "grad_norm": 0.9416896104812622, | |
| "learning_rate": 1.1513317787070002e-06, | |
| "loss": 0.0326, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 4.696428571428571, | |
| "grad_norm": 0.9042790532112122, | |
| "learning_rate": 1.1438063876254025e-06, | |
| "loss": 0.0376, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 4.700892857142857, | |
| "grad_norm": 1.1666338443756104, | |
| "learning_rate": 1.1363025150143508e-06, | |
| "loss": 0.0467, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 4.705357142857143, | |
| "grad_norm": 1.1811097860336304, | |
| "learning_rate": 1.1288202027417996e-06, | |
| "loss": 0.0456, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 4.709821428571429, | |
| "grad_norm": 1.1350547075271606, | |
| "learning_rate": 1.1213594925554027e-06, | |
| "loss": 0.0482, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 4.714285714285714, | |
| "grad_norm": 1.0363948345184326, | |
| "learning_rate": 1.1139204260822874e-06, | |
| "loss": 0.0516, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 4.71875, | |
| "grad_norm": 1.008671760559082, | |
| "learning_rate": 1.1065030448288196e-06, | |
| "loss": 0.0436, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 4.723214285714286, | |
| "grad_norm": 1.1995580196380615, | |
| "learning_rate": 1.0991073901803692e-06, | |
| "loss": 0.0464, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 4.727678571428571, | |
| "grad_norm": 1.1144418716430664, | |
| "learning_rate": 1.091733503401085e-06, | |
| "loss": 0.0439, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 4.732142857142857, | |
| "grad_norm": 1.1510818004608154, | |
| "learning_rate": 1.0843814256336622e-06, | |
| "loss": 0.0443, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 4.736607142857143, | |
| "grad_norm": 0.9981248378753662, | |
| "learning_rate": 1.0770511978991116e-06, | |
| "loss": 0.0414, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 4.741071428571429, | |
| "grad_norm": 1.0740482807159424, | |
| "learning_rate": 1.0697428610965275e-06, | |
| "loss": 0.0475, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 4.745535714285714, | |
| "grad_norm": 1.2632689476013184, | |
| "learning_rate": 1.0624564560028723e-06, | |
| "loss": 0.0464, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 4.75, | |
| "grad_norm": 0.9502501487731934, | |
| "learning_rate": 1.0551920232727309e-06, | |
| "loss": 0.0399, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 4.754464285714286, | |
| "grad_norm": 1.2378613948822021, | |
| "learning_rate": 1.0479496034380988e-06, | |
| "loss": 0.0504, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 4.758928571428571, | |
| "grad_norm": 1.0768165588378906, | |
| "learning_rate": 1.0407292369081479e-06, | |
| "loss": 0.0459, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 4.763392857142857, | |
| "grad_norm": 1.039918065071106, | |
| "learning_rate": 1.0335309639690041e-06, | |
| "loss": 0.0452, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 4.767857142857143, | |
| "grad_norm": 0.9375714659690857, | |
| "learning_rate": 1.0263548247835246e-06, | |
| "loss": 0.0419, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 4.772321428571429, | |
| "grad_norm": 0.9457955360412598, | |
| "learning_rate": 1.0192008593910643e-06, | |
| "loss": 0.0444, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 4.776785714285714, | |
| "grad_norm": 1.225278377532959, | |
| "learning_rate": 1.0120691077072643e-06, | |
| "loss": 0.0433, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 4.78125, | |
| "grad_norm": 1.281240463256836, | |
| "learning_rate": 1.004959609523824e-06, | |
| "loss": 0.0497, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 4.785714285714286, | |
| "grad_norm": 1.1898356676101685, | |
| "learning_rate": 9.978724045082772e-07, | |
| "loss": 0.0401, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 4.790178571428571, | |
| "grad_norm": 1.2565916776657104, | |
| "learning_rate": 9.908075322037738e-07, | |
| "loss": 0.0442, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 4.794642857142857, | |
| "grad_norm": 0.9693619012832642, | |
| "learning_rate": 9.83765032028858e-07, | |
| "loss": 0.0408, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 4.799107142857143, | |
| "grad_norm": 1.2235596179962158, | |
| "learning_rate": 9.767449432772485e-07, | |
| "loss": 0.0423, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 4.803571428571429, | |
| "grad_norm": 1.3905706405639648, | |
| "learning_rate": 9.697473051176173e-07, | |
| "loss": 0.0506, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 4.808035714285714, | |
| "grad_norm": 1.111341953277588, | |
| "learning_rate": 9.627721565933749e-07, | |
| "loss": 0.0409, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 4.8125, | |
| "grad_norm": 1.2336980104446411, | |
| "learning_rate": 9.558195366224508e-07, | |
| "loss": 0.0487, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 4.816964285714286, | |
| "grad_norm": 1.066314458847046, | |
| "learning_rate": 9.488894839970758e-07, | |
| "loss": 0.04, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 4.821428571428571, | |
| "grad_norm": 0.9892422556877136, | |
| "learning_rate": 9.419820373835668e-07, | |
| "loss": 0.0417, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 4.825892857142857, | |
| "grad_norm": 1.1529948711395264, | |
| "learning_rate": 9.350972353221052e-07, | |
| "loss": 0.0433, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 4.830357142857143, | |
| "grad_norm": 1.2176202535629272, | |
| "learning_rate": 9.282351162265363e-07, | |
| "loss": 0.0456, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 4.834821428571429, | |
| "grad_norm": 1.2151131629943848, | |
| "learning_rate": 9.213957183841355e-07, | |
| "loss": 0.0425, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 4.834821428571429, | |
| "eval_loss": 1.0552717447280884, | |
| "eval_runtime": 3.0996, | |
| "eval_samples_per_second": 19.357, | |
| "eval_steps_per_second": 1.29, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 4.839285714285714, | |
| "grad_norm": 1.2349426746368408, | |
| "learning_rate": 9.145790799554101e-07, | |
| "loss": 0.0467, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 4.84375, | |
| "grad_norm": 1.0061850547790527, | |
| "learning_rate": 9.077852389738817e-07, | |
| "loss": 0.0407, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 4.848214285714286, | |
| "grad_norm": 1.1742695569992065, | |
| "learning_rate": 9.010142333458698e-07, | |
| "loss": 0.0371, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 4.852678571428571, | |
| "grad_norm": 1.044801950454712, | |
| "learning_rate": 8.942661008502875e-07, | |
| "loss": 0.0409, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 4.857142857142857, | |
| "grad_norm": 1.2313170433044434, | |
| "learning_rate": 8.87540879138421e-07, | |
| "loss": 0.0474, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 4.861607142857143, | |
| "grad_norm": 1.2384284734725952, | |
| "learning_rate": 8.808386057337353e-07, | |
| "loss": 0.0502, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 4.866071428571429, | |
| "grad_norm": 1.2501604557037354, | |
| "learning_rate": 8.741593180316439e-07, | |
| "loss": 0.046, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 4.870535714285714, | |
| "grad_norm": 1.151362419128418, | |
| "learning_rate": 8.675030532993193e-07, | |
| "loss": 0.0411, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 4.875, | |
| "grad_norm": 1.1054540872573853, | |
| "learning_rate": 8.608698486754739e-07, | |
| "loss": 0.0453, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 4.879464285714286, | |
| "grad_norm": 1.1711599826812744, | |
| "learning_rate": 8.542597411701563e-07, | |
| "loss": 0.043, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 4.883928571428571, | |
| "grad_norm": 1.048905372619629, | |
| "learning_rate": 8.476727676645453e-07, | |
| "loss": 0.0417, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 4.888392857142857, | |
| "grad_norm": 0.9673648476600647, | |
| "learning_rate": 8.411089649107396e-07, | |
| "loss": 0.0388, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 4.892857142857143, | |
| "grad_norm": 1.09621000289917, | |
| "learning_rate": 8.345683695315633e-07, | |
| "loss": 0.0392, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 4.897321428571429, | |
| "grad_norm": 1.307114601135254, | |
| "learning_rate": 8.280510180203476e-07, | |
| "loss": 0.0491, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 4.901785714285714, | |
| "grad_norm": 1.013657569885254, | |
| "learning_rate": 8.215569467407386e-07, | |
| "loss": 0.0463, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 4.90625, | |
| "grad_norm": 1.0433632135391235, | |
| "learning_rate": 8.150861919264887e-07, | |
| "loss": 0.0415, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 4.910714285714286, | |
| "grad_norm": 1.2945899963378906, | |
| "learning_rate": 8.086387896812546e-07, | |
| "loss": 0.0503, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.915178571428571, | |
| "grad_norm": 1.0583516359329224, | |
| "learning_rate": 8.022147759784016e-07, | |
| "loss": 0.0426, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 4.919642857142857, | |
| "grad_norm": 1.0814223289489746, | |
| "learning_rate": 7.958141866607897e-07, | |
| "loss": 0.0438, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 4.924107142857143, | |
| "grad_norm": 1.4671058654785156, | |
| "learning_rate": 7.894370574405928e-07, | |
| "loss": 0.0489, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 4.928571428571429, | |
| "grad_norm": 0.9236575961112976, | |
| "learning_rate": 7.830834238990803e-07, | |
| "loss": 0.0416, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 4.933035714285714, | |
| "grad_norm": 1.0398495197296143, | |
| "learning_rate": 7.767533214864331e-07, | |
| "loss": 0.046, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 4.9375, | |
| "grad_norm": 1.139332890510559, | |
| "learning_rate": 7.70446785521533e-07, | |
| "loss": 0.0411, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 4.941964285714286, | |
| "grad_norm": 1.0621843338012695, | |
| "learning_rate": 7.641638511917806e-07, | |
| "loss": 0.0424, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 4.946428571428571, | |
| "grad_norm": 1.08333420753479, | |
| "learning_rate": 7.579045535528878e-07, | |
| "loss": 0.042, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 4.950892857142857, | |
| "grad_norm": 0.8833259344100952, | |
| "learning_rate": 7.516689275286813e-07, | |
| "loss": 0.0381, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 4.955357142857143, | |
| "grad_norm": 1.1298121213912964, | |
| "learning_rate": 7.454570079109164e-07, | |
| "loss": 0.0437, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 4.959821428571429, | |
| "grad_norm": 0.9060654044151306, | |
| "learning_rate": 7.392688293590767e-07, | |
| "loss": 0.037, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 4.964285714285714, | |
| "grad_norm": 0.8614628314971924, | |
| "learning_rate": 7.331044264001836e-07, | |
| "loss": 0.0341, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 4.96875, | |
| "grad_norm": 1.1477363109588623, | |
| "learning_rate": 7.269638334285973e-07, | |
| "loss": 0.0379, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 4.973214285714286, | |
| "grad_norm": 1.0552114248275757, | |
| "learning_rate": 7.208470847058387e-07, | |
| "loss": 0.052, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 4.977678571428571, | |
| "grad_norm": 1.0197445154190063, | |
| "learning_rate": 7.147542143603806e-07, | |
| "loss": 0.0381, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 4.982142857142857, | |
| "grad_norm": 1.0818407535552979, | |
| "learning_rate": 7.086852563874705e-07, | |
| "loss": 0.0451, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 4.986607142857143, | |
| "grad_norm": 1.2270269393920898, | |
| "learning_rate": 7.026402446489368e-07, | |
| "loss": 0.0405, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 4.991071428571429, | |
| "grad_norm": 1.078790307044983, | |
| "learning_rate": 6.966192128729969e-07, | |
| "loss": 0.0383, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 4.995535714285714, | |
| "grad_norm": 1.2608201503753662, | |
| "learning_rate": 6.906221946540765e-07, | |
| "loss": 0.0359, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.2368184328079224, | |
| "learning_rate": 6.846492234526105e-07, | |
| "loss": 0.0482, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 5.004464285714286, | |
| "grad_norm": 0.767040491104126, | |
| "learning_rate": 6.787003325948681e-07, | |
| "loss": 0.0239, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 5.008928571428571, | |
| "grad_norm": 0.5072701573371887, | |
| "learning_rate": 6.727755552727618e-07, | |
| "loss": 0.0208, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 5.013392857142857, | |
| "grad_norm": 0.5942046642303467, | |
| "learning_rate": 6.668749245436603e-07, | |
| "loss": 0.0231, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 5.017857142857143, | |
| "grad_norm": 0.6232619285583496, | |
| "learning_rate": 6.609984733302069e-07, | |
| "loss": 0.0291, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 5.022321428571429, | |
| "grad_norm": 0.7308699488639832, | |
| "learning_rate": 6.551462344201356e-07, | |
| "loss": 0.0275, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 5.026785714285714, | |
| "grad_norm": 0.6961386799812317, | |
| "learning_rate": 6.493182404660884e-07, | |
| "loss": 0.0287, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 5.03125, | |
| "grad_norm": 0.6047128438949585, | |
| "learning_rate": 6.435145239854279e-07, | |
| "loss": 0.0264, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 5.035714285714286, | |
| "grad_norm": 0.6986120939254761, | |
| "learning_rate": 6.377351173600649e-07, | |
| "loss": 0.0282, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 5.040178571428571, | |
| "grad_norm": 0.56222003698349, | |
| "learning_rate": 6.319800528362713e-07, | |
| "loss": 0.0205, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 5.044642857142857, | |
| "grad_norm": 0.6198891997337341, | |
| "learning_rate": 6.262493625245017e-07, | |
| "loss": 0.0212, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 5.049107142857143, | |
| "grad_norm": 0.7689740657806396, | |
| "learning_rate": 6.205430783992163e-07, | |
| "loss": 0.032, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 5.053571428571429, | |
| "grad_norm": 0.5175718665122986, | |
| "learning_rate": 6.148612322986963e-07, | |
| "loss": 0.021, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 5.058035714285714, | |
| "grad_norm": 0.551143229007721, | |
| "learning_rate": 6.092038559248772e-07, | |
| "loss": 0.0236, | |
| "step": 1133 | |
| }, | |
| { | |
| "epoch": 5.0625, | |
| "grad_norm": 0.5913689136505127, | |
| "learning_rate": 6.035709808431585e-07, | |
| "loss": 0.0207, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 5.066964285714286, | |
| "grad_norm": 0.6526609063148499, | |
| "learning_rate": 5.979626384822384e-07, | |
| "loss": 0.0267, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 5.071428571428571, | |
| "grad_norm": 0.5639797449111938, | |
| "learning_rate": 5.923788601339348e-07, | |
| "loss": 0.0195, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 5.075892857142857, | |
| "grad_norm": 0.5921444892883301, | |
| "learning_rate": 5.868196769530085e-07, | |
| "loss": 0.0211, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 5.080357142857143, | |
| "grad_norm": 0.6509416103363037, | |
| "learning_rate": 5.81285119956993e-07, | |
| "loss": 0.0211, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 5.084821428571429, | |
| "grad_norm": 0.7350341081619263, | |
| "learning_rate": 5.757752200260156e-07, | |
| "loss": 0.0256, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 5.089285714285714, | |
| "grad_norm": 0.581320583820343, | |
| "learning_rate": 5.702900079026365e-07, | |
| "loss": 0.0211, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 5.089285714285714, | |
| "eval_loss": 1.1169615983963013, | |
| "eval_runtime": 3.0991, | |
| "eval_samples_per_second": 19.361, | |
| "eval_steps_per_second": 1.291, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 5.09375, | |
| "grad_norm": 0.5776538252830505, | |
| "learning_rate": 5.648295141916629e-07, | |
| "loss": 0.021, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 5.098214285714286, | |
| "grad_norm": 0.6957246661186218, | |
| "learning_rate": 5.593937693599892e-07, | |
| "loss": 0.0226, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 5.102678571428571, | |
| "grad_norm": 0.6897474527359009, | |
| "learning_rate": 5.539828037364222e-07, | |
| "loss": 0.0239, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 5.107142857142857, | |
| "grad_norm": 0.6816636323928833, | |
| "learning_rate": 5.485966475115118e-07, | |
| "loss": 0.0254, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 5.111607142857143, | |
| "grad_norm": 0.7038552761077881, | |
| "learning_rate": 5.432353307373865e-07, | |
| "loss": 0.0248, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 5.116071428571429, | |
| "grad_norm": 0.8172852993011475, | |
| "learning_rate": 5.378988833275772e-07, | |
| "loss": 0.0304, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 5.120535714285714, | |
| "grad_norm": 0.6974570155143738, | |
| "learning_rate": 5.325873350568641e-07, | |
| "loss": 0.0236, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 5.125, | |
| "grad_norm": 0.7102988362312317, | |
| "learning_rate": 5.27300715561093e-07, | |
| "loss": 0.0264, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 5.129464285714286, | |
| "grad_norm": 0.7948593497276306, | |
| "learning_rate": 5.220390543370269e-07, | |
| "loss": 0.0272, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 5.133928571428571, | |
| "grad_norm": 0.659605085849762, | |
| "learning_rate": 5.16802380742169e-07, | |
| "loss": 0.0222, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 5.138392857142857, | |
| "grad_norm": 0.6647667288780212, | |
| "learning_rate": 5.115907239946071e-07, | |
| "loss": 0.0212, | |
| "step": 1151 | |
| }, | |
| { | |
| "epoch": 5.142857142857143, | |
| "grad_norm": 0.6717812418937683, | |
| "learning_rate": 5.064041131728456e-07, | |
| "loss": 0.0218, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 5.147321428571429, | |
| "grad_norm": 0.6710494756698608, | |
| "learning_rate": 5.012425772156433e-07, | |
| "loss": 0.0245, | |
| "step": 1153 | |
| }, | |
| { | |
| "epoch": 5.151785714285714, | |
| "grad_norm": 0.7254083752632141, | |
| "learning_rate": 4.961061449218561e-07, | |
| "loss": 0.0187, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 5.15625, | |
| "grad_norm": 0.7306315898895264, | |
| "learning_rate": 4.90994844950272e-07, | |
| "loss": 0.0235, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 5.160714285714286, | |
| "grad_norm": 0.8266925811767578, | |
| "learning_rate": 4.859087058194547e-07, | |
| "loss": 0.0237, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 5.165178571428571, | |
| "grad_norm": 0.9365939497947693, | |
| "learning_rate": 4.808477559075791e-07, | |
| "loss": 0.0265, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 5.169642857142857, | |
| "grad_norm": 0.7625580430030823, | |
| "learning_rate": 4.758120234522819e-07, | |
| "loss": 0.0249, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 5.174107142857143, | |
| "grad_norm": 0.7768296599388123, | |
| "learning_rate": 4.708015365504947e-07, | |
| "loss": 0.0272, | |
| "step": 1159 | |
| }, | |
| { | |
| "epoch": 5.178571428571429, | |
| "grad_norm": 0.6355034112930298, | |
| "learning_rate": 4.658163231582916e-07, | |
| "loss": 0.0195, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 5.183035714285714, | |
| "grad_norm": 0.7550672888755798, | |
| "learning_rate": 4.6085641109073313e-07, | |
| "loss": 0.0243, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 5.1875, | |
| "grad_norm": 0.8513064980506897, | |
| "learning_rate": 4.559218280217121e-07, | |
| "loss": 0.0268, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 5.191964285714286, | |
| "grad_norm": 0.6934846639633179, | |
| "learning_rate": 4.5101260148379735e-07, | |
| "loss": 0.0215, | |
| "step": 1163 | |
| }, | |
| { | |
| "epoch": 5.196428571428571, | |
| "grad_norm": 0.8758344054222107, | |
| "learning_rate": 4.461287588680783e-07, | |
| "loss": 0.0304, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 5.200892857142857, | |
| "grad_norm": 0.7218351364135742, | |
| "learning_rate": 4.4127032742401697e-07, | |
| "loss": 0.0252, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 5.205357142857143, | |
| "grad_norm": 0.7134067416191101, | |
| "learning_rate": 4.364373342592935e-07, | |
| "loss": 0.0256, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 5.209821428571429, | |
| "grad_norm": 0.7380569577217102, | |
| "learning_rate": 4.316298063396534e-07, | |
| "loss": 0.0213, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 5.214285714285714, | |
| "grad_norm": 0.7990237474441528, | |
| "learning_rate": 4.2684777048875913e-07, | |
| "loss": 0.0301, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 5.21875, | |
| "grad_norm": 0.8833650946617126, | |
| "learning_rate": 4.2209125338804007e-07, | |
| "loss": 0.0273, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 5.223214285714286, | |
| "grad_norm": 0.7409700155258179, | |
| "learning_rate": 4.173602815765447e-07, | |
| "loss": 0.0245, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 5.227678571428571, | |
| "grad_norm": 0.7785325050354004, | |
| "learning_rate": 4.126548814507876e-07, | |
| "loss": 0.0209, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 5.232142857142857, | |
| "grad_norm": 0.7405259609222412, | |
| "learning_rate": 4.079750792646085e-07, | |
| "loss": 0.0229, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 5.236607142857143, | |
| "grad_norm": 0.865561306476593, | |
| "learning_rate": 4.0332090112902294e-07, | |
| "loss": 0.0283, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 5.241071428571429, | |
| "grad_norm": 0.7885621786117554, | |
| "learning_rate": 3.98692373012076e-07, | |
| "loss": 0.0255, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 5.245535714285714, | |
| "grad_norm": 0.6892197728157043, | |
| "learning_rate": 3.940895207387007e-07, | |
| "loss": 0.021, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 5.25, | |
| "grad_norm": 0.819604754447937, | |
| "learning_rate": 3.89512369990565e-07, | |
| "loss": 0.0255, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 5.254464285714286, | |
| "grad_norm": 0.6747646331787109, | |
| "learning_rate": 3.849609463059437e-07, | |
| "loss": 0.02, | |
| "step": 1177 | |
| }, | |
| { | |
| "epoch": 5.258928571428571, | |
| "grad_norm": 0.7342539429664612, | |
| "learning_rate": 3.8043527507955926e-07, | |
| "loss": 0.0241, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 5.263392857142857, | |
| "grad_norm": 0.847237765789032, | |
| "learning_rate": 3.759353815624526e-07, | |
| "loss": 0.0251, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 5.267857142857143, | |
| "grad_norm": 0.9026116132736206, | |
| "learning_rate": 3.7146129086183547e-07, | |
| "loss": 0.0242, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 5.272321428571429, | |
| "grad_norm": 0.7088863253593445, | |
| "learning_rate": 3.6701302794095416e-07, | |
| "loss": 0.0208, | |
| "step": 1181 | |
| }, | |
| { | |
| "epoch": 5.276785714285714, | |
| "grad_norm": 0.8555396199226379, | |
| "learning_rate": 3.625906176189484e-07, | |
| "loss": 0.0264, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 5.28125, | |
| "grad_norm": 0.8181936740875244, | |
| "learning_rate": 3.5819408457070893e-07, | |
| "loss": 0.0233, | |
| "step": 1183 | |
| }, | |
| { | |
| "epoch": 5.285714285714286, | |
| "grad_norm": 0.7480681538581848, | |
| "learning_rate": 3.5382345332675154e-07, | |
| "loss": 0.0243, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 5.290178571428571, | |
| "grad_norm": 0.7232365012168884, | |
| "learning_rate": 3.494787482730647e-07, | |
| "loss": 0.0199, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 5.294642857142857, | |
| "grad_norm": 0.7661169171333313, | |
| "learning_rate": 3.4515999365098667e-07, | |
| "loss": 0.0208, | |
| "step": 1186 | |
| }, | |
| { | |
| "epoch": 5.299107142857143, | |
| "grad_norm": 0.8898442983627319, | |
| "learning_rate": 3.4086721355706303e-07, | |
| "loss": 0.0242, | |
| "step": 1187 | |
| }, | |
| { | |
| "epoch": 5.303571428571429, | |
| "grad_norm": 0.8835738897323608, | |
| "learning_rate": 3.366004319429139e-07, | |
| "loss": 0.0301, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 5.308035714285714, | |
| "grad_norm": 0.7845296859741211, | |
| "learning_rate": 3.323596726151021e-07, | |
| "loss": 0.0235, | |
| "step": 1189 | |
| }, | |
| { | |
| "epoch": 5.3125, | |
| "grad_norm": 0.723148763179779, | |
| "learning_rate": 3.2814495923499496e-07, | |
| "loss": 0.0185, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 5.316964285714286, | |
| "grad_norm": 0.8334882855415344, | |
| "learning_rate": 3.239563153186429e-07, | |
| "loss": 0.0211, | |
| "step": 1191 | |
| }, | |
| { | |
| "epoch": 5.321428571428571, | |
| "grad_norm": 1.186741828918457, | |
| "learning_rate": 3.197937642366339e-07, | |
| "loss": 0.0291, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 5.325892857142857, | |
| "grad_norm": 0.7373731136322021, | |
| "learning_rate": 3.1565732921397583e-07, | |
| "loss": 0.0232, | |
| "step": 1193 | |
| }, | |
| { | |
| "epoch": 5.330357142857143, | |
| "grad_norm": 0.7730535268783569, | |
| "learning_rate": 3.115470333299593e-07, | |
| "loss": 0.0247, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 5.334821428571429, | |
| "grad_norm": 0.7440225481987, | |
| "learning_rate": 3.0746289951803197e-07, | |
| "loss": 0.0231, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 5.339285714285714, | |
| "grad_norm": 0.8436987996101379, | |
| "learning_rate": 3.03404950565671e-07, | |
| "loss": 0.0259, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 5.34375, | |
| "grad_norm": 0.8651729822158813, | |
| "learning_rate": 2.9937320911425226e-07, | |
| "loss": 0.024, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 5.34375, | |
| "eval_loss": 1.1776899099349976, | |
| "eval_runtime": 3.1142, | |
| "eval_samples_per_second": 19.266, | |
| "eval_steps_per_second": 1.284, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 5.348214285714286, | |
| "grad_norm": 0.7915188074111938, | |
| "learning_rate": 2.953676976589278e-07, | |
| "loss": 0.0217, | |
| "step": 1198 | |
| }, | |
| { | |
| "epoch": 5.352678571428571, | |
| "grad_norm": 0.7155839204788208, | |
| "learning_rate": 2.9138843854849964e-07, | |
| "loss": 0.0218, | |
| "step": 1199 | |
| }, | |
| { | |
| "epoch": 5.357142857142857, | |
| "grad_norm": 0.7552268505096436, | |
| "learning_rate": 2.8743545398529436e-07, | |
| "loss": 0.0205, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.361607142857143, | |
| "grad_norm": 0.7332569360733032, | |
| "learning_rate": 2.8350876602503893e-07, | |
| "loss": 0.0242, | |
| "step": 1201 | |
| }, | |
| { | |
| "epoch": 5.366071428571429, | |
| "grad_norm": 0.7247368693351746, | |
| "learning_rate": 2.7960839657673844e-07, | |
| "loss": 0.0247, | |
| "step": 1202 | |
| }, | |
| { | |
| "epoch": 5.370535714285714, | |
| "grad_norm": 0.7671637535095215, | |
| "learning_rate": 2.7573436740255337e-07, | |
| "loss": 0.0262, | |
| "step": 1203 | |
| }, | |
| { | |
| "epoch": 5.375, | |
| "grad_norm": 0.8328778147697449, | |
| "learning_rate": 2.7188670011767715e-07, | |
| "loss": 0.0294, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 5.379464285714286, | |
| "grad_norm": 0.6904054284095764, | |
| "learning_rate": 2.680654161902189e-07, | |
| "loss": 0.0207, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 5.383928571428571, | |
| "grad_norm": 0.9550811648368835, | |
| "learning_rate": 2.6427053694107966e-07, | |
| "loss": 0.0179, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 5.388392857142857, | |
| "grad_norm": 0.6718007922172546, | |
| "learning_rate": 2.605020835438375e-07, | |
| "loss": 0.0201, | |
| "step": 1207 | |
| }, | |
| { | |
| "epoch": 5.392857142857143, | |
| "grad_norm": 0.6450316905975342, | |
| "learning_rate": 2.567600770246237e-07, | |
| "loss": 0.0208, | |
| "step": 1208 | |
| }, | |
| { | |
| "epoch": 5.397321428571429, | |
| "grad_norm": 0.669978678226471, | |
| "learning_rate": 2.5304453826201084e-07, | |
| "loss": 0.0205, | |
| "step": 1209 | |
| }, | |
| { | |
| "epoch": 5.401785714285714, | |
| "grad_norm": 0.7878643870353699, | |
| "learning_rate": 2.493554879868958e-07, | |
| "loss": 0.0211, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 5.40625, | |
| "grad_norm": 0.7882503271102905, | |
| "learning_rate": 2.4569294678237995e-07, | |
| "loss": 0.0258, | |
| "step": 1211 | |
| }, | |
| { | |
| "epoch": 5.410714285714286, | |
| "grad_norm": 0.6949124932289124, | |
| "learning_rate": 2.42056935083658e-07, | |
| "loss": 0.02, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 5.415178571428571, | |
| "grad_norm": 0.7191565632820129, | |
| "learning_rate": 2.38447473177903e-07, | |
| "loss": 0.0187, | |
| "step": 1213 | |
| }, | |
| { | |
| "epoch": 5.419642857142857, | |
| "grad_norm": 0.8001049757003784, | |
| "learning_rate": 2.3486458120415415e-07, | |
| "loss": 0.0228, | |
| "step": 1214 | |
| }, | |
| { | |
| "epoch": 5.424107142857143, | |
| "grad_norm": 0.6764441728591919, | |
| "learning_rate": 2.3130827915320015e-07, | |
| "loss": 0.0186, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 5.428571428571429, | |
| "grad_norm": 0.7706072330474854, | |
| "learning_rate": 2.2777858686747495e-07, | |
| "loss": 0.0268, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 5.433035714285714, | |
| "grad_norm": 0.6611673831939697, | |
| "learning_rate": 2.242755240409399e-07, | |
| "loss": 0.0196, | |
| "step": 1217 | |
| }, | |
| { | |
| "epoch": 5.4375, | |
| "grad_norm": 0.7625905275344849, | |
| "learning_rate": 2.20799110218979e-07, | |
| "loss": 0.0249, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 5.441964285714286, | |
| "grad_norm": 0.8830248713493347, | |
| "learning_rate": 2.173493647982873e-07, | |
| "loss": 0.0228, | |
| "step": 1219 | |
| }, | |
| { | |
| "epoch": 5.446428571428571, | |
| "grad_norm": 0.8279975652694702, | |
| "learning_rate": 2.139263070267605e-07, | |
| "loss": 0.0256, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 5.450892857142857, | |
| "grad_norm": 0.7940500378608704, | |
| "learning_rate": 2.105299560033954e-07, | |
| "loss": 0.0241, | |
| "step": 1221 | |
| }, | |
| { | |
| "epoch": 5.455357142857143, | |
| "grad_norm": 0.6114321947097778, | |
| "learning_rate": 2.0716033067817308e-07, | |
| "loss": 0.0211, | |
| "step": 1222 | |
| }, | |
| { | |
| "epoch": 5.459821428571429, | |
| "grad_norm": 0.724319577217102, | |
| "learning_rate": 2.03817449851962e-07, | |
| "loss": 0.0202, | |
| "step": 1223 | |
| }, | |
| { | |
| "epoch": 5.464285714285714, | |
| "grad_norm": 0.6841122508049011, | |
| "learning_rate": 2.0050133217640672e-07, | |
| "loss": 0.0227, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 5.46875, | |
| "grad_norm": 0.6772453784942627, | |
| "learning_rate": 1.972119961538277e-07, | |
| "loss": 0.0237, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 5.473214285714286, | |
| "grad_norm": 0.7682874798774719, | |
| "learning_rate": 1.9394946013711787e-07, | |
| "loss": 0.03, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 5.477678571428571, | |
| "grad_norm": 0.8899271488189697, | |
| "learning_rate": 1.9071374232963564e-07, | |
| "loss": 0.0255, | |
| "step": 1227 | |
| }, | |
| { | |
| "epoch": 5.482142857142857, | |
| "grad_norm": 0.759722888469696, | |
| "learning_rate": 1.8750486078511206e-07, | |
| "loss": 0.0266, | |
| "step": 1228 | |
| }, | |
| { | |
| "epoch": 5.486607142857143, | |
| "grad_norm": 0.8016492128372192, | |
| "learning_rate": 1.8432283340754e-07, | |
| "loss": 0.0246, | |
| "step": 1229 | |
| }, | |
| { | |
| "epoch": 5.491071428571429, | |
| "grad_norm": 0.7532429695129395, | |
| "learning_rate": 1.8116767795108175e-07, | |
| "loss": 0.0245, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 5.495535714285714, | |
| "grad_norm": 0.7854920029640198, | |
| "learning_rate": 1.780394120199672e-07, | |
| "loss": 0.0241, | |
| "step": 1231 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 0.7643800377845764, | |
| "learning_rate": 1.7493805306839532e-07, | |
| "loss": 0.0297, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 5.504464285714286, | |
| "grad_norm": 0.7853821516036987, | |
| "learning_rate": 1.718636184004378e-07, | |
| "loss": 0.0245, | |
| "step": 1233 | |
| }, | |
| { | |
| "epoch": 5.508928571428571, | |
| "grad_norm": 0.7025668025016785, | |
| "learning_rate": 1.688161251699405e-07, | |
| "loss": 0.0209, | |
| "step": 1234 | |
| }, | |
| { | |
| "epoch": 5.513392857142857, | |
| "grad_norm": 0.7605273723602295, | |
| "learning_rate": 1.6579559038043186e-07, | |
| "loss": 0.024, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 5.517857142857143, | |
