| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 20, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "completion_length": 46.0, | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 6.293609619140625, | |
| "kl": 0.0, | |
| "learning_rate": 5e-07, | |
| "loss": 0.0, | |
| "reward": 1.5549669550964609, | |
| "reward_std": 0.3723401267779991, | |
| "rewards/concensus_correctness_reward_func": 0.6185000129044056, | |
| "rewards/consensus_reward_func": 0.875, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.06146700005047023, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.0, | |
| "step": 2 | |
| }, | |
| { | |
| "completion_length": 16.416666666666668, | |
| "epoch": 1.0, | |
| "grad_norm": 27.29721450805664, | |
| "kl": 0.0007262630485153446, | |
| "learning_rate": 4.864543104251586e-07, | |
| "loss": 0.0, | |
| "reward": 1.4664433126648266, | |
| "reward_std": 0.390086072186629, | |
| "rewards/concensus_correctness_reward_func": 0.515666663646698, | |
| "rewards/consensus_reward_func": 0.8333333333333334, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.07577663698854546, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.041666666666666664, | |
| "step": 4 | |
| }, | |
| { | |
| "completion_length": 42.5, | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 28.40996742248535, | |
| "kl": 0.007182279427070171, | |
| "learning_rate": 4.472851273490984e-07, | |
| "loss": 0.0, | |
| "reward": 1.4908071565441787, | |
| "reward_std": 0.4558348673162982, | |
| "rewards/concensus_correctness_reward_func": 0.4921250082552433, | |
| "rewards/consensus_reward_func": 0.875, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.06899467878974974, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.0546875, | |
| "step": 6 | |
| }, | |
| { | |
| "completion_length": 48.416666666666664, | |
| "epoch": 2.0, | |
| "grad_norm": 6.576460361480713, | |
| "kl": 0.008331439225003123, | |
| "learning_rate": 3.867370395306068e-07, | |
| "loss": 0.0, | |
| "reward": 1.3414389540751774, | |
| "reward_std": 0.203631525238355, | |
| "rewards/concensus_correctness_reward_func": 0.4986666639645894, | |
| "rewards/consensus_reward_func": 0.6666666666666666, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.17210563784465194, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.004000000034769376, | |
| "step": 8 | |
| }, | |
| { | |
| "completion_length": 37.875, | |
| "epoch": 2.571428571428571, | |
| "grad_norm": 33.859840393066406, | |
| "kl": 0.019948258966905996, | |
| "learning_rate": 3.1137137178519977e-07, | |
| "loss": 0.0, | |
| "reward": 1.1040713414549828, | |
| "reward_std": 0.3551928735105321, | |
| "rewards/concensus_correctness_reward_func": 0.3880000002682209, | |
| "rewards/consensus_reward_func": 0.625, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.07738385046832263, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.013687499798834324, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 16.333333333333332, | |
| "epoch": 3.0, | |
| "grad_norm": 7.724001884460449, | |
| "kl": 0.013843087945133448, | |
| "learning_rate": 2.2935516363191693e-07, | |
| "loss": 0.0, | |
| "reward": 2.5858581960201263, | |
| "reward_std": 0.4463220114509265, | |
| "rewards/concensus_correctness_reward_func": 1.0366666813691456, | |
| "rewards/consensus_reward_func": 1.5, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.035691587487235665, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.01350000003973643, | |
| "step": 12 | |
| }, | |
| { | |
| "completion_length": 46.1875, | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 31.069372177124023, | |
| "kl": 0.030557939084246755, | |
| "learning_rate": 1.4957614383675767e-07, | |
| "loss": 0.0, | |
| "reward": 1.386018784251064, | |
| "reward_std": 0.6735579386004247, | |
| "rewards/concensus_correctness_reward_func": 0.5047500021755695, | |
| "rewards/consensus_reward_func": 0.75, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.13126881583593786, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.0, | |
| "step": 14 | |
| }, | |
| { | |
| "completion_length": 32.25, | |
| "epoch": 4.0, | |
| "grad_norm": 12.990033149719238, | |
| "kl": 0.013698793482035398, | |
| "learning_rate": 8.067960709356478e-08, | |
| "loss": 0.0, | |
| "reward": 2.2689745972553887, | |
| "reward_std": 0.006027515667180221, | |
| "rewards/concensus_correctness_reward_func": 0.8840000132719675, | |
| "rewards/consensus_reward_func": 1.3333333333333333, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.05164131483373543, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.0, | |
| "step": 16 | |
| }, | |
| { | |
| "completion_length": 38.375, | |
| "epoch": 4.571428571428571, | |
| "grad_norm": 8.09213924407959, | |
| "kl": 0.032939108728896827, | |
| "learning_rate": 3.013156219837776e-08, | |
| "loss": 0.0, | |
| "reward": 1.7274093367159367, | |
| "reward_std": 0.0716479024849832, | |
| "rewards/concensus_correctness_reward_func": 0.6040000058710575, | |
| "rewards/consensus_reward_func": 1.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.1402843581745401, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": -0.016875000670552254, | |
| "step": 18 | |
| }, | |
| { | |
| "completion_length": 34.833333333333336, | |
| "epoch": 5.0, | |
| "grad_norm": 10.715066909790039, | |
| "kl": 0.021114608040079474, | |
| "learning_rate": 3.4096741493194193e-09, | |
| "loss": 0.0, | |
| "reward": 1.5627532539268334, | |
| "reward_std": 0.34191878077884513, | |
| "rewards/concensus_correctness_reward_func": 0.6946666737397512, | |
| "rewards/consensus_reward_func": 0.8333333333333334, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.03475328969458739, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 20, | |
| "total_flos": 0.0, | |
| "train_loss": 1.3395203330901495e-05, | |
| "train_runtime": 31200.9923, | |
| "train_samples_per_second": 0.005, | |
| "train_steps_per_second": 0.001 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 20, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |