| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9904153354632586, | |
| "eval_steps": 500, | |
| "global_step": 312, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.009584664536741214, | |
| "grad_norm": 5.8170170344066925, | |
| "learning_rate": 3.125e-07, | |
| "loss": 0.8809, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.019169329073482427, | |
| "grad_norm": 5.964865941319853, | |
| "learning_rate": 6.25e-07, | |
| "loss": 0.8476, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.02875399361022364, | |
| "grad_norm": 5.949605467481716, | |
| "learning_rate": 9.375000000000001e-07, | |
| "loss": 0.8785, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.038338658146964855, | |
| "grad_norm": 5.816077371594469, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.8584, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.04792332268370607, | |
| "grad_norm": 5.680232856483124, | |
| "learning_rate": 1.5625e-06, | |
| "loss": 0.8521, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.05750798722044728, | |
| "grad_norm": 5.317946866795994, | |
| "learning_rate": 1.8750000000000003e-06, | |
| "loss": 0.8466, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0670926517571885, | |
| "grad_norm": 4.4135739401625935, | |
| "learning_rate": 2.1875000000000002e-06, | |
| "loss": 0.7985, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.07667731629392971, | |
| "grad_norm": 4.120851565588499, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.799, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.08626198083067092, | |
| "grad_norm": 2.3808611201037024, | |
| "learning_rate": 2.8125e-06, | |
| "loss": 0.7761, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.09584664536741214, | |
| "grad_norm": 2.3155842307573526, | |
| "learning_rate": 3.125e-06, | |
| "loss": 0.7929, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.10543130990415335, | |
| "grad_norm": 2.0363118182915594, | |
| "learning_rate": 3.4375e-06, | |
| "loss": 0.7765, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.11501597444089456, | |
| "grad_norm": 3.298839456416055, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.7405, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.12460063897763578, | |
| "grad_norm": 3.7898799065420046, | |
| "learning_rate": 4.0625000000000005e-06, | |
| "loss": 0.7637, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.134185303514377, | |
| "grad_norm": 3.736501456981756, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 0.7353, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.14376996805111822, | |
| "grad_norm": 3.7724559929311705, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "loss": 0.7413, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.15335463258785942, | |
| "grad_norm": 3.1164974040288467, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7451, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.16293929712460065, | |
| "grad_norm": 2.776350460867649, | |
| "learning_rate": 5.3125e-06, | |
| "loss": 0.7148, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.17252396166134185, | |
| "grad_norm": 2.220821680910024, | |
| "learning_rate": 5.625e-06, | |
| "loss": 0.7072, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.18210862619808307, | |
| "grad_norm": 1.610179260092372, | |
| "learning_rate": 5.9375e-06, | |
| "loss": 0.6814, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.19169329073482427, | |
| "grad_norm": 1.2735045602593746, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.6551, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2012779552715655, | |
| "grad_norm": 1.541215050318752, | |
| "learning_rate": 6.5625e-06, | |
| "loss": 0.6493, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2108626198083067, | |
| "grad_norm": 1.6511266062967132, | |
| "learning_rate": 6.875e-06, | |
| "loss": 0.6532, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.22044728434504793, | |
| "grad_norm": 1.5541355580486804, | |
| "learning_rate": 7.1875e-06, | |
| "loss": 0.6301, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.23003194888178913, | |
| "grad_norm": 1.1270462341122904, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.6401, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.23961661341853036, | |
| "grad_norm": 1.0580481479196262, | |
| "learning_rate": 7.8125e-06, | |
| "loss": 0.6146, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.24920127795527156, | |
| "grad_norm": 1.2181186083491524, | |
| "learning_rate": 8.125000000000001e-06, | |
| "loss": 0.5944, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.25878594249201275, | |
| "grad_norm": 1.1331957612427834, | |
| "learning_rate": 8.4375e-06, | |
| "loss": 0.5969, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.268370607028754, | |
| "grad_norm": 0.7984145124760821, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.5704, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2779552715654952, | |
| "grad_norm": 0.764654167830464, | |
| "learning_rate": 9.0625e-06, | |
| "loss": 0.6107, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.28753993610223644, | |
| "grad_norm": 1.0122629597873543, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.5865, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2971246006389776, | |
| "grad_norm": 0.9744973968078775, | |
| "learning_rate": 9.6875e-06, | |
| "loss": 0.5823, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.30670926517571884, | |
| "grad_norm": 0.7470316548473647, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6105, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.31629392971246006, | |
| "grad_norm": 0.7124516902926219, | |
| "learning_rate": 9.999685283773504e-06, | |
| "loss": 0.5756, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3258785942492013, | |
| "grad_norm": 0.8013820997171508, | |
| "learning_rate": 9.998741174712534e-06, | |
| "loss": 0.5986, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.3354632587859425, | |
| "grad_norm": 0.792615180495177, | |
| "learning_rate": 9.997167791667668e-06, | |
| "loss": 0.5737, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3450479233226837, | |
| "grad_norm": 0.560597995453546, | |
| "learning_rate": 9.994965332706574e-06, | |
| "loss": 0.5788, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3546325878594249, | |
| "grad_norm": 0.577619417855181, | |
| "learning_rate": 9.992134075089085e-06, | |
| "loss": 0.5884, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.36421725239616615, | |
| "grad_norm": 0.7535402082009444, | |
| "learning_rate": 9.98867437523228e-06, | |
| "loss": 0.6038, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3738019169329074, | |
| "grad_norm": 0.631420386055008, | |
| "learning_rate": 9.984586668665641e-06, | |
| "loss": 0.5251, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.38338658146964855, | |
| "grad_norm": 0.5138239688062902, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.5577, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3929712460063898, | |
| "grad_norm": 0.6832561481924041, | |
| "learning_rate": 9.974529372743762e-06, | |
| "loss": 0.5743, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.402555910543131, | |
| "grad_norm": 0.6316465186524606, | |
| "learning_rate": 9.968561049466214e-06, | |
| "loss": 0.598, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.41214057507987223, | |
| "grad_norm": 0.6253306326370084, | |
| "learning_rate": 9.961967251474823e-06, | |
| "loss": 0.5583, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4217252396166134, | |
| "grad_norm": 0.5663333091253309, | |
| "learning_rate": 9.954748808839675e-06, | |
| "loss": 0.5705, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.43130990415335463, | |
| "grad_norm": 0.6403188899550863, | |
| "learning_rate": 9.946906630265184e-06, | |
| "loss": 0.5658, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.44089456869009586, | |
| "grad_norm": 0.5843819870750462, | |
| "learning_rate": 9.938441702975689e-06, | |
| "loss": 0.5614, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4504792332268371, | |
| "grad_norm": 0.6004701145280968, | |
| "learning_rate": 9.92935509259118e-06, | |
| "loss": 0.5787, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.46006389776357826, | |
| "grad_norm": 0.6535770247403028, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.5734, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4696485623003195, | |
| "grad_norm": 0.6171849783449403, | |
| "learning_rate": 9.909321476180594e-06, | |
| "loss": 0.5662, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.4792332268370607, | |
| "grad_norm": 0.5326540447085683, | |
| "learning_rate": 9.898376992116179e-06, | |
| "loss": 0.5599, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.48881789137380194, | |
| "grad_norm": 0.5757099019141989, | |
| "learning_rate": 9.886815868562596e-06, | |
| "loss": 0.5484, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4984025559105431, | |
| "grad_norm": 0.5344280073992373, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.5665, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5079872204472844, | |
| "grad_norm": 0.4635656351194223, | |
| "learning_rate": 9.861849601988384e-06, | |
| "loss": 0.5103, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5175718849840255, | |
| "grad_norm": 0.5034112087938782, | |
| "learning_rate": 9.848447601883436e-06, | |
| "loss": 0.5503, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5271565495207667, | |
| "grad_norm": 0.5284355726123227, | |
| "learning_rate": 9.834435247725032e-06, | |
| "loss": 0.5587, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.536741214057508, | |
| "grad_norm": 0.5262305666228788, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.5228, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5463258785942492, | |
| "grad_norm": 0.5609147310404334, | |
| "learning_rate": 9.804586609725499e-06, | |
| "loss": 0.5642, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5559105431309904, | |
| "grad_norm": 0.5120344081292038, | |
| "learning_rate": 9.788754083424654e-06, | |
| "loss": 0.5552, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5654952076677316, | |
| "grad_norm": 0.5146361935249247, | |
| "learning_rate": 9.772318717677905e-06, | |
| "loss": 0.5383, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.5750798722044729, | |
| "grad_norm": 0.5371214730445336, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.5392, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5846645367412141, | |
| "grad_norm": 0.5964691491945227, | |
| "learning_rate": 9.737647819437645e-06, | |
| "loss": 0.5362, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5942492012779552, | |
| "grad_norm": 0.48313239695565535, | |
| "learning_rate": 9.719416651541839e-06, | |
| "loss": 0.5144, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6038338658146964, | |
| "grad_norm": 0.6221345651275353, | |
| "learning_rate": 9.700591372846096e-06, | |
| "loss": 0.5127, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6134185303514377, | |
| "grad_norm": 0.5006680181156226, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.5117, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.6230031948881789, | |
| "grad_norm": 0.5382179363120568, | |
| "learning_rate": 9.661168036940071e-06, | |
| "loss": 0.5454, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6325878594249201, | |
| "grad_norm": 0.5385518159897789, | |
| "learning_rate": 9.640574942595195e-06, | |
| "loss": 0.5225, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6421725239616614, | |
| "grad_norm": 0.4689043255414553, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.5327, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6517571884984026, | |
| "grad_norm": 0.5292222732521383, | |
| "learning_rate": 9.597638862757255e-06, | |
| "loss": 0.5283, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6613418530351438, | |
| "grad_norm": 0.43950376287362897, | |
| "learning_rate": 9.5753012823366e-06, | |
| "loss": 0.5186, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.670926517571885, | |
| "grad_norm": 0.47233451081042727, | |
| "learning_rate": 9.552387733294081e-06, | |
| "loss": 0.5458, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6805111821086262, | |
| "grad_norm": 0.4628025571404574, | |
| "learning_rate": 9.528901100135971e-06, | |
| "loss": 0.5356, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6900958466453674, | |
| "grad_norm": 0.5126394234748157, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.5309, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6996805111821086, | |
| "grad_norm": 0.5308767674253421, | |
| "learning_rate": 9.480220479843627e-06, | |
| "loss": 0.5654, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7092651757188498, | |
| "grad_norm": 0.42179503602433727, | |
| "learning_rate": 9.45503262094184e-06, | |
| "loss": 0.541, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.7188498402555911, | |
| "grad_norm": 0.4798669525125099, | |
| "learning_rate": 9.4292839336179e-06, | |
| "loss": 0.5032, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7284345047923323, | |
| "grad_norm": 0.437998826621754, | |
| "learning_rate": 9.40297765928369e-06, | |
| "loss": 0.5409, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.7380191693290735, | |
| "grad_norm": 0.44824174730233135, | |
| "learning_rate": 9.376117109543769e-06, | |
| "loss": 0.536, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7476038338658147, | |
| "grad_norm": 0.4727220888171027, | |
| "learning_rate": 9.348705665778479e-06, | |
| "loss": 0.5268, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7571884984025559, | |
| "grad_norm": 0.4270490814671495, | |
| "learning_rate": 9.320746778718274e-06, | |
| "loss": 0.5163, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.7667731629392971, | |
| "grad_norm": 0.4809618776769443, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.5149, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7763578274760383, | |
| "grad_norm": 0.46662577657144244, | |
| "learning_rate": 9.263200821770462e-06, | |
| "loss": 0.5118, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7859424920127795, | |
| "grad_norm": 0.4962951829249085, | |
| "learning_rate": 9.233620996141421e-06, | |
| "loss": 0.546, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7955271565495208, | |
| "grad_norm": 0.4317332039379004, | |
| "learning_rate": 9.203508214822652e-06, | |
| "loss": 0.5441, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.805111821086262, | |
| "grad_norm": 0.5586390992151423, | |
| "learning_rate": 9.172866268606514e-06, | |
| "loss": 0.509, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.8146964856230032, | |
| "grad_norm": 0.4209710566538431, | |
| "learning_rate": 9.141699014900084e-06, | |
| "loss": 0.5024, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8242811501597445, | |
| "grad_norm": 0.5743724056993619, | |
| "learning_rate": 9.110010377239552e-06, | |
| "loss": 0.5221, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8338658146964856, | |
| "grad_norm": 0.5051105188903774, | |
| "learning_rate": 9.077804344796302e-06, | |
| "loss": 0.5288, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.8434504792332268, | |
| "grad_norm": 0.539069215163949, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.5514, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.853035143769968, | |
| "grad_norm": 0.5623984067731996, | |
| "learning_rate": 9.011856377401891e-06, | |
| "loss": 0.5018, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.8626198083067093, | |
| "grad_norm": 0.4588364024339851, | |
| "learning_rate": 8.978122744408905e-06, | |
| "loss": 0.5159, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8722044728434505, | |
| "grad_norm": 0.4725351294006038, | |
| "learning_rate": 8.943888319504456e-06, | |
| "loss": 0.5366, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8817891373801917, | |
| "grad_norm": 0.4885275771439361, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.5153, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8913738019169329, | |
| "grad_norm": 0.46982573258410526, | |
| "learning_rate": 8.873934395068006e-06, | |
| "loss": 0.5143, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.9009584664536742, | |
| "grad_norm": 0.5064009096592067, | |
| "learning_rate": 8.838223701790057e-06, | |
| "loss": 0.5283, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.9105431309904153, | |
| "grad_norm": 0.5071967330914865, | |
| "learning_rate": 8.802029828000157e-06, | |
| "loss": 0.5209, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9201277955271565, | |
| "grad_norm": 0.4620880317657483, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.5133, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9297124600638977, | |
| "grad_norm": 0.6120424568919531, | |
| "learning_rate": 8.728210824415829e-06, | |
| "loss": 0.5071, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.939297124600639, | |
| "grad_norm": 0.4957134561945823, | |
| "learning_rate": 8.690594987436705e-06, | |
| "loss": 0.5154, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9488817891373802, | |
| "grad_norm": 0.5020460143670394, | |
| "learning_rate": 8.652514554406388e-06, | |
| "loss": 0.5312, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.9584664536741214, | |
| "grad_norm": 0.6055544143836603, | |
| "learning_rate": 8.613974319136959e-06, | |
| "loss": 0.5176, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9680511182108626, | |
| "grad_norm": 0.4818922080620888, | |
| "learning_rate": 8.574979133323378e-06, | |
| "loss": 0.5019, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9776357827476039, | |
| "grad_norm": 0.4425519331117039, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.4947, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.987220447284345, | |
| "grad_norm": 0.5691939631842999, | |
| "learning_rate": 8.495643602586287e-06, | |
| "loss": 0.521, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9968051118210862, | |
| "grad_norm": 0.4827354946336946, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.5222, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.0063897763578276, | |
| "grad_norm": 0.9314354209643376, | |
| "learning_rate": 8.414547910024035e-06, | |
| "loss": 0.915, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.0159744408945688, | |
| "grad_norm": 0.6349106724675688, | |
| "learning_rate": 8.373352729660373e-06, | |
| "loss": 0.4384, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0255591054313098, | |
| "grad_norm": 0.6104761029257652, | |
| "learning_rate": 8.331732889760021e-06, | |
| "loss": 0.4684, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.035143769968051, | |
| "grad_norm": 0.4776108567051273, | |
| "learning_rate": 8.289693629698564e-06, | |
| "loss": 0.4708, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.0447284345047922, | |
| "grad_norm": 0.7614191288323594, | |
| "learning_rate": 8.247240241650918e-06, | |
| "loss": 0.5193, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.0543130990415335, | |
| "grad_norm": 0.4589450486506878, | |
| "learning_rate": 8.204378069925121e-06, | |
| "loss": 0.4573, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0638977635782747, | |
| "grad_norm": 0.5910279451821843, | |
| "learning_rate": 8.16111251028955e-06, | |
| "loss": 0.5152, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.073482428115016, | |
| "grad_norm": 0.44037431067948174, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.3978, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.0830670926517572, | |
| "grad_norm": 0.43462818393290725, | |
| "learning_rate": 8.073393063582386e-06, | |
| "loss": 0.4922, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0926517571884984, | |
| "grad_norm": 0.45022504859694273, | |
| "learning_rate": 8.0289502192041e-06, | |
| "loss": 0.4813, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.1022364217252396, | |
| "grad_norm": 0.4815299760405501, | |
| "learning_rate": 7.984126070912519e-06, | |
| "loss": 0.5044, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.1118210862619808, | |
| "grad_norm": 0.42846411195285283, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.46, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.121405750798722, | |
| "grad_norm": 0.516147601544945, | |
| "learning_rate": 7.89335648089903e-06, | |
| "loss": 0.4915, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.1309904153354633, | |
| "grad_norm": 0.3928542530235746, | |
| "learning_rate": 7.84742246584226e-06, | |
| "loss": 0.4433, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.1405750798722045, | |
| "grad_norm": 0.46662363198847323, | |