| "grad_norm": 0.6690748333930969, | |
| "learning_rate": 1.6280203088502275e-07, | |
| "loss": 0.0214, | |
| "step": 1236 | |
| }, | |
| { | |
| "epoch": 5.522321428571429, | |
| "grad_norm": 0.6744450926780701, | |
| "learning_rate": 1.5983546338631578e-07, | |
| "loss": 0.0194, | |
| "step": 1237 | |
| }, | |
| { | |
| "epoch": 5.526785714285714, | |
| "grad_norm": 0.6696710586547852, | |
| "learning_rate": 1.5689590443631178e-07, | |
| "loss": 0.0216, | |
| "step": 1238 | |
| }, | |
| { | |
| "epoch": 5.53125, | |
| "grad_norm": 0.9168437123298645, | |
| "learning_rate": 1.5398337043631723e-07, | |
| "loss": 0.0214, | |
| "step": 1239 | |
| }, | |
| { | |
| "epoch": 5.535714285714286, | |
| "grad_norm": 0.7635628581047058, | |
| "learning_rate": 1.5109787763685323e-07, | |
| "loss": 0.0251, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 5.540178571428571, | |
| "grad_norm": 0.6954474449157715, | |
| "learning_rate": 1.4823944213756056e-07, | |
| "loss": 0.0218, | |
| "step": 1241 | |
| }, | |
| { | |
| "epoch": 5.544642857142857, | |
| "grad_norm": 0.7508897185325623, | |
| "learning_rate": 1.4540807988711857e-07, | |
| "loss": 0.0231, | |
| "step": 1242 | |
| }, | |
| { | |
| "epoch": 5.549107142857143, | |
| "grad_norm": 0.8188397288322449, | |
| "learning_rate": 1.4260380668314764e-07, | |
| "loss": 0.0205, | |
| "step": 1243 | |
| }, | |
| { | |
| "epoch": 5.553571428571429, | |
| "grad_norm": 0.7487647533416748, | |
| "learning_rate": 1.3982663817212475e-07, | |
| "loss": 0.025, | |
| "step": 1244 | |
| }, | |
| { | |
| "epoch": 5.558035714285714, | |
| "grad_norm": 0.7735338807106018, | |
| "learning_rate": 1.3707658984929738e-07, | |
| "loss": 0.0228, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 5.5625, | |
| "grad_norm": 0.8641656041145325, | |
| "learning_rate": 1.3435367705859475e-07, | |
| "loss": 0.0266, | |
| "step": 1246 | |
| }, | |
| { | |
| "epoch": 5.566964285714286, | |
| "grad_norm": 0.7950751781463623, | |
| "learning_rate": 1.3165791499254294e-07, | |
| "loss": 0.0269, | |
| "step": 1247 | |
| }, | |
| { | |
| "epoch": 5.571428571428571, | |
| "grad_norm": 0.9540092945098877, | |
| "learning_rate": 1.2898931869218046e-07, | |
| "loss": 0.0239, | |
| "step": 1248 | |
| }, | |
| { | |
| "epoch": 5.575892857142857, | |
| "grad_norm": 0.8364283442497253, | |
| "learning_rate": 1.2634790304697283e-07, | |
| "loss": 0.0247, | |
| "step": 1249 | |
| }, | |
| { | |
| "epoch": 5.580357142857143, | |
| "grad_norm": 0.7901514768600464, | |
| "learning_rate": 1.2373368279473364e-07, | |
| "loss": 0.0228, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 5.584821428571429, | |
| "grad_norm": 0.8159101605415344, | |
| "learning_rate": 1.2114667252153644e-07, | |
| "loss": 0.0263, | |
| "step": 1251 | |
| }, | |
| { | |
| "epoch": 5.589285714285714, | |
| "grad_norm": 0.6788504719734192, | |
| "learning_rate": 1.1858688666163752e-07, | |
| "loss": 0.0186, | |
| "step": 1252 | |
| }, | |
| { | |
| "epoch": 5.59375, | |
| "grad_norm": 0.8900226354598999, | |
| "learning_rate": 1.1605433949739546e-07, | |
| "loss": 0.0286, | |
| "step": 1253 | |
| }, | |
| { | |
| "epoch": 5.598214285714286, | |
| "grad_norm": 0.7049673795700073, | |
| "learning_rate": 1.1354904515918834e-07, | |
| "loss": 0.0244, | |
| "step": 1254 | |
| }, | |
| { | |
| "epoch": 5.598214285714286, | |
| "eval_loss": 1.183043122291565, | |
| "eval_runtime": 3.1011, | |
| "eval_samples_per_second": 19.348, | |
| "eval_steps_per_second": 1.29, | |
| "step": 1254 | |
| }, | |
| { | |
| "epoch": 5.602678571428571, | |
| "grad_norm": 0.7123156785964966, | |
| "learning_rate": 1.1107101762533725e-07, | |
| "loss": 0.0236, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 5.607142857142857, | |
| "grad_norm": 0.6948925256729126, | |
| "learning_rate": 1.0862027072202796e-07, | |
| "loss": 0.02, | |
| "step": 1256 | |
| }, | |
| { | |
| "epoch": 5.611607142857143, | |
| "grad_norm": 0.8849415183067322, | |
| "learning_rate": 1.0619681812323437e-07, | |
| "loss": 0.0278, | |
| "step": 1257 | |
| }, | |
| { | |
| "epoch": 5.616071428571429, | |
| "grad_norm": 0.7653726935386658, | |
| "learning_rate": 1.0380067335064019e-07, | |
| "loss": 0.0256, | |
| "step": 1258 | |
| }, | |
| { | |
| "epoch": 5.620535714285714, | |
| "grad_norm": 0.7977951765060425, | |
| "learning_rate": 1.0143184977356513e-07, | |
| "loss": 0.0227, | |
| "step": 1259 | |
| }, | |
| { | |
| "epoch": 5.625, | |
| "grad_norm": 0.7583563327789307, | |
| "learning_rate": 9.909036060889063e-08, | |
| "loss": 0.0237, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 5.629464285714286, | |
| "grad_norm": 0.7033082842826843, | |
| "learning_rate": 9.677621892098471e-08, | |
| "loss": 0.0215, | |
| "step": 1261 | |
| }, | |
| { | |
| "epoch": 5.633928571428571, | |
| "grad_norm": 0.7182819843292236, | |
| "learning_rate": 9.448943762163063e-08, | |
| "loss": 0.0225, | |
| "step": 1262 | |
| }, | |
| { | |
| "epoch": 5.638392857142857, | |
| "grad_norm": 0.7455835938453674, | |
| "learning_rate": 9.223002946995223e-08, | |
| "loss": 0.0232, | |
| "step": 1263 | |
| }, | |
| { | |
| "epoch": 5.642857142857143, | |
| "grad_norm": 0.7556766271591187, | |
| "learning_rate": 8.999800707234651e-08, | |
| "loss": 0.0216, | |
| "step": 1264 | |
| }, | |
| { | |
| "epoch": 5.647321428571429, | |
| "grad_norm": 0.7347747087478638, | |
| "learning_rate": 8.77933828824112e-08, | |
| "loss": 0.0192, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 5.651785714285714, | |
| "grad_norm": 0.6410439610481262, | |
| "learning_rate": 8.561616920087338e-08, | |
| "loss": 0.0171, | |
| "step": 1266 | |
| }, | |
| { | |
| "epoch": 5.65625, | |
| "grad_norm": 0.8440278768539429, | |
| "learning_rate": 8.346637817552435e-08, | |
| "loss": 0.0298, | |
| "step": 1267 | |
| }, | |
| { | |
| "epoch": 5.660714285714286, | |
| "grad_norm": 0.8332878351211548, | |
| "learning_rate": 8.134402180115097e-08, | |
| "loss": 0.0267, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 5.665178571428571, | |
| "grad_norm": 0.7442850470542908, | |
| "learning_rate": 7.924911191946728e-08, | |
| "loss": 0.0291, | |
| "step": 1269 | |
| }, | |
| { | |
| "epoch": 5.669642857142857, | |
| "grad_norm": 0.7406393885612488, | |
| "learning_rate": 7.718166021904903e-08, | |
| "loss": 0.0234, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 5.674107142857143, | |
| "grad_norm": 0.7578692436218262, | |
| "learning_rate": 7.514167823526817e-08, | |
| "loss": 0.0273, | |
| "step": 1271 | |
| }, | |
| { | |
| "epoch": 5.678571428571429, | |
| "grad_norm": 0.6509707570075989, | |
| "learning_rate": 7.31291773502313e-08, | |
| "loss": 0.0193, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 5.683035714285714, | |
| "grad_norm": 0.7130553126335144, | |
| "learning_rate": 7.114416879271188e-08, | |
| "loss": 0.0223, | |
| "step": 1273 | |
| }, | |
| { | |
| "epoch": 5.6875, | |
| "grad_norm": 0.6997455954551697, | |
| "learning_rate": 6.918666363808976e-08, | |
| "loss": 0.0196, | |
| "step": 1274 | |
| }, | |
| { | |
| "epoch": 5.691964285714286, | |
| "grad_norm": 0.7897157073020935, | |
| "learning_rate": 6.725667280828959e-08, | |
| "loss": 0.0238, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 5.696428571428571, | |
| "grad_norm": 0.7304858565330505, | |
| "learning_rate": 6.535420707172025e-08, | |
| "loss": 0.0248, | |
| "step": 1276 | |
| }, | |
| { | |
| "epoch": 5.700892857142857, | |
| "grad_norm": 0.6769036650657654, | |
| "learning_rate": 6.347927704321335e-08, | |
| "loss": 0.0193, | |
| "step": 1277 | |
| }, | |
| { | |
| "epoch": 5.705357142857143, | |
| "grad_norm": 0.713193416595459, | |
| "learning_rate": 6.163189318396261e-08, | |
| "loss": 0.0222, | |
| "step": 1278 | |
| }, | |
| { | |
| "epoch": 5.709821428571429, | |
| "grad_norm": 0.798915684223175, | |
| "learning_rate": 5.981206580147232e-08, | |
| "loss": 0.0202, | |
| "step": 1279 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "grad_norm": 0.865748941898346, | |
| "learning_rate": 5.8019805049490143e-08, | |
| "loss": 0.0263, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 5.71875, | |
| "grad_norm": 0.7367980480194092, | |
| "learning_rate": 5.625512092795828e-08, | |
| "loss": 0.0193, | |
| "step": 1281 | |
| }, | |
| { | |
| "epoch": 5.723214285714286, | |
| "grad_norm": 0.6353578567504883, | |
| "learning_rate": 5.451802328295408e-08, | |
| "loss": 0.0201, | |
| "step": 1282 | |
| }, | |
| { | |
| "epoch": 5.727678571428571, | |
| "grad_norm": 0.8302894830703735, | |
| "learning_rate": 5.2808521806635646e-08, | |
| "loss": 0.03, | |
| "step": 1283 | |
| }, | |
| { | |
| "epoch": 5.732142857142857, | |
| "grad_norm": 0.746477484703064, | |
| "learning_rate": 5.1126626037188537e-08, | |
| "loss": 0.0216, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 5.736607142857143, | |
| "grad_norm": 0.655388593673706, | |
| "learning_rate": 4.9472345358769714e-08, | |
| "loss": 0.0197, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 5.741071428571429, | |
| "grad_norm": 1.365317702293396, | |
| "learning_rate": 4.784568900146095e-08, | |
| "loss": 0.0213, | |
| "step": 1286 | |
| }, | |
| { | |
| "epoch": 5.745535714285714, | |
| "grad_norm": 0.6960539221763611, | |
| "learning_rate": 4.624666604121047e-08, | |
| "loss": 0.019, | |
| "step": 1287 | |
| }, | |
| { | |
| "epoch": 5.75, | |
| "grad_norm": 0.6388968229293823, | |
| "learning_rate": 4.4675285399787523e-08, | |
| "loss": 0.0195, | |
| "step": 1288 | |
| }, | |
| { | |
| "epoch": 5.754464285714286, | |
| "grad_norm": 0.8190224170684814, | |
| "learning_rate": 4.3131555844730135e-08, | |
| "loss": 0.0231, | |
| "step": 1289 | |
| }, | |
| { | |
| "epoch": 5.758928571428571, | |
| "grad_norm": 0.7648060917854309, | |
| "learning_rate": 4.161548598929689e-08, | |
| "loss": 0.0219, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 5.763392857142857, | |
| "grad_norm": 0.6431726217269897, | |
| "learning_rate": 4.01270842924191e-08, | |
| "loss": 0.0215, | |
| "step": 1291 | |
| }, | |
| { | |
| "epoch": 5.767857142857143, | |
| "grad_norm": 0.6700044274330139, | |
| "learning_rate": 3.8666359058652064e-08, | |
| "loss": 0.0189, | |
| "step": 1292 | |
| }, | |
| { | |
| "epoch": 5.772321428571429, | |
| "grad_norm": 0.6487048268318176, | |
| "learning_rate": 3.7233318438130064e-08, | |
| "loss": 0.0211, | |
| "step": 1293 | |
| }, | |
| { | |
| "epoch": 5.776785714285714, | |
| "grad_norm": 0.7264946103096008, | |
| "learning_rate": 3.582797042652248e-08, | |
| "loss": 0.0243, | |
| "step": 1294 | |
| }, | |
| { | |
| "epoch": 5.78125, | |
| "grad_norm": 1.036159873008728, | |
| "learning_rate": 3.4450322864986106e-08, | |
| "loss": 0.0301, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 5.785714285714286, | |
| "grad_norm": 0.7695539593696594, | |
| "learning_rate": 3.310038344012184e-08, | |
| "loss": 0.0235, | |
| "step": 1296 | |
| }, | |
| { | |
| "epoch": 5.790178571428571, | |
| "grad_norm": 0.655296266078949, | |
| "learning_rate": 3.17781596839347e-08, | |
| "loss": 0.0183, | |
| "step": 1297 | |
| }, | |
| { | |
| "epoch": 5.794642857142857, | |
| "grad_norm": 0.8248305916786194, | |
| "learning_rate": 3.0483658973788894e-08, | |
| "loss": 0.0212, | |
| "step": 1298 | |
| }, | |
| { | |
| "epoch": 5.799107142857143, | |
| "grad_norm": 0.7949267029762268, | |
| "learning_rate": 2.92168885323662e-08, | |
| "loss": 0.0177, | |
| "step": 1299 | |
| }, | |
| { | |
| "epoch": 5.803571428571429, | |
| "grad_norm": 0.7954432368278503, | |
| "learning_rate": 2.797785542762927e-08, | |
| "loss": 0.0232, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 5.808035714285714, | |
| "grad_norm": 0.8179627060890198, | |
| "learning_rate": 2.67665665727784e-08, | |
| "loss": 0.026, | |
| "step": 1301 | |
| }, | |
| { | |
| "epoch": 5.8125, | |
| "grad_norm": 0.7565423846244812, | |
| "learning_rate": 2.5583028726215427e-08, | |
| "loss": 0.0193, | |
| "step": 1302 | |
| }, | |
| { | |
| "epoch": 5.816964285714286, | |
| "grad_norm": 0.9201167225837708, | |
| "learning_rate": 2.442724849150487e-08, | |
| "loss": 0.0229, | |
| "step": 1303 | |
| }, | |
| { | |
| "epoch": 5.821428571428571, | |
| "grad_norm": 0.7921238541603088, | |
| "learning_rate": 2.3299232317335643e-08, | |
| "loss": 0.024, | |
| "step": 1304 | |
| }, | |
| { | |
| "epoch": 5.825892857142857, | |
| "grad_norm": 0.8329811692237854, | |
| "learning_rate": 2.2198986497489963e-08, | |
| "loss": 0.0238, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 5.830357142857143, | |
| "grad_norm": 0.6987408399581909, | |
| "learning_rate": 2.112651717080228e-08, | |
| "loss": 0.0191, | |
| "step": 1306 | |
| }, | |
| { | |
| "epoch": 5.834821428571429, | |
| "grad_norm": 0.8431302309036255, | |
| "learning_rate": 2.0081830321129298e-08, | |
| "loss": 0.0251, | |
| "step": 1307 | |
| }, | |
| { | |
| "epoch": 5.839285714285714, | |
| "grad_norm": 0.7029537558555603, | |
| "learning_rate": 1.906493177731391e-08, | |
| "loss": 0.0193, | |
| "step": 1308 | |
| }, | |
| { | |
| "epoch": 5.84375, | |
| "grad_norm": 0.7598433494567871, | |
| "learning_rate": 1.80758272131541e-08, | |
| "loss": 0.0224, | |
| "step": 1309 | |
| }, | |
| { | |
| "epoch": 5.848214285714286, | |
| "grad_norm": 3.8866190910339355, | |
| "learning_rate": 1.711452214737187e-08, | |
| "loss": 0.0216, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 5.852678571428571, | |
| "grad_norm": 0.7216530442237854, | |
| "learning_rate": 1.6181021943580477e-08, | |
| "loss": 0.0219, | |
| "step": 1311 | |
| }, | |
| { | |
| "epoch": 5.852678571428571, | |
| "eval_loss": 1.1876667737960815, | |
| "eval_runtime": 3.1035, | |
| "eval_samples_per_second": 19.333, | |
| "eval_steps_per_second": 1.289, | |
| "step": 1311 | |
| }, | |
| { | |
| "epoch": 5.857142857142857, | |
| "grad_norm": 0.7401444911956787, | |
| "learning_rate": 1.5275331810256708e-08, | |
| "loss": 0.0217, | |
| "step": 1312 | |
| }, | |
| { | |
| "epoch": 5.861607142857143, | |
| "grad_norm": 0.7781946659088135, | |
| "learning_rate": 1.439745680070921e-08, | |
| "loss": 0.0263, | |
| "step": 1313 | |
| }, | |
| { | |
| "epoch": 5.866071428571429, | |
| "grad_norm": 0.6831724047660828, | |
| "learning_rate": 1.3547401813053533e-08, | |
| "loss": 0.0232, | |
| "step": 1314 | |
| }, | |
| { | |
| "epoch": 5.870535714285714, | |
| "grad_norm": 0.7941478490829468, | |
| "learning_rate": 1.2725171590181043e-08, | |
| "loss": 0.0196, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 5.875, | |
| "grad_norm": 0.7862687706947327, | |
| "learning_rate": 1.1930770719736715e-08, | |
| "loss": 0.0258, | |
| "step": 1316 | |
| }, | |
| { | |
| "epoch": 5.879464285714286, | |
| "grad_norm": 0.5605817437171936, | |
| "learning_rate": 1.1164203634089718e-08, | |
| "loss": 0.0164, | |
| "step": 1317 | |
| }, | |
| { | |
| "epoch": 5.883928571428571, | |
| "grad_norm": 0.7842698097229004, | |
| "learning_rate": 1.0425474610310654e-08, | |
| "loss": 0.0265, | |
| "step": 1318 | |
| }, | |
| { | |
| "epoch": 5.888392857142857, | |
| "grad_norm": 0.6489875912666321, | |
| "learning_rate": 9.714587770147148e-09, | |
| "loss": 0.0196, | |
| "step": 1319 | |
| }, | |
| { | |
| "epoch": 5.892857142857143, | |
| "grad_norm": 0.6310600638389587, | |
| "learning_rate": 9.031547080002185e-09, | |
| "loss": 0.0188, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 5.897321428571429, | |
| "grad_norm": 0.6858588457107544, | |
| "learning_rate": 8.376356350909694e-09, | |
| "loss": 0.0211, | |
| "step": 1321 | |
| }, | |
| { | |
| "epoch": 5.901785714285714, | |
| "grad_norm": 0.8310582637786865, | |
| "learning_rate": 7.749019238513461e-09, | |
| "loss": 0.0305, | |
| "step": 1322 | |
| }, | |
| { | |
| "epoch": 5.90625, | |
| "grad_norm": 0.8480810523033142, | |
| "learning_rate": 7.149539243050468e-09, | |
| "loss": 0.0241, | |
| "step": 1323 | |
| }, | |
| { | |
| "epoch": 5.910714285714286, | |
| "grad_norm": 0.7383941411972046, | |
| "learning_rate": 6.577919709325367e-09, | |
| "loss": 0.0216, | |
| "step": 1324 | |
| }, | |
| { | |
| "epoch": 5.915178571428571, | |
| "grad_norm": 0.6655860543251038, | |
| "learning_rate": 6.034163826697711e-09, | |
| "loss": 0.0204, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 5.919642857142857, | |
| "grad_norm": 0.7747806906700134, | |
| "learning_rate": 5.518274629059199e-09, | |
| "loss": 0.0229, | |
| "step": 1326 | |
| }, | |
| { | |
| "epoch": 5.924107142857143, | |
| "grad_norm": 0.7398366928100586, | |
| "learning_rate": 5.030254994820907e-09, | |
| "loss": 0.0228, | |
| "step": 1327 | |
| }, | |
| { | |
| "epoch": 5.928571428571429, | |
| "grad_norm": 0.7458826303482056, | |
| "learning_rate": 4.570107646894414e-09, | |
| "loss": 0.023, | |
| "step": 1328 | |
| }, | |
| { | |
| "epoch": 5.933035714285714, | |
| "grad_norm": 0.7212875485420227, | |
| "learning_rate": 4.137835152677938e-09, | |
| "loss": 0.02, | |
| "step": 1329 | |
| }, | |
| { | |
| "epoch": 5.9375, | |
| "grad_norm": 0.714472770690918, | |
| "learning_rate": 3.7334399240402185e-09, | |
| "loss": 0.0203, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 5.941964285714286, | |
| "grad_norm": 0.7454721331596375, | |
| "learning_rate": 3.356924217310546e-09, | |
| "loss": 0.0264, | |
| "step": 1331 | |
| }, | |
| { | |
| "epoch": 5.946428571428571, | |
| "grad_norm": 0.6192952990531921, | |
| "learning_rate": 3.008290133262653e-09, | |
| "loss": 0.0172, | |
| "step": 1332 | |
| }, | |
| { | |
| "epoch": 5.950892857142857, | |
| "grad_norm": 0.6783339381217957, | |
| "learning_rate": 2.687539617105282e-09, | |
| "loss": 0.0186, | |
| "step": 1333 | |
| }, | |
| { | |
| "epoch": 5.955357142857143, | |
| "grad_norm": 0.8283798098564148, | |
| "learning_rate": 2.3946744584688626e-09, | |
| "loss": 0.0252, | |
| "step": 1334 | |
| }, | |
| { | |
| "epoch": 5.959821428571429, | |
| "grad_norm": 0.5973673462867737, | |
| "learning_rate": 2.1296962913994105e-09, | |
| "loss": 0.0181, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 5.964285714285714, | |
| "grad_norm": 0.6673908233642578, | |
| "learning_rate": 1.892606594345199e-09, | |
| "loss": 0.0192, | |
| "step": 1336 | |
| }, | |
| { | |
| "epoch": 5.96875, | |
| "grad_norm": 0.7677484154701233, | |
| "learning_rate": 1.6834066901512136e-09, | |
| "loss": 0.0248, | |
| "step": 1337 | |
| }, | |
| { | |
| "epoch": 5.973214285714286, | |
| "grad_norm": 0.6924158334732056, | |
| "learning_rate": 1.5020977460513809e-09, | |
| "loss": 0.023, | |
| "step": 1338 | |
| }, | |
| { | |
| "epoch": 5.977678571428571, | |
| "grad_norm": 0.6726967096328735, | |
| "learning_rate": 1.3486807736613498e-09, | |
| "loss": 0.0192, | |
| "step": 1339 | |
| }, | |
| { | |
| "epoch": 5.982142857142857, | |
| "grad_norm": 0.6165555715560913, | |
| "learning_rate": 1.2231566289723888e-09, | |
| "loss": 0.0163, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 5.986607142857143, | |
| "grad_norm": 0.6672031283378601, | |
| "learning_rate": 1.1255260123486095e-09, | |
| "loss": 0.022, | |
| "step": 1341 | |
| }, | |
| { | |
| "epoch": 5.991071428571429, | |
| "grad_norm": 0.8877815008163452, | |
| "learning_rate": 1.0557894685208617e-09, | |
| "loss": 0.0245, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 5.995535714285714, | |
| "grad_norm": 0.7125616073608398, | |
| "learning_rate": 1.013947386585067e-09, | |
| "loss": 0.025, | |
| "step": 1343 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.6855891942977905, | |
| "learning_rate": 1e-09, | |
| "loss": 0.0186, | |
| "step": 1344 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "step": 1344, | |
| "total_flos": 2.1132791208917074e+18, | |
| "train_loss": 0.19653707026043862, | |
| "train_runtime": 12053.73, | |
| "train_samples_per_second": 1.784, | |
| "train_steps_per_second": 0.112 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 1344, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 57, | |
| "total_flos": 2.1132791208917074e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |