| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.3555555555555556, | |
| "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": 133.5625, | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 13.52751636505127, | |
| "kl": 0.0, | |
| "learning_rate": 5e-07, | |
| "loss": 0.0, | |
| "reward": 0.1565040776040405, | |
| "reward_std": 0.10902803647331893, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.0625, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.09400407574139535, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.0, | |
| "step": 2 | |
| }, | |
| { | |
| "completion_length": 148.59375, | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 6.099397659301758, | |
| "kl": 0.0013341379481062177, | |
| "learning_rate": 4.864543104251586e-07, | |
| "loss": 0.0, | |
| "reward": 0.06237883912399411, | |
| "reward_std": 0.014348274113217485, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.06237883912399411, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.0, | |
| "step": 4 | |
| }, | |
| { | |
| "completion_length": 107.73076923076923, | |
| "epoch": 1.0, | |
| "grad_norm": 24.062419891357422, | |
| "kl": 0.0014511766102469454, | |
| "learning_rate": 4.472851273490984e-07, | |
| "loss": 0.0, | |
| "reward": 0.09928419011143538, | |
| "reward_std": 0.05120771519768123, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.09928419011143538, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.0, | |
| "step": 6 | |
| }, | |
| { | |
| "completion_length": 158.9375, | |
| "epoch": 1.3555555555555556, | |
| "grad_norm": 17.330768585205078, | |
| "kl": 0.0014418280325116939, | |
| "learning_rate": 3.8673703953060673e-07, | |
| "loss": 0.0, | |
| "reward": 0.0868330531520769, | |
| "reward_std": 0.0413361593564332, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.0868330531520769, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.0, | |
| "step": 8 | |
| }, | |
| { | |
| "completion_length": 162.0, | |
| "epoch": 1.7111111111111112, | |
| "grad_norm": 16.9694766998291, | |
| "kl": 0.005620427723442845, | |
| "learning_rate": 3.1137137178519977e-07, | |
| "loss": 0.0, | |
| "reward": 0.09794915979728103, | |
| "reward_std": 0.04119475433253683, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.09794915979728103, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.0, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 163.73076923076923, | |
| "epoch": 2.0, | |
| "grad_norm": 9.495556831359863, | |
| "kl": 0.013490407373935271, | |
| "learning_rate": 2.2935516363191693e-07, | |
| "loss": 0.0, | |
| "reward": 0.059179484844207764, | |
| "reward_std": 0.016660264437982384, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.059179484844207764, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.0, | |
| "step": 12 | |
| }, | |
| { | |
| "completion_length": 153.03125, | |
| "epoch": 2.3555555555555556, | |
| "grad_norm": 17.99853515625, | |
| "kl": 0.010461385358212283, | |
| "learning_rate": 1.4957614383675767e-07, | |
| "loss": 0.0, | |
| "reward": 0.08466288115596399, | |
| "reward_std": 0.014161967958898458, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.08466288115596399, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.0, | |
| "step": 14 | |
| }, | |
| { | |
| "completion_length": 152.1875, | |
| "epoch": 2.7111111111111112, | |
| "grad_norm": 4.608132839202881, | |
| "kl": 0.00154107587877661, | |
| "learning_rate": 8.067960709356475e-08, | |
| "loss": 0.0, | |
| "reward": 0.13246500422246754, | |
| "reward_std": 0.050385024725983385, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.13246500422246754, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.0, | |
| "step": 16 | |
| }, | |
| { | |
| "completion_length": 107.53846153846153, | |
| "epoch": 3.0, | |
| "grad_norm": 6.680927753448486, | |
| "kl": 0.012647272033903461, | |
| "learning_rate": 3.013156219837776e-08, | |
| "loss": 0.0, | |
| "reward": 0.07309502997220708, | |
| "reward_std": 0.028329174410408504, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.07309502997220708, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.0, | |
| "step": 18 | |
| }, | |
| { | |
| "completion_length": 222.625, | |
| "epoch": 3.3555555555555556, | |
| "grad_norm": 6.675609588623047, | |
| "kl": 0.0028270833790884353, | |
| "learning_rate": 3.4096741493194193e-09, | |
| "loss": 0.0, | |
| "reward": 0.08211358141852543, | |
| "reward_std": 0.036704360514704604, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.08211358141852543, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.3555555555555556, | |
| "step": 20, | |
| "total_flos": 0.0, | |
| "train_loss": 4.571605620640184e-06, | |
| "train_runtime": 8225.7678, | |
| "train_samples_per_second": 0.039, | |
| "train_steps_per_second": 0.002 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 20, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "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 | |
| } | |