| "learning_rate": 7.801129998764014e-06, | |
| "loss": 0.5104, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.1501597444089458, | |
| "grad_norm": 0.426809172642443, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.4243, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.159744408945687, | |
| "grad_norm": 0.4804615577287185, | |
| "learning_rate": 7.70749306331863e-06, | |
| "loss": 0.5034, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.1693290734824282, | |
| "grad_norm": 0.5269310380967154, | |
| "learning_rate": 7.660160382576683e-06, | |
| "loss": 0.4491, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.1789137380191694, | |
| "grad_norm": 0.5315217742513717, | |
| "learning_rate": 7.612492823579744e-06, | |
| "loss": 0.4824, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.1884984025559104, | |
| "grad_norm": 0.49718868613651707, | |
| "learning_rate": 7.564496387029532e-06, | |
| "loss": 0.4901, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.1980830670926517, | |
| "grad_norm": 0.5176541476716205, | |
| "learning_rate": 7.516177115029002e-06, | |
| "loss": 0.5084, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.207667731629393, | |
| "grad_norm": 0.4380644233453936, | |
| "learning_rate": 7.467541090321735e-06, | |
| "loss": 0.466, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.2172523961661341, | |
| "grad_norm": 0.5415498882546443, | |
| "learning_rate": 7.4185944355261996e-06, | |
| "loss": 0.5182, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.2268370607028753, | |
| "grad_norm": 0.4070643826100599, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.4533, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.2364217252396166, | |
| "grad_norm": 0.4914075211952379, | |
| "learning_rate": 7.319793920889171e-06, | |
| "loss": 0.5024, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.2460063897763578, | |
| "grad_norm": 0.4951508354388962, | |
| "learning_rate": 7.269952498697734e-06, | |
| "loss": 0.5173, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.255591054313099, | |
| "grad_norm": 0.3985003578811613, | |
| "learning_rate": 7.219825320152411e-06, | |
| "loss": 0.4858, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.2651757188498403, | |
| "grad_norm": 0.524302859103588, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.4805, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.2747603833865815, | |
| "grad_norm": 0.4608934698481493, | |
| "learning_rate": 7.118738970516944e-06, | |
| "loss": 0.4751, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.2843450479233227, | |
| "grad_norm": 0.44246670089704204, | |
| "learning_rate": 7.067792524832604e-06, | |
| "loss": 0.5015, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.293929712460064, | |
| "grad_norm": 0.4620940578555168, | |
| "learning_rate": 7.016585772004026e-06, | |
| "loss": 0.5022, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.3035143769968052, | |
| "grad_norm": 0.42748829503391783, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.4763, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.3130990415335464, | |
| "grad_norm": 0.39365635758180245, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.4756, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.3226837060702876, | |
| "grad_norm": 0.4321184464708107, | |
| "learning_rate": 6.8614682920097265e-06, | |
| "loss": 0.4785, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.3322683706070286, | |
| "grad_norm": 0.44742588860045274, | |
| "learning_rate": 6.809285088483361e-06, | |
| "loss": 0.495, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.34185303514377, | |
| "grad_norm": 0.3939535232372297, | |
| "learning_rate": 6.7568741204067145e-06, | |
| "loss": 0.5052, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.351437699680511, | |
| "grad_norm": 0.390354646900369, | |
| "learning_rate": 6.704241985612625e-06, | |
| "loss": 0.4554, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.3610223642172525, | |
| "grad_norm": 0.5179114114410603, | |
| "learning_rate": 6.651395309775837e-06, | |
| "loss": 0.5017, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.3706070287539935, | |
| "grad_norm": 0.46077297707284576, | |
| "learning_rate": 6.598340745578908e-06, | |
| "loss": 0.4734, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.3801916932907348, | |
| "grad_norm": 0.3846892682729467, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.4403, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.389776357827476, | |
| "grad_norm": 0.5274229571998926, | |
| "learning_rate": 6.491634692845781e-06, | |
| "loss": 0.5301, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.3993610223642172, | |
| "grad_norm": 0.42055539348810306, | |
| "learning_rate": 6.437996637160086e-06, | |
| "loss": 0.4242, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.4089456869009584, | |
| "grad_norm": 0.5159236263141203, | |
| "learning_rate": 6.384177557124247e-06, | |
| "loss": 0.5033, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.4185303514376997, | |
| "grad_norm": 0.39286311206913027, | |
| "learning_rate": 6.330184227833376e-06, | |
| "loss": 0.4689, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.428115015974441, | |
| "grad_norm": 0.44762005973934715, | |
| "learning_rate": 6.276023446318214e-06, | |
| "loss": 0.4375, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.4376996805111821, | |
| "grad_norm": 0.5172846241311203, | |
| "learning_rate": 6.2217020306894705e-06, | |
| "loss": 0.5133, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4472843450479234, | |
| "grad_norm": 0.3994110285680104, | |
| "learning_rate": 6.1672268192795285e-06, | |
| "loss": 0.4392, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.4568690095846646, | |
| "grad_norm": 0.42379414167911805, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.5202, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.4664536741214058, | |
| "grad_norm": 0.3634842545965251, | |
| "learning_rate": 6.057842458386315e-06, | |
| "loss": 0.4381, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.476038338658147, | |
| "grad_norm": 0.3998915875264103, | |
| "learning_rate": 6.002947078916365e-06, | |
| "loss": 0.4308, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.4856230031948883, | |
| "grad_norm": 0.44449181661322107, | |
| "learning_rate": 5.947925441958393e-06, | |
| "loss": 0.5083, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.4952076677316293, | |
| "grad_norm": 0.3591024398917291, | |
| "learning_rate": 5.892784473993184e-06, | |
| "loss": 0.4524, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.5047923322683707, | |
| "grad_norm": 0.42938429215604684, | |
| "learning_rate": 5.837531116523683e-06, | |
| "loss": 0.452, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.5143769968051117, | |
| "grad_norm": 0.44760410123755834, | |
| "learning_rate": 5.782172325201155e-06, | |
| "loss": 0.4976, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.5239616613418532, | |
| "grad_norm": 0.36214497361824977, | |
| "learning_rate": 5.726715068949564e-06, | |
| "loss": 0.4006, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.5335463258785942, | |
| "grad_norm": 0.434866771300161, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.464, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5431309904153354, | |
| "grad_norm": 0.45200708389798794, | |
| "learning_rate": 5.615533098453215e-06, | |
| "loss": 0.5127, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.5527156549520766, | |
| "grad_norm": 0.4473591428543371, | |
| "learning_rate": 5.559822380516539e-06, | |
| "loss": 0.4287, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.5623003194888179, | |
| "grad_norm": 0.458282143104649, | |
| "learning_rate": 5.504041188505022e-06, | |
| "loss": 0.5366, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.571884984025559, | |
| "grad_norm": 0.39722618824348216, | |
| "learning_rate": 5.448196544517168e-06, | |
| "loss": 0.3943, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.5814696485623003, | |
| "grad_norm": 0.505628724820421, | |
| "learning_rate": 5.392295478639226e-06, | |
| "loss": 0.4858, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.5910543130990416, | |
| "grad_norm": 0.4132986111453938, | |
| "learning_rate": 5.336345028060199e-06, | |
| "loss": 0.4927, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.6006389776357828, | |
| "grad_norm": 0.45183647170475516, | |
| "learning_rate": 5.2803522361859596e-06, | |
| "loss": 0.5138, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.610223642172524, | |
| "grad_norm": 0.43519120543226025, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.4179, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.619808306709265, | |
| "grad_norm": 0.4748354840271485, | |
| "learning_rate": 5.168267827938971e-06, | |
| "loss": 0.5093, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.6293929712460065, | |
| "grad_norm": 0.351498291180111, | |
| "learning_rate": 5.112190321479026e-06, | |
| "loss": 0.3929, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6389776357827475, | |
| "grad_norm": 0.45576743121968355, | |
| "learning_rate": 5.05609869177323e-06, | |
| "loss": 0.5245, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.648562300319489, | |
| "grad_norm": 0.4286118858222971, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4593, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.65814696485623, | |
| "grad_norm": 0.3854186952560733, | |
| "learning_rate": 4.943901308226771e-06, | |
| "loss": 0.4332, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.6677316293929714, | |
| "grad_norm": 0.4261934687721202, | |
| "learning_rate": 4.887809678520976e-06, | |
| "loss": 0.4812, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.6773162939297124, | |
| "grad_norm": 0.4031374826535253, | |
| "learning_rate": 4.831732172061032e-06, | |
| "loss": 0.4362, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.6869009584664538, | |
| "grad_norm": 0.39268123732155447, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.4891, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.6964856230031948, | |
| "grad_norm": 0.37807443637067384, | |
| "learning_rate": 4.719647763814041e-06, | |
| "loss": 0.4691, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.706070287539936, | |
| "grad_norm": 0.4264428355090927, | |
| "learning_rate": 4.663654971939802e-06, | |
| "loss": 0.5074, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.7156549520766773, | |
| "grad_norm": 0.4105909401464241, | |
| "learning_rate": 4.6077045213607765e-06, | |
| "loss": 0.4566, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.7252396166134185, | |
| "grad_norm": 0.44481989652869997, | |
| "learning_rate": 4.551803455482833e-06, | |
| "loss": 0.4726, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.7348242811501597, | |
| "grad_norm": 0.39434949832508126, | |
| "learning_rate": 4.4959588114949785e-06, | |
| "loss": 0.4979, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.744408945686901, | |
| "grad_norm": 0.4015105577293125, | |
| "learning_rate": 4.4401776194834615e-06, | |
| "loss": 0.4598, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.7539936102236422, | |
| "grad_norm": 0.4072866915116779, | |
| "learning_rate": 4.384466901546786e-06, | |
| "loss": 0.4683, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.7635782747603834, | |
| "grad_norm": 0.43006771507548924, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.4778, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.7731629392971247, | |
| "grad_norm": 0.36887337855307, | |
| "learning_rate": 4.273284931050438e-06, | |
| "loss": 0.4609, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.7827476038338657, | |
| "grad_norm": 0.4323198930499473, | |
| "learning_rate": 4.217827674798845e-06, | |
| "loss": 0.4735, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.792332268370607, | |
| "grad_norm": 0.3965356425033133, | |
| "learning_rate": 4.162468883476319e-06, | |
| "loss": 0.4641, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.8019169329073481, | |
| "grad_norm": 0.42534997339395686, | |
| "learning_rate": 4.107215526006818e-06, | |
| "loss": 0.4563, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.8115015974440896, | |
| "grad_norm": 0.41383446059928714, | |
| "learning_rate": 4.052074558041608e-06, | |
| "loss": 0.4848, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.8210862619808306, | |
| "grad_norm": 0.364275091281534, | |
| "learning_rate": 3.997052921083637e-06, | |
| "loss": 0.4648, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.830670926517572, | |
| "grad_norm": 0.43527796346056263, | |
| "learning_rate": 3.9421575416136866e-06, | |
| "loss": 0.502, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.840255591054313, | |
| "grad_norm": 0.41590733735533003, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.4705, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.8498402555910545, | |
| "grad_norm": 0.4202363258362791, | |
| "learning_rate": 3.832773180720475e-06, | |
| "loss": 0.4707, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.8594249201277955, | |
| "grad_norm": 0.4137149217601345, | |
| "learning_rate": 3.778297969310529e-06, | |
| "loss": 0.4478, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.8690095846645367, | |
| "grad_norm": 0.37533390746112527, | |
| "learning_rate": 3.723976553681787e-06, | |
| "loss": 0.435, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.878594249201278, | |
| "grad_norm": 0.40035552994094187, | |
| "learning_rate": 3.669815772166625e-06, | |
| "loss": 0.4378, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.8881789137380192, | |
| "grad_norm": 0.4593014845015901, | |
| "learning_rate": 3.6158224428757538e-06, | |
| "loss": 0.4758, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.8977635782747604, | |
| "grad_norm": 0.40352135973956654, | |
| "learning_rate": 3.562003362839914e-06, | |
| "loss": 0.494, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.9073482428115016, | |
| "grad_norm": 0.3677042327773077, | |
| "learning_rate": 3.50836530715422e-06, | |
| "loss": 0.4644, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.9169329073482428, | |
| "grad_norm": 0.4584617447317549, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.5103, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.926517571884984, | |
| "grad_norm": 0.4006678101981588, | |
| "learning_rate": 3.4016592544210937e-06, | |
| "loss": 0.4887, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.9361022364217253, | |
| "grad_norm": 0.37463082710114304, | |
| "learning_rate": 3.3486046902241663e-06, | |
| "loss": 0.46, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.9456869009584663, | |
| "grad_norm": 0.3717196797809635, | |
| "learning_rate": 3.295758014387375e-06, | |
| "loss": 0.4612, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.9552715654952078, | |
| "grad_norm": 0.45125297588813473, | |
| "learning_rate": 3.2431258795932863e-06, | |
| "loss": 0.4954, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.9648562300319488, | |
| "grad_norm": 0.4049386598381838, | |
| "learning_rate": 3.1907149115166403e-06, | |
| "loss": 0.4873, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.9744408945686902, | |
| "grad_norm": 0.41325288292429, | |
| "learning_rate": 3.1385317079902743e-06, | |
| "loss": 0.4661, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.9840255591054312, | |
| "grad_norm": 0.4335324460553787, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.457, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.9936102236421727, | |
| "grad_norm": 0.4396120820992226, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.5035, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.0031948881789137, | |
| "grad_norm": 0.6917669610680982, | |
| "learning_rate": 2.9834142279959754e-06, | |
| "loss": 0.6896, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.012779552715655, | |
| "grad_norm": 0.4412739849029467, | |
| "learning_rate": 2.932207475167398e-06, | |
| "loss": 0.4463, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.022364217252396, | |
| "grad_norm": 0.40961457427380926, | |
| "learning_rate": 2.8812610294830568e-06, | |
| "loss": 0.4534, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.0319488817891376, | |
| "grad_norm": 0.36828997966826105, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.4202, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.0415335463258786, | |
| "grad_norm": 0.40781329630746305, | |
| "learning_rate": 2.7801746798475905e-06, | |
| "loss": 0.4707, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.0511182108626196, | |
| "grad_norm": 0.41211693092573937, | |
| "learning_rate": 2.7300475013022666e-06, | |
| "loss": 0.4331, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.060702875399361, | |
| "grad_norm": 0.38367657052142934, | |
| "learning_rate": 2.6802060791108304e-06, | |
| "loss": 0.414, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.070287539936102, | |
| "grad_norm": 0.40439847386598615, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.4774, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.0798722044728435, | |
| "grad_norm": 0.39238906175963617, | |
| "learning_rate": 2.5814055644738013e-06, | |
| "loss": 0.4315, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.0894568690095845, | |
| "grad_norm": 0.379711586064365, | |
| "learning_rate": 2.532458909678266e-06, | |
| "loss": 0.4439, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.099041533546326, | |
| "grad_norm": 0.3857028125193212, | |
| "learning_rate": 2.483822884971e-06, | |
| "loss": 0.4014, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.108626198083067, | |
| "grad_norm": 0.39697984988186674, | |
| "learning_rate": 2.43550361297047e-06, | |
| "loss": 0.432, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.1182108626198084, | |
| "grad_norm": 0.43185067273414596, | |
| "learning_rate": 2.387507176420256e-06, | |
| "loss": 0.448, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.1277955271565494, | |
| "grad_norm": 0.36811647776061696, | |
| "learning_rate": 2.339839617423318e-06, | |
| "loss": 0.4039, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.137380191693291, | |
| "grad_norm": 0.3851766988271702, | |
| "learning_rate": 2.2925069366813718e-06, | |
| "loss": 0.4392, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.146964856230032, | |
| "grad_norm": 0.3824630039068114, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.4331, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.1565495207667733, | |
| "grad_norm": 0.41526298733575523, | |
| "learning_rate": 2.1988700012359865e-06, | |
| "loss": 0.4751, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.1661341853035143, | |
| "grad_norm": 0.3713218550182188, | |
| "learning_rate": 2.1525775341577404e-06, | |
| "loss": 0.3917, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.1757188498402558, | |
| "grad_norm": 0.3843189913444912, | |
| "learning_rate": 2.1066435191009717e-06, | |
| "loss": 0.4214, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.1853035143769968, | |
| "grad_norm": 0.3897613915388535, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.4672, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.194888178913738, | |
| "grad_norm": 0.35116315860464736, | |
| "learning_rate": 2.0158739290874822e-06, | |
| "loss": 0.4295, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.2044728434504792, | |
| "grad_norm": 0.38459525840293196, | |
| "learning_rate": 1.971049780795901e-06, | |
| "loss": 0.4691, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.2140575079872207, | |
| "grad_norm": 0.3478489653448972, | |
| "learning_rate": 1.9266069364176144e-06, | |
| "loss": 0.3848, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.2236421725239617, | |
| "grad_norm": 0.3788719346551245, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.4645, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.2332268370607027, | |
| "grad_norm": 0.3799875682505591, | |
| "learning_rate": 1.838887489710452e-06, | |
| "loss": 0.4481, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.242811501597444, | |
| "grad_norm": 0.3554435625384451, | |
| "learning_rate": 1.7956219300748796e-06, | |
| "loss": 0.423, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.252396166134185, | |
| "grad_norm": 0.35697587226247235, | |
| "learning_rate": 1.7527597583490825e-06, | |
| "loss": 0.4513, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.2619808306709266, | |
| "grad_norm": 0.3445045795998756, | |
| "learning_rate": 1.7103063703014372e-06, | |
| "loss": 0.3933, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.2715654952076676, | |
| "grad_norm": 0.33493068570219264, | |
| "learning_rate": 1.6682671102399806e-06, | |
| "loss": 0.4241, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.281150159744409, | |
| "grad_norm": 0.3775328797092562, | |
| "learning_rate": 1.6266472703396286e-06, | |
| "loss": 0.4194, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.29073482428115, | |
| "grad_norm": 0.39057219804229737, | |
| "learning_rate": 1.5854520899759656e-06, | |
| "loss": 0.4568, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.3003194888178915, | |
| "grad_norm": 0.37467812273702855, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.4108, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.3099041533546325, | |
| "grad_norm": 0.3170699590955595, | |
| "learning_rate": 1.5043563974137132e-06, | |
| "loss": 0.3627, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.319488817891374, | |
| "grad_norm": 0.35938177592803017, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.4612, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.329073482428115, | |
| "grad_norm": 0.32332197878849234, | |
| "learning_rate": 1.4250208666766235e-06, | |
| "loss": 0.3903, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.3386581469648564, | |
| "grad_norm": 0.35410314975398555, | |
| "learning_rate": 1.3860256808630429e-06, | |
| "loss": 0.4426, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.3482428115015974, | |
| "grad_norm": 0.35578224407251896, | |
| "learning_rate": 1.3474854455936126e-06, | |
| "loss": 0.4374, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.357827476038339, | |
| "grad_norm": 0.3661378086886109, | |
| "learning_rate": 1.3094050125632973e-06, | |
| "loss": 0.4555, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.36741214057508, | |
| "grad_norm": 0.33396832853545655, | |
| "learning_rate": 1.2717891755841722e-06, | |
| "loss": 0.3781, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.376996805111821, | |
| "grad_norm": 0.3662542560305177, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.4462, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.3865814696485623, | |
| "grad_norm": 0.35308360226374075, | |
| "learning_rate": 1.1979701719998454e-06, | |
| "loss": 0.4445, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.3961661341853033, | |
| "grad_norm": 0.33493300536805287, | |
| "learning_rate": 1.1617762982099446e-06, | |
| "loss": 0.4371, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.405750798722045, | |
| "grad_norm": 0.358400742170288, | |
| "learning_rate": 1.1260656049319957e-06, | |
| "loss": 0.4408, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.415335463258786, | |
| "grad_norm": 0.3352245283886859, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.468, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.4249201277955272, | |
| "grad_norm": 0.3674441639041068, | |
| "learning_rate": 1.0561116804955451e-06, | |
| "loss": 0.4251, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.4345047923322682, | |
| "grad_norm": 0.35571070097051444, | |
| "learning_rate": 1.0218772555910955e-06, | |
| "loss": 0.3993, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.4440894568690097, | |
| "grad_norm": 0.3314576767115369, | |
| "learning_rate": 9.881436225981107e-07, | |
| "loss": 0.4545, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.4536741214057507, | |
| "grad_norm": 0.3193938077931927, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.4266, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.463258785942492, | |
| "grad_norm": 0.3156721980778057, | |
| "learning_rate": 9.221956552036992e-07, | |
| "loss": 0.4171, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.472843450479233, | |
| "grad_norm": 0.3284367383055379, | |
| "learning_rate": 8.899896227604509e-07, | |
| "loss": 0.3893, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.4824281150159746, | |
| "grad_norm": 0.36445938982620985, | |
| "learning_rate": 8.58300985099918e-07, | |
| "loss": 0.4389, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.4920127795527156, | |
| "grad_norm": 0.35617212056141584, | |
| "learning_rate": 8.271337313934869e-07, | |
| "loss": 0.4863, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.501597444089457, | |
| "grad_norm": 0.3515264363535228, | |
| "learning_rate": 7.964917851773496e-07, | |
| "loss": 0.4392, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.511182108626198, | |
| "grad_norm": 0.33265459392905583, | |
| "learning_rate": 7.663790038585794e-07, | |
| "loss": 0.4208, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.520766773162939, | |
| "grad_norm": 0.34820794755597795, | |
| "learning_rate": 7.367991782295392e-07, | |
| "loss": 0.4424, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.5303514376996805, | |
| "grad_norm": 0.3481132770441028, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.4282, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.539936102236422, | |
| "grad_norm": 0.3164816747033596, | |
| "learning_rate": 6.792532212817271e-07, | |
| "loss": 0.4192, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.549520766773163, | |
| "grad_norm": 0.3322445817366458, | |
| "learning_rate": 6.512943342215234e-07, | |
| "loss": 0.4672, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.559105431309904, | |
| "grad_norm": 0.3276953525154931, | |
| "learning_rate": 6.238828904562316e-07, | |
| "loss": 0.4359, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.5686900958466454, | |
| "grad_norm": 0.35146245455025726, | |
| "learning_rate": 5.9702234071631e-07, | |
| "loss": 0.4338, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.5782747603833864, | |
| "grad_norm": 0.3413515111649245, | |
| "learning_rate": 5.707160663821009e-07, | |
| "loss": 0.3954, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.587859424920128, | |
| "grad_norm": 0.3344424262221673, | |
| "learning_rate": 5.449673790581611e-07, | |
| "loss": 0.4193, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.597444089456869, | |
| "grad_norm": 0.35641204717794256, | |
| "learning_rate": 5.197795201563744e-07, | |
| "loss": 0.4431, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.6070287539936103, | |
| "grad_norm": 0.30996088276643385, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.4221, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.6166134185303513, | |
| "grad_norm": 0.3367375510355728, | |
| "learning_rate": 4.710988998640298e-07, | |
| "loss": 0.4535, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.626198083067093, | |
| "grad_norm": 0.33760776912528123, | |
| "learning_rate": 4.4761226670592074e-07, | |
| "loss": 0.4426, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.635782747603834, | |
| "grad_norm": 0.33822421672650754, | |
| "learning_rate": 4.2469871766340096e-07, | |
| "loss": 0.4564, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.6453674121405752, | |
| "grad_norm": 0.3165958916410871, | |
| "learning_rate": 4.0236113724274716e-07, | |
| "loss": 0.4387, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.6549520766773163, | |
| "grad_norm": 0.3166114641674283, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.4355, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.6645367412140573, | |
| "grad_norm": 0.32357628063325744, | |
| "learning_rate": 3.5942505740480583e-07, | |
| "loss": 0.4079, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.6741214057507987, | |
| "grad_norm": 0.3674207025641318, | |
| "learning_rate": 3.3883196305992906e-07, | |
| "loss": 0.474, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.68370607028754, | |
| "grad_norm": 0.3227250777202084, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.4111, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.693290734824281, | |
| "grad_norm": 0.3412839969181316, | |
| "learning_rate": 2.9940862715390483e-07, | |
| "loss": 0.4383, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.702875399361022, | |
| "grad_norm": 0.31992392443231005, | |
| "learning_rate": 2.8058334845816214e-07, | |
| "loss": 0.4313, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.7124600638977636, | |
| "grad_norm": 0.3439074759306473, | |
| "learning_rate": 2.6235218056235633e-07, | |
| "loss": 0.4198, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.722044728434505, | |
| "grad_norm": 0.29603175883785177, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.3918, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.731629392971246, | |
| "grad_norm": 0.3343645398366866, | |
| "learning_rate": 2.276812823220964e-07, | |
| "loss": 0.4532, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.741214057507987, | |
| "grad_norm": 0.32261947554889003, | |
| "learning_rate": 2.1124591657534776e-07, | |
| "loss": 0.412, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.7507987220447285, | |
| "grad_norm": 0.3396612301497462, | |
| "learning_rate": 1.9541339027450256e-07, | |
| "loss": 0.4587, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.7603833865814695, | |
| "grad_norm": 0.33472142260056553, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.4595, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.769968051118211, | |
| "grad_norm": 0.3242842872825828, | |
| "learning_rate": 1.6556475227496816e-07, | |
| "loss": 0.4236, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.779552715654952, | |
| "grad_norm": 0.3175778106446133, | |
| "learning_rate": 1.5155239811656562e-07, | |
| "loss": 0.4257, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.7891373801916934, | |
| "grad_norm": 0.3289614099966931, | |
| "learning_rate": 1.3815039801161723e-07, | |
| "loss": 0.4913, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.7987220447284344, | |
| "grad_norm": 0.30791769668762514, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.3783, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.8083067092651754, | |
| "grad_norm": 0.3320693876956837, | |
| "learning_rate": 1.1318413143740436e-07, | |
| "loss": 0.432, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.817891373801917, | |
| "grad_norm": 0.3074885619997424, | |
| "learning_rate": 1.0162300788382263e-07, | |
| "loss": 0.4224, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.8274760383386583, | |
| "grad_norm": 0.34636412261842403, | |
| "learning_rate": 9.0678523819408e-08, | |
| "loss": 0.501, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.8370607028753994, | |
| "grad_norm": 0.3151556818210804, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.396, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.8466453674121404, | |
| "grad_norm": 0.3230604897560115, | |
| "learning_rate": 7.064490740882057e-08, | |
| "loss": 0.4608, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.856230031948882, | |
| "grad_norm": 0.31345682144863746, | |
| "learning_rate": 6.15582970243117e-08, | |
| "loss": 0.4249, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.8658146964856233, | |
| "grad_norm": 0.3340356755237237, | |
| "learning_rate": 5.3093369734816824e-08, | |
| "loss": 0.4764, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.8753993610223643, | |
| "grad_norm": 0.29707978963053755, | |
| "learning_rate": 4.52511911603265e-08, | |
| "loss": 0.3852, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.8849840255591053, | |
| "grad_norm": 0.317923141392632, | |
| "learning_rate": 3.8032748525179684e-08, | |
| "loss": 0.4481, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.8945686900958467, | |
| "grad_norm": 0.3289514478243867, | |
| "learning_rate": 3.143895053378698e-08, | |
| "loss": 0.4728, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.9041533546325877, | |
| "grad_norm": 0.3063433105982972, | |
| "learning_rate": 2.547062725623828e-08, | |
| "loss": 0.4466, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.913738019169329, | |
| "grad_norm": 0.32532721274765475, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.477, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.92332268370607, | |
| "grad_norm": 0.2989852424499799, | |
| "learning_rate": 1.541333133436018e-08, | |
| "loss": 0.3868, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.9329073482428116, | |
| "grad_norm": 0.30827875699085777, | |
| "learning_rate": 1.132562476771959e-08, | |
| "loss": 0.413, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.9424920127795526, | |
| "grad_norm": 0.3623133712269534, | |
| "learning_rate": 7.865924910916977e-09, | |
| "loss": 0.5252, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.952076677316294, | |
| "grad_norm": 0.30425224919672916, | |
| "learning_rate": 5.034667293427053e-09, | |
| "loss": 0.408, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.961661341853035, | |
| "grad_norm": 0.31234286411700085, | |
| "learning_rate": 2.8322083323334417e-09, | |
| "loss": 0.4236, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.9712460063897765, | |
| "grad_norm": 0.31975718528398644, | |
| "learning_rate": 1.2588252874673469e-09, | |
| "loss": 0.4601, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.9808306709265175, | |
| "grad_norm": 0.2990705875145571, | |
| "learning_rate": 3.147162264971471e-10, | |
| "loss": 0.4209, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.9904153354632586, | |
| "grad_norm": 0.31441532508599973, | |
| "learning_rate": 0.0, | |
| "loss": 0.4239, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.9904153354632586, | |
| "step": 312, | |
| "total_flos": 4.4115474684667494e+17, | |
| "train_loss": 0.5030276312086827, | |
| "train_runtime": 10179.8811, | |
| "train_samples_per_second": 2.946, | |
| "train_steps_per_second": 0.031 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 312, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.4115474684667494e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |