| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.853519340519975, | |
| "eval_steps": 500, | |
| "global_step": 4500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03170577045022194, | |
| "grad_norm": 306.2328796386719, | |
| "learning_rate": 4.94715704924963e-05, | |
| "loss": 5.2387, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.06341154090044387, | |
| "grad_norm": 2.433814525604248, | |
| "learning_rate": 4.8943140984992603e-05, | |
| "loss": 1.131, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09511731135066583, | |
| "grad_norm": 1.5357435941696167, | |
| "learning_rate": 4.8414711477488904e-05, | |
| "loss": 0.5924, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.12682308180088775, | |
| "grad_norm": 1.2382903099060059, | |
| "learning_rate": 4.788628196998521e-05, | |
| "loss": 0.6336, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.15852885225110971, | |
| "grad_norm": 1.0461214780807495, | |
| "learning_rate": 4.7357852462481505e-05, | |
| "loss": 0.5381, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.19023462270133165, | |
| "grad_norm": 1.6466267108917236, | |
| "learning_rate": 4.6829422954977806e-05, | |
| "loss": 0.6359, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2219403931515536, | |
| "grad_norm": 0.8996474146842957, | |
| "learning_rate": 4.630099344747411e-05, | |
| "loss": 0.6849, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.2536461636017755, | |
| "grad_norm": 0.6775561571121216, | |
| "learning_rate": 4.577256393997041e-05, | |
| "loss": 0.5673, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.2853519340519975, | |
| "grad_norm": 0.6857176423072815, | |
| "learning_rate": 4.5244134432466714e-05, | |
| "loss": 0.4975, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.31705770450221943, | |
| "grad_norm": 0.7963148951530457, | |
| "learning_rate": 4.4715704924963015e-05, | |
| "loss": 0.6324, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.34876347495244137, | |
| "grad_norm": 0.7400671243667603, | |
| "learning_rate": 4.4187275417459315e-05, | |
| "loss": 0.5257, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.3804692454026633, | |
| "grad_norm": 0.8451323509216309, | |
| "learning_rate": 4.3658845909955616e-05, | |
| "loss": 0.5694, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.41217501585288524, | |
| "grad_norm": 0.8110406398773193, | |
| "learning_rate": 4.3130416402451916e-05, | |
| "loss": 0.5675, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.4438807863031072, | |
| "grad_norm": 1.4713784456253052, | |
| "learning_rate": 4.260198689494822e-05, | |
| "loss": 0.6193, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.4755865567533291, | |
| "grad_norm": 0.8999515175819397, | |
| "learning_rate": 4.207355738744452e-05, | |
| "loss": 0.5402, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.507292327203551, | |
| "grad_norm": 1.0108189582824707, | |
| "learning_rate": 4.154512787994082e-05, | |
| "loss": 0.5411, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.5389980976537729, | |
| "grad_norm": 0.8886571526527405, | |
| "learning_rate": 4.101669837243712e-05, | |
| "loss": 0.5962, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.570703868103995, | |
| "grad_norm": 1.0816487073898315, | |
| "learning_rate": 4.048826886493342e-05, | |
| "loss": 0.5302, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.6024096385542169, | |
| "grad_norm": 0.6238551735877991, | |
| "learning_rate": 3.995983935742972e-05, | |
| "loss": 0.5216, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.6341154090044389, | |
| "grad_norm": 1.0108839273452759, | |
| "learning_rate": 3.943140984992602e-05, | |
| "loss": 0.522, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6658211794546608, | |
| "grad_norm": 0.6171120405197144, | |
| "learning_rate": 3.890298034242232e-05, | |
| "loss": 0.4833, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.6975269499048827, | |
| "grad_norm": 0.634709894657135, | |
| "learning_rate": 3.837455083491862e-05, | |
| "loss": 0.5404, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.7292327203551047, | |
| "grad_norm": 0.6804570555686951, | |
| "learning_rate": 3.784612132741492e-05, | |
| "loss": 0.5194, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.7609384908053266, | |
| "grad_norm": 1.224907636642456, | |
| "learning_rate": 3.731769181991123e-05, | |
| "loss": 0.6098, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.7926442612555485, | |
| "grad_norm": 0.6107101440429688, | |
| "learning_rate": 3.678926231240752e-05, | |
| "loss": 0.5721, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.8243500317057705, | |
| "grad_norm": 0.672882080078125, | |
| "learning_rate": 3.626083280490383e-05, | |
| "loss": 0.5276, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.8560558021559924, | |
| "grad_norm": 0.7011030316352844, | |
| "learning_rate": 3.573240329740013e-05, | |
| "loss": 0.6049, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.8877615726062144, | |
| "grad_norm": 0.4889984726905823, | |
| "learning_rate": 3.5203973789896424e-05, | |
| "loss": 0.4657, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.9194673430564363, | |
| "grad_norm": 1.1118626594543457, | |
| "learning_rate": 3.467554428239273e-05, | |
| "loss": 0.5328, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.9511731135066582, | |
| "grad_norm": 1.2527000904083252, | |
| "learning_rate": 3.414711477488903e-05, | |
| "loss": 0.5355, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.9828788839568802, | |
| "grad_norm": 0.8558881878852844, | |
| "learning_rate": 3.361868526738533e-05, | |
| "loss": 0.6129, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.014584654407102, | |
| "grad_norm": 0.5950063467025757, | |
| "learning_rate": 3.309025575988163e-05, | |
| "loss": 0.5058, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.046290424857324, | |
| "grad_norm": 0.5575762987136841, | |
| "learning_rate": 3.2561826252377934e-05, | |
| "loss": 0.5286, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.0779961953075459, | |
| "grad_norm": 0.9821068644523621, | |
| "learning_rate": 3.2033396744874234e-05, | |
| "loss": 0.4943, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.1097019657577678, | |
| "grad_norm": 0.9642919301986694, | |
| "learning_rate": 3.1504967237370535e-05, | |
| "loss": 0.5102, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.1414077362079897, | |
| "grad_norm": 0.6422669887542725, | |
| "learning_rate": 3.097653772986684e-05, | |
| "loss": 0.5178, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.1731135066582117, | |
| "grad_norm": 0.6244462728500366, | |
| "learning_rate": 3.0448108222363136e-05, | |
| "loss": 0.5699, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.2048192771084336, | |
| "grad_norm": 0.6992059946060181, | |
| "learning_rate": 2.991967871485944e-05, | |
| "loss": 0.5031, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.2365250475586556, | |
| "grad_norm": 0.609976589679718, | |
| "learning_rate": 2.939124920735574e-05, | |
| "loss": 0.4497, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.2682308180088775, | |
| "grad_norm": 1.18268883228302, | |
| "learning_rate": 2.886281969985204e-05, | |
| "loss": 0.6237, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.2999365884590994, | |
| "grad_norm": 0.467398077249527, | |
| "learning_rate": 2.833439019234834e-05, | |
| "loss": 0.6139, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.3316423589093214, | |
| "grad_norm": 0.7419503927230835, | |
| "learning_rate": 2.7805960684844646e-05, | |
| "loss": 0.4893, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.3633481293595433, | |
| "grad_norm": 1.152620792388916, | |
| "learning_rate": 2.7277531177340943e-05, | |
| "loss": 0.4538, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 1.3950538998097652, | |
| "grad_norm": 0.8632370829582214, | |
| "learning_rate": 2.6749101669837247e-05, | |
| "loss": 0.4809, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.4267596702599874, | |
| "grad_norm": 0.8322456479072571, | |
| "learning_rate": 2.6220672162333547e-05, | |
| "loss": 0.4329, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.4584654407102093, | |
| "grad_norm": 0.8780176639556885, | |
| "learning_rate": 2.5692242654829844e-05, | |
| "loss": 0.4623, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.4901712111604313, | |
| "grad_norm": 1.8087239265441895, | |
| "learning_rate": 2.5163813147326148e-05, | |
| "loss": 0.4098, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 1.521876981610653, | |
| "grad_norm": 0.6666421890258789, | |
| "learning_rate": 2.463538363982245e-05, | |
| "loss": 0.4768, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.553582752060875, | |
| "grad_norm": 0.7871991991996765, | |
| "learning_rate": 2.410695413231875e-05, | |
| "loss": 0.4899, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 1.5852885225110969, | |
| "grad_norm": 0.6651085615158081, | |
| "learning_rate": 2.357852462481505e-05, | |
| "loss": 0.5141, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.6169942929613188, | |
| "grad_norm": 0.5895043611526489, | |
| "learning_rate": 2.305009511731135e-05, | |
| "loss": 0.4217, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 1.6487000634115407, | |
| "grad_norm": 0.6137962341308594, | |
| "learning_rate": 2.252166560980765e-05, | |
| "loss": 0.4758, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.6804058338617627, | |
| "grad_norm": 0.686687707901001, | |
| "learning_rate": 2.1993236102303955e-05, | |
| "loss": 0.5715, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 1.7121116043119848, | |
| "grad_norm": 0.7576509118080139, | |
| "learning_rate": 2.1464806594800255e-05, | |
| "loss": 0.494, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.7438173747622068, | |
| "grad_norm": 0.9644522666931152, | |
| "learning_rate": 2.0936377087296556e-05, | |
| "loss": 0.5243, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 1.7755231452124287, | |
| "grad_norm": 1.0673723220825195, | |
| "learning_rate": 2.0407947579792857e-05, | |
| "loss": 0.5254, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.8072289156626506, | |
| "grad_norm": 0.4408644735813141, | |
| "learning_rate": 1.9879518072289157e-05, | |
| "loss": 0.5063, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 1.8389346861128726, | |
| "grad_norm": 0.6671141386032104, | |
| "learning_rate": 1.9351088564785458e-05, | |
| "loss": 0.4384, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.8706404565630945, | |
| "grad_norm": 0.9607345461845398, | |
| "learning_rate": 1.882265905728176e-05, | |
| "loss": 0.6554, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 1.9023462270133165, | |
| "grad_norm": 1.385923981666565, | |
| "learning_rate": 1.8294229549778062e-05, | |
| "loss": 0.499, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.9340519974635384, | |
| "grad_norm": 0.6698076725006104, | |
| "learning_rate": 1.776580004227436e-05, | |
| "loss": 0.4753, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 1.9657577679137603, | |
| "grad_norm": 0.7723073363304138, | |
| "learning_rate": 1.7237370534770663e-05, | |
| "loss": 0.462, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.9974635383639823, | |
| "grad_norm": 0.5628081560134888, | |
| "learning_rate": 1.6708941027266964e-05, | |
| "loss": 0.5225, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 2.029169308814204, | |
| "grad_norm": 1.0150758028030396, | |
| "learning_rate": 1.6180511519763264e-05, | |
| "loss": 0.3888, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 2.060875079264426, | |
| "grad_norm": 0.6272999048233032, | |
| "learning_rate": 1.5652082012259565e-05, | |
| "loss": 0.4211, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 2.092580849714648, | |
| "grad_norm": 0.672383725643158, | |
| "learning_rate": 1.5123652504755867e-05, | |
| "loss": 0.4215, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 2.12428662016487, | |
| "grad_norm": 1.0365363359451294, | |
| "learning_rate": 1.4595222997252168e-05, | |
| "loss": 0.4371, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 2.1559923906150917, | |
| "grad_norm": 1.0515477657318115, | |
| "learning_rate": 1.4066793489748466e-05, | |
| "loss": 0.544, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 2.187698161065314, | |
| "grad_norm": 1.1331499814987183, | |
| "learning_rate": 1.353836398224477e-05, | |
| "loss": 0.5532, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 2.2194039315155356, | |
| "grad_norm": 0.7524999976158142, | |
| "learning_rate": 1.3009934474741071e-05, | |
| "loss": 0.4989, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.251109701965758, | |
| "grad_norm": 1.2541077136993408, | |
| "learning_rate": 1.2481504967237371e-05, | |
| "loss": 0.4576, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 2.2828154724159795, | |
| "grad_norm": 0.7316697835922241, | |
| "learning_rate": 1.1953075459733672e-05, | |
| "loss": 0.4228, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 2.314521242866202, | |
| "grad_norm": 0.680396318435669, | |
| "learning_rate": 1.1424645952229973e-05, | |
| "loss": 0.5042, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 2.3462270133164234, | |
| "grad_norm": 0.5581865310668945, | |
| "learning_rate": 1.0896216444726275e-05, | |
| "loss": 0.4988, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 2.3779327837666457, | |
| "grad_norm": 0.7080293893814087, | |
| "learning_rate": 1.0367786937222575e-05, | |
| "loss": 0.4647, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 2.4096385542168672, | |
| "grad_norm": 1.065440058708191, | |
| "learning_rate": 9.839357429718876e-06, | |
| "loss": 0.4572, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 2.4413443246670896, | |
| "grad_norm": 0.6128831505775452, | |
| "learning_rate": 9.310927922215176e-06, | |
| "loss": 0.4352, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 2.473050095117311, | |
| "grad_norm": 0.6767387390136719, | |
| "learning_rate": 8.782498414711479e-06, | |
| "loss": 0.4352, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 2.5047558655675335, | |
| "grad_norm": 0.56428062915802, | |
| "learning_rate": 8.254068907207779e-06, | |
| "loss": 0.5202, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 2.536461636017755, | |
| "grad_norm": 0.9430966973304749, | |
| "learning_rate": 7.72563939970408e-06, | |
| "loss": 0.5099, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.5681674064679774, | |
| "grad_norm": 1.0679271221160889, | |
| "learning_rate": 7.197209892200381e-06, | |
| "loss": 0.5605, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 2.599873176918199, | |
| "grad_norm": 0.5181096792221069, | |
| "learning_rate": 6.6687803846966825e-06, | |
| "loss": 0.4438, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "grad_norm": 0.6481731534004211, | |
| "learning_rate": 6.140350877192982e-06, | |
| "loss": 0.5255, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 2.6632847178186427, | |
| "grad_norm": 0.6905925869941711, | |
| "learning_rate": 5.6119213696892836e-06, | |
| "loss": 0.5113, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 2.694990488268865, | |
| "grad_norm": 0.7039913535118103, | |
| "learning_rate": 5.083491862185585e-06, | |
| "loss": 0.5378, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 2.7266962587190866, | |
| "grad_norm": 0.6583430171012878, | |
| "learning_rate": 4.5550623546818855e-06, | |
| "loss": 0.4796, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 2.758402029169309, | |
| "grad_norm": 1.0837993621826172, | |
| "learning_rate": 4.026632847178187e-06, | |
| "loss": 0.4655, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 2.7901077996195305, | |
| "grad_norm": 0.7365415692329407, | |
| "learning_rate": 3.498203339674488e-06, | |
| "loss": 0.3938, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 2.821813570069753, | |
| "grad_norm": 0.5105493664741516, | |
| "learning_rate": 2.9697738321707884e-06, | |
| "loss": 0.4043, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 2.853519340519975, | |
| "grad_norm": 0.9684469103813171, | |
| "learning_rate": 2.4413443246670898e-06, | |
| "loss": 0.4907, | |
| "step": 4500 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 4731, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4702733991936000.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |