| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.2, | |
| "eval_steps": 500, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "completion_length": 253.0375, | |
| "epoch": 0.08, | |
| "grad_norm": 0.34916529059410095, | |
| "kl": 0.00048295065280399283, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 0.0, | |
| "reward": 0.6274130865931511, | |
| "reward_std": 0.03938147491426207, | |
| "rewards/<lambda>": 0.6274130865931511, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 253.225, | |
| "epoch": 0.16, | |
| "grad_norm": 0.41210684180259705, | |
| "kl": 0.000624238243472064, | |
| "learning_rate": 7.600000000000001e-06, | |
| "loss": 0.0, | |
| "reward": 0.6234804779291153, | |
| "reward_std": 0.039024536509532484, | |
| "rewards/<lambda>": 0.6234804779291153, | |
| "step": 20 | |
| }, | |
| { | |
| "completion_length": 254.025, | |
| "epoch": 0.24, | |
| "grad_norm": 0.24894775450229645, | |
| "kl": 0.000612898166582454, | |
| "learning_rate": 1.16e-05, | |
| "loss": 0.0, | |
| "reward": 0.6120496317744255, | |
| "reward_std": 0.03366693327843677, | |
| "rewards/<lambda>": 0.6120496317744255, | |
| "step": 30 | |
| }, | |
| { | |
| "completion_length": 252.25, | |
| "epoch": 0.32, | |
| "grad_norm": 0.38096368312835693, | |
| "kl": 0.0006592169374926016, | |
| "learning_rate": 1.5600000000000003e-05, | |
| "loss": 0.0, | |
| "reward": 0.6164205402135849, | |
| "reward_std": 0.042428855784237385, | |
| "rewards/<lambda>": 0.6164205402135849, | |
| "step": 40 | |
| }, | |
| { | |
| "completion_length": 251.275, | |
| "epoch": 0.4, | |
| "grad_norm": 0.33936625719070435, | |
| "kl": 0.0006831091290223412, | |
| "learning_rate": 1.9600000000000002e-05, | |
| "loss": 0.0, | |
| "reward": 0.6328483298420906, | |
| "reward_std": 0.035810240707360205, | |
| "rewards/<lambda>": 0.6328483298420906, | |
| "step": 50 | |
| }, | |
| { | |
| "completion_length": 254.4125, | |
| "epoch": 0.48, | |
| "grad_norm": 0.4410642385482788, | |
| "kl": 0.0009464225804549642, | |
| "learning_rate": 1.9600000000000002e-05, | |
| "loss": 0.0, | |
| "reward": 0.6187735944986343, | |
| "reward_std": 0.03256268355762586, | |
| "rewards/<lambda>": 0.6187735944986343, | |
| "step": 60 | |
| }, | |
| { | |
| "completion_length": 255.4625, | |
| "epoch": 0.56, | |
| "grad_norm": 0.45320507884025574, | |
| "kl": 0.0017867269460111856, | |
| "learning_rate": 1.9155555555555558e-05, | |
| "loss": 0.0001, | |
| "reward": 0.619798681139946, | |
| "reward_std": 0.02807733681984246, | |
| "rewards/<lambda>": 0.619798681139946, | |
| "step": 70 | |
| }, | |
| { | |
| "completion_length": 253.925, | |
| "epoch": 0.64, | |
| "grad_norm": 0.43050986528396606, | |
| "kl": 0.0034164128388511016, | |
| "learning_rate": 1.8711111111111113e-05, | |
| "loss": 0.0001, | |
| "reward": 0.6333043664693833, | |
| "reward_std": 0.03672210559016094, | |
| "rewards/<lambda>": 0.6333043664693833, | |
| "step": 80 | |
| }, | |
| { | |
| "completion_length": 251.0625, | |
| "epoch": 0.72, | |
| "grad_norm": 0.39146214723587036, | |
| "kl": 0.0044442670623539016, | |
| "learning_rate": 1.826666666666667e-05, | |
| "loss": 0.0002, | |
| "reward": 0.6258660212159157, | |
| "reward_std": 0.03840005891397595, | |
| "rewards/<lambda>": 0.6258660212159157, | |
| "step": 90 | |
| }, | |
| { | |
| "completion_length": 245.4875, | |
| "epoch": 0.8, | |
| "grad_norm": 0.5947660207748413, | |
| "kl": 0.009713826864026488, | |
| "learning_rate": 1.782222222222222e-05, | |
| "loss": 0.0004, | |
| "reward": 0.647130835056305, | |
| "reward_std": 0.033473254850832745, | |
| "rewards/<lambda>": 0.647130835056305, | |
| "step": 100 | |
| }, | |
| { | |
| "completion_length": 253.6625, | |
| "epoch": 0.88, | |
| "grad_norm": 0.487610399723053, | |
| "kl": 0.015179099002853035, | |
| "learning_rate": 1.737777777777778e-05, | |
| "loss": 0.0006, | |
| "reward": 0.6606008350849152, | |
| "reward_std": 0.043368036416359244, | |
| "rewards/<lambda>": 0.6606008350849152, | |
| "step": 110 | |
| }, | |
| { | |
| "completion_length": 247.3875, | |
| "epoch": 0.96, | |
| "grad_norm": 0.4823659062385559, | |
| "kl": 0.023328896961174905, | |
| "learning_rate": 1.6933333333333336e-05, | |
| "loss": 0.0009, | |
| "reward": 0.6569811046123505, | |
| "reward_std": 0.03937327671446837, | |
| "rewards/<lambda>": 0.6569811046123505, | |
| "step": 120 | |
| }, | |
| { | |
| "completion_length": 240.6125, | |
| "epoch": 1.04, | |
| "grad_norm": 0.6331102848052979, | |
| "kl": 0.04820071374997496, | |
| "learning_rate": 1.648888888888889e-05, | |
| "loss": 0.0019, | |
| "reward": 0.678092373907566, | |
| "reward_std": 0.025905024044914172, | |
| "rewards/<lambda>": 0.678092373907566, | |
| "step": 130 | |
| }, | |
| { | |
| "completion_length": 243.7125, | |
| "epoch": 1.12, | |
| "grad_norm": 0.8691033720970154, | |
| "kl": 0.06938900966197252, | |
| "learning_rate": 1.6044444444444444e-05, | |
| "loss": 0.0028, | |
| "reward": 0.6855655789375306, | |
| "reward_std": 0.021212380210636185, | |
| "rewards/<lambda>": 0.6855655789375306, | |
| "step": 140 | |
| }, | |
| { | |
| "completion_length": 254.3125, | |
| "epoch": 1.2, | |
| "grad_norm": 0.4045356214046478, | |
| "kl": 0.06224800143390894, | |
| "learning_rate": 1.5600000000000003e-05, | |
| "loss": 0.0025, | |
| "reward": 0.7006005674600602, | |
| "reward_std": 0.013145922217518091, | |
| "rewards/<lambda>": 0.7006005674600602, | |
| "step": 150 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |