| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 7.0, | |
| "eval_steps": 500, | |
| "global_step": 105, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.3463203463203463, | |
| "grad_norm": 19.692244939100938, | |
| "learning_rate": 7.272727272727273e-06, | |
| "loss": 1.5151, | |
| "mean_token_accuracy": 0.6845420658588409, | |
| "num_tokens": 2984558.0, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6926406926406926, | |
| "grad_norm": 3.220764064970589, | |
| "learning_rate": 1.6363636363636366e-05, | |
| "loss": 0.9695, | |
| "mean_token_accuracy": 0.7747866742312908, | |
| "num_tokens": 5908318.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.6575383904754147, | |
| "learning_rate": 1.994977815088504e-05, | |
| "loss": 0.5114, | |
| "mean_token_accuracy": 0.8788952617578103, | |
| "num_tokens": 8399102.0, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.3463203463203464, | |
| "grad_norm": 0.33840481163574176, | |
| "learning_rate": 1.964469175054377e-05, | |
| "loss": 0.4654, | |
| "mean_token_accuracy": 0.8905634626746177, | |
| "num_tokens": 11312902.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.6926406926406927, | |
| "grad_norm": 0.29844939490798694, | |
| "learning_rate": 1.907090913734341e-05, | |
| "loss": 0.4231, | |
| "mean_token_accuracy": 0.9002685286104679, | |
| "num_tokens": 14259671.0, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.3013905351728782, | |
| "learning_rate": 1.8244415603417603e-05, | |
| "loss": 0.4446, | |
| "mean_token_accuracy": 0.8957275404057032, | |
| "num_tokens": 16800284.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.346320346320346, | |
| "grad_norm": 0.2360117652157928, | |
| "learning_rate": 1.7188236838779297e-05, | |
| "loss": 0.3648, | |
| "mean_token_accuracy": 0.9100905917584896, | |
| "num_tokens": 19728099.0, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 2.6926406926406927, | |
| "grad_norm": 0.2391244050549276, | |
| "learning_rate": 1.5931797447293553e-05, | |
| "loss": 0.3774, | |
| "mean_token_accuracy": 0.9097285665571689, | |
| "num_tokens": 22686649.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.4994457433638163, | |
| "learning_rate": 1.451010119216102e-05, | |
| "loss": 0.3629, | |
| "mean_token_accuracy": 0.9135821478467592, | |
| "num_tokens": 25201452.0, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 3.346320346320346, | |
| "grad_norm": 0.25158347609311765, | |
| "learning_rate": 1.2962755808856341e-05, | |
| "loss": 0.3194, | |
| "mean_token_accuracy": 0.924391707777977, | |
| "num_tokens": 28025121.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 3.6926406926406927, | |
| "grad_norm": 0.23233646369262503, | |
| "learning_rate": 1.133286955373779e-05, | |
| "loss": 0.3128, | |
| "mean_token_accuracy": 0.9216921903192997, | |
| "num_tokens": 31045393.0, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.21067819664553092, | |
| "learning_rate": 9.665850229923258e-06, | |
| "loss": 0.2902, | |
| "mean_token_accuracy": 0.9269120584071522, | |
| "num_tokens": 33595239.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 4.346320346320346, | |
| "grad_norm": 0.2221024058682983, | |
| "learning_rate": 8.008140148961642e-06, | |
| "loss": 0.2857, | |
| "mean_token_accuracy": 0.9298652648925781, | |
| "num_tokens": 36607286.0, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 4.692640692640692, | |
| "grad_norm": 0.2279645495215411, | |
| "learning_rate": 6.405922271624874e-06, | |
| "loss": 0.267, | |
| "mean_token_accuracy": 0.9332851983606816, | |
| "num_tokens": 39636669.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.33858416341824765, | |
| "learning_rate": 4.903833574080825e-06, | |
| "loss": 0.2359, | |
| "mean_token_accuracy": 0.9388179233376409, | |
| "num_tokens": 42016856.0, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 5.346320346320346, | |
| "grad_norm": 0.2550466927281677, | |
| "learning_rate": 3.543721484411976e-06, | |
| "loss": 0.2445, | |
| "mean_token_accuracy": 0.9396580524742604, | |
| "num_tokens": 44947028.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 5.692640692640692, | |
| "grad_norm": 0.2195997254176569, | |
| "learning_rate": 2.3634780345266805e-06, | |
| "loss": 0.2333, | |
| "mean_token_accuracy": 0.9418695524334908, | |
| "num_tokens": 47918431.0, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.22859511699176785, | |
| "learning_rate": 1.3959842073986085e-06, | |
| "loss": 0.2149, | |
| "mean_token_accuracy": 0.9448052300533778, | |
| "num_tokens": 50428344.0, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 6.346320346320346, | |
| "grad_norm": 0.20971033814817522, | |
| "learning_rate": 6.681938895839746e-07, | |
| "loss": 0.2256, | |
| "mean_token_accuracy": 0.9433787778019905, | |
| "num_tokens": 53450457.0, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 6.692640692640692, | |
| "grad_norm": 0.2087888768198865, | |
| "learning_rate": 2.0038294963413251e-07, | |
| "loss": 0.2195, | |
| "mean_token_accuracy": 0.9459246933460236, | |
| "num_tokens": 56378770.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.3392676914048232, | |
| "learning_rate": 5.584362697453882e-09, | |
| "loss": 0.2138, | |
| "mean_token_accuracy": 0.9467127667346471, | |
| "num_tokens": 58830536.0, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "step": 105, | |
| "total_flos": 118916339728384.0, | |
| "train_loss": 0.40460834616706487, | |
| "train_runtime": 2384.1607, | |
| "train_samples_per_second": 5.417, | |
| "train_steps_per_second": 0.044 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 105, | |
| "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": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 118916339728384.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |