| { | |
| "best_metric": 2.310253143310547, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 0.24837007140639553, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0012418503570319776, | |
| "grad_norm": 55.993560791015625, | |
| "learning_rate": 1e-05, | |
| "loss": 12.8678, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0012418503570319776, | |
| "eval_loss": 3.4761929512023926, | |
| "eval_runtime": 120.7304, | |
| "eval_samples_per_second": 11.232, | |
| "eval_steps_per_second": 2.808, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0024837007140639552, | |
| "grad_norm": 45.881263732910156, | |
| "learning_rate": 2e-05, | |
| "loss": 13.1715, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.003725551071095933, | |
| "grad_norm": 30.949663162231445, | |
| "learning_rate": 3e-05, | |
| "loss": 12.6575, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0049674014281279105, | |
| "grad_norm": 37.812198638916016, | |
| "learning_rate": 4e-05, | |
| "loss": 12.3519, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.006209251785159888, | |
| "grad_norm": 21.051006317138672, | |
| "learning_rate": 5e-05, | |
| "loss": 11.8649, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.007451102142191866, | |
| "grad_norm": 21.31414222717285, | |
| "learning_rate": 6e-05, | |
| "loss": 12.0152, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.008692952499223843, | |
| "grad_norm": 10.278292655944824, | |
| "learning_rate": 7e-05, | |
| "loss": 12.1357, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.009934802856255821, | |
| "grad_norm": 5.365834712982178, | |
| "learning_rate": 8e-05, | |
| "loss": 12.1292, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.011176653213287799, | |
| "grad_norm": 15.813675880432129, | |
| "learning_rate": 9e-05, | |
| "loss": 12.3614, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.012418503570319777, | |
| "grad_norm": 12.399641036987305, | |
| "learning_rate": 0.0001, | |
| "loss": 12.643, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.013660353927351754, | |
| "grad_norm": 5.817680835723877, | |
| "learning_rate": 9.99983777858264e-05, | |
| "loss": 12.9288, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.014902204284383732, | |
| "grad_norm": 39.127357482910156, | |
| "learning_rate": 9.999351124856874e-05, | |
| "loss": 11.9864, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.01614405464141571, | |
| "grad_norm": 8.296622276306152, | |
| "learning_rate": 9.998540070400966e-05, | |
| "loss": 12.1663, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.017385904998447686, | |
| "grad_norm": 3.8759069442749023, | |
| "learning_rate": 9.997404667843075e-05, | |
| "loss": 11.9151, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.018627755355479666, | |
| "grad_norm": 2.11600399017334, | |
| "learning_rate": 9.995944990857849e-05, | |
| "loss": 11.8678, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.019869605712511642, | |
| "grad_norm": 2.760542631149292, | |
| "learning_rate": 9.994161134161634e-05, | |
| "loss": 11.8883, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.02111145606954362, | |
| "grad_norm": 2.460388422012329, | |
| "learning_rate": 9.992053213506334e-05, | |
| "loss": 12.0369, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.022353306426575598, | |
| "grad_norm": 2.8548941612243652, | |
| "learning_rate": 9.989621365671902e-05, | |
| "loss": 11.8477, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.023595156783607574, | |
| "grad_norm": 3.272533655166626, | |
| "learning_rate": 9.986865748457457e-05, | |
| "loss": 11.5853, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.024837007140639553, | |
| "grad_norm": 9.008646011352539, | |
| "learning_rate": 9.983786540671051e-05, | |
| "loss": 11.9561, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02607885749767153, | |
| "grad_norm": 10.807662963867188, | |
| "learning_rate": 9.980383942118066e-05, | |
| "loss": 11.7804, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.02732070785470351, | |
| "grad_norm": 21.672609329223633, | |
| "learning_rate": 9.976658173588244e-05, | |
| "loss": 12.346, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.028562558211735485, | |
| "grad_norm": 12.436042785644531, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 12.2528, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.029804408568767465, | |
| "grad_norm": 14.396573066711426, | |
| "learning_rate": 9.968238114591566e-05, | |
| "loss": 12.0058, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.03104625892579944, | |
| "grad_norm": 4.903384208679199, | |
| "learning_rate": 9.96354437049027e-05, | |
| "loss": 11.9545, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.03228810928283142, | |
| "grad_norm": 3.9211270809173584, | |
| "learning_rate": 9.95852854910781e-05, | |
| "loss": 12.0176, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.033529959639863396, | |
| "grad_norm": 26.65906524658203, | |
| "learning_rate": 9.953190975913647e-05, | |
| "loss": 12.0961, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.03477180999689537, | |
| "grad_norm": 20.78506851196289, | |
| "learning_rate": 9.947531997255256e-05, | |
| "loss": 12.0395, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.03601366035392735, | |
| "grad_norm": 2.8185300827026367, | |
| "learning_rate": 9.941551980335652e-05, | |
| "loss": 11.7895, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.03725551071095933, | |
| "grad_norm": 9.578609466552734, | |
| "learning_rate": 9.935251313189564e-05, | |
| "loss": 12.0755, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03849736106799131, | |
| "grad_norm": 3.954073429107666, | |
| "learning_rate": 9.928630404658255e-05, | |
| "loss": 11.6921, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.039739211425023284, | |
| "grad_norm": 8.112652778625488, | |
| "learning_rate": 9.921689684362989e-05, | |
| "loss": 12.04, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.04098106178205526, | |
| "grad_norm": 3.787114143371582, | |
| "learning_rate": 9.914429602677162e-05, | |
| "loss": 11.9404, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.04222291213908724, | |
| "grad_norm": 9.006427764892578, | |
| "learning_rate": 9.906850630697068e-05, | |
| "loss": 11.8041, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.04346476249611922, | |
| "grad_norm": 2.9068195819854736, | |
| "learning_rate": 9.898953260211338e-05, | |
| "loss": 11.8365, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.044706612853151195, | |
| "grad_norm": 12.303101539611816, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 12.2312, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.04594846321018317, | |
| "grad_norm": 7.705003261566162, | |
| "learning_rate": 9.882205394146361e-05, | |
| "loss": 12.1321, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.04719031356721515, | |
| "grad_norm": 3.9026267528533936, | |
| "learning_rate": 9.87335598531214e-05, | |
| "loss": 11.7614, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.04843216392424713, | |
| "grad_norm": 4.167300701141357, | |
| "learning_rate": 9.864190351391822e-05, | |
| "loss": 11.9624, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.04967401428127911, | |
| "grad_norm": 6.323347568511963, | |
| "learning_rate": 9.85470908713026e-05, | |
| "loss": 11.8362, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.05091586463831108, | |
| "grad_norm": 5.323836803436279, | |
| "learning_rate": 9.844912807753104e-05, | |
| "loss": 11.8706, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.05215771499534306, | |
| "grad_norm": 5.965713024139404, | |
| "learning_rate": 9.834802148926882e-05, | |
| "loss": 11.5284, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.05339956535237504, | |
| "grad_norm": 4.599259853363037, | |
| "learning_rate": 9.824377766717759e-05, | |
| "loss": 11.6113, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.05464141570940702, | |
| "grad_norm": 17.384479522705078, | |
| "learning_rate": 9.813640337548954e-05, | |
| "loss": 11.5999, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.055883266066438994, | |
| "grad_norm": 9.93163013458252, | |
| "learning_rate": 9.802590558156862e-05, | |
| "loss": 11.7976, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.05712511642347097, | |
| "grad_norm": 8.654122352600098, | |
| "learning_rate": 9.791229145545831e-05, | |
| "loss": 11.1107, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.058366966780502946, | |
| "grad_norm": 10.904827117919922, | |
| "learning_rate": 9.779556836941645e-05, | |
| "loss": 11.3452, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.05960881713753493, | |
| "grad_norm": 8.286592483520508, | |
| "learning_rate": 9.767574389743682e-05, | |
| "loss": 11.0275, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.060850667494566905, | |
| "grad_norm": 6.991479396820068, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 9.8946, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.06209251785159888, | |
| "grad_norm": 14.688076972961426, | |
| "learning_rate": 9.742682209735727e-05, | |
| "loss": 11.1673, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.06209251785159888, | |
| "eval_loss": 2.852301836013794, | |
| "eval_runtime": 121.8017, | |
| "eval_samples_per_second": 11.133, | |
| "eval_steps_per_second": 2.783, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.06333436820863086, | |
| "grad_norm": 20.07223129272461, | |
| "learning_rate": 9.729774092143627e-05, | |
| "loss": 10.9665, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.06457621856566284, | |
| "grad_norm": 19.648399353027344, | |
| "learning_rate": 9.716559066288715e-05, | |
| "loss": 11.0387, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.06581806892269482, | |
| "grad_norm": 9.359997749328613, | |
| "learning_rate": 9.703037989675087e-05, | |
| "loss": 11.5768, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.06705991927972679, | |
| "grad_norm": 11.937564849853516, | |
| "learning_rate": 9.689211739666023e-05, | |
| "loss": 11.7724, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.06830176963675877, | |
| "grad_norm": 7.769301891326904, | |
| "learning_rate": 9.675081213427076e-05, | |
| "loss": 11.7561, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.06954361999379075, | |
| "grad_norm": 3.720431327819824, | |
| "learning_rate": 9.66064732786784e-05, | |
| "loss": 12.0489, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.07078547035082272, | |
| "grad_norm": 4.992816925048828, | |
| "learning_rate": 9.645911019582467e-05, | |
| "loss": 12.0206, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.0720273207078547, | |
| "grad_norm": 3.545199155807495, | |
| "learning_rate": 9.630873244788883e-05, | |
| "loss": 11.9533, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.07326917106488669, | |
| "grad_norm": 12.412578582763672, | |
| "learning_rate": 9.615534979266745e-05, | |
| "loss": 12.097, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.07451102142191866, | |
| "grad_norm": 17.594213485717773, | |
| "learning_rate": 9.599897218294122e-05, | |
| "loss": 12.6395, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.07575287177895064, | |
| "grad_norm": 12.056193351745605, | |
| "learning_rate": 9.583960976582913e-05, | |
| "loss": 12.5772, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.07699472213598262, | |
| "grad_norm": 11.694857597351074, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 11.6403, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.07823657249301459, | |
| "grad_norm": 4.9527387619018555, | |
| "learning_rate": 9.551197206565173e-05, | |
| "loss": 11.3746, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.07947842285004657, | |
| "grad_norm": 7.177594184875488, | |
| "learning_rate": 9.534371804252728e-05, | |
| "loss": 12.0748, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.08072027320707854, | |
| "grad_norm": 5.73252010345459, | |
| "learning_rate": 9.517252173051911e-05, | |
| "loss": 11.5529, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.08196212356411052, | |
| "grad_norm": 5.650352478027344, | |
| "learning_rate": 9.49983942383106e-05, | |
| "loss": 11.7497, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.0832039739211425, | |
| "grad_norm": 8.362493515014648, | |
| "learning_rate": 9.482134686478519e-05, | |
| "loss": 11.8166, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.08444582427817449, | |
| "grad_norm": 7.898835182189941, | |
| "learning_rate": 9.464139109829321e-05, | |
| "loss": 11.4327, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.08568767463520646, | |
| "grad_norm": 8.106947898864746, | |
| "learning_rate": 9.445853861590647e-05, | |
| "loss": 10.6585, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.08692952499223844, | |
| "grad_norm": 5.23832893371582, | |
| "learning_rate": 9.42728012826605e-05, | |
| "loss": 11.172, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.08817137534927041, | |
| "grad_norm": 6.251285076141357, | |
| "learning_rate": 9.408419115078471e-05, | |
| "loss": 10.9628, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.08941322570630239, | |
| "grad_norm": 4.437039375305176, | |
| "learning_rate": 9.389272045892024e-05, | |
| "loss": 11.5214, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.09065507606333437, | |
| "grad_norm": 3.935523509979248, | |
| "learning_rate": 9.36984016313259e-05, | |
| "loss": 11.6944, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.09189692642036634, | |
| "grad_norm": 5.158120632171631, | |
| "learning_rate": 9.350124727707197e-05, | |
| "loss": 11.274, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.09313877677739832, | |
| "grad_norm": 6.234116554260254, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 11.3354, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.0943806271344303, | |
| "grad_norm": 6.205196380615234, | |
| "learning_rate": 9.309848334400246e-05, | |
| "loss": 11.2995, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.09562247749146228, | |
| "grad_norm": 8.435301780700684, | |
| "learning_rate": 9.289289989996133e-05, | |
| "loss": 11.0732, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.09686432784849426, | |
| "grad_norm": 8.505961418151855, | |
| "learning_rate": 9.268453319711363e-05, | |
| "loss": 11.6349, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.09810617820552624, | |
| "grad_norm": 7.8026509284973145, | |
| "learning_rate": 9.247339675607605e-05, | |
| "loss": 11.8426, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.09934802856255821, | |
| "grad_norm": 8.800403594970703, | |
| "learning_rate": 9.225950427718975e-05, | |
| "loss": 10.6523, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.10058987891959019, | |
| "grad_norm": 8.991898536682129, | |
| "learning_rate": 9.204286963963111e-05, | |
| "loss": 11.1845, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.10183172927662217, | |
| "grad_norm": 11.886899948120117, | |
| "learning_rate": 9.182350690051133e-05, | |
| "loss": 10.9825, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.10307357963365414, | |
| "grad_norm": 9.59271240234375, | |
| "learning_rate": 9.160143029396422e-05, | |
| "loss": 11.1428, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.10431542999068612, | |
| "grad_norm": 8.187872886657715, | |
| "learning_rate": 9.13766542302225e-05, | |
| "loss": 10.9437, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.1055572803477181, | |
| "grad_norm": 8.21299934387207, | |
| "learning_rate": 9.114919329468282e-05, | |
| "loss": 10.8253, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.10679913070475008, | |
| "grad_norm": 9.130791664123535, | |
| "learning_rate": 9.091906224695935e-05, | |
| "loss": 10.7161, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.10804098106178206, | |
| "grad_norm": 8.906632423400879, | |
| "learning_rate": 9.068627601992598e-05, | |
| "loss": 11.6251, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.10928283141881404, | |
| "grad_norm": 11.780891418457031, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 11.2236, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.11052468177584601, | |
| "grad_norm": 10.903058052062988, | |
| "learning_rate": 9.021279861989885e-05, | |
| "loss": 11.036, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.11176653213287799, | |
| "grad_norm": 16.120445251464844, | |
| "learning_rate": 8.997213817017507e-05, | |
| "loss": 9.8846, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.11300838248990996, | |
| "grad_norm": 13.418635368347168, | |
| "learning_rate": 8.972888398568772e-05, | |
| "loss": 10.8782, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.11425023284694194, | |
| "grad_norm": 11.932881355285645, | |
| "learning_rate": 8.948305185085225e-05, | |
| "loss": 10.7951, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.11549208320397392, | |
| "grad_norm": 10.037215232849121, | |
| "learning_rate": 8.92346577173636e-05, | |
| "loss": 9.9729, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.11673393356100589, | |
| "grad_norm": 13.088410377502441, | |
| "learning_rate": 8.898371770316111e-05, | |
| "loss": 9.0308, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.11797578391803788, | |
| "grad_norm": 17.424819946289062, | |
| "learning_rate": 8.873024809138272e-05, | |
| "loss": 10.7438, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.11921763427506986, | |
| "grad_norm": 27.27771759033203, | |
| "learning_rate": 8.847426532930831e-05, | |
| "loss": 9.2174, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.12045948463210183, | |
| "grad_norm": 12.472175598144531, | |
| "learning_rate": 8.821578602729242e-05, | |
| "loss": 10.2176, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.12170133498913381, | |
| "grad_norm": 14.316574096679688, | |
| "learning_rate": 8.795482695768658e-05, | |
| "loss": 8.6067, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.12294318534616579, | |
| "grad_norm": 17.61221694946289, | |
| "learning_rate": 8.769140505375085e-05, | |
| "loss": 9.455, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.12418503570319776, | |
| "grad_norm": 9.281057357788086, | |
| "learning_rate": 8.742553740855506e-05, | |
| "loss": 11.2471, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.12418503570319776, | |
| "eval_loss": 2.644759178161621, | |
| "eval_runtime": 121.8734, | |
| "eval_samples_per_second": 11.126, | |
| "eval_steps_per_second": 2.782, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.12542688606022975, | |
| "grad_norm": 13.796201705932617, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 10.2913, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.12666873641726173, | |
| "grad_norm": 8.96422290802002, | |
| "learning_rate": 8.688653405904652e-05, | |
| "loss": 11.0569, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.1279105867742937, | |
| "grad_norm": 6.833736419677734, | |
| "learning_rate": 8.661343332988869e-05, | |
| "loss": 11.1132, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.12915243713132568, | |
| "grad_norm": 5.519217014312744, | |
| "learning_rate": 8.633795680751116e-05, | |
| "loss": 10.945, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.13039428748835766, | |
| "grad_norm": 6.731013298034668, | |
| "learning_rate": 8.606012236719073e-05, | |
| "loss": 11.1413, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.13163613784538963, | |
| "grad_norm": 5.802521228790283, | |
| "learning_rate": 8.577994803720606e-05, | |
| "loss": 11.5632, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.1328779882024216, | |
| "grad_norm": 9.574117660522461, | |
| "learning_rate": 8.549745199766792e-05, | |
| "loss": 11.9083, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.13411983855945359, | |
| "grad_norm": 11.746011734008789, | |
| "learning_rate": 8.521265257933948e-05, | |
| "loss": 11.4296, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.13536168891648556, | |
| "grad_norm": 9.897080421447754, | |
| "learning_rate": 8.492556826244687e-05, | |
| "loss": 11.1493, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.13660353927351754, | |
| "grad_norm": 15.780738830566406, | |
| "learning_rate": 8.463621767547998e-05, | |
| "loss": 12.4325, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.13784538963054951, | |
| "grad_norm": 13.746973037719727, | |
| "learning_rate": 8.434461959398376e-05, | |
| "loss": 10.7841, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.1390872399875815, | |
| "grad_norm": 13.816986083984375, | |
| "learning_rate": 8.405079293933986e-05, | |
| "loss": 10.959, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.14032909034461347, | |
| "grad_norm": 15.745361328125, | |
| "learning_rate": 8.375475677753881e-05, | |
| "loss": 10.696, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.14157094070164544, | |
| "grad_norm": 5.254451274871826, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 11.1161, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.14281279105867742, | |
| "grad_norm": 5.749464988708496, | |
| "learning_rate": 8.315613291203976e-05, | |
| "loss": 11.2368, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.1440546414157094, | |
| "grad_norm": 3.0821642875671387, | |
| "learning_rate": 8.285358405218655e-05, | |
| "loss": 11.3437, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.1452964917727414, | |
| "grad_norm": 3.3653647899627686, | |
| "learning_rate": 8.25489033703452e-05, | |
| "loss": 11.3482, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.14653834212977337, | |
| "grad_norm": 8.493938446044922, | |
| "learning_rate": 8.224211063680853e-05, | |
| "loss": 10.4739, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.14778019248680535, | |
| "grad_norm": 7.391737937927246, | |
| "learning_rate": 8.19332257589174e-05, | |
| "loss": 10.4445, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.14902204284383733, | |
| "grad_norm": 10.750813484191895, | |
| "learning_rate": 8.162226877976887e-05, | |
| "loss": 10.7364, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1502638932008693, | |
| "grad_norm": 8.89720344543457, | |
| "learning_rate": 8.130925987691569e-05, | |
| "loss": 11.4644, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.15150574355790128, | |
| "grad_norm": 7.023311614990234, | |
| "learning_rate": 8.099421936105702e-05, | |
| "loss": 10.9645, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.15274759391493326, | |
| "grad_norm": 6.37396764755249, | |
| "learning_rate": 8.067716767472045e-05, | |
| "loss": 10.3982, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.15398944427196523, | |
| "grad_norm": 10.189886093139648, | |
| "learning_rate": 8.035812539093557e-05, | |
| "loss": 11.0702, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.1552312946289972, | |
| "grad_norm": 11.05607795715332, | |
| "learning_rate": 8.003711321189895e-05, | |
| "loss": 10.7855, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.15647314498602918, | |
| "grad_norm": 9.561419486999512, | |
| "learning_rate": 7.971415196763088e-05, | |
| "loss": 9.2603, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.15771499534306116, | |
| "grad_norm": 7.216650485992432, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 11.0279, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.15895684570009314, | |
| "grad_norm": 8.75108528137207, | |
| "learning_rate": 7.906246623448183e-05, | |
| "loss": 9.2971, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.1601986960571251, | |
| "grad_norm": 7.954449653625488, | |
| "learning_rate": 7.873378403255419e-05, | |
| "loss": 10.6976, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.1614405464141571, | |
| "grad_norm": 8.621119499206543, | |
| "learning_rate": 7.840323733655778e-05, | |
| "loss": 9.9067, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.16268239677118906, | |
| "grad_norm": 9.431071281433105, | |
| "learning_rate": 7.807084759519405e-05, | |
| "loss": 10.666, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.16392424712822104, | |
| "grad_norm": 11.671051025390625, | |
| "learning_rate": 7.773663637675694e-05, | |
| "loss": 7.8493, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.16516609748525302, | |
| "grad_norm": 8.27625560760498, | |
| "learning_rate": 7.740062536773352e-05, | |
| "loss": 8.3422, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.166407947842285, | |
| "grad_norm": 16.25979995727539, | |
| "learning_rate": 7.706283637139658e-05, | |
| "loss": 8.9211, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.167649798199317, | |
| "grad_norm": 13.576790809631348, | |
| "learning_rate": 7.672329130639005e-05, | |
| "loss": 9.4351, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.16889164855634897, | |
| "grad_norm": 10.519822120666504, | |
| "learning_rate": 7.638201220530665e-05, | |
| "loss": 8.0853, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.17013349891338095, | |
| "grad_norm": 12.99472713470459, | |
| "learning_rate": 7.603902121325813e-05, | |
| "loss": 11.3433, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.17137534927041292, | |
| "grad_norm": 9.731450080871582, | |
| "learning_rate": 7.569434058643844e-05, | |
| "loss": 9.3824, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.1726171996274449, | |
| "grad_norm": 9.284636497497559, | |
| "learning_rate": 7.534799269067953e-05, | |
| "loss": 10.1964, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.17385904998447688, | |
| "grad_norm": 8.576375961303711, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 8.4821, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.17510090034150885, | |
| "grad_norm": 9.682588577270508, | |
| "learning_rate": 7.465038509514688e-05, | |
| "loss": 9.3556, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.17634275069854083, | |
| "grad_norm": 9.57217788696289, | |
| "learning_rate": 7.42991706621303e-05, | |
| "loss": 8.6373, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.1775846010555728, | |
| "grad_norm": 12.602686882019043, | |
| "learning_rate": 7.394637949075154e-05, | |
| "loss": 7.3971, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.17882645141260478, | |
| "grad_norm": 15.801133155822754, | |
| "learning_rate": 7.35920344731241e-05, | |
| "loss": 9.2575, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.18006830176963676, | |
| "grad_norm": 18.355545043945312, | |
| "learning_rate": 7.323615860218843e-05, | |
| "loss": 10.2104, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.18131015212666873, | |
| "grad_norm": 19.0859317779541, | |
| "learning_rate": 7.287877497021978e-05, | |
| "loss": 9.2487, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.1825520024837007, | |
| "grad_norm": 15.948431015014648, | |
| "learning_rate": 7.251990676732984e-05, | |
| "loss": 9.6023, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.18379385284073269, | |
| "grad_norm": 10.660468101501465, | |
| "learning_rate": 7.215957727996207e-05, | |
| "loss": 9.7277, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.18503570319776466, | |
| "grad_norm": 8.521693229675293, | |
| "learning_rate": 7.179780988938051e-05, | |
| "loss": 8.7685, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.18627755355479664, | |
| "grad_norm": 14.70208740234375, | |
| "learning_rate": 7.143462807015271e-05, | |
| "loss": 7.8579, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.18627755355479664, | |
| "eval_loss": 2.472458839416504, | |
| "eval_runtime": 121.706, | |
| "eval_samples_per_second": 11.142, | |
| "eval_steps_per_second": 2.785, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1875194039118286, | |
| "grad_norm": 6.803307056427002, | |
| "learning_rate": 7.107005538862646e-05, | |
| "loss": 10.561, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.1887612542688606, | |
| "grad_norm": 8.229329109191895, | |
| "learning_rate": 7.07041155014006e-05, | |
| "loss": 10.2723, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.1900031046258926, | |
| "grad_norm": 8.097644805908203, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 10.7189, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.19124495498292457, | |
| "grad_norm": 9.763362884521484, | |
| "learning_rate": 6.996822917828477e-05, | |
| "loss": 10.8785, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.19248680533995655, | |
| "grad_norm": 8.936421394348145, | |
| "learning_rate": 6.959833049300377e-05, | |
| "loss": 10.7473, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.19372865569698852, | |
| "grad_norm": 14.159116744995117, | |
| "learning_rate": 6.922716010014255e-05, | |
| "loss": 11.4862, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.1949705060540205, | |
| "grad_norm": 10.795000076293945, | |
| "learning_rate": 6.885474208441603e-05, | |
| "loss": 10.2663, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.19621235641105247, | |
| "grad_norm": 12.574379920959473, | |
| "learning_rate": 6.848110061149556e-05, | |
| "loss": 10.6066, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.19745420676808445, | |
| "grad_norm": 13.208904266357422, | |
| "learning_rate": 6.810625992644085e-05, | |
| "loss": 10.2697, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.19869605712511643, | |
| "grad_norm": 12.457254409790039, | |
| "learning_rate": 6.773024435212678e-05, | |
| "loss": 12.1264, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.1999379074821484, | |
| "grad_norm": 11.590801239013672, | |
| "learning_rate": 6.735307828766515e-05, | |
| "loss": 11.8134, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.20117975783918038, | |
| "grad_norm": 8.898362159729004, | |
| "learning_rate": 6.697478620682137e-05, | |
| "loss": 10.6429, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.20242160819621235, | |
| "grad_norm": 9.280109405517578, | |
| "learning_rate": 6.659539265642643e-05, | |
| "loss": 10.5507, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.20366345855324433, | |
| "grad_norm": 11.889431953430176, | |
| "learning_rate": 6.621492225478414e-05, | |
| "loss": 11.0789, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.2049053089102763, | |
| "grad_norm": 5.70781946182251, | |
| "learning_rate": 6.583339969007363e-05, | |
| "loss": 10.9493, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.20614715926730828, | |
| "grad_norm": 8.63260269165039, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 10.3209, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.20738900962434026, | |
| "grad_norm": 7.887843132019043, | |
| "learning_rate": 6.506729716392481e-05, | |
| "loss": 8.8037, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.20863085998137224, | |
| "grad_norm": 11.230926513671875, | |
| "learning_rate": 6.468276691378155e-05, | |
| "loss": 10.7108, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.2098727103384042, | |
| "grad_norm": 11.069745063781738, | |
| "learning_rate": 6.429728391993446e-05, | |
| "loss": 10.1439, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.2111145606954362, | |
| "grad_norm": 15.257699966430664, | |
| "learning_rate": 6.391087319582264e-05, | |
| "loss": 10.3077, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2123564110524682, | |
| "grad_norm": 13.242205619812012, | |
| "learning_rate": 6.35235598150842e-05, | |
| "loss": 10.6351, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.21359826140950017, | |
| "grad_norm": 18.31103515625, | |
| "learning_rate": 6.313536890992935e-05, | |
| "loss": 10.1206, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.21484011176653214, | |
| "grad_norm": 14.47295093536377, | |
| "learning_rate": 6.274632566950967e-05, | |
| "loss": 8.8579, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.21608196212356412, | |
| "grad_norm": 17.428049087524414, | |
| "learning_rate": 6.235645533828349e-05, | |
| "loss": 9.6196, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.2173238124805961, | |
| "grad_norm": 16.387075424194336, | |
| "learning_rate": 6.19657832143779e-05, | |
| "loss": 10.0083, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.21856566283762807, | |
| "grad_norm": 15.025436401367188, | |
| "learning_rate": 6.157433464794716e-05, | |
| "loss": 10.7442, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.21980751319466005, | |
| "grad_norm": 12.375956535339355, | |
| "learning_rate": 6.118213503952779e-05, | |
| "loss": 10.269, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.22104936355169202, | |
| "grad_norm": 10.693756103515625, | |
| "learning_rate": 6.078920983839031e-05, | |
| "loss": 9.8473, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.222291213908724, | |
| "grad_norm": 9.529464721679688, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 10.1938, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.22353306426575598, | |
| "grad_norm": 8.467986106872559, | |
| "learning_rate": 6.0001284688802226e-05, | |
| "loss": 9.6416, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.22477491462278795, | |
| "grad_norm": 8.995750427246094, | |
| "learning_rate": 5.960633586768543e-05, | |
| "loss": 8.5713, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.22601676497981993, | |
| "grad_norm": 9.754181861877441, | |
| "learning_rate": 5.921076370520058e-05, | |
| "loss": 9.4782, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.2272586153368519, | |
| "grad_norm": 8.055115699768066, | |
| "learning_rate": 5.8814593869458455e-05, | |
| "loss": 9.0996, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.22850046569388388, | |
| "grad_norm": 10.501749038696289, | |
| "learning_rate": 5.841785206735192e-05, | |
| "loss": 10.3803, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.22974231605091586, | |
| "grad_norm": 13.05416488647461, | |
| "learning_rate": 5.8020564042888015e-05, | |
| "loss": 9.4394, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.23098416640794783, | |
| "grad_norm": 11.48218822479248, | |
| "learning_rate": 5.762275557551727e-05, | |
| "loss": 8.1828, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.2322260167649798, | |
| "grad_norm": 9.102746963500977, | |
| "learning_rate": 5.7224452478461064e-05, | |
| "loss": 9.1802, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.23346786712201179, | |
| "grad_norm": 13.362663269042969, | |
| "learning_rate": 5.682568059703659e-05, | |
| "loss": 10.5763, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.2347097174790438, | |
| "grad_norm": 14.221250534057617, | |
| "learning_rate": 5.642646580697973e-05, | |
| "loss": 8.3783, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.23595156783607577, | |
| "grad_norm": 17.283653259277344, | |
| "learning_rate": 5.602683401276615e-05, | |
| "loss": 8.8715, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.23719341819310774, | |
| "grad_norm": 20.920652389526367, | |
| "learning_rate": 5.562681114593028e-05, | |
| "loss": 7.1134, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.23843526855013972, | |
| "grad_norm": 20.802019119262695, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 8.4245, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.2396771189071717, | |
| "grad_norm": 16.49863052368164, | |
| "learning_rate": 5.482569604572576e-05, | |
| "loss": 7.6499, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.24091896926420367, | |
| "grad_norm": 12.959442138671875, | |
| "learning_rate": 5.442465579556793e-05, | |
| "loss": 8.1491, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.24216081962123565, | |
| "grad_norm": 14.34575080871582, | |
| "learning_rate": 5.402332843583631e-05, | |
| "loss": 9.1671, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.24340266997826762, | |
| "grad_norm": 14.149574279785156, | |
| "learning_rate": 5.3621740008088126e-05, | |
| "loss": 7.9398, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.2446445203352996, | |
| "grad_norm": 13.768582344055176, | |
| "learning_rate": 5.321991657082097e-05, | |
| "loss": 7.9755, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.24588637069233157, | |
| "grad_norm": 13.12028980255127, | |
| "learning_rate": 5.281788419778187e-05, | |
| "loss": 5.7906, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.24712822104936355, | |
| "grad_norm": 18.232118606567383, | |
| "learning_rate": 5.2415668976275355e-05, | |
| "loss": 6.1965, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.24837007140639553, | |
| "grad_norm": 16.104488372802734, | |
| "learning_rate": 5.201329700547076e-05, | |
| "loss": 9.2066, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.24837007140639553, | |
| "eval_loss": 2.310253143310547, | |
| "eval_runtime": 121.7061, | |
| "eval_samples_per_second": 11.142, | |
| "eval_steps_per_second": 2.785, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.4987901319747994e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |