| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 141420, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07071135624381275, | |
| "grad_norm": 1.3512808084487915, | |
| "learning_rate": 9.375e-06, | |
| "loss": 5.8359, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.1414227124876255, | |
| "grad_norm": 1.6514761447906494, | |
| "learning_rate": 1.875e-05, | |
| "loss": 3.8313, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.21213406873143828, | |
| "grad_norm": 1.4630589485168457, | |
| "learning_rate": 2.8125e-05, | |
| "loss": 3.4288, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.282845424975251, | |
| "grad_norm": 1.2321913242340088, | |
| "learning_rate": 3.75e-05, | |
| "loss": 3.202, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.3535567812190638, | |
| "grad_norm": 1.084225058555603, | |
| "learning_rate": 4.6874999999999994e-05, | |
| "loss": 3.0637, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.42426813746287656, | |
| "grad_norm": 0.9784078598022461, | |
| "learning_rate": 5.625e-05, | |
| "loss": 2.9404, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.4949794937066893, | |
| "grad_norm": 0.9349497556686401, | |
| "learning_rate": 6.5625e-05, | |
| "loss": 2.8482, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.565690849950502, | |
| "grad_norm": 0.7968702912330627, | |
| "learning_rate": 7.5e-05, | |
| "loss": 2.7644, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.6364022061943148, | |
| "grad_norm": 0.7970491647720337, | |
| "learning_rate": 8.437499999999999e-05, | |
| "loss": 2.7035, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.7071135624381276, | |
| "grad_norm": 0.808237612247467, | |
| "learning_rate": 9.374999999999999e-05, | |
| "loss": 2.6559, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.7778249186819404, | |
| "grad_norm": 0.6886675953865051, | |
| "learning_rate": 0.00010312499999999999, | |
| "loss": 2.5948, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.8485362749257531, | |
| "grad_norm": 0.7730551362037659, | |
| "learning_rate": 0.0001125, | |
| "loss": 2.5558, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.9192476311695659, | |
| "grad_norm": 0.6982184052467346, | |
| "learning_rate": 0.000121865625, | |
| "loss": 2.5352, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.9899589874133786, | |
| "grad_norm": 0.666688084602356, | |
| "learning_rate": 0.000131240625, | |
| "loss": 2.5065, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.49123160368602076, | |
| "eval_loss": 2.7265632152557373, | |
| "eval_runtime": 100.7986, | |
| "eval_samples_per_second": 464.967, | |
| "eval_steps_per_second": 7.272, | |
| "step": 14142 | |
| }, | |
| { | |
| "epoch": 1.0606703436571914, | |
| "grad_norm": 0.6360883116722107, | |
| "learning_rate": 0.00014060625, | |
| "loss": 2.449, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.131381699901004, | |
| "grad_norm": 0.659411609172821, | |
| "learning_rate": 0.00014998125, | |
| "loss": 2.4338, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.2020930561448169, | |
| "grad_norm": 0.6129982471466064, | |
| "learning_rate": 0.000159346875, | |
| "loss": 2.4172, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.2728044123886297, | |
| "grad_norm": 0.6230875849723816, | |
| "learning_rate": 0.00016872187499999999, | |
| "loss": 2.4099, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.3435157686324424, | |
| "grad_norm": 0.5862132906913757, | |
| "learning_rate": 0.00017808749999999999, | |
| "loss": 2.403, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.414227124876255, | |
| "grad_norm": 0.5603518486022949, | |
| "learning_rate": 0.00018746249999999998, | |
| "loss": 2.3744, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.4849384811200679, | |
| "grad_norm": 0.5778409242630005, | |
| "learning_rate": 0.00019682812499999998, | |
| "loss": 2.3861, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.5556498373638807, | |
| "grad_norm": 0.5504118800163269, | |
| "learning_rate": 0.00020620312499999998, | |
| "loss": 2.3661, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 1.6263611936076934, | |
| "grad_norm": 0.5299841165542603, | |
| "learning_rate": 0.00021556874999999998, | |
| "loss": 2.371, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 1.697072549851506, | |
| "grad_norm": 0.4794081747531891, | |
| "learning_rate": 0.00022494374999999998, | |
| "loss": 2.3571, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.7677839060953189, | |
| "grad_norm": 0.5159916877746582, | |
| "learning_rate": 0.00023430937499999997, | |
| "loss": 2.3483, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.8384952623391317, | |
| "grad_norm": 0.4809582233428955, | |
| "learning_rate": 0.00024368437499999997, | |
| "loss": 2.3403, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 1.9092066185829444, | |
| "grad_norm": 0.4775105118751526, | |
| "learning_rate": 0.00025305, | |
| "loss": 2.3318, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 1.979917974826757, | |
| "grad_norm": 0.46707651019096375, | |
| "learning_rate": 0.000262425, | |
| "loss": 2.3323, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5073636729232759, | |
| "eval_loss": 2.5641860961914062, | |
| "eval_runtime": 101.8267, | |
| "eval_samples_per_second": 460.272, | |
| "eval_steps_per_second": 7.199, | |
| "step": 28284 | |
| }, | |
| { | |
| "epoch": 2.05062933107057, | |
| "grad_norm": 0.4617663025856018, | |
| "learning_rate": 0.000271790625, | |
| "loss": 2.2778, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 2.1213406873143827, | |
| "grad_norm": 0.4646793007850647, | |
| "learning_rate": 0.000281165625, | |
| "loss": 2.2697, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.1920520435581956, | |
| "grad_norm": 0.4711323082447052, | |
| "learning_rate": 0.00029053124999999994, | |
| "loss": 2.2687, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 2.262763399802008, | |
| "grad_norm": 0.4392153322696686, | |
| "learning_rate": 0.000299896875, | |
| "loss": 2.263, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 2.333474756045821, | |
| "grad_norm": 0.4188326299190521, | |
| "learning_rate": 0.00029728842990312557, | |
| "loss": 2.2572, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 2.4041861122896337, | |
| "grad_norm": 0.4100128710269928, | |
| "learning_rate": 0.0002945467007859623, | |
| "loss": 2.2616, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 2.4748974685334466, | |
| "grad_norm": 0.3786410391330719, | |
| "learning_rate": 0.0002918049716687991, | |
| "loss": 2.2527, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 2.5456088247772595, | |
| "grad_norm": 0.40753409266471863, | |
| "learning_rate": 0.00028906598428075306, | |
| "loss": 2.2385, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 2.616320181021072, | |
| "grad_norm": 0.3649774193763733, | |
| "learning_rate": 0.000286326996892707, | |
| "loss": 2.2364, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 2.6870315372648848, | |
| "grad_norm": 0.35535377264022827, | |
| "learning_rate": 0.00028358526777554375, | |
| "loss": 2.2312, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 2.7577428935086976, | |
| "grad_norm": 0.3544033169746399, | |
| "learning_rate": 0.0002808462803874977, | |
| "loss": 2.2245, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 2.82845424975251, | |
| "grad_norm": 0.3812442719936371, | |
| "learning_rate": 0.00027810455127033444, | |
| "loss": 2.2202, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 2.899165605996323, | |
| "grad_norm": 0.3528529405593872, | |
| "learning_rate": 0.00027536282215317125, | |
| "loss": 2.2127, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 2.9698769622401358, | |
| "grad_norm": 0.36804184317588806, | |
| "learning_rate": 0.000272621093036008, | |
| "loss": 2.2158, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.518365513011292, | |
| "eval_loss": 2.4669742584228516, | |
| "eval_runtime": 101.8636, | |
| "eval_samples_per_second": 460.105, | |
| "eval_steps_per_second": 7.196, | |
| "step": 42426 | |
| }, | |
| { | |
| "epoch": 3.0405883184839486, | |
| "grad_norm": 0.4022282063961029, | |
| "learning_rate": 0.00026988210564796194, | |
| "loss": 2.1561, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 3.1112996747277615, | |
| "grad_norm": 0.36605340242385864, | |
| "learning_rate": 0.00026714037653079875, | |
| "loss": 2.1226, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 3.182011030971574, | |
| "grad_norm": 0.34304675459861755, | |
| "learning_rate": 0.0002644013891427527, | |
| "loss": 2.1217, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 3.2527223872153868, | |
| "grad_norm": 0.35023054480552673, | |
| "learning_rate": 0.00026165966002558944, | |
| "loss": 2.1274, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 3.3234337434591996, | |
| "grad_norm": 0.3564451038837433, | |
| "learning_rate": 0.00025891793090842625, | |
| "loss": 2.1314, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 3.3941450997030125, | |
| "grad_norm": 0.3799506425857544, | |
| "learning_rate": 0.000256176201791263, | |
| "loss": 2.1193, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 3.464856455946825, | |
| "grad_norm": 0.3519129455089569, | |
| "learning_rate": 0.00025343721440321693, | |
| "loss": 2.1219, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 3.5355678121906378, | |
| "grad_norm": 0.3426956534385681, | |
| "learning_rate": 0.00025069548528605374, | |
| "loss": 2.1172, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 3.6062791684344506, | |
| "grad_norm": 0.3629501760005951, | |
| "learning_rate": 0.0002479564978980076, | |
| "loss": 2.1177, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 3.6769905246782635, | |
| "grad_norm": 0.3377913534641266, | |
| "learning_rate": 0.00024521476878084443, | |
| "loss": 2.1175, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 3.747701880922076, | |
| "grad_norm": 0.339662104845047, | |
| "learning_rate": 0.00024247303966368121, | |
| "loss": 2.1186, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 3.8184132371658888, | |
| "grad_norm": 0.34301745891571045, | |
| "learning_rate": 0.00023973405227563515, | |
| "loss": 2.1155, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 3.8891245934097016, | |
| "grad_norm": 0.3543142080307007, | |
| "learning_rate": 0.00023699232315847193, | |
| "loss": 2.1056, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 3.9598359496535145, | |
| "grad_norm": 0.3132970929145813, | |
| "learning_rate": 0.0002342533357704259, | |
| "loss": 2.1109, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5248612794140942, | |
| "eval_loss": 2.4178478717803955, | |
| "eval_runtime": 102.1506, | |
| "eval_samples_per_second": 458.813, | |
| "eval_steps_per_second": 7.176, | |
| "step": 56568 | |
| }, | |
| { | |
| "epoch": 4.030547305897327, | |
| "grad_norm": 0.3517058491706848, | |
| "learning_rate": 0.00023151160665326262, | |
| "loss": 2.055, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 4.10125866214114, | |
| "grad_norm": 0.35281458497047424, | |
| "learning_rate": 0.00022877261926521658, | |
| "loss": 2.0072, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 4.171970018384952, | |
| "grad_norm": 0.3426903188228607, | |
| "learning_rate": 0.00022603089014805336, | |
| "loss": 2.0172, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 4.2426813746287655, | |
| "grad_norm": 0.35942381620407104, | |
| "learning_rate": 0.0002232919027600073, | |
| "loss": 2.0161, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 4.313392730872578, | |
| "grad_norm": 0.36592814326286316, | |
| "learning_rate": 0.00022055017364284408, | |
| "loss": 2.0166, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 4.384104087116391, | |
| "grad_norm": 0.3915741741657257, | |
| "learning_rate": 0.000217811186254798, | |
| "loss": 2.0206, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 4.454815443360204, | |
| "grad_norm": 0.3634374439716339, | |
| "learning_rate": 0.0002150694571376348, | |
| "loss": 2.0227, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 4.525526799604016, | |
| "grad_norm": 0.38214239478111267, | |
| "learning_rate": 0.00021233046974958875, | |
| "loss": 2.0201, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 4.596238155847829, | |
| "grad_norm": 0.33417147397994995, | |
| "learning_rate": 0.00020958874063242548, | |
| "loss": 2.0242, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 4.666949512091642, | |
| "grad_norm": 0.3682495057582855, | |
| "learning_rate": 0.00020684975324437944, | |
| "loss": 2.0351, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 4.737660868335455, | |
| "grad_norm": 0.36803096532821655, | |
| "learning_rate": 0.00020410802412721622, | |
| "loss": 2.0257, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 4.8083722245792675, | |
| "grad_norm": 0.3615093529224396, | |
| "learning_rate": 0.00020136903673917015, | |
| "loss": 2.033, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 4.87908358082308, | |
| "grad_norm": 0.3575078845024109, | |
| "learning_rate": 0.00019862730762200694, | |
| "loss": 2.0305, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 4.949794937066893, | |
| "grad_norm": 0.34489330649375916, | |
| "learning_rate": 0.00019588832023396087, | |
| "loss": 2.0194, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5280184201411432, | |
| "eval_loss": 2.400068998336792, | |
| "eval_runtime": 102.0615, | |
| "eval_samples_per_second": 459.213, | |
| "eval_steps_per_second": 7.182, | |
| "step": 70710 | |
| }, | |
| { | |
| "epoch": 5.020506293310706, | |
| "grad_norm": 0.3643249273300171, | |
| "learning_rate": 0.00019314659111679765, | |
| "loss": 1.9864, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 5.091217649554518, | |
| "grad_norm": 0.4295579195022583, | |
| "learning_rate": 0.00019040760372875156, | |
| "loss": 1.9059, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 5.161929005798331, | |
| "grad_norm": 0.3856581151485443, | |
| "learning_rate": 0.00018766587461158834, | |
| "loss": 1.9127, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 5.232640362042144, | |
| "grad_norm": 0.39465850591659546, | |
| "learning_rate": 0.0001849268872235423, | |
| "loss": 1.918, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 5.303351718285957, | |
| "grad_norm": 0.37704503536224365, | |
| "learning_rate": 0.00018218515810637905, | |
| "loss": 1.9239, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 5.3740630745297695, | |
| "grad_norm": 0.3656931221485138, | |
| "learning_rate": 0.00017944342898921584, | |
| "loss": 1.9284, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 5.444774430773582, | |
| "grad_norm": 0.3781563341617584, | |
| "learning_rate": 0.0001767044416011698, | |
| "loss": 1.9319, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 5.515485787017395, | |
| "grad_norm": 0.38474106788635254, | |
| "learning_rate": 0.00017396271248400655, | |
| "loss": 1.9363, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 5.586197143261208, | |
| "grad_norm": 0.35963472723960876, | |
| "learning_rate": 0.0001712237250959605, | |
| "loss": 1.9504, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 5.65690849950502, | |
| "grad_norm": 0.38337036967277527, | |
| "learning_rate": 0.0001684819959787973, | |
| "loss": 1.9397, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 5.727619855748833, | |
| "grad_norm": 0.41345345973968506, | |
| "learning_rate": 0.00016574300859075123, | |
| "loss": 1.9517, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 5.798331211992646, | |
| "grad_norm": 0.4042370021343231, | |
| "learning_rate": 0.000163001279473588, | |
| "loss": 1.9455, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 5.869042568236459, | |
| "grad_norm": 0.3603546619415283, | |
| "learning_rate": 0.00016026229208554192, | |
| "loss": 1.9401, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 5.9397539244802715, | |
| "grad_norm": 0.3735804557800293, | |
| "learning_rate": 0.0001575205629683787, | |
| "loss": 1.938, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.5290693059969069, | |
| "eval_loss": 2.4066717624664307, | |
| "eval_runtime": 102.8194, | |
| "eval_samples_per_second": 455.828, | |
| "eval_steps_per_second": 7.129, | |
| "step": 84852 | |
| }, | |
| { | |
| "epoch": 6.010465280724084, | |
| "grad_norm": 0.3881630599498749, | |
| "learning_rate": 0.00015478157558033266, | |
| "loss": 1.9349, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 6.081176636967897, | |
| "grad_norm": 0.41385406255722046, | |
| "learning_rate": 0.0001520398464631694, | |
| "loss": 1.8133, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 6.15188799321171, | |
| "grad_norm": 0.4141051173210144, | |
| "learning_rate": 0.0001492981173460062, | |
| "loss": 1.8269, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 6.222599349455523, | |
| "grad_norm": 0.43294477462768555, | |
| "learning_rate": 0.00014655912995796013, | |
| "loss": 1.8367, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 6.293310705699335, | |
| "grad_norm": 0.40869224071502686, | |
| "learning_rate": 0.0001438174008407969, | |
| "loss": 1.8283, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 6.364022061943148, | |
| "grad_norm": 0.43988147377967834, | |
| "learning_rate": 0.00014107841345275087, | |
| "loss": 1.8398, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 6.434733418186961, | |
| "grad_norm": 0.4311169385910034, | |
| "learning_rate": 0.00013833668433558763, | |
| "loss": 1.8467, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 6.5054447744307735, | |
| "grad_norm": 0.4262569844722748, | |
| "learning_rate": 0.00013559769694754156, | |
| "loss": 1.8504, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 6.576156130674587, | |
| "grad_norm": 0.4127192199230194, | |
| "learning_rate": 0.00013285596783037834, | |
| "loss": 1.8553, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 6.646867486918399, | |
| "grad_norm": 0.4223586916923523, | |
| "learning_rate": 0.0001301169804423323, | |
| "loss": 1.8472, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 6.717578843162212, | |
| "grad_norm": 0.40577736496925354, | |
| "learning_rate": 0.00012737525132516906, | |
| "loss": 1.8497, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 6.788290199406025, | |
| "grad_norm": 0.42064401507377625, | |
| "learning_rate": 0.000124636263937123, | |
| "loss": 1.8625, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 6.859001555649837, | |
| "grad_norm": 0.4312768578529358, | |
| "learning_rate": 0.00012189453481995978, | |
| "loss": 1.8687, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 6.92971291189365, | |
| "grad_norm": 0.41861584782600403, | |
| "learning_rate": 0.00011915554743191372, | |
| "loss": 1.8569, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.5282999459696228, | |
| "eval_loss": 2.4313058853149414, | |
| "eval_runtime": 102.0616, | |
| "eval_samples_per_second": 459.213, | |
| "eval_steps_per_second": 7.182, | |
| "step": 98994 | |
| }, | |
| { | |
| "epoch": 7.000424268137463, | |
| "grad_norm": 0.40982282161712646, | |
| "learning_rate": 0.00011641381831475049, | |
| "loss": 1.8623, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 7.0711356243812755, | |
| "grad_norm": 0.5019235610961914, | |
| "learning_rate": 0.00011367208919758727, | |
| "loss": 1.7161, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 7.141846980625088, | |
| "grad_norm": 0.44862595200538635, | |
| "learning_rate": 0.00011093310180954122, | |
| "loss": 1.7258, | |
| "step": 101000 | |
| }, | |
| { | |
| "epoch": 7.212558336868901, | |
| "grad_norm": 0.515501856803894, | |
| "learning_rate": 0.00010819137269237798, | |
| "loss": 1.7396, | |
| "step": 102000 | |
| }, | |
| { | |
| "epoch": 7.283269693112714, | |
| "grad_norm": 0.49428591132164, | |
| "learning_rate": 0.00010545238530433192, | |
| "loss": 1.7388, | |
| "step": 103000 | |
| }, | |
| { | |
| "epoch": 7.353981049356527, | |
| "grad_norm": 0.5093473196029663, | |
| "learning_rate": 0.0001027106561871687, | |
| "loss": 1.7468, | |
| "step": 104000 | |
| }, | |
| { | |
| "epoch": 7.424692405600339, | |
| "grad_norm": 0.4532296061515808, | |
| "learning_rate": 9.997166879912265e-05, | |
| "loss": 1.7588, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 7.495403761844152, | |
| "grad_norm": 0.4764035940170288, | |
| "learning_rate": 9.722993968195941e-05, | |
| "loss": 1.7627, | |
| "step": 106000 | |
| }, | |
| { | |
| "epoch": 7.566115118087965, | |
| "grad_norm": 0.46253088116645813, | |
| "learning_rate": 9.449095229391335e-05, | |
| "loss": 1.7655, | |
| "step": 107000 | |
| }, | |
| { | |
| "epoch": 7.6368264743317775, | |
| "grad_norm": 0.5122313499450684, | |
| "learning_rate": 9.174922317675013e-05, | |
| "loss": 1.7677, | |
| "step": 108000 | |
| }, | |
| { | |
| "epoch": 7.707537830575591, | |
| "grad_norm": 0.4764326214790344, | |
| "learning_rate": 8.901023578870408e-05, | |
| "loss": 1.7715, | |
| "step": 109000 | |
| }, | |
| { | |
| "epoch": 7.778249186819403, | |
| "grad_norm": 0.47460654377937317, | |
| "learning_rate": 8.626850667154085e-05, | |
| "loss": 1.7665, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 7.848960543063216, | |
| "grad_norm": 0.4747351109981537, | |
| "learning_rate": 8.352951928349478e-05, | |
| "loss": 1.7796, | |
| "step": 111000 | |
| }, | |
| { | |
| "epoch": 7.919671899307029, | |
| "grad_norm": 0.47340136766433716, | |
| "learning_rate": 8.078779016633157e-05, | |
| "loss": 1.7741, | |
| "step": 112000 | |
| }, | |
| { | |
| "epoch": 7.990383255550841, | |
| "grad_norm": 0.47387179732322693, | |
| "learning_rate": 7.804880277828549e-05, | |
| "loss": 1.7668, | |
| "step": 113000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.5260272330369687, | |
| "eval_loss": 2.4765777587890625, | |
| "eval_runtime": 102.0203, | |
| "eval_samples_per_second": 459.399, | |
| "eval_steps_per_second": 7.185, | |
| "step": 113136 | |
| }, | |
| { | |
| "epoch": 8.061094611794655, | |
| "grad_norm": 0.5383424162864685, | |
| "learning_rate": 7.530707366112228e-05, | |
| "loss": 1.6391, | |
| "step": 114000 | |
| }, | |
| { | |
| "epoch": 8.131805968038467, | |
| "grad_norm": 0.5288543105125427, | |
| "learning_rate": 7.256808627307622e-05, | |
| "loss": 1.6373, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 8.20251732428228, | |
| "grad_norm": 0.5492115616798401, | |
| "learning_rate": 6.982635715591299e-05, | |
| "loss": 1.6506, | |
| "step": 116000 | |
| }, | |
| { | |
| "epoch": 8.273228680526092, | |
| "grad_norm": 0.579088568687439, | |
| "learning_rate": 6.708736976786694e-05, | |
| "loss": 1.6573, | |
| "step": 117000 | |
| }, | |
| { | |
| "epoch": 8.343940036769904, | |
| "grad_norm": 0.5820958614349365, | |
| "learning_rate": 6.43456406507037e-05, | |
| "loss": 1.6593, | |
| "step": 118000 | |
| }, | |
| { | |
| "epoch": 8.414651393013719, | |
| "grad_norm": 0.4972964823246002, | |
| "learning_rate": 6.160391153354047e-05, | |
| "loss": 1.6636, | |
| "step": 119000 | |
| }, | |
| { | |
| "epoch": 8.485362749257531, | |
| "grad_norm": 0.537187933921814, | |
| "learning_rate": 5.886492414549442e-05, | |
| "loss": 1.6654, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 8.556074105501343, | |
| "grad_norm": 0.5469970107078552, | |
| "learning_rate": 5.61231950283312e-05, | |
| "loss": 1.6739, | |
| "step": 121000 | |
| }, | |
| { | |
| "epoch": 8.626785461745156, | |
| "grad_norm": 0.5517230033874512, | |
| "learning_rate": 5.3384207640285136e-05, | |
| "loss": 1.676, | |
| "step": 122000 | |
| }, | |
| { | |
| "epoch": 8.697496817988968, | |
| "grad_norm": 0.5250716805458069, | |
| "learning_rate": 5.064247852312191e-05, | |
| "loss": 1.6689, | |
| "step": 123000 | |
| }, | |
| { | |
| "epoch": 8.768208174232782, | |
| "grad_norm": 0.5485174059867859, | |
| "learning_rate": 4.790349113507585e-05, | |
| "loss": 1.6701, | |
| "step": 124000 | |
| }, | |
| { | |
| "epoch": 8.838919530476595, | |
| "grad_norm": 0.5471927523612976, | |
| "learning_rate": 4.516176201791263e-05, | |
| "loss": 1.6723, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 8.909630886720407, | |
| "grad_norm": 0.4902847409248352, | |
| "learning_rate": 4.2422774629866566e-05, | |
| "loss": 1.6744, | |
| "step": 126000 | |
| }, | |
| { | |
| "epoch": 8.98034224296422, | |
| "grad_norm": 0.5765844583511353, | |
| "learning_rate": 3.968104551270334e-05, | |
| "loss": 1.6733, | |
| "step": 127000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.5228845380469165, | |
| "eval_loss": 2.5416789054870605, | |
| "eval_runtime": 102.2087, | |
| "eval_samples_per_second": 458.552, | |
| "eval_steps_per_second": 7.172, | |
| "step": 127278 | |
| }, | |
| { | |
| "epoch": 9.051053599208032, | |
| "grad_norm": 0.5943803191184998, | |
| "learning_rate": 3.693931639554012e-05, | |
| "loss": 1.586, | |
| "step": 128000 | |
| }, | |
| { | |
| "epoch": 9.121764955451846, | |
| "grad_norm": 0.5351667404174805, | |
| "learning_rate": 3.420032900749406e-05, | |
| "loss": 1.5515, | |
| "step": 129000 | |
| }, | |
| { | |
| "epoch": 9.192476311695659, | |
| "grad_norm": 0.6331674456596375, | |
| "learning_rate": 3.145859989033083e-05, | |
| "loss": 1.5732, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 9.263187667939471, | |
| "grad_norm": 0.5774397253990173, | |
| "learning_rate": 2.8719612502284773e-05, | |
| "loss": 1.5673, | |
| "step": 131000 | |
| }, | |
| { | |
| "epoch": 9.333899024183284, | |
| "grad_norm": 0.5716825723648071, | |
| "learning_rate": 2.5977883385121545e-05, | |
| "loss": 1.5681, | |
| "step": 132000 | |
| }, | |
| { | |
| "epoch": 9.404610380427096, | |
| "grad_norm": 0.5614861249923706, | |
| "learning_rate": 2.3238895997075488e-05, | |
| "loss": 1.5691, | |
| "step": 133000 | |
| }, | |
| { | |
| "epoch": 9.47532173667091, | |
| "grad_norm": 0.5972283482551575, | |
| "learning_rate": 2.0497166879912264e-05, | |
| "loss": 1.5756, | |
| "step": 134000 | |
| }, | |
| { | |
| "epoch": 9.546033092914723, | |
| "grad_norm": 0.6400578022003174, | |
| "learning_rate": 1.7758179491866203e-05, | |
| "loss": 1.5708, | |
| "step": 135000 | |
| }, | |
| { | |
| "epoch": 9.616744449158535, | |
| "grad_norm": 0.5687015056610107, | |
| "learning_rate": 1.5016450374702977e-05, | |
| "loss": 1.5724, | |
| "step": 136000 | |
| }, | |
| { | |
| "epoch": 9.687455805402347, | |
| "grad_norm": 0.6058946251869202, | |
| "learning_rate": 1.2277462986656918e-05, | |
| "loss": 1.5676, | |
| "step": 137000 | |
| }, | |
| { | |
| "epoch": 9.75816716164616, | |
| "grad_norm": 0.6315460801124573, | |
| "learning_rate": 9.535733869493694e-06, | |
| "loss": 1.5771, | |
| "step": 138000 | |
| }, | |
| { | |
| "epoch": 9.828878517889972, | |
| "grad_norm": 0.5992266535758972, | |
| "learning_rate": 6.796746481447632e-06, | |
| "loss": 1.5753, | |
| "step": 139000 | |
| }, | |
| { | |
| "epoch": 9.899589874133786, | |
| "grad_norm": 0.5792600512504578, | |
| "learning_rate": 4.055017364284408e-06, | |
| "loss": 1.5747, | |
| "step": 140000 | |
| }, | |
| { | |
| "epoch": 9.970301230377599, | |
| "grad_norm": 0.5985902547836304, | |
| "learning_rate": 1.3132882471211842e-06, | |
| "loss": 1.579, | |
| "step": 141000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.5183717396220663, | |
| "eval_loss": 2.6186234951019287, | |
| "eval_runtime": 102.0754, | |
| "eval_samples_per_second": 459.151, | |
| "eval_steps_per_second": 7.181, | |
| "step": 141420 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 141420, | |
| "total_flos": 6.171008476428288e+17, | |
| "train_loss": 2.062552438357513, | |
| "train_runtime": 24003.715, | |
| "train_samples_per_second": 188.522, | |
| "train_steps_per_second": 5.892 | |
| } | |
| ], | |
| "logging_steps": 1000, | |
| "max_steps": 141420, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 5000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.171008476428288e+17, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |