| { | |
| "best_global_step": 2080, | |
| "best_metric": 0.12297015637159348, | |
| "best_model_checkpoint": "./llama3-ft-deepspeed/checkpoint-2080", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 2080, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.024052916416115455, | |
| "grad_norm": 1.146120548248291, | |
| "learning_rate": 0.0001995673076923077, | |
| "loss": 1.6924, | |
| "mean_token_accuracy": 0.6730133682489395, | |
| "num_tokens": 10240.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04810583283223091, | |
| "grad_norm": 1.4788161516189575, | |
| "learning_rate": 0.00019908653846153847, | |
| "loss": 0.7696, | |
| "mean_token_accuracy": 0.835153691470623, | |
| "num_tokens": 20480.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07215874924834637, | |
| "grad_norm": 1.395352840423584, | |
| "learning_rate": 0.00019860576923076922, | |
| "loss": 0.4512, | |
| "mean_token_accuracy": 0.8917673289775848, | |
| "num_tokens": 30720.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.09621166566446182, | |
| "grad_norm": 1.7515279054641724, | |
| "learning_rate": 0.000198125, | |
| "loss": 0.3253, | |
| "mean_token_accuracy": 0.9200380504131317, | |
| "num_tokens": 40960.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.12026458208057728, | |
| "grad_norm": 0.7726802229881287, | |
| "learning_rate": 0.00019764423076923079, | |
| "loss": 0.2675, | |
| "mean_token_accuracy": 0.9355374664068222, | |
| "num_tokens": 51200.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.14431749849669273, | |
| "grad_norm": 0.6123030781745911, | |
| "learning_rate": 0.00019716346153846154, | |
| "loss": 0.2365, | |
| "mean_token_accuracy": 0.9364401072263717, | |
| "num_tokens": 61440.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1683704149128082, | |
| "grad_norm": 0.6146537065505981, | |
| "learning_rate": 0.00019668269230769232, | |
| "loss": 0.2189, | |
| "mean_token_accuracy": 0.9391294255852699, | |
| "num_tokens": 71680.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.19242333132892364, | |
| "grad_norm": 0.5948600769042969, | |
| "learning_rate": 0.0001962019230769231, | |
| "loss": 0.2081, | |
| "mean_token_accuracy": 0.9405477479100227, | |
| "num_tokens": 81920.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2164762477450391, | |
| "grad_norm": 0.5457659363746643, | |
| "learning_rate": 0.00019572115384615385, | |
| "loss": 0.2019, | |
| "mean_token_accuracy": 0.9453761458396912, | |
| "num_tokens": 92160.0, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.24052916416115455, | |
| "grad_norm": 0.49840784072875977, | |
| "learning_rate": 0.00019524038461538463, | |
| "loss": 0.2039, | |
| "mean_token_accuracy": 0.9441700100898742, | |
| "num_tokens": 102400.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.26458208057727, | |
| "grad_norm": 0.46384331583976746, | |
| "learning_rate": 0.00019475961538461541, | |
| "loss": 0.2028, | |
| "mean_token_accuracy": 0.9462413385510444, | |
| "num_tokens": 112640.0, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.28863499699338546, | |
| "grad_norm": 0.44089949131011963, | |
| "learning_rate": 0.00019427884615384617, | |
| "loss": 0.1871, | |
| "mean_token_accuracy": 0.9464857250452041, | |
| "num_tokens": 122880.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3126879134095009, | |
| "grad_norm": 0.3707946240901947, | |
| "learning_rate": 0.00019379807692307695, | |
| "loss": 0.177, | |
| "mean_token_accuracy": 0.9484001085162163, | |
| "num_tokens": 133120.0, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3367408298256164, | |
| "grad_norm": 0.3341757655143738, | |
| "learning_rate": 0.0001933173076923077, | |
| "loss": 0.1847, | |
| "mean_token_accuracy": 0.9493804201483727, | |
| "num_tokens": 143360.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3607937462417318, | |
| "grad_norm": 0.4739825129508972, | |
| "learning_rate": 0.00019283653846153845, | |
| "loss": 0.1631, | |
| "mean_token_accuracy": 0.9556754723191261, | |
| "num_tokens": 153600.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3848466626578473, | |
| "grad_norm": 0.3651292622089386, | |
| "learning_rate": 0.00019235576923076923, | |
| "loss": 0.1597, | |
| "mean_token_accuracy": 0.9511243969202041, | |
| "num_tokens": 163840.0, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.40889957907396274, | |
| "grad_norm": 0.3972345292568207, | |
| "learning_rate": 0.00019187500000000002, | |
| "loss": 0.19, | |
| "mean_token_accuracy": 0.9461832985281944, | |
| "num_tokens": 174080.0, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.4329524954900782, | |
| "grad_norm": 0.281608521938324, | |
| "learning_rate": 0.00019139423076923077, | |
| "loss": 0.1693, | |
| "mean_token_accuracy": 0.9498817101120949, | |
| "num_tokens": 184320.0, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.45700541190619365, | |
| "grad_norm": 0.3879983723163605, | |
| "learning_rate": 0.00019091346153846155, | |
| "loss": 0.1549, | |
| "mean_token_accuracy": 0.9546341434121132, | |
| "num_tokens": 194560.0, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4810583283223091, | |
| "grad_norm": 0.31266576051712036, | |
| "learning_rate": 0.00019043269230769233, | |
| "loss": 0.1689, | |
| "mean_token_accuracy": 0.95065086632967, | |
| "num_tokens": 204800.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5051112447384245, | |
| "grad_norm": 0.41939643025398254, | |
| "learning_rate": 0.00018995192307692308, | |
| "loss": 0.1798, | |
| "mean_token_accuracy": 0.9457689806818962, | |
| "num_tokens": 215040.0, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.52916416115454, | |
| "grad_norm": 0.45524027943611145, | |
| "learning_rate": 0.00018947115384615386, | |
| "loss": 0.1708, | |
| "mean_token_accuracy": 0.9461533144116402, | |
| "num_tokens": 225280.0, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5532170775706554, | |
| "grad_norm": 0.3720427453517914, | |
| "learning_rate": 0.00018899038461538462, | |
| "loss": 0.162, | |
| "mean_token_accuracy": 0.9516115784645081, | |
| "num_tokens": 235520.0, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5772699939867709, | |
| "grad_norm": 0.3217442035675049, | |
| "learning_rate": 0.0001885096153846154, | |
| "loss": 0.163, | |
| "mean_token_accuracy": 0.9550325214862824, | |
| "num_tokens": 245760.0, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.6013229104028863, | |
| "grad_norm": 0.38512468338012695, | |
| "learning_rate": 0.00018802884615384618, | |
| "loss": 0.1607, | |
| "mean_token_accuracy": 0.9519079640507698, | |
| "num_tokens": 256000.0, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6253758268190018, | |
| "grad_norm": 0.31379130482673645, | |
| "learning_rate": 0.00018754807692307693, | |
| "loss": 0.1379, | |
| "mean_token_accuracy": 0.9568226426839829, | |
| "num_tokens": 266240.0, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.6494287432351172, | |
| "grad_norm": 0.3925431966781616, | |
| "learning_rate": 0.00018706730769230768, | |
| "loss": 0.1527, | |
| "mean_token_accuracy": 0.9536312386393547, | |
| "num_tokens": 276480.0, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6734816596512327, | |
| "grad_norm": 0.3034338355064392, | |
| "learning_rate": 0.00018658653846153847, | |
| "loss": 0.1648, | |
| "mean_token_accuracy": 0.9480760931968689, | |
| "num_tokens": 286720.0, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6975345760673481, | |
| "grad_norm": 0.36241453886032104, | |
| "learning_rate": 0.00018610576923076925, | |
| "loss": 0.1757, | |
| "mean_token_accuracy": 0.9463522598147392, | |
| "num_tokens": 296960.0, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.7215874924834637, | |
| "grad_norm": 0.3405214846134186, | |
| "learning_rate": 0.000185625, | |
| "loss": 0.1631, | |
| "mean_token_accuracy": 0.949561494588852, | |
| "num_tokens": 307200.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.745640408899579, | |
| "grad_norm": 0.2847406268119812, | |
| "learning_rate": 0.00018514423076923078, | |
| "loss": 0.1533, | |
| "mean_token_accuracy": 0.9555268749594689, | |
| "num_tokens": 317440.0, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.7696933253156946, | |
| "grad_norm": 0.26305094361305237, | |
| "learning_rate": 0.00018466346153846153, | |
| "loss": 0.1611, | |
| "mean_token_accuracy": 0.9532571867108345, | |
| "num_tokens": 327680.0, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.79374624173181, | |
| "grad_norm": 0.2989274263381958, | |
| "learning_rate": 0.0001841826923076923, | |
| "loss": 0.1441, | |
| "mean_token_accuracy": 0.9533735632896423, | |
| "num_tokens": 337920.0, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.8177991581479255, | |
| "grad_norm": 0.254688560962677, | |
| "learning_rate": 0.0001837019230769231, | |
| "loss": 0.1495, | |
| "mean_token_accuracy": 0.9542877942323684, | |
| "num_tokens": 348160.0, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.8418520745640409, | |
| "grad_norm": 0.26552721858024597, | |
| "learning_rate": 0.00018322115384615385, | |
| "loss": 0.149, | |
| "mean_token_accuracy": 0.9527715161442757, | |
| "num_tokens": 358400.0, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.8659049909801564, | |
| "grad_norm": 0.2920278012752533, | |
| "learning_rate": 0.00018274038461538463, | |
| "loss": 0.1506, | |
| "mean_token_accuracy": 0.9556835174560547, | |
| "num_tokens": 368640.0, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8899579073962718, | |
| "grad_norm": 0.34060242772102356, | |
| "learning_rate": 0.0001822596153846154, | |
| "loss": 0.1332, | |
| "mean_token_accuracy": 0.959219790995121, | |
| "num_tokens": 378880.0, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.9140108238123873, | |
| "grad_norm": 0.3549470603466034, | |
| "learning_rate": 0.00018177884615384616, | |
| "loss": 0.1553, | |
| "mean_token_accuracy": 0.9529827669262886, | |
| "num_tokens": 389120.0, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.9380637402285027, | |
| "grad_norm": 0.26259645819664, | |
| "learning_rate": 0.00018129807692307694, | |
| "loss": 0.1355, | |
| "mean_token_accuracy": 0.9546903222799301, | |
| "num_tokens": 399360.0, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.9621166566446182, | |
| "grad_norm": 0.28768494725227356, | |
| "learning_rate": 0.0001808173076923077, | |
| "loss": 0.1314, | |
| "mean_token_accuracy": 0.9578663051128388, | |
| "num_tokens": 409600.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9861695730607336, | |
| "grad_norm": 0.23446713387966156, | |
| "learning_rate": 0.00018033653846153848, | |
| "loss": 0.1433, | |
| "mean_token_accuracy": 0.9555784195661545, | |
| "num_tokens": 419840.0, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.14436590671539307, | |
| "eval_mean_token_accuracy": 0.9552223542903332, | |
| "eval_num_tokens": 425728.0, | |
| "eval_runtime": 25.1182, | |
| "eval_samples_per_second": 14.73, | |
| "eval_steps_per_second": 1.871, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.009621166566446, | |
| "grad_norm": 0.2881328761577606, | |
| "learning_rate": 0.00017985576923076923, | |
| "loss": 0.1424, | |
| "mean_token_accuracy": 0.9537300452207907, | |
| "num_tokens": 429824.0, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.0336740829825617, | |
| "grad_norm": 1.1323237419128418, | |
| "learning_rate": 0.000179375, | |
| "loss": 0.117, | |
| "mean_token_accuracy": 0.960891704261303, | |
| "num_tokens": 440064.0, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.057726999398677, | |
| "grad_norm": 0.3971687853336334, | |
| "learning_rate": 0.00017889423076923076, | |
| "loss": 0.1233, | |
| "mean_token_accuracy": 0.9594320252537727, | |
| "num_tokens": 450304.0, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.0817799158147925, | |
| "grad_norm": 0.235861137509346, | |
| "learning_rate": 0.00017841346153846154, | |
| "loss": 0.1322, | |
| "mean_token_accuracy": 0.9574261009693146, | |
| "num_tokens": 460544.0, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.1058328322309081, | |
| "grad_norm": 0.2572536766529083, | |
| "learning_rate": 0.00017793269230769232, | |
| "loss": 0.1288, | |
| "mean_token_accuracy": 0.9600686863064766, | |
| "num_tokens": 470784.0, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.1298857486470235, | |
| "grad_norm": 0.33651918172836304, | |
| "learning_rate": 0.00017745192307692308, | |
| "loss": 0.1367, | |
| "mean_token_accuracy": 0.9558201640844345, | |
| "num_tokens": 481024.0, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.153938665063139, | |
| "grad_norm": 0.2679121196269989, | |
| "learning_rate": 0.00017697115384615386, | |
| "loss": 0.1218, | |
| "mean_token_accuracy": 0.9592645660042762, | |
| "num_tokens": 491264.0, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.1779915814792543, | |
| "grad_norm": 0.24321414530277252, | |
| "learning_rate": 0.00017649038461538464, | |
| "loss": 0.1338, | |
| "mean_token_accuracy": 0.958324646949768, | |
| "num_tokens": 501504.0, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.2020444978953697, | |
| "grad_norm": 0.29272493720054626, | |
| "learning_rate": 0.0001760096153846154, | |
| "loss": 0.1252, | |
| "mean_token_accuracy": 0.9605333045125007, | |
| "num_tokens": 511744.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.2260974143114853, | |
| "grad_norm": 0.24240529537200928, | |
| "learning_rate": 0.00017552884615384617, | |
| "loss": 0.1208, | |
| "mean_token_accuracy": 0.960656826198101, | |
| "num_tokens": 521984.0, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.2501503307276007, | |
| "grad_norm": 0.3100011944770813, | |
| "learning_rate": 0.00017504807692307695, | |
| "loss": 0.1254, | |
| "mean_token_accuracy": 0.9609130263328552, | |
| "num_tokens": 532224.0, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.2742032471437161, | |
| "grad_norm": 0.28045424818992615, | |
| "learning_rate": 0.00017456730769230768, | |
| "loss": 0.1304, | |
| "mean_token_accuracy": 0.9579013884067535, | |
| "num_tokens": 542464.0, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.2982561635598318, | |
| "grad_norm": 0.3832496702671051, | |
| "learning_rate": 0.00017408653846153846, | |
| "loss": 0.1427, | |
| "mean_token_accuracy": 0.9550067007541656, | |
| "num_tokens": 552704.0, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.3223090799759472, | |
| "grad_norm": 0.32106658816337585, | |
| "learning_rate": 0.00017360576923076924, | |
| "loss": 0.129, | |
| "mean_token_accuracy": 0.956248240172863, | |
| "num_tokens": 562944.0, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.3463619963920626, | |
| "grad_norm": 0.3141852915287018, | |
| "learning_rate": 0.000173125, | |
| "loss": 0.1145, | |
| "mean_token_accuracy": 0.9619206488132477, | |
| "num_tokens": 573184.0, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.370414912808178, | |
| "grad_norm": 0.31353870034217834, | |
| "learning_rate": 0.00017264423076923077, | |
| "loss": 0.126, | |
| "mean_token_accuracy": 0.9610635504126549, | |
| "num_tokens": 583424.0, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.3944678292242934, | |
| "grad_norm": 0.24016191065311432, | |
| "learning_rate": 0.00017216346153846155, | |
| "loss": 0.1318, | |
| "mean_token_accuracy": 0.9551108077168464, | |
| "num_tokens": 593664.0, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.418520745640409, | |
| "grad_norm": 0.2362779825925827, | |
| "learning_rate": 0.0001716826923076923, | |
| "loss": 0.1163, | |
| "mean_token_accuracy": 0.9617116883397102, | |
| "num_tokens": 603904.0, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.4425736620565244, | |
| "grad_norm": 0.2012677937746048, | |
| "learning_rate": 0.0001712019230769231, | |
| "loss": 0.1378, | |
| "mean_token_accuracy": 0.9564808994531632, | |
| "num_tokens": 614144.0, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.4666265784726398, | |
| "grad_norm": 0.33629047870635986, | |
| "learning_rate": 0.00017072115384615387, | |
| "loss": 0.1215, | |
| "mean_token_accuracy": 0.9621520712971687, | |
| "num_tokens": 624384.0, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.4906794948887554, | |
| "grad_norm": 0.30020079016685486, | |
| "learning_rate": 0.00017024038461538462, | |
| "loss": 0.1318, | |
| "mean_token_accuracy": 0.9570673331618309, | |
| "num_tokens": 634624.0, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.5147324113048706, | |
| "grad_norm": 0.2715687155723572, | |
| "learning_rate": 0.0001697596153846154, | |
| "loss": 0.1248, | |
| "mean_token_accuracy": 0.9611581727862358, | |
| "num_tokens": 644864.0, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.5387853277209862, | |
| "grad_norm": 0.2919619381427765, | |
| "learning_rate": 0.00016927884615384618, | |
| "loss": 0.116, | |
| "mean_token_accuracy": 0.9624506369233131, | |
| "num_tokens": 655104.0, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.5628382441371016, | |
| "grad_norm": 0.26573172211647034, | |
| "learning_rate": 0.00016879807692307694, | |
| "loss": 0.1253, | |
| "mean_token_accuracy": 0.9609674572944641, | |
| "num_tokens": 665344.0, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.586891160553217, | |
| "grad_norm": 0.3533133566379547, | |
| "learning_rate": 0.0001683173076923077, | |
| "loss": 0.129, | |
| "mean_token_accuracy": 0.9564981684088707, | |
| "num_tokens": 675584.0, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.6109440769693326, | |
| "grad_norm": 0.23616884648799896, | |
| "learning_rate": 0.00016783653846153847, | |
| "loss": 0.1198, | |
| "mean_token_accuracy": 0.9604200229048729, | |
| "num_tokens": 685824.0, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.634996993385448, | |
| "grad_norm": 0.29362770915031433, | |
| "learning_rate": 0.00016735576923076922, | |
| "loss": 0.1248, | |
| "mean_token_accuracy": 0.9565964505076409, | |
| "num_tokens": 696064.0, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.6590499098015634, | |
| "grad_norm": 0.32813316583633423, | |
| "learning_rate": 0.000166875, | |
| "loss": 0.1404, | |
| "mean_token_accuracy": 0.9544960707426071, | |
| "num_tokens": 706304.0, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.683102826217679, | |
| "grad_norm": 0.23794318735599518, | |
| "learning_rate": 0.00016639423076923078, | |
| "loss": 0.1239, | |
| "mean_token_accuracy": 0.9590726107358932, | |
| "num_tokens": 716544.0, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.7071557426337942, | |
| "grad_norm": 0.3290494680404663, | |
| "learning_rate": 0.00016591346153846154, | |
| "loss": 0.1136, | |
| "mean_token_accuracy": 0.9635635167360306, | |
| "num_tokens": 726784.0, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.7312086590499098, | |
| "grad_norm": 0.2637302577495575, | |
| "learning_rate": 0.00016543269230769232, | |
| "loss": 0.1211, | |
| "mean_token_accuracy": 0.957936991751194, | |
| "num_tokens": 737024.0, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.7552615754660252, | |
| "grad_norm": 0.2866067588329315, | |
| "learning_rate": 0.0001649519230769231, | |
| "loss": 0.1188, | |
| "mean_token_accuracy": 0.9595590904355049, | |
| "num_tokens": 747264.0, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.7793144918821406, | |
| "grad_norm": 0.31299787759780884, | |
| "learning_rate": 0.00016447115384615385, | |
| "loss": 0.1303, | |
| "mean_token_accuracy": 0.9595273941755295, | |
| "num_tokens": 757504.0, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.8033674082982563, | |
| "grad_norm": 0.25704699754714966, | |
| "learning_rate": 0.00016399038461538463, | |
| "loss": 0.12, | |
| "mean_token_accuracy": 0.9622810766100883, | |
| "num_tokens": 767744.0, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.8274203247143717, | |
| "grad_norm": 0.20122775435447693, | |
| "learning_rate": 0.0001635096153846154, | |
| "loss": 0.1153, | |
| "mean_token_accuracy": 0.961222605407238, | |
| "num_tokens": 777984.0, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.851473241130487, | |
| "grad_norm": 0.21299275755882263, | |
| "learning_rate": 0.00016302884615384617, | |
| "loss": 0.114, | |
| "mean_token_accuracy": 0.9608678832650185, | |
| "num_tokens": 788224.0, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.8755261575466027, | |
| "grad_norm": 0.25124669075012207, | |
| "learning_rate": 0.00016254807692307695, | |
| "loss": 0.122, | |
| "mean_token_accuracy": 0.962291096150875, | |
| "num_tokens": 798464.0, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.8995790739627179, | |
| "grad_norm": 0.2945667803287506, | |
| "learning_rate": 0.0001620673076923077, | |
| "loss": 0.1183, | |
| "mean_token_accuracy": 0.9604008629918098, | |
| "num_tokens": 808704.0, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.9236319903788335, | |
| "grad_norm": 0.2346087247133255, | |
| "learning_rate": 0.00016158653846153845, | |
| "loss": 0.1136, | |
| "mean_token_accuracy": 0.963145537674427, | |
| "num_tokens": 818944.0, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.9476849067949489, | |
| "grad_norm": 0.22119984030723572, | |
| "learning_rate": 0.00016110576923076923, | |
| "loss": 0.1235, | |
| "mean_token_accuracy": 0.9591810956597329, | |
| "num_tokens": 829184.0, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.9717378232110643, | |
| "grad_norm": 0.2481156885623932, | |
| "learning_rate": 0.00016062500000000001, | |
| "loss": 0.1208, | |
| "mean_token_accuracy": 0.9615909501910209, | |
| "num_tokens": 839424.0, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.99579073962718, | |
| "grad_norm": 0.24333243072032928, | |
| "learning_rate": 0.00016014423076923077, | |
| "loss": 0.121, | |
| "mean_token_accuracy": 0.9590676620602607, | |
| "num_tokens": 849664.0, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.1264333575963974, | |
| "eval_mean_token_accuracy": 0.9596163808031285, | |
| "eval_num_tokens": 851456.0, | |
| "eval_runtime": 24.9078, | |
| "eval_samples_per_second": 14.855, | |
| "eval_steps_per_second": 1.887, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 2.019242333132892, | |
| "grad_norm": 0.25562378764152527, | |
| "learning_rate": 0.00015966346153846155, | |
| "loss": 0.1082, | |
| "mean_token_accuracy": 0.9661401877036462, | |
| "num_tokens": 859648.0, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.0432952495490078, | |
| "grad_norm": 0.2851448059082031, | |
| "learning_rate": 0.00015918269230769233, | |
| "loss": 0.0988, | |
| "mean_token_accuracy": 0.9670166626572609, | |
| "num_tokens": 869888.0, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.0673481659651234, | |
| "grad_norm": 0.3552910387516022, | |
| "learning_rate": 0.00015870192307692308, | |
| "loss": 0.1058, | |
| "mean_token_accuracy": 0.961455948650837, | |
| "num_tokens": 880128.0, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.0914010823812386, | |
| "grad_norm": 0.18810558319091797, | |
| "learning_rate": 0.00015822115384615386, | |
| "loss": 0.0962, | |
| "mean_token_accuracy": 0.965205217897892, | |
| "num_tokens": 890368.0, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.115453998797354, | |
| "grad_norm": 0.24357327818870544, | |
| "learning_rate": 0.00015774038461538462, | |
| "loss": 0.1121, | |
| "mean_token_accuracy": 0.9618488609790802, | |
| "num_tokens": 900608.0, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.13950691521347, | |
| "grad_norm": 0.24040566384792328, | |
| "learning_rate": 0.0001572596153846154, | |
| "loss": 0.1064, | |
| "mean_token_accuracy": 0.9660862430930137, | |
| "num_tokens": 910848.0, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.163559831629585, | |
| "grad_norm": 0.24545449018478394, | |
| "learning_rate": 0.00015677884615384618, | |
| "loss": 0.0965, | |
| "mean_token_accuracy": 0.9677914634346962, | |
| "num_tokens": 921088.0, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.1876127480457006, | |
| "grad_norm": 0.30429190397262573, | |
| "learning_rate": 0.00015629807692307693, | |
| "loss": 0.105, | |
| "mean_token_accuracy": 0.9644672557711601, | |
| "num_tokens": 931328.0, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.2116656644618162, | |
| "grad_norm": 0.2934926450252533, | |
| "learning_rate": 0.00015581730769230768, | |
| "loss": 0.0968, | |
| "mean_token_accuracy": 0.965617573261261, | |
| "num_tokens": 941568.0, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.2357185808779314, | |
| "grad_norm": 0.3533467948436737, | |
| "learning_rate": 0.00015533653846153846, | |
| "loss": 0.1073, | |
| "mean_token_accuracy": 0.9617443799972534, | |
| "num_tokens": 951808.0, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.259771497294047, | |
| "grad_norm": 0.15458804368972778, | |
| "learning_rate": 0.00015485576923076924, | |
| "loss": 0.1069, | |
| "mean_token_accuracy": 0.9629932105541229, | |
| "num_tokens": 962048.0, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.283824413710162, | |
| "grad_norm": 0.6568232774734497, | |
| "learning_rate": 0.000154375, | |
| "loss": 0.1077, | |
| "mean_token_accuracy": 0.9618542537093162, | |
| "num_tokens": 972288.0, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.307877330126278, | |
| "grad_norm": 0.3730143904685974, | |
| "learning_rate": 0.00015389423076923078, | |
| "loss": 0.1139, | |
| "mean_token_accuracy": 0.9590795397758484, | |
| "num_tokens": 982528.0, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.3319302465423934, | |
| "grad_norm": 0.26199349761009216, | |
| "learning_rate": 0.00015341346153846153, | |
| "loss": 0.115, | |
| "mean_token_accuracy": 0.9604379132390022, | |
| "num_tokens": 992768.0, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.3559831629585086, | |
| "grad_norm": 0.3791094720363617, | |
| "learning_rate": 0.0001529326923076923, | |
| "loss": 0.1066, | |
| "mean_token_accuracy": 0.9632134348154068, | |
| "num_tokens": 1003008.0, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.3800360793746242, | |
| "grad_norm": 0.1697998046875, | |
| "learning_rate": 0.0001524519230769231, | |
| "loss": 0.1024, | |
| "mean_token_accuracy": 0.9637576416134834, | |
| "num_tokens": 1013248.0, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.4040889957907394, | |
| "grad_norm": 0.3326902389526367, | |
| "learning_rate": 0.00015197115384615385, | |
| "loss": 0.1066, | |
| "mean_token_accuracy": 0.9614829778671264, | |
| "num_tokens": 1023488.0, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.428141912206855, | |
| "grad_norm": 0.22747723758220673, | |
| "learning_rate": 0.00015149038461538463, | |
| "loss": 0.0924, | |
| "mean_token_accuracy": 0.9677571937441826, | |
| "num_tokens": 1033728.0, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.4521948286229707, | |
| "grad_norm": 0.25410860776901245, | |
| "learning_rate": 0.0001510096153846154, | |
| "loss": 0.1097, | |
| "mean_token_accuracy": 0.963656097650528, | |
| "num_tokens": 1043968.0, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.476247745039086, | |
| "grad_norm": 0.2268102765083313, | |
| "learning_rate": 0.00015052884615384616, | |
| "loss": 0.1018, | |
| "mean_token_accuracy": 0.9653624445199966, | |
| "num_tokens": 1054208.0, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.5003006614552015, | |
| "grad_norm": 0.23511724174022675, | |
| "learning_rate": 0.00015004807692307694, | |
| "loss": 0.0969, | |
| "mean_token_accuracy": 0.9658154919743538, | |
| "num_tokens": 1064448.0, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.5243535778713166, | |
| "grad_norm": 0.23023869097232819, | |
| "learning_rate": 0.0001495673076923077, | |
| "loss": 0.1053, | |
| "mean_token_accuracy": 0.96552574634552, | |
| "num_tokens": 1074688.0, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.5484064942874323, | |
| "grad_norm": 0.25081828236579895, | |
| "learning_rate": 0.00014908653846153845, | |
| "loss": 0.1044, | |
| "mean_token_accuracy": 0.9639346837997437, | |
| "num_tokens": 1084928.0, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.572459410703548, | |
| "grad_norm": 0.2557835280895233, | |
| "learning_rate": 0.00014860576923076923, | |
| "loss": 0.106, | |
| "mean_token_accuracy": 0.9623663991689682, | |
| "num_tokens": 1095168.0, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.5965123271196635, | |
| "grad_norm": 0.1407734751701355, | |
| "learning_rate": 0.000148125, | |
| "loss": 0.1092, | |
| "mean_token_accuracy": 0.9625474750995636, | |
| "num_tokens": 1105408.0, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.6205652435357787, | |
| "grad_norm": 0.2862281799316406, | |
| "learning_rate": 0.00014764423076923076, | |
| "loss": 0.1053, | |
| "mean_token_accuracy": 0.9632299616932869, | |
| "num_tokens": 1115648.0, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.6446181599518943, | |
| "grad_norm": 0.26334843039512634, | |
| "learning_rate": 0.00014716346153846154, | |
| "loss": 0.1209, | |
| "mean_token_accuracy": 0.9601381734013558, | |
| "num_tokens": 1125888.0, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.6686710763680095, | |
| "grad_norm": 0.20137883722782135, | |
| "learning_rate": 0.00014668269230769232, | |
| "loss": 0.1057, | |
| "mean_token_accuracy": 0.9626315057277679, | |
| "num_tokens": 1136128.0, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.692723992784125, | |
| "grad_norm": 0.3017769157886505, | |
| "learning_rate": 0.00014620192307692308, | |
| "loss": 0.104, | |
| "mean_token_accuracy": 0.9653211057186126, | |
| "num_tokens": 1146368.0, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.7167769092002407, | |
| "grad_norm": 0.24447724223136902, | |
| "learning_rate": 0.00014572115384615386, | |
| "loss": 0.111, | |
| "mean_token_accuracy": 0.9617329522967338, | |
| "num_tokens": 1156608.0, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.740829825616356, | |
| "grad_norm": 0.22615140676498413, | |
| "learning_rate": 0.00014524038461538464, | |
| "loss": 0.113, | |
| "mean_token_accuracy": 0.9617576941847801, | |
| "num_tokens": 1166848.0, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.7648827420324715, | |
| "grad_norm": 0.20445455610752106, | |
| "learning_rate": 0.0001447596153846154, | |
| "loss": 0.1064, | |
| "mean_token_accuracy": 0.965658301115036, | |
| "num_tokens": 1177088.0, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.7889356584485867, | |
| "grad_norm": 0.2821711003780365, | |
| "learning_rate": 0.00014427884615384617, | |
| "loss": 0.1088, | |
| "mean_token_accuracy": 0.9635483458638191, | |
| "num_tokens": 1187328.0, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.8129885748647023, | |
| "grad_norm": 0.21199429035186768, | |
| "learning_rate": 0.00014379807692307695, | |
| "loss": 0.1071, | |
| "mean_token_accuracy": 0.9626901999115944, | |
| "num_tokens": 1197568.0, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.837041491280818, | |
| "grad_norm": 0.23910683393478394, | |
| "learning_rate": 0.00014331730769230768, | |
| "loss": 0.1095, | |
| "mean_token_accuracy": 0.9642450660467148, | |
| "num_tokens": 1207808.0, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.861094407696933, | |
| "grad_norm": 0.22157999873161316, | |
| "learning_rate": 0.00014283653846153846, | |
| "loss": 0.0968, | |
| "mean_token_accuracy": 0.964881993830204, | |
| "num_tokens": 1218048.0, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.8851473241130488, | |
| "grad_norm": 0.25650596618652344, | |
| "learning_rate": 0.00014235576923076924, | |
| "loss": 0.1027, | |
| "mean_token_accuracy": 0.9612545475363732, | |
| "num_tokens": 1228288.0, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.909200240529164, | |
| "grad_norm": 0.32955703139305115, | |
| "learning_rate": 0.000141875, | |
| "loss": 0.1096, | |
| "mean_token_accuracy": 0.9617963835597039, | |
| "num_tokens": 1238528.0, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.9332531569452795, | |
| "grad_norm": 0.21780824661254883, | |
| "learning_rate": 0.00014139423076923077, | |
| "loss": 0.1033, | |
| "mean_token_accuracy": 0.9646632343530654, | |
| "num_tokens": 1248768.0, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.957306073361395, | |
| "grad_norm": 0.2339126020669937, | |
| "learning_rate": 0.00014091346153846155, | |
| "loss": 0.1031, | |
| "mean_token_accuracy": 0.964007930457592, | |
| "num_tokens": 1259008.0, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.981358989777511, | |
| "grad_norm": 0.31986188888549805, | |
| "learning_rate": 0.0001404326923076923, | |
| "loss": 0.1018, | |
| "mean_token_accuracy": 0.9668645352125168, | |
| "num_tokens": 1269248.0, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.12327124178409576, | |
| "eval_mean_token_accuracy": 0.9614017339462929, | |
| "eval_num_tokens": 1277184.0, | |
| "eval_runtime": 25.0119, | |
| "eval_samples_per_second": 14.793, | |
| "eval_steps_per_second": 1.879, | |
| "step": 1248 | |
| }, | |
| { | |
| "epoch": 3.004810583283223, | |
| "grad_norm": 0.20351627469062805, | |
| "learning_rate": 0.0001399519230769231, | |
| "loss": 0.0923, | |
| "mean_token_accuracy": 0.9675439733725327, | |
| "num_tokens": 1279232.0, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 3.0288634996993387, | |
| "grad_norm": 0.3330595791339874, | |
| "learning_rate": 0.00013947115384615387, | |
| "loss": 0.0906, | |
| "mean_token_accuracy": 0.9674146190285683, | |
| "num_tokens": 1289472.0, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 3.052916416115454, | |
| "grad_norm": 0.2875508666038513, | |
| "learning_rate": 0.00013899038461538462, | |
| "loss": 0.0953, | |
| "mean_token_accuracy": 0.9661522597074509, | |
| "num_tokens": 1299712.0, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 3.0769693325315695, | |
| "grad_norm": 0.26365041732788086, | |
| "learning_rate": 0.0001385096153846154, | |
| "loss": 0.0898, | |
| "mean_token_accuracy": 0.9656254693865776, | |
| "num_tokens": 1309952.0, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 3.101022248947685, | |
| "grad_norm": 0.22576822340488434, | |
| "learning_rate": 0.00013802884615384618, | |
| "loss": 0.0911, | |
| "mean_token_accuracy": 0.9678579106926918, | |
| "num_tokens": 1320192.0, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 3.1250751653638003, | |
| "grad_norm": 0.2698352038860321, | |
| "learning_rate": 0.00013754807692307694, | |
| "loss": 0.098, | |
| "mean_token_accuracy": 0.9671777725219727, | |
| "num_tokens": 1330432.0, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.149128081779916, | |
| "grad_norm": 0.29429781436920166, | |
| "learning_rate": 0.0001370673076923077, | |
| "loss": 0.0883, | |
| "mean_token_accuracy": 0.9666375696659089, | |
| "num_tokens": 1340672.0, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 3.173180998196031, | |
| "grad_norm": 0.25614482164382935, | |
| "learning_rate": 0.00013658653846153847, | |
| "loss": 0.0915, | |
| "mean_token_accuracy": 0.967202040553093, | |
| "num_tokens": 1350912.0, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 3.1972339146121467, | |
| "grad_norm": 0.2897791862487793, | |
| "learning_rate": 0.00013610576923076922, | |
| "loss": 0.0956, | |
| "mean_token_accuracy": 0.965019428730011, | |
| "num_tokens": 1361152.0, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 3.2212868310282623, | |
| "grad_norm": 0.1813499927520752, | |
| "learning_rate": 0.000135625, | |
| "loss": 0.0941, | |
| "mean_token_accuracy": 0.9646437510848045, | |
| "num_tokens": 1371392.0, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 3.2453397474443775, | |
| "grad_norm": 0.19243107736110687, | |
| "learning_rate": 0.00013514423076923078, | |
| "loss": 0.091, | |
| "mean_token_accuracy": 0.9650159657001496, | |
| "num_tokens": 1381632.0, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 3.269392663860493, | |
| "grad_norm": 0.22238536179065704, | |
| "learning_rate": 0.00013466346153846154, | |
| "loss": 0.0924, | |
| "mean_token_accuracy": 0.9653570145368576, | |
| "num_tokens": 1391872.0, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 3.2934455802766087, | |
| "grad_norm": 0.15673308074474335, | |
| "learning_rate": 0.00013418269230769232, | |
| "loss": 0.0892, | |
| "mean_token_accuracy": 0.9664774596691131, | |
| "num_tokens": 1402112.0, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 3.317498496692724, | |
| "grad_norm": 0.2173730581998825, | |
| "learning_rate": 0.0001337019230769231, | |
| "loss": 0.0978, | |
| "mean_token_accuracy": 0.9646672755479813, | |
| "num_tokens": 1412352.0, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 3.3415514131088395, | |
| "grad_norm": 0.17452774941921234, | |
| "learning_rate": 0.00013322115384615385, | |
| "loss": 0.0988, | |
| "mean_token_accuracy": 0.9650528088212014, | |
| "num_tokens": 1422592.0, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 3.365604329524955, | |
| "grad_norm": 0.19794107973575592, | |
| "learning_rate": 0.00013274038461538463, | |
| "loss": 0.0906, | |
| "mean_token_accuracy": 0.9670823276042938, | |
| "num_tokens": 1432832.0, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.3896572459410703, | |
| "grad_norm": 0.17740049958229065, | |
| "learning_rate": 0.00013225961538461539, | |
| "loss": 0.0917, | |
| "mean_token_accuracy": 0.9663808658719063, | |
| "num_tokens": 1443072.0, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 3.413710162357186, | |
| "grad_norm": 0.24540650844573975, | |
| "learning_rate": 0.00013177884615384617, | |
| "loss": 0.0889, | |
| "mean_token_accuracy": 0.9661073669791221, | |
| "num_tokens": 1453312.0, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 3.437763078773301, | |
| "grad_norm": 0.2300598919391632, | |
| "learning_rate": 0.00013129807692307695, | |
| "loss": 0.0865, | |
| "mean_token_accuracy": 0.9688371613621711, | |
| "num_tokens": 1463552.0, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 3.4618159951894167, | |
| "grad_norm": 0.18420913815498352, | |
| "learning_rate": 0.0001308173076923077, | |
| "loss": 0.0963, | |
| "mean_token_accuracy": 0.9648661985993385, | |
| "num_tokens": 1473792.0, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 3.4858689116055324, | |
| "grad_norm": 0.27637794613838196, | |
| "learning_rate": 0.00013033653846153845, | |
| "loss": 0.0984, | |
| "mean_token_accuracy": 0.9647489741444588, | |
| "num_tokens": 1484032.0, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 3.5099218280216475, | |
| "grad_norm": 0.28515446186065674, | |
| "learning_rate": 0.00012985576923076923, | |
| "loss": 0.0915, | |
| "mean_token_accuracy": 0.9642152190208435, | |
| "num_tokens": 1494272.0, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 3.533974744437763, | |
| "grad_norm": 0.19887416064739227, | |
| "learning_rate": 0.00012937500000000001, | |
| "loss": 0.0963, | |
| "mean_token_accuracy": 0.9659793645143508, | |
| "num_tokens": 1504512.0, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 3.5580276608538783, | |
| "grad_norm": 0.2664344906806946, | |
| "learning_rate": 0.00012889423076923077, | |
| "loss": 0.0896, | |
| "mean_token_accuracy": 0.9680359989404679, | |
| "num_tokens": 1514752.0, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 3.582080577269994, | |
| "grad_norm": 0.23911890387535095, | |
| "learning_rate": 0.00012841346153846155, | |
| "loss": 0.0966, | |
| "mean_token_accuracy": 0.9650948256254196, | |
| "num_tokens": 1524992.0, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 3.6061334936861096, | |
| "grad_norm": 0.21762046217918396, | |
| "learning_rate": 0.0001279326923076923, | |
| "loss": 0.0952, | |
| "mean_token_accuracy": 0.9666063264012337, | |
| "num_tokens": 1535232.0, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.6301864101022248, | |
| "grad_norm": 0.263232558965683, | |
| "learning_rate": 0.00012745192307692308, | |
| "loss": 0.0993, | |
| "mean_token_accuracy": 0.9661560922861099, | |
| "num_tokens": 1545472.0, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 3.6542393265183404, | |
| "grad_norm": 0.20630541443824768, | |
| "learning_rate": 0.00012697115384615386, | |
| "loss": 0.0918, | |
| "mean_token_accuracy": 0.9665474951267242, | |
| "num_tokens": 1555712.0, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 3.6782922429344556, | |
| "grad_norm": 0.21548768877983093, | |
| "learning_rate": 0.00012649038461538462, | |
| "loss": 0.0942, | |
| "mean_token_accuracy": 0.9646090790629387, | |
| "num_tokens": 1565952.0, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 3.702345159350571, | |
| "grad_norm": 0.21956314146518707, | |
| "learning_rate": 0.0001260096153846154, | |
| "loss": 0.0963, | |
| "mean_token_accuracy": 0.9646272197365761, | |
| "num_tokens": 1576192.0, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 3.726398075766687, | |
| "grad_norm": 0.22929218411445618, | |
| "learning_rate": 0.00012552884615384618, | |
| "loss": 0.0946, | |
| "mean_token_accuracy": 0.9660244584083557, | |
| "num_tokens": 1586432.0, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 3.7504509921828024, | |
| "grad_norm": 0.20628125965595245, | |
| "learning_rate": 0.00012504807692307693, | |
| "loss": 0.0939, | |
| "mean_token_accuracy": 0.9671428605914116, | |
| "num_tokens": 1596672.0, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 3.7745039085989176, | |
| "grad_norm": 0.1761239767074585, | |
| "learning_rate": 0.00012456730769230768, | |
| "loss": 0.0985, | |
| "mean_token_accuracy": 0.9652422875165939, | |
| "num_tokens": 1606912.0, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 3.7985568250150332, | |
| "grad_norm": 0.20456120371818542, | |
| "learning_rate": 0.00012408653846153846, | |
| "loss": 0.0992, | |
| "mean_token_accuracy": 0.9635955929756165, | |
| "num_tokens": 1617152.0, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 3.8226097414311484, | |
| "grad_norm": 0.24852736294269562, | |
| "learning_rate": 0.00012360576923076922, | |
| "loss": 0.0969, | |
| "mean_token_accuracy": 0.9656632468104362, | |
| "num_tokens": 1627392.0, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 3.846662657847264, | |
| "grad_norm": 0.2334773987531662, | |
| "learning_rate": 0.000123125, | |
| "loss": 0.0945, | |
| "mean_token_accuracy": 0.9682855114340783, | |
| "num_tokens": 1637632.0, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.8707155742633796, | |
| "grad_norm": 0.2439630627632141, | |
| "learning_rate": 0.00012264423076923078, | |
| "loss": 0.0945, | |
| "mean_token_accuracy": 0.9671804904937744, | |
| "num_tokens": 1647872.0, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 3.894768490679495, | |
| "grad_norm": 0.21514557301998138, | |
| "learning_rate": 0.00012216346153846153, | |
| "loss": 0.0952, | |
| "mean_token_accuracy": 0.9666045814752579, | |
| "num_tokens": 1658112.0, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 3.9188214070956104, | |
| "grad_norm": 0.16191978752613068, | |
| "learning_rate": 0.00012168269230769231, | |
| "loss": 0.097, | |
| "mean_token_accuracy": 0.966627161204815, | |
| "num_tokens": 1668352.0, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 3.9428743235117256, | |
| "grad_norm": 0.17086118459701538, | |
| "learning_rate": 0.00012120192307692308, | |
| "loss": 0.0923, | |
| "mean_token_accuracy": 0.9672711491584778, | |
| "num_tokens": 1678592.0, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 3.9669272399278412, | |
| "grad_norm": 0.19103848934173584, | |
| "learning_rate": 0.00012072115384615386, | |
| "loss": 0.0921, | |
| "mean_token_accuracy": 0.9661801844835282, | |
| "num_tokens": 1688832.0, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 3.990980156343957, | |
| "grad_norm": 0.22176852822303772, | |
| "learning_rate": 0.00012024038461538463, | |
| "loss": 0.0917, | |
| "mean_token_accuracy": 0.9642520830035209, | |
| "num_tokens": 1699072.0, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.12371965497732162, | |
| "eval_mean_token_accuracy": 0.9598764219182603, | |
| "eval_num_tokens": 1702912.0, | |
| "eval_runtime": 25.0613, | |
| "eval_samples_per_second": 14.764, | |
| "eval_steps_per_second": 1.875, | |
| "step": 1664 | |
| }, | |
| { | |
| "epoch": 4.0144317498496696, | |
| "grad_norm": 0.1616247445344925, | |
| "learning_rate": 0.0001197596153846154, | |
| "loss": 0.0846, | |
| "mean_token_accuracy": 0.9689040917616624, | |
| "num_tokens": 1709056.0, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 4.038484666265784, | |
| "grad_norm": 0.1974208503961563, | |
| "learning_rate": 0.00011927884615384617, | |
| "loss": 0.0902, | |
| "mean_token_accuracy": 0.9659950643777847, | |
| "num_tokens": 1719296.0, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 4.0625375826819, | |
| "grad_norm": 0.19191399216651917, | |
| "learning_rate": 0.00011879807692307694, | |
| "loss": 0.0876, | |
| "mean_token_accuracy": 0.9678742095828057, | |
| "num_tokens": 1729536.0, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 4.0865904990980155, | |
| "grad_norm": 0.17952245473861694, | |
| "learning_rate": 0.0001183173076923077, | |
| "loss": 0.0935, | |
| "mean_token_accuracy": 0.9667584493756294, | |
| "num_tokens": 1739776.0, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 4.110643415514131, | |
| "grad_norm": 0.2491922825574875, | |
| "learning_rate": 0.00011783653846153846, | |
| "loss": 0.0829, | |
| "mean_token_accuracy": 0.9714985772967338, | |
| "num_tokens": 1750016.0, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 4.134696331930247, | |
| "grad_norm": 0.23436853289604187, | |
| "learning_rate": 0.00011735576923076923, | |
| "loss": 0.0884, | |
| "mean_token_accuracy": 0.9674542516469955, | |
| "num_tokens": 1760256.0, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 4.158749248346362, | |
| "grad_norm": 0.21396978199481964, | |
| "learning_rate": 0.000116875, | |
| "loss": 0.0871, | |
| "mean_token_accuracy": 0.9674944341182709, | |
| "num_tokens": 1770496.0, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 4.182802164762477, | |
| "grad_norm": 0.15618039667606354, | |
| "learning_rate": 0.00011639423076923078, | |
| "loss": 0.0762, | |
| "mean_token_accuracy": 0.97198745906353, | |
| "num_tokens": 1780736.0, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 4.206855081178593, | |
| "grad_norm": 0.16134659945964813, | |
| "learning_rate": 0.00011591346153846154, | |
| "loss": 0.093, | |
| "mean_token_accuracy": 0.9636217251420021, | |
| "num_tokens": 1790976.0, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 4.230907997594708, | |
| "grad_norm": 0.12519049644470215, | |
| "learning_rate": 0.00011543269230769231, | |
| "loss": 0.0817, | |
| "mean_token_accuracy": 0.9709686204791069, | |
| "num_tokens": 1801216.0, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 4.254960914010824, | |
| "grad_norm": 0.17119652032852173, | |
| "learning_rate": 0.00011495192307692309, | |
| "loss": 0.0821, | |
| "mean_token_accuracy": 0.9688223898410797, | |
| "num_tokens": 1811456.0, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 4.27901383042694, | |
| "grad_norm": 0.13762086629867554, | |
| "learning_rate": 0.00011447115384615386, | |
| "loss": 0.0877, | |
| "mean_token_accuracy": 0.9673933520913124, | |
| "num_tokens": 1821696.0, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 4.303066746843054, | |
| "grad_norm": 0.22409266233444214, | |
| "learning_rate": 0.00011399038461538462, | |
| "loss": 0.0907, | |
| "mean_token_accuracy": 0.9678563743829727, | |
| "num_tokens": 1831936.0, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 4.32711966325917, | |
| "grad_norm": 0.21305358409881592, | |
| "learning_rate": 0.0001135096153846154, | |
| "loss": 0.0845, | |
| "mean_token_accuracy": 0.9678771600127221, | |
| "num_tokens": 1842176.0, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.351172579675286, | |
| "grad_norm": 0.24969004094600677, | |
| "learning_rate": 0.00011302884615384617, | |
| "loss": 0.0896, | |
| "mean_token_accuracy": 0.967383436858654, | |
| "num_tokens": 1852416.0, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 4.375225496091401, | |
| "grad_norm": 0.22428163886070251, | |
| "learning_rate": 0.00011254807692307694, | |
| "loss": 0.0895, | |
| "mean_token_accuracy": 0.9665916368365288, | |
| "num_tokens": 1862656.0, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 4.399278412507517, | |
| "grad_norm": 0.2337762862443924, | |
| "learning_rate": 0.00011206730769230769, | |
| "loss": 0.0926, | |
| "mean_token_accuracy": 0.964175409078598, | |
| "num_tokens": 1872896.0, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 4.4233313289236325, | |
| "grad_norm": 0.24770590662956238, | |
| "learning_rate": 0.00011158653846153846, | |
| "loss": 0.0916, | |
| "mean_token_accuracy": 0.9662819102406501, | |
| "num_tokens": 1883136.0, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 4.447384245339747, | |
| "grad_norm": 0.2134462147951126, | |
| "learning_rate": 0.00011110576923076923, | |
| "loss": 0.0881, | |
| "mean_token_accuracy": 0.9670575857162476, | |
| "num_tokens": 1893376.0, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 4.471437161755863, | |
| "grad_norm": 0.16387952864170074, | |
| "learning_rate": 0.000110625, | |
| "loss": 0.0871, | |
| "mean_token_accuracy": 0.9688689664006234, | |
| "num_tokens": 1903616.0, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 4.495490078171978, | |
| "grad_norm": 0.1718788743019104, | |
| "learning_rate": 0.00011014423076923077, | |
| "loss": 0.0887, | |
| "mean_token_accuracy": 0.9679651662707329, | |
| "num_tokens": 1913856.0, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 4.519542994588094, | |
| "grad_norm": 0.13868968188762665, | |
| "learning_rate": 0.00010966346153846154, | |
| "loss": 0.0869, | |
| "mean_token_accuracy": 0.9675926879048348, | |
| "num_tokens": 1924096.0, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 4.543595911004209, | |
| "grad_norm": 0.23226922750473022, | |
| "learning_rate": 0.00010918269230769232, | |
| "loss": 0.0893, | |
| "mean_token_accuracy": 0.9671968877315521, | |
| "num_tokens": 1934336.0, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 4.567648827420324, | |
| "grad_norm": 0.2364928126335144, | |
| "learning_rate": 0.00010870192307692309, | |
| "loss": 0.0868, | |
| "mean_token_accuracy": 0.9687703013420105, | |
| "num_tokens": 1944576.0, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.59170174383644, | |
| "grad_norm": 0.46209245920181274, | |
| "learning_rate": 0.00010822115384615385, | |
| "loss": 0.0908, | |
| "mean_token_accuracy": 0.9664398193359375, | |
| "num_tokens": 1954816.0, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 4.615754660252556, | |
| "grad_norm": 0.164833664894104, | |
| "learning_rate": 0.00010774038461538462, | |
| "loss": 0.0848, | |
| "mean_token_accuracy": 0.9686171919107437, | |
| "num_tokens": 1965056.0, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 4.639807576668671, | |
| "grad_norm": 0.2000456005334854, | |
| "learning_rate": 0.0001072596153846154, | |
| "loss": 0.0847, | |
| "mean_token_accuracy": 0.9683021917939186, | |
| "num_tokens": 1975296.0, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 4.663860493084787, | |
| "grad_norm": 0.1978590190410614, | |
| "learning_rate": 0.00010677884615384617, | |
| "loss": 0.0881, | |
| "mean_token_accuracy": 0.9670929208397865, | |
| "num_tokens": 1985536.0, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 4.687913409500902, | |
| "grad_norm": 0.3182586133480072, | |
| "learning_rate": 0.00010629807692307694, | |
| "loss": 0.0913, | |
| "mean_token_accuracy": 0.9663867846131324, | |
| "num_tokens": 1995776.0, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 4.711966325917017, | |
| "grad_norm": 0.18092605471611023, | |
| "learning_rate": 0.00010581730769230769, | |
| "loss": 0.0905, | |
| "mean_token_accuracy": 0.9664200291037559, | |
| "num_tokens": 2006016.0, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 4.736019242333133, | |
| "grad_norm": 0.17231811583042145, | |
| "learning_rate": 0.00010533653846153846, | |
| "loss": 0.0859, | |
| "mean_token_accuracy": 0.968836921453476, | |
| "num_tokens": 2016256.0, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 4.7600721587492485, | |
| "grad_norm": 0.15652543306350708, | |
| "learning_rate": 0.00010485576923076924, | |
| "loss": 0.0834, | |
| "mean_token_accuracy": 0.9684307157993317, | |
| "num_tokens": 2026496.0, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 4.784125075165364, | |
| "grad_norm": 0.13773974776268005, | |
| "learning_rate": 0.000104375, | |
| "loss": 0.0868, | |
| "mean_token_accuracy": 0.970444829761982, | |
| "num_tokens": 2036736.0, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 4.808177991581479, | |
| "grad_norm": 0.239529550075531, | |
| "learning_rate": 0.00010389423076923077, | |
| "loss": 0.0824, | |
| "mean_token_accuracy": 0.9694223091006279, | |
| "num_tokens": 2046976.0, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.8322309079975945, | |
| "grad_norm": 0.18709523975849152, | |
| "learning_rate": 0.00010341346153846154, | |
| "loss": 0.0949, | |
| "mean_token_accuracy": 0.965866394340992, | |
| "num_tokens": 2057216.0, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 4.85628382441371, | |
| "grad_norm": 0.2615446448326111, | |
| "learning_rate": 0.00010293269230769232, | |
| "loss": 0.083, | |
| "mean_token_accuracy": 0.9698115825653076, | |
| "num_tokens": 2067456.0, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 4.880336740829826, | |
| "grad_norm": 0.1910138875246048, | |
| "learning_rate": 0.00010245192307692308, | |
| "loss": 0.0907, | |
| "mean_token_accuracy": 0.9668553739786148, | |
| "num_tokens": 2077696.0, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 4.904389657245941, | |
| "grad_norm": 0.19025610387325287, | |
| "learning_rate": 0.00010197115384615385, | |
| "loss": 0.093, | |
| "mean_token_accuracy": 0.9664333924651146, | |
| "num_tokens": 2087936.0, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 4.928442573662057, | |
| "grad_norm": 0.35547563433647156, | |
| "learning_rate": 0.00010149038461538463, | |
| "loss": 0.0863, | |
| "mean_token_accuracy": 0.967856514453888, | |
| "num_tokens": 2098176.0, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 4.952495490078172, | |
| "grad_norm": 0.15040579438209534, | |
| "learning_rate": 0.0001010096153846154, | |
| "loss": 0.0823, | |
| "mean_token_accuracy": 0.9676536738872528, | |
| "num_tokens": 2108416.0, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 4.976548406494287, | |
| "grad_norm": 0.24159961938858032, | |
| "learning_rate": 0.00010052884615384617, | |
| "loss": 0.0853, | |
| "mean_token_accuracy": 0.9684896349906922, | |
| "num_tokens": 2118656.0, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.21156367659568787, | |
| "learning_rate": 0.00010004807692307693, | |
| "loss": 0.0872, | |
| "mean_token_accuracy": 0.9677492471841666, | |
| "num_tokens": 2128640.0, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.12297015637159348, | |
| "eval_mean_token_accuracy": 0.960414043132295, | |
| "eval_num_tokens": 2128640.0, | |
| "eval_runtime": 26.101, | |
| "eval_samples_per_second": 14.176, | |
| "eval_steps_per_second": 1.801, | |
| "step": 2080 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 4160, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.631112336965632e+16, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |