| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 3.0, |
| "eval_steps": 500, |
| "global_step": 1245, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.060350030175015085, |
| "grad_norm": 0.2716701030731201, |
| "learning_rate": 0.00013963636363636362, |
| "loss": 1.7977, |
| "mean_token_accuracy": 0.6328366297483444, |
| "num_tokens": 156326.0, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.12070006035003017, |
| "grad_norm": 0.2526017725467682, |
| "learning_rate": 0.00028509090909090905, |
| "loss": 0.8743, |
| "mean_token_accuracy": 0.7787316054105758, |
| "num_tokens": 282083.0, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.18105009052504525, |
| "grad_norm": 0.19058053195476532, |
| "learning_rate": 0.00043054545454545456, |
| "loss": 0.6236, |
| "mean_token_accuracy": 0.8299106657505035, |
| "num_tokens": 439671.0, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.24140012070006034, |
| "grad_norm": 0.21313656866550446, |
| "learning_rate": 0.000511980737353935, |
| "loss": 0.496, |
| "mean_token_accuracy": 0.8596594697237014, |
| "num_tokens": 567445.0, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.30175015087507545, |
| "grad_norm": 0.19823844730854034, |
| "learning_rate": 0.0005117937078575993, |
| "loss": 0.3458, |
| "mean_token_accuracy": 0.9001932269334794, |
| "num_tokens": 722851.0, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.3621001810500905, |
| "grad_norm": 0.2446502149105072, |
| "learning_rate": 0.0005114078548252564, |
| "loss": 0.3064, |
| "mean_token_accuracy": 0.911021100282669, |
| "num_tokens": 849117.0, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.4224502112251056, |
| "grad_norm": 0.1619698852300644, |
| "learning_rate": 0.0005108234781730571, |
| "loss": 0.2402, |
| "mean_token_accuracy": 0.930964640378952, |
| "num_tokens": 1007875.0, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.4828002414001207, |
| "grad_norm": 0.27491629123687744, |
| "learning_rate": 0.0005100410321257489, |
| "loss": 0.2161, |
| "mean_token_accuracy": 0.938733617067337, |
| "num_tokens": 1135042.0, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.5431502715751357, |
| "grad_norm": 0.18663127720355988, |
| "learning_rate": 0.0005090611248636156, |
| "loss": 0.1748, |
| "mean_token_accuracy": 0.9507552969455719, |
| "num_tokens": 1294935.0, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.6035003017501509, |
| "grad_norm": 0.23655210435390472, |
| "learning_rate": 0.0005078845180497507, |
| "loss": 0.1534, |
| "mean_token_accuracy": 0.9558097630739212, |
| "num_tokens": 1421875.0, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.663850331925166, |
| "grad_norm": 0.1793680489063263, |
| "learning_rate": 0.0005065121262380311, |
| "loss": 0.1339, |
| "mean_token_accuracy": 0.9624842357635498, |
| "num_tokens": 1582177.0, |
| "step": 275 |
| }, |
| { |
| "epoch": 0.724200362100181, |
| "grad_norm": 0.21529488265514374, |
| "learning_rate": 0.0005049450161622521, |
| "loss": 0.1082, |
| "mean_token_accuracy": 0.9684510946273803, |
| "num_tokens": 1709697.0, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.7845503922751962, |
| "grad_norm": 0.10737641900777817, |
| "learning_rate": 0.0005031844059069758, |
| "loss": 0.1139, |
| "mean_token_accuracy": 0.9678166025876999, |
| "num_tokens": 1866595.0, |
| "step": 325 |
| }, |
| { |
| "epoch": 0.8449004224502112, |
| "grad_norm": 0.25487855076789856, |
| "learning_rate": 0.0005012316639607367, |
| "loss": 0.1061, |
| "mean_token_accuracy": 0.9704786598682403, |
| "num_tokens": 1992513.0, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.9052504526252263, |
| "grad_norm": 0.11293583363294601, |
| "learning_rate": 0.000499088308152343, |
| "loss": 0.1044, |
| "mean_token_accuracy": 0.9708359962701798, |
| "num_tokens": 2148616.0, |
| "step": 375 |
| }, |
| { |
| "epoch": 0.9656004828002414, |
| "grad_norm": 0.24543644487857819, |
| "learning_rate": 0.0004967560044710963, |
| "loss": 0.0913, |
| "mean_token_accuracy": 0.9756689429283142, |
| "num_tokens": 2275044.0, |
| "step": 400 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_loss": 0.08796942234039307, |
| "eval_mean_token_accuracy": 0.9756118339461249, |
| "eval_num_tokens": 2354180.0, |
| "eval_runtime": 15.8439, |
| "eval_samples_per_second": 23.29, |
| "eval_steps_per_second": 11.676, |
| "step": 415 |
| }, |
| { |
| "epoch": 1.024140012070006, |
| "grad_norm": 0.11087150126695633, |
| "learning_rate": 0.0004942365657718508, |
| "loss": 0.086, |
| "mean_token_accuracy": 0.9767568179012573, |
| "num_tokens": 2421983.0, |
| "step": 425 |
| }, |
| { |
| "epoch": 1.0844900422450212, |
| "grad_norm": 0.13107234239578247, |
| "learning_rate": 0.0004915319503659158, |
| "loss": 0.0582, |
| "mean_token_accuracy": 0.9829362910985947, |
| "num_tokens": 2563714.0, |
| "step": 450 |
| }, |
| { |
| "epoch": 1.1448400724200363, |
| "grad_norm": 0.08740212023258209, |
| "learning_rate": 0.0004886442604988977, |
| "loss": 0.0729, |
| "mean_token_accuracy": 0.978913226723671, |
| "num_tokens": 2705420.0, |
| "step": 475 |
| }, |
| { |
| "epoch": 1.2051901025950513, |
| "grad_norm": 0.106731116771698, |
| "learning_rate": 0.0004855757407166656, |
| "loss": 0.0571, |
| "mean_token_accuracy": 0.9836699557304383, |
| "num_tokens": 2846774.0, |
| "step": 500 |
| }, |
| { |
| "epoch": 1.2655401327700664, |
| "grad_norm": 0.08578667789697647, |
| "learning_rate": 0.00048232877612070814, |
| "loss": 0.0772, |
| "mean_token_accuracy": 0.9792935186624527, |
| "num_tokens": 2988445.0, |
| "step": 525 |
| }, |
| { |
| "epoch": 1.3258901629450814, |
| "grad_norm": 0.07904712110757828, |
| "learning_rate": 0.0004789058905142402, |
| "loss": 0.0509, |
| "mean_token_accuracy": 0.985254094004631, |
| "num_tokens": 3131701.0, |
| "step": 550 |
| }, |
| { |
| "epoch": 1.3862401931200965, |
| "grad_norm": 0.08005515486001968, |
| "learning_rate": 0.0004753097444404994, |
| "loss": 0.0672, |
| "mean_token_accuracy": 0.9816042059659957, |
| "num_tokens": 3273618.0, |
| "step": 575 |
| }, |
| { |
| "epoch": 1.4465902232951118, |
| "grad_norm": 0.11021016538143158, |
| "learning_rate": 0.0004715431331147565, |
| "loss": 0.0522, |
| "mean_token_accuracy": 0.9851613295078278, |
| "num_tokens": 3417492.0, |
| "step": 600 |
| }, |
| { |
| "epoch": 1.5069402534701268, |
| "grad_norm": 0.12960106134414673, |
| "learning_rate": 0.0004676089842516484, |
| "loss": 0.0725, |
| "mean_token_accuracy": 0.9796771186590195, |
| "num_tokens": 3560878.0, |
| "step": 625 |
| }, |
| { |
| "epoch": 1.567290283645142, |
| "grad_norm": 0.14702805876731873, |
| "learning_rate": 0.00046351035578952207, |
| "loss": 0.0469, |
| "mean_token_accuracy": 0.986492536664009, |
| "num_tokens": 3704626.0, |
| "step": 650 |
| }, |
| { |
| "epoch": 1.627640313820157, |
| "grad_norm": 0.07580998539924622, |
| "learning_rate": 0.0004592504335135573, |
| "loss": 0.0634, |
| "mean_token_accuracy": 0.9821663647890091, |
| "num_tokens": 3847508.0, |
| "step": 675 |
| }, |
| { |
| "epoch": 1.687990343995172, |
| "grad_norm": 0.10504812747240067, |
| "learning_rate": 0.0004548325285795178, |
| "loss": 0.0478, |
| "mean_token_accuracy": 0.9862841880321502, |
| "num_tokens": 3989632.0, |
| "step": 700 |
| }, |
| { |
| "epoch": 1.748340374170187, |
| "grad_norm": 0.09579284489154816, |
| "learning_rate": 0.00045026007494005255, |
| "loss": 0.0627, |
| "mean_token_accuracy": 0.9824772262573243, |
| "num_tokens": 4130651.0, |
| "step": 725 |
| }, |
| { |
| "epoch": 1.8086904043452021, |
| "grad_norm": 0.09132307767868042, |
| "learning_rate": 0.0004455366266755509, |
| "loss": 0.0453, |
| "mean_token_accuracy": 0.9872566330432891, |
| "num_tokens": 4272783.0, |
| "step": 750 |
| }, |
| { |
| "epoch": 1.8690404345202172, |
| "grad_norm": 0.11937957257032394, |
| "learning_rate": 0.00044066585523162415, |
| "loss": 0.0623, |
| "mean_token_accuracy": 0.9824487656354904, |
| "num_tokens": 4413897.0, |
| "step": 775 |
| }, |
| { |
| "epoch": 1.9293904646952322, |
| "grad_norm": 0.10511502623558044, |
| "learning_rate": 0.0004356515465653613, |
| "loss": 0.0445, |
| "mean_token_accuracy": 0.98703648686409, |
| "num_tokens": 4556054.0, |
| "step": 800 |
| }, |
| { |
| "epoch": 1.9897404948702473, |
| "grad_norm": 0.07509177178144455, |
| "learning_rate": 0.000430497598202577, |
| "loss": 0.0534, |
| "mean_token_accuracy": 0.9852619272470474, |
| "num_tokens": 4688140.0, |
| "step": 825 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_loss": 0.05571499839425087, |
| "eval_mean_token_accuracy": 0.9846324215064177, |
| "eval_num_tokens": 4708360.0, |
| "eval_runtime": 15.8265, |
| "eval_samples_per_second": 23.315, |
| "eval_steps_per_second": 11.689, |
| "step": 830 |
| }, |
| { |
| "epoch": 2.048280024140012, |
| "grad_norm": 0.10296711325645447, |
| "learning_rate": 0.0004252080162083404, |
| "loss": 0.0488, |
| "mean_token_accuracy": 0.9856051372498581, |
| "num_tokens": 4834012.0, |
| "step": 850 |
| }, |
| { |
| "epoch": 2.1086300543150274, |
| "grad_norm": 0.13229815661907196, |
| "learning_rate": 0.0004197869120731367, |
| "loss": 0.0341, |
| "mean_token_accuracy": 0.9902466797828674, |
| "num_tokens": 4965876.0, |
| "step": 875 |
| }, |
| { |
| "epoch": 2.1689800844900424, |
| "grad_norm": 0.10297457873821259, |
| "learning_rate": 0.0004142384995170855, |
| "loss": 0.0452, |
| "mean_token_accuracy": 0.986401349902153, |
| "num_tokens": 5119378.0, |
| "step": 900 |
| }, |
| { |
| "epoch": 2.2293301146650575, |
| "grad_norm": 0.04365512356162071, |
| "learning_rate": 0.0004085670912146964, |
| "loss": 0.031, |
| "mean_token_accuracy": 0.9906719970703125, |
| "num_tokens": 5252167.0, |
| "step": 925 |
| }, |
| { |
| "epoch": 2.2896801448400725, |
| "grad_norm": 0.04382198303937912, |
| "learning_rate": 0.0004027770954427105, |
| "loss": 0.0487, |
| "mean_token_accuracy": 0.9857749968767167, |
| "num_tokens": 5403450.0, |
| "step": 950 |
| }, |
| { |
| "epoch": 2.3500301750150876, |
| "grad_norm": 0.10280822962522507, |
| "learning_rate": 0.00039687301265363086, |
| "loss": 0.033, |
| "mean_token_accuracy": 0.99033636033535, |
| "num_tokens": 5535695.0, |
| "step": 975 |
| }, |
| { |
| "epoch": 2.4103802051901027, |
| "grad_norm": 0.04137711599469185, |
| "learning_rate": 0.00039085943197760806, |
| "loss": 0.0446, |
| "mean_token_accuracy": 0.9868948411941528, |
| "num_tokens": 5686536.0, |
| "step": 1000 |
| }, |
| { |
| "epoch": 2.4707302353651177, |
| "grad_norm": 0.04095058888196945, |
| "learning_rate": 0.00038474102765539666, |
| "loss": 0.0324, |
| "mean_token_accuracy": 0.990428722500801, |
| "num_tokens": 5819478.0, |
| "step": 1025 |
| }, |
| { |
| "epoch": 2.5310802655401328, |
| "grad_norm": 0.06750275939702988, |
| "learning_rate": 0.0003785225554051583, |
| "loss": 0.0484, |
| "mean_token_accuracy": 0.9855219483375549, |
| "num_tokens": 5973334.0, |
| "step": 1050 |
| }, |
| { |
| "epoch": 2.591430295715148, |
| "grad_norm": 0.03156557306647301, |
| "learning_rate": 0.0003722088487259329, |
| "loss": 0.0301, |
| "mean_token_accuracy": 0.9906777936220169, |
| "num_tokens": 6106945.0, |
| "step": 1075 |
| }, |
| { |
| "epoch": 2.651780325890163, |
| "grad_norm": 0.0798354223370552, |
| "learning_rate": 0.00036580481514065285, |
| "loss": 0.041, |
| "mean_token_accuracy": 0.9879147779941558, |
| "num_tokens": 6258955.0, |
| "step": 1100 |
| }, |
| { |
| "epoch": 2.712130356065178, |
| "grad_norm": 0.0534820593893528, |
| "learning_rate": 0.00035931543238161963, |
| "loss": 0.031, |
| "mean_token_accuracy": 0.9908862364292145, |
| "num_tokens": 6391949.0, |
| "step": 1125 |
| }, |
| { |
| "epoch": 2.772480386240193, |
| "grad_norm": 0.0678093433380127, |
| "learning_rate": 0.0003527457445214084, |
| "loss": 0.0416, |
| "mean_token_accuracy": 0.9876690626144409, |
| "num_tokens": 6542859.0, |
| "step": 1150 |
| }, |
| { |
| "epoch": 2.832830416415208, |
| "grad_norm": 0.058269429951906204, |
| "learning_rate": 0.00034610085805220666, |
| "loss": 0.0288, |
| "mean_token_accuracy": 0.9910433840751648, |
| "num_tokens": 6676094.0, |
| "step": 1175 |
| }, |
| { |
| "epoch": 2.8931804465902236, |
| "grad_norm": 0.044238071888685226, |
| "learning_rate": 0.00033938593791663655, |
| "loss": 0.041, |
| "mean_token_accuracy": 0.9880746394395828, |
| "num_tokens": 6826469.0, |
| "step": 1200 |
| }, |
| { |
| "epoch": 2.9535304767652386, |
| "grad_norm": 0.02994498796761036, |
| "learning_rate": 0.0003326062034931439, |
| "loss": 0.0276, |
| "mean_token_accuracy": 0.9915395766496659, |
| "num_tokens": 6958335.0, |
| "step": 1225 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_loss": 0.04431217536330223, |
| "eval_mean_token_accuracy": 0.9878581050279979, |
| "eval_num_tokens": 7062540.0, |
| "eval_runtime": 15.7984, |
| "eval_samples_per_second": 23.357, |
| "eval_steps_per_second": 11.71, |
| "step": 1245 |
| } |
| ], |
| "logging_steps": 25, |
| "max_steps": 2905, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 7, |
| "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": 3.068266812188928e+17, |
| "train_batch_size": 2, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
