| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 6.7272727272727275, | |
| "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": 482.3125, | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.6279861927032471, | |
| "kl": 0.0009842957915680017, | |
| "learning_rate": 5e-07, | |
| "loss": 0.0, | |
| "reward": 0.6330392956733704, | |
| "reward_std": 1.1256224997341633, | |
| "rewards/concensus_correctness_reward_func": 0.015124999918043613, | |
| "rewards/consensus_reward_func": 0.1875, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.5232893172651529, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.015625, | |
| "rewards/xmlcount_reward_func": -0.10850000288337469, | |
| "step": 2 | |
| }, | |
| { | |
| "completion_length": 422.89285714285717, | |
| "epoch": 1.3636363636363638, | |
| "grad_norm": 0.6565085053443909, | |
| "kl": 0.0014388576881693943, | |
| "learning_rate": 4.864543104251586e-07, | |
| "loss": 0.0, | |
| "reward": 0.8594981346811567, | |
| "reward_std": 1.1113629596573966, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.2857142857142857, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.5231409924370902, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.017857142857142856, | |
| "rewards/xmlcount_reward_func": 0.03278571793011257, | |
| "step": 4 | |
| }, | |
| { | |
| "completion_length": 533.6785714285714, | |
| "epoch": 2.0, | |
| "grad_norm": 0.32601848244667053, | |
| "kl": 0.0011336713297558682, | |
| "learning_rate": 4.472851273490984e-07, | |
| "loss": 0.0, | |
| "reward": 1.016930375780378, | |
| "reward_std": 0.86872969354902, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.2857142857142857, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.5773232643093381, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.15389286247747286, | |
| "step": 6 | |
| }, | |
| { | |
| "completion_length": 513.0, | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 0.4944967031478882, | |
| "kl": 0.0013540829895646311, | |
| "learning_rate": 3.867370395306068e-07, | |
| "loss": 0.0, | |
| "reward": 0.9314682334661484, | |
| "reward_std": 1.133194461464882, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.1875, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.598999485373497, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.03125, | |
| "rewards/xmlcount_reward_func": 0.11371875554323196, | |
| "step": 8 | |
| }, | |
| { | |
| "completion_length": 617.1428571428571, | |
| "epoch": 3.3636363636363638, | |
| "grad_norm": 0.38184112310409546, | |
| "kl": 0.0010740802307347102, | |
| "learning_rate": 3.1137137178519977e-07, | |
| "loss": 0.0, | |
| "reward": 1.5167566197259086, | |
| "reward_std": 1.177451640367508, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.7142857142857143, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.5730780448232379, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.22939286274569376, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 527.3571428571429, | |
| "epoch": 4.0, | |
| "grad_norm": 0.28677794337272644, | |
| "kl": 0.001259099051821977, | |
| "learning_rate": 2.2935516363191693e-07, | |
| "loss": 0.0, | |
| "reward": 0.703001109617097, | |
| "reward_std": 0.8764347093445914, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.07142857142857142, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.5151082554033825, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.017857142857142856, | |
| "rewards/xmlcount_reward_func": 0.09860714418547493, | |
| "step": 12 | |
| }, | |
| { | |
| "completion_length": 502.65625, | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 7.51918363571167, | |
| "kl": 0.0011182346497662365, | |
| "learning_rate": 1.4957614383675767e-07, | |
| "loss": 0.0, | |
| "reward": 1.0275223618373275, | |
| "reward_std": 0.9978118315339088, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.3125, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.5700535625219345, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.1449687611311674, | |
| "step": 14 | |
| }, | |
| { | |
| "completion_length": 450.17857142857144, | |
| "epoch": 5.363636363636363, | |
| "grad_norm": 2.624408721923828, | |
| "kl": 0.0018549877318686672, | |
| "learning_rate": 8.067960709356478e-08, | |
| "loss": 0.0, | |
| "reward": 0.9029674737581185, | |
| "reward_std": 0.8381789156368801, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.2857142857142857, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.07142857142857142, | |
| "rewards/question_recreation_reward_func": 0.5387888678482601, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.017857142857142856, | |
| "rewards/xmlcount_reward_func": -0.010821425489016942, | |
| "step": 16 | |
| }, | |
| { | |
| "completion_length": 378.7142857142857, | |
| "epoch": 6.0, | |
| "grad_norm": 29.219697952270508, | |
| "kl": 0.0020587185031867455, | |
| "learning_rate": 3.013156219837776e-08, | |
| "loss": 0.0, | |
| "reward": 1.33051621062415, | |
| "reward_std": 1.2022017240524292, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.5, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.627409074987684, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.03571428571428571, | |
| "rewards/xmlcount_reward_func": 0.16739285843712942, | |
| "step": 18 | |
| }, | |
| { | |
| "completion_length": 499.46875, | |
| "epoch": 6.7272727272727275, | |
| "grad_norm": 2.2562317848205566, | |
| "kl": 0.0015057353666634299, | |
| "learning_rate": 3.4096741493194193e-09, | |
| "loss": 0.0, | |
| "reward": 1.1141281686723232, | |
| "reward_std": 1.3231780752539635, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.375, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.6001906581223011, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.015625, | |
| "rewards/xmlcount_reward_func": 0.12331249937415123, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 6.7272727272727275, | |
| "step": 20, | |
| "total_flos": 0.0, | |
| "train_loss": 1.2697431884589604e-06, | |
| "train_runtime": 659.566, | |
| "train_samples_per_second": 0.485, | |
| "train_steps_per_second": 0.03 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 20, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "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": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |