| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 9760, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.23359630724948005, | |
| "eval_loss": 4.781733989715576, | |
| "eval_runtime": 7.3547, | |
| "eval_samples_per_second": 30.321, | |
| "eval_steps_per_second": 0.952, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 1.0245901639344261, | |
| "grad_norm": 13917.33203125, | |
| "learning_rate": 0.0005993999999999999, | |
| "loss": 1.2461, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.2654778724561881, | |
| "eval_loss": 4.399641990661621, | |
| "eval_runtime": 6.6539, | |
| "eval_samples_per_second": 33.514, | |
| "eval_steps_per_second": 1.052, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 2.0491803278688523, | |
| "grad_norm": 10209.078125, | |
| "learning_rate": 0.0005315753424657533, | |
| "loss": 1.0362, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.28501224188919994, | |
| "eval_loss": 4.204015254974365, | |
| "eval_runtime": 6.7337, | |
| "eval_samples_per_second": 33.117, | |
| "eval_steps_per_second": 1.04, | |
| "step": 2928 | |
| }, | |
| { | |
| "epoch": 3.0737704918032787, | |
| "grad_norm": 9973.97265625, | |
| "learning_rate": 0.0004630821917808219, | |
| "loss": 0.9715, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.29392819846778934, | |
| "eval_loss": 4.097615718841553, | |
| "eval_runtime": 6.7811, | |
| "eval_samples_per_second": 32.886, | |
| "eval_steps_per_second": 1.032, | |
| "step": 3904 | |
| }, | |
| { | |
| "epoch": 4.098360655737705, | |
| "grad_norm": 10912.2734375, | |
| "learning_rate": 0.00039458904109589035, | |
| "loss": 0.9346, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.3033619123673795, | |
| "eval_loss": 4.011116027832031, | |
| "eval_runtime": 6.6583, | |
| "eval_samples_per_second": 33.492, | |
| "eval_steps_per_second": 1.051, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 5.122950819672131, | |
| "grad_norm": 9932.109375, | |
| "learning_rate": 0.0003260958904109589, | |
| "loss": 0.9069, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.3087500987249129, | |
| "eval_loss": 3.9493002891540527, | |
| "eval_runtime": 6.7082, | |
| "eval_samples_per_second": 33.243, | |
| "eval_steps_per_second": 1.043, | |
| "step": 5856 | |
| }, | |
| { | |
| "epoch": 6.147540983606557, | |
| "grad_norm": 9450.361328125, | |
| "learning_rate": 0.00025760273972602734, | |
| "loss": 0.8835, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.3171044202434337, | |
| "eval_loss": 3.8856163024902344, | |
| "eval_runtime": 6.6331, | |
| "eval_samples_per_second": 33.619, | |
| "eval_steps_per_second": 1.055, | |
| "step": 6832 | |
| }, | |
| { | |
| "epoch": 7.172131147540983, | |
| "grad_norm": 9493.173828125, | |
| "learning_rate": 0.00018910958904109587, | |
| "loss": 0.8622, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.32337893693013786, | |
| "eval_loss": 3.8232619762420654, | |
| "eval_runtime": 6.7593, | |
| "eval_samples_per_second": 32.991, | |
| "eval_steps_per_second": 1.036, | |
| "step": 7808 | |
| }, | |
| { | |
| "epoch": 8.19672131147541, | |
| "grad_norm": 9608.576171875, | |
| "learning_rate": 0.00012061643835616437, | |
| "loss": 0.8415, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.329653453616842, | |
| "eval_loss": 3.7797963619232178, | |
| "eval_runtime": 6.6962, | |
| "eval_samples_per_second": 33.303, | |
| "eval_steps_per_second": 1.045, | |
| "step": 8784 | |
| }, | |
| { | |
| "epoch": 9.221311475409836, | |
| "grad_norm": 9442.396484375, | |
| "learning_rate": 5.212328767123287e-05, | |
| "loss": 0.8265, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.3321018314568287, | |
| "eval_loss": 3.758513927459717, | |
| "eval_runtime": 6.7113, | |
| "eval_samples_per_second": 33.228, | |
| "eval_steps_per_second": 1.043, | |
| "step": 9760 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 9760, | |
| "total_flos": 8.16015015936e+16, | |
| "train_loss": 0.9351713461954085, | |
| "train_runtime": 10046.3917, | |
| "train_samples_per_second": 31.086, | |
| "train_steps_per_second": 0.971 | |
| } | |
| ], | |
| "logging_steps": 1000, | |
| "max_steps": 9760, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "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": 8.16015015936e+16, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |