| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.861614129219763, | |
| "eval_steps": 50, | |
| "global_step": 800, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.035770176615247035, | |
| "grad_norm": 0.2605220377445221, | |
| "learning_rate": 1.1904761904761905e-05, | |
| "loss": 1.9177, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07154035323049407, | |
| "grad_norm": 0.29095056653022766, | |
| "learning_rate": 2.380952380952381e-05, | |
| "loss": 1.8015, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.10731052984574112, | |
| "grad_norm": 0.2734232544898987, | |
| "learning_rate": 3.571428571428572e-05, | |
| "loss": 1.5329, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.14308070646098814, | |
| "grad_norm": 0.3937128186225891, | |
| "learning_rate": 4.761904761904762e-05, | |
| "loss": 1.0377, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1788508830762352, | |
| "grad_norm": 0.08446656167507172, | |
| "learning_rate": 5.9523809523809524e-05, | |
| "loss": 0.3894, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1788508830762352, | |
| "eval_loss": 0.2632354497909546, | |
| "eval_runtime": 681.8071, | |
| "eval_samples_per_second": 2.916, | |
| "eval_steps_per_second": 0.729, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.21462105969148224, | |
| "grad_norm": 0.03653671219944954, | |
| "learning_rate": 7.142857142857143e-05, | |
| "loss": 0.2474, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.25039123630672927, | |
| "grad_norm": 0.04479149729013443, | |
| "learning_rate": 8.333333333333334e-05, | |
| "loss": 0.2377, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2861614129219763, | |
| "grad_norm": 0.025527365505695343, | |
| "learning_rate": 9.523809523809524e-05, | |
| "loss": 0.2238, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.32193158953722334, | |
| "grad_norm": 0.02618996612727642, | |
| "learning_rate": 9.920318725099602e-05, | |
| "loss": 0.2207, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3577017661524704, | |
| "grad_norm": 0.02924993820488453, | |
| "learning_rate": 9.787516600265605e-05, | |
| "loss": 0.214, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3577017661524704, | |
| "eval_loss": 0.211310014128685, | |
| "eval_runtime": 681.6444, | |
| "eval_samples_per_second": 2.916, | |
| "eval_steps_per_second": 0.729, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3934719427677174, | |
| "grad_norm": 0.024481266736984253, | |
| "learning_rate": 9.654714475431607e-05, | |
| "loss": 0.2125, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.42924211938296447, | |
| "grad_norm": 0.023656228557229042, | |
| "learning_rate": 9.52191235059761e-05, | |
| "loss": 0.2034, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4650122959982115, | |
| "grad_norm": 0.023205537348985672, | |
| "learning_rate": 9.389110225763613e-05, | |
| "loss": 0.2023, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5007824726134585, | |
| "grad_norm": 0.025412462651729584, | |
| "learning_rate": 9.256308100929614e-05, | |
| "loss": 0.2017, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5365526492287056, | |
| "grad_norm": 0.025770917534828186, | |
| "learning_rate": 9.123505976095618e-05, | |
| "loss": 0.1987, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5365526492287056, | |
| "eval_loss": 0.19897614419460297, | |
| "eval_runtime": 681.7987, | |
| "eval_samples_per_second": 2.916, | |
| "eval_steps_per_second": 0.729, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5723228258439526, | |
| "grad_norm": 0.05282260850071907, | |
| "learning_rate": 8.990703851261621e-05, | |
| "loss": 0.1981, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.6080930024591996, | |
| "grad_norm": 0.03017415478825569, | |
| "learning_rate": 8.857901726427624e-05, | |
| "loss": 0.1961, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6438631790744467, | |
| "grad_norm": 0.032814837992191315, | |
| "learning_rate": 8.725099601593627e-05, | |
| "loss": 0.1972, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6796333556896937, | |
| "grad_norm": 0.03899679705500603, | |
| "learning_rate": 8.592297476759629e-05, | |
| "loss": 0.1915, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.7154035323049408, | |
| "grad_norm": 0.037111785262823105, | |
| "learning_rate": 8.459495351925632e-05, | |
| "loss": 0.1917, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7154035323049408, | |
| "eval_loss": 0.18888385593891144, | |
| "eval_runtime": 681.8781, | |
| "eval_samples_per_second": 2.915, | |
| "eval_steps_per_second": 0.729, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7511737089201878, | |
| "grad_norm": 0.03370217606425285, | |
| "learning_rate": 8.326693227091635e-05, | |
| "loss": 0.1883, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7869438855354348, | |
| "grad_norm": 0.038315266370773315, | |
| "learning_rate": 8.193891102257637e-05, | |
| "loss": 0.1871, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.8227140621506819, | |
| "grad_norm": 0.039963819086551666, | |
| "learning_rate": 8.061088977423639e-05, | |
| "loss": 0.1833, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.8584842387659289, | |
| "grad_norm": 0.03971581161022186, | |
| "learning_rate": 7.928286852589642e-05, | |
| "loss": 0.1813, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8942544153811759, | |
| "grad_norm": 0.0481211394071579, | |
| "learning_rate": 7.795484727755644e-05, | |
| "loss": 0.1791, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8942544153811759, | |
| "eval_loss": 0.17786787450313568, | |
| "eval_runtime": 681.7787, | |
| "eval_samples_per_second": 2.916, | |
| "eval_steps_per_second": 0.729, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.930024591996423, | |
| "grad_norm": 0.04264125972986221, | |
| "learning_rate": 7.662682602921647e-05, | |
| "loss": 0.1737, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.96579476861167, | |
| "grad_norm": 0.04976845905184746, | |
| "learning_rate": 7.52988047808765e-05, | |
| "loss": 0.1749, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.001564945226917, | |
| "grad_norm": 0.05254867300391197, | |
| "learning_rate": 7.397078353253653e-05, | |
| "loss": 0.1719, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.037335121842164, | |
| "grad_norm": 0.055027466267347336, | |
| "learning_rate": 7.264276228419655e-05, | |
| "loss": 0.1651, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.0731052984574112, | |
| "grad_norm": 0.07110296189785004, | |
| "learning_rate": 7.131474103585658e-05, | |
| "loss": 0.1612, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.0731052984574112, | |
| "eval_loss": 0.16210857033729553, | |
| "eval_runtime": 680.436, | |
| "eval_samples_per_second": 2.922, | |
| "eval_steps_per_second": 0.73, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.1088754750726582, | |
| "grad_norm": 0.06281454861164093, | |
| "learning_rate": 6.998671978751661e-05, | |
| "loss": 0.1575, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.144645651687905, | |
| "grad_norm": 0.07971349358558655, | |
| "learning_rate": 6.865869853917662e-05, | |
| "loss": 0.1534, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.1804158283031523, | |
| "grad_norm": 0.08370919525623322, | |
| "learning_rate": 6.733067729083665e-05, | |
| "loss": 0.1487, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.2161860049183992, | |
| "grad_norm": 0.08752310276031494, | |
| "learning_rate": 6.600265604249668e-05, | |
| "loss": 0.1461, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.2519561815336462, | |
| "grad_norm": 0.09238439798355103, | |
| "learning_rate": 6.46746347941567e-05, | |
| "loss": 0.1394, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.2519561815336462, | |
| "eval_loss": 0.13786284625530243, | |
| "eval_runtime": 680.5234, | |
| "eval_samples_per_second": 2.921, | |
| "eval_steps_per_second": 0.73, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.2877263581488934, | |
| "grad_norm": 0.10012154281139374, | |
| "learning_rate": 6.334661354581673e-05, | |
| "loss": 0.1359, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.3234965347641405, | |
| "grad_norm": 0.11639019101858139, | |
| "learning_rate": 6.201859229747676e-05, | |
| "loss": 0.1309, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.3592667113793875, | |
| "grad_norm": 0.10917984694242477, | |
| "learning_rate": 6.069057104913679e-05, | |
| "loss": 0.123, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.3950368879946344, | |
| "grad_norm": 0.1265992373228073, | |
| "learning_rate": 5.936254980079682e-05, | |
| "loss": 0.1161, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.4308070646098816, | |
| "grad_norm": 0.12816202640533447, | |
| "learning_rate": 5.803452855245685e-05, | |
| "loss": 0.1115, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.4308070646098816, | |
| "eval_loss": 0.10932095348834991, | |
| "eval_runtime": 680.5519, | |
| "eval_samples_per_second": 2.921, | |
| "eval_steps_per_second": 0.73, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.4665772412251286, | |
| "grad_norm": 0.13660216331481934, | |
| "learning_rate": 5.6706507304116875e-05, | |
| "loss": 0.105, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.5023474178403755, | |
| "grad_norm": 0.14816845953464508, | |
| "learning_rate": 5.537848605577689e-05, | |
| "loss": 0.1016, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.5381175944556227, | |
| "grad_norm": 0.16880381107330322, | |
| "learning_rate": 5.405046480743692e-05, | |
| "loss": 0.0949, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.5738877710708696, | |
| "grad_norm": 0.15608805418014526, | |
| "learning_rate": 5.2722443559096944e-05, | |
| "loss": 0.09, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.6096579476861166, | |
| "grad_norm": 0.15114428102970123, | |
| "learning_rate": 5.139442231075697e-05, | |
| "loss": 0.0852, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.6096579476861166, | |
| "eval_loss": 0.08125962316989899, | |
| "eval_runtime": 680.5355, | |
| "eval_samples_per_second": 2.921, | |
| "eval_steps_per_second": 0.73, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.6454281243013638, | |
| "grad_norm": 0.1671249121427536, | |
| "learning_rate": 5.0066401062417e-05, | |
| "loss": 0.0793, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.681198300916611, | |
| "grad_norm": 0.16212376952171326, | |
| "learning_rate": 4.8738379814077026e-05, | |
| "loss": 0.0745, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.7169684775318577, | |
| "grad_norm": 0.1631159633398056, | |
| "learning_rate": 4.7410358565737054e-05, | |
| "loss": 0.0699, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.7527386541471048, | |
| "grad_norm": 0.17926354706287384, | |
| "learning_rate": 4.608233731739708e-05, | |
| "loss": 0.0621, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.788508830762352, | |
| "grad_norm": 0.1659461259841919, | |
| "learning_rate": 4.475431606905711e-05, | |
| "loss": 0.06, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.788508830762352, | |
| "eval_loss": 0.058878425508737564, | |
| "eval_runtime": 680.5096, | |
| "eval_samples_per_second": 2.921, | |
| "eval_steps_per_second": 0.73, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.824279007377599, | |
| "grad_norm": 0.1780182421207428, | |
| "learning_rate": 4.3426294820717136e-05, | |
| "loss": 0.0564, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.860049183992846, | |
| "grad_norm": 0.17848838865756989, | |
| "learning_rate": 4.2098273572377163e-05, | |
| "loss": 0.0522, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.895819360608093, | |
| "grad_norm": 0.17067597806453705, | |
| "learning_rate": 4.077025232403719e-05, | |
| "loss": 0.0505, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.93158953722334, | |
| "grad_norm": 0.16824504733085632, | |
| "learning_rate": 3.944223107569721e-05, | |
| "loss": 0.0464, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.967359713838587, | |
| "grad_norm": 0.17104943096637726, | |
| "learning_rate": 3.811420982735724e-05, | |
| "loss": 0.043, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.967359713838587, | |
| "eval_loss": 0.04172534495592117, | |
| "eval_runtime": 680.4722, | |
| "eval_samples_per_second": 2.922, | |
| "eval_steps_per_second": 0.73, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.003129890453834, | |
| "grad_norm": 0.1792486160993576, | |
| "learning_rate": 3.6786188579017266e-05, | |
| "loss": 0.0399, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.0389000670690813, | |
| "grad_norm": 0.13952085375785828, | |
| "learning_rate": 3.5458167330677294e-05, | |
| "loss": 0.0327, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.074670243684328, | |
| "grad_norm": 0.1532105654478073, | |
| "learning_rate": 3.4130146082337314e-05, | |
| "loss": 0.0308, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.1104404202995752, | |
| "grad_norm": 0.16048988699913025, | |
| "learning_rate": 3.280212483399734e-05, | |
| "loss": 0.0295, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.1462105969148224, | |
| "grad_norm": 0.13789258897304535, | |
| "learning_rate": 3.147410358565737e-05, | |
| "loss": 0.0294, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.1462105969148224, | |
| "eval_loss": 0.030796948820352554, | |
| "eval_runtime": 681.0206, | |
| "eval_samples_per_second": 2.919, | |
| "eval_steps_per_second": 0.73, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.181980773530069, | |
| "grad_norm": 0.1437063366174698, | |
| "learning_rate": 3.01460823373174e-05, | |
| "loss": 0.0273, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.2177509501453163, | |
| "grad_norm": 0.14592672884464264, | |
| "learning_rate": 2.8818061088977427e-05, | |
| "loss": 0.0255, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.2535211267605635, | |
| "grad_norm": 0.14989186823368073, | |
| "learning_rate": 2.7490039840637448e-05, | |
| "loss": 0.0238, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.28929130337581, | |
| "grad_norm": 0.13284747302532196, | |
| "learning_rate": 2.6162018592297476e-05, | |
| "loss": 0.0227, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.3250614799910574, | |
| "grad_norm": 0.13112632930278778, | |
| "learning_rate": 2.4833997343957506e-05, | |
| "loss": 0.0222, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.3250614799910574, | |
| "eval_loss": 0.023032352328300476, | |
| "eval_runtime": 681.8193, | |
| "eval_samples_per_second": 2.916, | |
| "eval_steps_per_second": 0.729, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.3608316566063046, | |
| "grad_norm": 0.12297944724559784, | |
| "learning_rate": 2.3505976095617534e-05, | |
| "loss": 0.0206, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.3966018332215517, | |
| "grad_norm": 0.13545510172843933, | |
| "learning_rate": 2.2177954847277558e-05, | |
| "loss": 0.0197, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.4323720098367985, | |
| "grad_norm": 0.11164336651563644, | |
| "learning_rate": 2.0849933598937585e-05, | |
| "loss": 0.0192, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.4681421864520456, | |
| "grad_norm": 0.11971808969974518, | |
| "learning_rate": 1.952191235059761e-05, | |
| "loss": 0.0187, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.5039123630672924, | |
| "grad_norm": 0.08939900994300842, | |
| "learning_rate": 1.8193891102257637e-05, | |
| "loss": 0.0177, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.5039123630672924, | |
| "eval_loss": 0.018428906798362732, | |
| "eval_runtime": 681.7832, | |
| "eval_samples_per_second": 2.916, | |
| "eval_steps_per_second": 0.729, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.5396825396825395, | |
| "grad_norm": 0.09480512142181396, | |
| "learning_rate": 1.6865869853917664e-05, | |
| "loss": 0.017, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.5754527162977867, | |
| "grad_norm": 0.11313159018754959, | |
| "learning_rate": 1.553784860557769e-05, | |
| "loss": 0.0164, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.611222892913034, | |
| "grad_norm": 0.10267580300569534, | |
| "learning_rate": 1.4209827357237717e-05, | |
| "loss": 0.0161, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.646993069528281, | |
| "grad_norm": 0.11123546212911606, | |
| "learning_rate": 1.2881806108897743e-05, | |
| "loss": 0.0154, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.682763246143528, | |
| "grad_norm": 0.09317459166049957, | |
| "learning_rate": 1.1553784860557769e-05, | |
| "loss": 0.0152, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.682763246143528, | |
| "eval_loss": 0.015589827671647072, | |
| "eval_runtime": 681.7975, | |
| "eval_samples_per_second": 2.916, | |
| "eval_steps_per_second": 0.729, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.718533422758775, | |
| "grad_norm": 0.0923512727022171, | |
| "learning_rate": 1.0225763612217796e-05, | |
| "loss": 0.015, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.7543035993740217, | |
| "grad_norm": 0.09502092003822327, | |
| "learning_rate": 8.897742363877823e-06, | |
| "loss": 0.0143, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.790073775989269, | |
| "grad_norm": 0.09042944759130478, | |
| "learning_rate": 7.569721115537849e-06, | |
| "loss": 0.0143, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.825843952604516, | |
| "grad_norm": 0.08655345439910889, | |
| "learning_rate": 6.241699867197876e-06, | |
| "loss": 0.0139, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.861614129219763, | |
| "grad_norm": 0.07545724511146545, | |
| "learning_rate": 4.9136786188579014e-06, | |
| "loss": 0.0137, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.861614129219763, | |
| "eval_loss": 0.014045392163097858, | |
| "eval_runtime": 681.381, | |
| "eval_samples_per_second": 2.918, | |
| "eval_steps_per_second": 0.729, | |
| "step": 800 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 837, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.194702349626769e+19, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |