| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.7874015748031497, |
| "eval_steps": 50, |
| "global_step": 400, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "entropy": 1.437505340576172, |
| "epoch": 0.01968503937007874, |
| "grad_norm": 0.63671875, |
| "learning_rate": 3.91304347826087e-05, |
| "loss": 1.657536506652832, |
| "mean_token_accuracy": 0.6597658038139343, |
| "num_tokens": 115439.0, |
| "step": 10 |
| }, |
| { |
| "entropy": 1.5124324679374694, |
| "epoch": 0.03937007874015748, |
| "grad_norm": 0.4453125, |
| "learning_rate": 8.260869565217392e-05, |
| "loss": 1.4648273468017579, |
| "mean_token_accuracy": 0.683241218328476, |
| "num_tokens": 225264.0, |
| "step": 20 |
| }, |
| { |
| "entropy": 1.2266616582870484, |
| "epoch": 0.05905511811023622, |
| "grad_norm": 0.6796875, |
| "learning_rate": 0.00012608695652173915, |
| "loss": 1.113858699798584, |
| "mean_token_accuracy": 0.7449506461620331, |
| "num_tokens": 330923.0, |
| "step": 30 |
| }, |
| { |
| "entropy": 1.05804543197155, |
| "epoch": 0.07874015748031496, |
| "grad_norm": 0.47265625, |
| "learning_rate": 0.00016956521739130436, |
| "loss": 1.076363468170166, |
| "mean_token_accuracy": 0.7455369174480438, |
| "num_tokens": 437622.0, |
| "step": 40 |
| }, |
| { |
| "entropy": 0.9502384155988693, |
| "epoch": 0.0984251968503937, |
| "grad_norm": 0.5, |
| "learning_rate": 0.00019999796688511753, |
| "loss": 0.9587571144104003, |
| "mean_token_accuracy": 0.7683711260557174, |
| "num_tokens": 551175.0, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.0984251968503937, |
| "eval_entropy": 1.0894114132280703, |
| "eval_loss": 1.0480613708496094, |
| "eval_mean_token_accuracy": 0.745736007337217, |
| "eval_num_tokens": 551175.0, |
| "eval_runtime": 15.9708, |
| "eval_samples_per_second": 26.736, |
| "eval_steps_per_second": 3.381, |
| "step": 50 |
| }, |
| { |
| "entropy": 0.9499408066272735, |
| "epoch": 0.11811023622047244, |
| "grad_norm": 0.5859375, |
| "learning_rate": 0.0001999618249203163, |
| "loss": 0.9569917678833008, |
| "mean_token_accuracy": 0.7686831414699554, |
| "num_tokens": 658006.0, |
| "step": 60 |
| }, |
| { |
| "entropy": 1.0497896373271942, |
| "epoch": 0.1377952755905512, |
| "grad_norm": 0.52734375, |
| "learning_rate": 0.00019988052141977201, |
| "loss": 0.9973426818847656, |
| "mean_token_accuracy": 0.7564341962337494, |
| "num_tokens": 771278.0, |
| "step": 70 |
| }, |
| { |
| "entropy": 0.8503248453140259, |
| "epoch": 0.15748031496062992, |
| "grad_norm": 0.5, |
| "learning_rate": 0.0001997540931154166, |
| "loss": 0.839632225036621, |
| "mean_token_accuracy": 0.7869657546281814, |
| "num_tokens": 886071.0, |
| "step": 80 |
| }, |
| { |
| "entropy": 0.9232195496559144, |
| "epoch": 0.17716535433070865, |
| "grad_norm": 0.482421875, |
| "learning_rate": 0.00019958259712601903, |
| "loss": 0.8961009025573731, |
| "mean_token_accuracy": 0.7736886650323868, |
| "num_tokens": 996361.0, |
| "step": 90 |
| }, |
| { |
| "entropy": 0.8409020006656647, |
| "epoch": 0.1968503937007874, |
| "grad_norm": 0.515625, |
| "learning_rate": 0.00019936611093137983, |
| "loss": 0.839316463470459, |
| "mean_token_accuracy": 0.787284591794014, |
| "num_tokens": 1103999.0, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.1968503937007874, |
| "eval_entropy": 0.988924115896225, |
| "eval_loss": 0.9491098523139954, |
| "eval_mean_token_accuracy": 0.7609603978969433, |
| "eval_num_tokens": 1103999.0, |
| "eval_runtime": 15.8986, |
| "eval_samples_per_second": 26.858, |
| "eval_steps_per_second": 3.397, |
| "step": 100 |
| }, |
| { |
| "entropy": 0.8768177807331086, |
| "epoch": 0.21653543307086615, |
| "grad_norm": 0.41015625, |
| "learning_rate": 0.00019910473233732668, |
| "loss": 0.8348051071166992, |
| "mean_token_accuracy": 0.7883853554725647, |
| "num_tokens": 1216586.0, |
| "step": 110 |
| }, |
| { |
| "entropy": 0.8739962369203568, |
| "epoch": 0.23622047244094488, |
| "grad_norm": 0.59375, |
| "learning_rate": 0.00019879857943152674, |
| "loss": 0.8352558135986328, |
| "mean_token_accuracy": 0.7789304822683334, |
| "num_tokens": 1324767.0, |
| "step": 120 |
| }, |
| { |
| "entropy": 0.8777758985757828, |
| "epoch": 0.2559055118110236, |
| "grad_norm": 0.54296875, |
| "learning_rate": 0.00019844779053013637, |
| "loss": 0.8331022262573242, |
| "mean_token_accuracy": 0.7784219890832901, |
| "num_tokens": 1432356.0, |
| "step": 130 |
| }, |
| { |
| "entropy": 0.7877890735864639, |
| "epoch": 0.2755905511811024, |
| "grad_norm": 0.55078125, |
| "learning_rate": 0.00019805252411531126, |
| "loss": 0.7960677146911621, |
| "mean_token_accuracy": 0.7926767230033874, |
| "num_tokens": 1542545.0, |
| "step": 140 |
| }, |
| { |
| "entropy": 0.8658760845661163, |
| "epoch": 0.2952755905511811, |
| "grad_norm": 0.5625, |
| "learning_rate": 0.0001976129587636064, |
| "loss": 0.8095477104187012, |
| "mean_token_accuracy": 0.7904855847358704, |
| "num_tokens": 1654462.0, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.2952755905511811, |
| "eval_entropy": 0.9384135085123556, |
| "eval_loss": 0.9150573015213013, |
| "eval_mean_token_accuracy": 0.7675120333830515, |
| "eval_num_tokens": 1654462.0, |
| "eval_runtime": 15.91, |
| "eval_samples_per_second": 26.839, |
| "eval_steps_per_second": 3.394, |
| "step": 150 |
| }, |
| { |
| "entropy": 0.771521957218647, |
| "epoch": 0.31496062992125984, |
| "grad_norm": 0.470703125, |
| "learning_rate": 0.00019712929306529723, |
| "loss": 0.7662525177001953, |
| "mean_token_accuracy": 0.8057839542627334, |
| "num_tokens": 1760181.0, |
| "step": 160 |
| }, |
| { |
| "entropy": 0.8810683578252793, |
| "epoch": 0.3346456692913386, |
| "grad_norm": 0.625, |
| "learning_rate": 0.000196601745534659, |
| "loss": 0.8402512550354004, |
| "mean_token_accuracy": 0.7819562494754791, |
| "num_tokens": 1864683.0, |
| "step": 170 |
| }, |
| { |
| "entropy": 0.8320274949073792, |
| "epoch": 0.3543307086614173, |
| "grad_norm": 0.474609375, |
| "learning_rate": 0.0001960305545112445, |
| "loss": 0.8194713592529297, |
| "mean_token_accuracy": 0.7880559414625168, |
| "num_tokens": 1978641.0, |
| "step": 180 |
| }, |
| { |
| "entropy": 0.8050856739282608, |
| "epoch": 0.37401574803149606, |
| "grad_norm": 0.4921875, |
| "learning_rate": 0.0001954159780522056, |
| "loss": 0.8002090454101562, |
| "mean_token_accuracy": 0.791279673576355, |
| "num_tokens": 2087989.0, |
| "step": 190 |
| }, |
| { |
| "entropy": 0.8930813580751419, |
| "epoch": 0.3937007874015748, |
| "grad_norm": 0.6171875, |
| "learning_rate": 0.00019475829381570566, |
| "loss": 0.8300093650817871, |
| "mean_token_accuracy": 0.7832900315523148, |
| "num_tokens": 2195494.0, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.3937007874015748, |
| "eval_entropy": 0.9506367754053187, |
| "eval_loss": 0.8978024125099182, |
| "eval_mean_token_accuracy": 0.7694783839914534, |
| "eval_num_tokens": 2195494.0, |
| "eval_runtime": 15.9025, |
| "eval_samples_per_second": 26.851, |
| "eval_steps_per_second": 3.396, |
| "step": 200 |
| }, |
| { |
| "entropy": 0.7895667225122451, |
| "epoch": 0.41338582677165353, |
| "grad_norm": 0.41015625, |
| "learning_rate": 0.0001940577989354773, |
| "loss": 0.7946149826049804, |
| "mean_token_accuracy": 0.7988882720470428, |
| "num_tokens": 2314228.0, |
| "step": 210 |
| }, |
| { |
| "entropy": 0.8575397372245789, |
| "epoch": 0.4330708661417323, |
| "grad_norm": 0.6640625, |
| "learning_rate": 0.00019331480988658083, |
| "loss": 0.812718391418457, |
| "mean_token_accuracy": 0.7836957842111587, |
| "num_tokens": 2419122.0, |
| "step": 220 |
| }, |
| { |
| "entropy": 0.8544128388166428, |
| "epoch": 0.452755905511811, |
| "grad_norm": 0.462890625, |
| "learning_rate": 0.00019252966234242468, |
| "loss": 0.8198953628540039, |
| "mean_token_accuracy": 0.7827243804931641, |
| "num_tokens": 2524623.0, |
| "step": 230 |
| }, |
| { |
| "entropy": 0.9001998215913772, |
| "epoch": 0.47244094488188976, |
| "grad_norm": 0.59375, |
| "learning_rate": 0.00019170271102311238, |
| "loss": 0.9015755653381348, |
| "mean_token_accuracy": 0.7709000498056412, |
| "num_tokens": 2631031.0, |
| "step": 240 |
| }, |
| { |
| "entropy": 0.7926925078034401, |
| "epoch": 0.4921259842519685, |
| "grad_norm": 0.49609375, |
| "learning_rate": 0.00019083432953518418, |
| "loss": 0.8107254028320312, |
| "mean_token_accuracy": 0.7990583658218384, |
| "num_tokens": 2741031.0, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.4921259842519685, |
| "eval_entropy": 0.9023421671655443, |
| "eval_loss": 0.884621262550354, |
| "eval_mean_token_accuracy": 0.7719593346118927, |
| "eval_num_tokens": 2741031.0, |
| "eval_runtime": 15.8929, |
| "eval_samples_per_second": 26.867, |
| "eval_steps_per_second": 3.398, |
| "step": 250 |
| }, |
| { |
| "entropy": 0.7738933145999909, |
| "epoch": 0.5118110236220472, |
| "grad_norm": 0.58203125, |
| "learning_rate": 0.00018992491020282618, |
| "loss": 0.7572973251342774, |
| "mean_token_accuracy": 0.7949966192245483, |
| "num_tokens": 2853547.0, |
| "step": 260 |
| }, |
| { |
| "entropy": 0.8569424450397491, |
| "epoch": 0.531496062992126, |
| "grad_norm": 1.375, |
| "learning_rate": 0.0001889748638906231, |
| "loss": 0.8083616256713867, |
| "mean_token_accuracy": 0.791676989197731, |
| "num_tokens": 2968367.0, |
| "step": 270 |
| }, |
| { |
| "entropy": 0.889449554681778, |
| "epoch": 0.5511811023622047, |
| "grad_norm": 0.44921875, |
| "learning_rate": 0.00018798461981793462, |
| "loss": 0.8472497940063477, |
| "mean_token_accuracy": 0.7779456943273544, |
| "num_tokens": 3075963.0, |
| "step": 280 |
| }, |
| { |
| "entropy": 0.8320091620087624, |
| "epoch": 0.5708661417322834, |
| "grad_norm": 0.453125, |
| "learning_rate": 0.00018695462536497934, |
| "loss": 0.8240900993347168, |
| "mean_token_accuracy": 0.7846117675304413, |
| "num_tokens": 3185899.0, |
| "step": 290 |
| }, |
| { |
| "entropy": 0.8333972141146659, |
| "epoch": 0.5905511811023622, |
| "grad_norm": 0.486328125, |
| "learning_rate": 0.00018588534587071385, |
| "loss": 0.8117186546325683, |
| "mean_token_accuracy": 0.7908233135938645, |
| "num_tokens": 3297319.0, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.5905511811023622, |
| "eval_entropy": 0.8452967874429844, |
| "eval_loss": 0.8709425926208496, |
| "eval_mean_token_accuracy": 0.7752477537702631, |
| "eval_num_tokens": 3297319.0, |
| "eval_runtime": 15.8962, |
| "eval_samples_per_second": 26.862, |
| "eval_steps_per_second": 3.397, |
| "step": 300 |
| }, |
| { |
| "entropy": 0.8324314504861832, |
| "epoch": 0.610236220472441, |
| "grad_norm": 0.39453125, |
| "learning_rate": 0.00018477726442259806, |
| "loss": 0.801210880279541, |
| "mean_token_accuracy": 0.7885541886091232, |
| "num_tokens": 3415477.0, |
| "step": 310 |
| }, |
| { |
| "entropy": 0.8624329552054405, |
| "epoch": 0.6299212598425197, |
| "grad_norm": 0.400390625, |
| "learning_rate": 0.00018363088163834252, |
| "loss": 0.8126995086669921, |
| "mean_token_accuracy": 0.785911500453949, |
| "num_tokens": 3524759.0, |
| "step": 320 |
| }, |
| { |
| "entropy": 0.8507865130901336, |
| "epoch": 0.6496062992125984, |
| "grad_norm": 0.458984375, |
| "learning_rate": 0.00018244671543973513, |
| "loss": 0.8206624031066895, |
| "mean_token_accuracy": 0.7833890795707703, |
| "num_tokens": 3640838.0, |
| "step": 330 |
| }, |
| { |
| "entropy": 0.7785286724567413, |
| "epoch": 0.6692913385826772, |
| "grad_norm": 0.51953125, |
| "learning_rate": 0.0001812253008186505, |
| "loss": 0.753394889831543, |
| "mean_token_accuracy": 0.8014622688293457, |
| "num_tokens": 3756735.0, |
| "step": 340 |
| }, |
| { |
| "entropy": 0.8389557063579559, |
| "epoch": 0.6889763779527559, |
| "grad_norm": 0.50390625, |
| "learning_rate": 0.0001799671895953471, |
| "loss": 0.8116916656494141, |
| "mean_token_accuracy": 0.7909021496772766, |
| "num_tokens": 3870377.0, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.6889763779527559, |
| "eval_entropy": 0.8793082656683745, |
| "eval_loss": 0.8570566177368164, |
| "eval_mean_token_accuracy": 0.7760916120476193, |
| "eval_num_tokens": 3870377.0, |
| "eval_runtime": 15.9217, |
| "eval_samples_per_second": 26.819, |
| "eval_steps_per_second": 3.392, |
| "step": 350 |
| }, |
| { |
| "entropy": 0.7998805820941925, |
| "epoch": 0.7086614173228346, |
| "grad_norm": 0.494140625, |
| "learning_rate": 0.00017867295016916167, |
| "loss": 0.7578703880310058, |
| "mean_token_accuracy": 0.7953435629606247, |
| "num_tokens": 3975692.0, |
| "step": 360 |
| }, |
| { |
| "entropy": 0.9054980486631393, |
| "epoch": 0.7283464566929134, |
| "grad_norm": 0.4765625, |
| "learning_rate": 0.0001773431672617132, |
| "loss": 0.8652525901794433, |
| "mean_token_accuracy": 0.7738145262002945, |
| "num_tokens": 4087256.0, |
| "step": 370 |
| }, |
| { |
| "entropy": 0.7607795268297195, |
| "epoch": 0.7480314960629921, |
| "grad_norm": 0.498046875, |
| "learning_rate": 0.00017597844165273286, |
| "loss": 0.7365350723266602, |
| "mean_token_accuracy": 0.7966376423835755, |
| "num_tokens": 4192629.0, |
| "step": 380 |
| }, |
| { |
| "entropy": 0.8066679537296295, |
| "epoch": 0.7677165354330708, |
| "grad_norm": 0.443359375, |
| "learning_rate": 0.00017457938990863908, |
| "loss": 0.7610409736633301, |
| "mean_token_accuracy": 0.8000116467475891, |
| "num_tokens": 4303838.0, |
| "step": 390 |
| }, |
| { |
| "entropy": 0.8927708208560944, |
| "epoch": 0.7874015748031497, |
| "grad_norm": 0.494140625, |
| "learning_rate": 0.00017314664410398033, |
| "loss": 0.8361794471740722, |
| "mean_token_accuracy": 0.7784365385770797, |
| "num_tokens": 4414002.0, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.7874015748031497, |
| "eval_entropy": 0.8607326823252218, |
| "eval_loss": 0.8459087610244751, |
| "eval_mean_token_accuracy": 0.7791097947844753, |
| "eval_num_tokens": 4414002.0, |
| "eval_runtime": 16.176, |
| "eval_samples_per_second": 26.397, |
| "eval_steps_per_second": 3.338, |
| "step": 400 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 1524, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 50, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 6.197880255742771e+16, |
| "train_batch_size": 8, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
