| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 3000, | |
| "global_step": 44343, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 6.76544212164265e-05, | |
| "grad_norm": 5.40625, | |
| "learning_rate": 0.0002999932345578783, | |
| "loss": 4.4062, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.20296326364927947, | |
| "grad_norm": 0.8125, | |
| "learning_rate": 0.000279703673635072, | |
| "loss": 2.8418, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.20296326364927947, | |
| "eval_loss": 2.390815496444702, | |
| "eval_runtime": 59.2474, | |
| "eval_samples_per_second": 1593.404, | |
| "eval_steps_per_second": 6.228, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.40592652729855894, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 0.0002594073472701441, | |
| "loss": 2.7413, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.40592652729855894, | |
| "eval_loss": 2.3675475120544434, | |
| "eval_runtime": 59.0908, | |
| "eval_samples_per_second": 1597.625, | |
| "eval_steps_per_second": 6.245, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.6088897909478385, | |
| "grad_norm": 0.98828125, | |
| "learning_rate": 0.0002391110209052161, | |
| "loss": 2.7179, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.6088897909478385, | |
| "eval_loss": 2.3553311824798584, | |
| "eval_runtime": 59.096, | |
| "eval_samples_per_second": 1597.485, | |
| "eval_steps_per_second": 6.244, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.8118530545971179, | |
| "grad_norm": 0.89453125, | |
| "learning_rate": 0.0002188146945402882, | |
| "loss": 2.7088, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.8118530545971179, | |
| "eval_loss": 2.3488736152648926, | |
| "eval_runtime": 59.0776, | |
| "eval_samples_per_second": 1597.984, | |
| "eval_steps_per_second": 6.246, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.0148163182463974, | |
| "grad_norm": 0.82421875, | |
| "learning_rate": 0.00019851836817536025, | |
| "loss": 2.7028, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.0148163182463974, | |
| "eval_loss": 2.345126152038574, | |
| "eval_runtime": 59.0813, | |
| "eval_samples_per_second": 1597.884, | |
| "eval_steps_per_second": 6.246, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.217779581895677, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 0.0001782220418104323, | |
| "loss": 2.697, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.217779581895677, | |
| "eval_loss": 2.3421409130096436, | |
| "eval_runtime": 59.2296, | |
| "eval_samples_per_second": 1593.882, | |
| "eval_steps_per_second": 6.23, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.4207428455449564, | |
| "grad_norm": 0.96484375, | |
| "learning_rate": 0.00015792571544550436, | |
| "loss": 2.6955, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.4207428455449564, | |
| "eval_loss": 2.3397908210754395, | |
| "eval_runtime": 59.0689, | |
| "eval_samples_per_second": 1598.219, | |
| "eval_steps_per_second": 6.247, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.6237061091942357, | |
| "grad_norm": 0.85546875, | |
| "learning_rate": 0.0001376293890805764, | |
| "loss": 2.693, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.6237061091942357, | |
| "eval_loss": 2.3400025367736816, | |
| "eval_runtime": 59.0531, | |
| "eval_samples_per_second": 1598.647, | |
| "eval_steps_per_second": 6.249, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.8266693728435153, | |
| "grad_norm": 0.8046875, | |
| "learning_rate": 0.00011733306271564845, | |
| "loss": 2.6911, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 1.8266693728435153, | |
| "eval_loss": 2.3384146690368652, | |
| "eval_runtime": 59.0762, | |
| "eval_samples_per_second": 1598.022, | |
| "eval_steps_per_second": 6.246, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.029632636492795, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 9.703673635072052e-05, | |
| "loss": 2.6925, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.029632636492795, | |
| "eval_loss": 2.3384358882904053, | |
| "eval_runtime": 59.0594, | |
| "eval_samples_per_second": 1598.476, | |
| "eval_steps_per_second": 6.248, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.232595900142074, | |
| "grad_norm": 1.0625, | |
| "learning_rate": 7.674040998579256e-05, | |
| "loss": 2.6897, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 2.232595900142074, | |
| "eval_loss": 2.3386263847351074, | |
| "eval_runtime": 59.0986, | |
| "eval_samples_per_second": 1597.416, | |
| "eval_steps_per_second": 6.244, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 2.435559163791354, | |
| "grad_norm": 0.91796875, | |
| "learning_rate": 5.644408362086462e-05, | |
| "loss": 2.6921, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 2.435559163791354, | |
| "eval_loss": 2.3384146690368652, | |
| "eval_runtime": 59.0899, | |
| "eval_samples_per_second": 1597.649, | |
| "eval_steps_per_second": 6.245, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 2.638522427440633, | |
| "grad_norm": 0.8359375, | |
| "learning_rate": 3.614775725593667e-05, | |
| "loss": 2.6916, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 2.638522427440633, | |
| "eval_loss": 2.338075876235962, | |
| "eval_runtime": 58.9643, | |
| "eval_samples_per_second": 1601.054, | |
| "eval_steps_per_second": 6.258, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 2.841485691089913, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 1.5851430891008727e-05, | |
| "loss": 2.691, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 2.841485691089913, | |
| "eval_loss": 2.3384780883789062, | |
| "eval_runtime": 59.0786, | |
| "eval_samples_per_second": 1597.956, | |
| "eval_steps_per_second": 6.246, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 44343, | |
| "total_flos": 4.3758285331164365e+17, | |
| "train_loss": 2.7094475170827415, | |
| "train_runtime": 21883.4885, | |
| "train_samples_per_second": 518.723, | |
| "train_steps_per_second": 2.026 | |
| } | |
| ], | |
| "logging_steps": 3000, | |
| "max_steps": 44343, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 3000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.3758285331164365e+17, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |