| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9996261682242991, | |
| "eval_steps": 250, | |
| "global_step": 1337, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05009345794392523, | |
| "grad_norm": 12.282962799072266, | |
| "learning_rate": 4.925373134328357e-07, | |
| "loss": 2.005, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.10018691588785046, | |
| "grad_norm": 16.077674865722656, | |
| "learning_rate": 9.925373134328357e-07, | |
| "loss": 1.9012, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.1502803738317757, | |
| "grad_norm": 5.729518413543701, | |
| "learning_rate": 9.451371571072319e-07, | |
| "loss": 1.1013, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.18691588785046728, | |
| "eval_loss": 0.3121011257171631, | |
| "eval_runtime": 26.5581, | |
| "eval_samples_per_second": 18.827, | |
| "eval_steps_per_second": 9.413, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.20037383177570092, | |
| "grad_norm": 1.5927001237869263, | |
| "learning_rate": 8.894430590191188e-07, | |
| "loss": 0.391, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.2504672897196262, | |
| "grad_norm": 1.310795545578003, | |
| "learning_rate": 8.337489609310058e-07, | |
| "loss": 0.2437, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.3005607476635514, | |
| "grad_norm": 0.8947069048881531, | |
| "learning_rate": 7.780548628428927e-07, | |
| "loss": 0.2037, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.3506542056074766, | |
| "grad_norm": 1.539252758026123, | |
| "learning_rate": 7.223607647547797e-07, | |
| "loss": 0.1918, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.37383177570093457, | |
| "eval_loss": 0.18135082721710205, | |
| "eval_runtime": 26.5479, | |
| "eval_samples_per_second": 18.834, | |
| "eval_steps_per_second": 9.417, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.40074766355140184, | |
| "grad_norm": 0.8958560228347778, | |
| "learning_rate": 6.666666666666666e-07, | |
| "loss": 0.1797, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.4508411214953271, | |
| "grad_norm": 1.4521379470825195, | |
| "learning_rate": 6.109725685785536e-07, | |
| "loss": 0.1618, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.5009345794392523, | |
| "grad_norm": 1.4709662199020386, | |
| "learning_rate": 5.552784704904405e-07, | |
| "loss": 0.1627, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.5510280373831775, | |
| "grad_norm": 1.4817917346954346, | |
| "learning_rate": 4.995843724023275e-07, | |
| "loss": 0.1654, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 0.5607476635514018, | |
| "eval_loss": 0.15894462168216705, | |
| "eval_runtime": 26.5439, | |
| "eval_samples_per_second": 18.837, | |
| "eval_steps_per_second": 9.418, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.6011214953271028, | |
| "grad_norm": 1.6152915954589844, | |
| "learning_rate": 4.438902743142144e-07, | |
| "loss": 0.1507, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.6512149532710281, | |
| "grad_norm": 1.5369535684585571, | |
| "learning_rate": 3.881961762261014e-07, | |
| "loss": 0.1596, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 0.7013084112149532, | |
| "grad_norm": 1.1234904527664185, | |
| "learning_rate": 3.3250207813798835e-07, | |
| "loss": 0.1499, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 0.7476635514018691, | |
| "eval_loss": 0.14763584733009338, | |
| "eval_runtime": 26.5517, | |
| "eval_samples_per_second": 18.831, | |
| "eval_steps_per_second": 9.416, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7514018691588785, | |
| "grad_norm": 0.9679731726646423, | |
| "learning_rate": 2.7680798004987534e-07, | |
| "loss": 0.1237, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.8014953271028037, | |
| "grad_norm": 0.9841839075088501, | |
| "learning_rate": 2.2111388196176226e-07, | |
| "loss": 0.1281, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 0.851588785046729, | |
| "grad_norm": 2.0309741497039795, | |
| "learning_rate": 1.6541978387364923e-07, | |
| "loss": 0.1394, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 0.9016822429906542, | |
| "grad_norm": 0.7312546968460083, | |
| "learning_rate": 1.0972568578553615e-07, | |
| "loss": 0.1227, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 0.9345794392523364, | |
| "eval_loss": 0.14381863176822662, | |
| "eval_runtime": 26.5473, | |
| "eval_samples_per_second": 18.834, | |
| "eval_steps_per_second": 9.417, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.9517757009345794, | |
| "grad_norm": 1.2973382472991943, | |
| "learning_rate": 5.403158769742311e-08, | |
| "loss": 0.1298, | |
| "step": 1273 | |
| }, | |
| { | |
| "epoch": 0.9996261682242991, | |
| "step": 1337, | |
| "total_flos": 8.688878626214707e+16, | |
| "train_loss": 0.3979023610288382, | |
| "train_runtime": 1906.0981, | |
| "train_samples_per_second": 5.614, | |
| "train_steps_per_second": 0.701 | |
| } | |
| ], | |
| "logging_steps": 67, | |
| "max_steps": 1337, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "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": 8.688878626214707e+16, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |