| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.6811409110259685, |
| "eval_steps": 500, |
| "global_step": 200, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.017028522775649212, |
| "grad_norm": 6.522716999053955, |
| "learning_rate": 0.0001988623435722412, |
| "loss": 2.7211, |
| "mean_token_accuracy": 0.5789041496813297, |
| "num_tokens": 4727.0, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.034057045551298425, |
| "grad_norm": 1.0070050954818726, |
| "learning_rate": 0.00019772468714448238, |
| "loss": 0.9998, |
| "mean_token_accuracy": 0.8499793156981468, |
| "num_tokens": 9471.0, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.05108556832694764, |
| "grad_norm": 0.7373425960540771, |
| "learning_rate": 0.00019658703071672356, |
| "loss": 0.8303, |
| "mean_token_accuracy": 0.8681916877627373, |
| "num_tokens": 14013.0, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.06811409110259685, |
| "grad_norm": 0.8468236327171326, |
| "learning_rate": 0.00019544937428896475, |
| "loss": 0.8404, |
| "mean_token_accuracy": 0.8622620105743408, |
| "num_tokens": 18725.0, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.08514261387824607, |
| "grad_norm": 0.8178804516792297, |
| "learning_rate": 0.00019431171786120593, |
| "loss": 0.9073, |
| "mean_token_accuracy": 0.8524447843432427, |
| "num_tokens": 23497.0, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.10217113665389528, |
| "grad_norm": 0.890177309513092, |
| "learning_rate": 0.00019317406143344712, |
| "loss": 0.8029, |
| "mean_token_accuracy": 0.8642359614372254, |
| "num_tokens": 28192.0, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.11919965942954448, |
| "grad_norm": 1.0545469522476196, |
| "learning_rate": 0.0001920364050056883, |
| "loss": 0.7458, |
| "mean_token_accuracy": 0.8732351213693619, |
| "num_tokens": 32805.0, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.1362281822051937, |
| "grad_norm": 1.0384492874145508, |
| "learning_rate": 0.00019089874857792946, |
| "loss": 0.9235, |
| "mean_token_accuracy": 0.8408558666706085, |
| "num_tokens": 37714.0, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.1532567049808429, |
| "grad_norm": 0.6860374808311462, |
| "learning_rate": 0.00018976109215017067, |
| "loss": 0.7623, |
| "mean_token_accuracy": 0.8656536340713501, |
| "num_tokens": 42429.0, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.17028522775649213, |
| "grad_norm": 0.7284501791000366, |
| "learning_rate": 0.00018862343572241183, |
| "loss": 0.7588, |
| "mean_token_accuracy": 0.8688850715756417, |
| "num_tokens": 47099.0, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.18731375053214133, |
| "grad_norm": 0.8899862170219421, |
| "learning_rate": 0.00018748577929465302, |
| "loss": 0.7065, |
| "mean_token_accuracy": 0.8717748820781708, |
| "num_tokens": 51795.0, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.20434227330779056, |
| "grad_norm": 0.5350505709648132, |
| "learning_rate": 0.0001863481228668942, |
| "loss": 0.6696, |
| "mean_token_accuracy": 0.8851396456360817, |
| "num_tokens": 56369.0, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.22137079608343976, |
| "grad_norm": 0.6087013483047485, |
| "learning_rate": 0.0001852104664391354, |
| "loss": 0.7204, |
| "mean_token_accuracy": 0.8682332798838616, |
| "num_tokens": 61018.0, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.23839931885908897, |
| "grad_norm": 0.7879019379615784, |
| "learning_rate": 0.00018407281001137657, |
| "loss": 0.7692, |
| "mean_token_accuracy": 0.8653983056545258, |
| "num_tokens": 65741.0, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.2554278416347382, |
| "grad_norm": 0.6546444296836853, |
| "learning_rate": 0.00018293515358361776, |
| "loss": 0.7119, |
| "mean_token_accuracy": 0.8726310178637504, |
| "num_tokens": 70425.0, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.2724563644103874, |
| "grad_norm": 0.770902156829834, |
| "learning_rate": 0.00018179749715585894, |
| "loss": 0.7035, |
| "mean_token_accuracy": 0.8711811751127243, |
| "num_tokens": 75101.0, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.2894848871860366, |
| "grad_norm": 0.8873187899589539, |
| "learning_rate": 0.00018065984072810013, |
| "loss": 0.7419, |
| "mean_token_accuracy": 0.8616803497076034, |
| "num_tokens": 79888.0, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.3065134099616858, |
| "grad_norm": 0.7061370015144348, |
| "learning_rate": 0.0001795221843003413, |
| "loss": 0.66, |
| "mean_token_accuracy": 0.8744896531105042, |
| "num_tokens": 84562.0, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.32354193273733506, |
| "grad_norm": 0.8232033252716064, |
| "learning_rate": 0.0001783845278725825, |
| "loss": 0.7403, |
| "mean_token_accuracy": 0.8659066379070282, |
| "num_tokens": 89238.0, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.34057045551298426, |
| "grad_norm": 0.7313310503959656, |
| "learning_rate": 0.00017724687144482368, |
| "loss": 0.7157, |
| "mean_token_accuracy": 0.8661555901169777, |
| "num_tokens": 93987.0, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.35759897828863346, |
| "grad_norm": 0.7119390964508057, |
| "learning_rate": 0.00017610921501706487, |
| "loss": 0.7016, |
| "mean_token_accuracy": 0.8678551658987999, |
| "num_tokens": 98722.0, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.37462750106428266, |
| "grad_norm": 0.7126006484031677, |
| "learning_rate": 0.00017497155858930602, |
| "loss": 0.7009, |
| "mean_token_accuracy": 0.867291408777237, |
| "num_tokens": 103439.0, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.39165602383993187, |
| "grad_norm": 0.6158230304718018, |
| "learning_rate": 0.00017383390216154724, |
| "loss": 0.7236, |
| "mean_token_accuracy": 0.8636128515005111, |
| "num_tokens": 108209.0, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.4086845466155811, |
| "grad_norm": 0.6498322486877441, |
| "learning_rate": 0.0001726962457337884, |
| "loss": 0.6813, |
| "mean_token_accuracy": 0.8757071048021317, |
| "num_tokens": 112855.0, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.4257130693912303, |
| "grad_norm": 0.6594287157058716, |
| "learning_rate": 0.0001715585893060296, |
| "loss": 0.7062, |
| "mean_token_accuracy": 0.8650311172008515, |
| "num_tokens": 117653.0, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.4427415921668795, |
| "grad_norm": 0.7375237941741943, |
| "learning_rate": 0.00017042093287827076, |
| "loss": 0.6477, |
| "mean_token_accuracy": 0.8733970731496811, |
| "num_tokens": 122397.0, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.45977011494252873, |
| "grad_norm": 0.5958442687988281, |
| "learning_rate": 0.00016928327645051198, |
| "loss": 0.691, |
| "mean_token_accuracy": 0.8681387722492218, |
| "num_tokens": 127136.0, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.47679863771817793, |
| "grad_norm": 0.7407785654067993, |
| "learning_rate": 0.00016814562002275313, |
| "loss": 0.6477, |
| "mean_token_accuracy": 0.8752307429909706, |
| "num_tokens": 131833.0, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.49382716049382713, |
| "grad_norm": 0.8190199732780457, |
| "learning_rate": 0.00016700796359499432, |
| "loss": 0.6608, |
| "mean_token_accuracy": 0.8762851104140281, |
| "num_tokens": 136441.0, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.5108556832694764, |
| "grad_norm": 0.7571837306022644, |
| "learning_rate": 0.0001658703071672355, |
| "loss": 0.673, |
| "mean_token_accuracy": 0.8691665843129158, |
| "num_tokens": 141186.0, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.5278842060451255, |
| "grad_norm": 0.6653426885604858, |
| "learning_rate": 0.0001647326507394767, |
| "loss": 0.7126, |
| "mean_token_accuracy": 0.8647830635309219, |
| "num_tokens": 145971.0, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.5449127288207748, |
| "grad_norm": 0.7333567142486572, |
| "learning_rate": 0.00016359499431171787, |
| "loss": 0.6287, |
| "mean_token_accuracy": 0.878324082493782, |
| "num_tokens": 150669.0, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.561941251596424, |
| "grad_norm": 0.8214460015296936, |
| "learning_rate": 0.00016245733788395906, |
| "loss": 0.6608, |
| "mean_token_accuracy": 0.8729922890663147, |
| "num_tokens": 155374.0, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.5789697743720732, |
| "grad_norm": 0.6251844167709351, |
| "learning_rate": 0.00016131968145620024, |
| "loss": 0.6556, |
| "mean_token_accuracy": 0.8770609870553017, |
| "num_tokens": 160068.0, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.5959982971477225, |
| "grad_norm": 0.804166853427887, |
| "learning_rate": 0.00016018202502844143, |
| "loss": 0.6366, |
| "mean_token_accuracy": 0.8760894432663917, |
| "num_tokens": 164696.0, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.6130268199233716, |
| "grad_norm": 0.7013877034187317, |
| "learning_rate": 0.0001590443686006826, |
| "loss": 0.6328, |
| "mean_token_accuracy": 0.8779815405607223, |
| "num_tokens": 169356.0, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.6300553426990209, |
| "grad_norm": 1.0623070001602173, |
| "learning_rate": 0.0001579067121729238, |
| "loss": 0.6734, |
| "mean_token_accuracy": 0.8710577815771103, |
| "num_tokens": 174041.0, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.6470838654746701, |
| "grad_norm": 0.7858535647392273, |
| "learning_rate": 0.00015676905574516496, |
| "loss": 0.6775, |
| "mean_token_accuracy": 0.8692880481481552, |
| "num_tokens": 178858.0, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.6641123882503193, |
| "grad_norm": 0.8943309187889099, |
| "learning_rate": 0.00015563139931740617, |
| "loss": 0.6874, |
| "mean_token_accuracy": 0.868885512650013, |
| "num_tokens": 183669.0, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.6811409110259685, |
| "grad_norm": 0.5898745656013489, |
| "learning_rate": 0.00015449374288964733, |
| "loss": 0.6907, |
| "mean_token_accuracy": 0.8661940798163414, |
| "num_tokens": 188473.0, |
| "step": 200 |
| } |
| ], |
| "logging_steps": 5, |
| "max_steps": 879, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 200, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 8534284789702656.0, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
