| { | |
| "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": 504.25, | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.6294240951538086, | |
| "kl": 0.0008423507533734664, | |
| "learning_rate": 5e-07, | |
| "loss": 0.0, | |
| "reward": 1.2705396972596645, | |
| "reward_std": 0.8699165377765894, | |
| "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": 0.6449146866798401, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.03125, | |
| "rewards/xmlcount_reward_func": 0.406875004991889, | |
| "step": 2 | |
| }, | |
| { | |
| "completion_length": 460.60714285714283, | |
| "epoch": 1.3636363636363638, | |
| "grad_norm": 0.39938339591026306, | |
| "kl": 0.0009498974208586983, | |
| "learning_rate": 4.864543104251586e-07, | |
| "loss": 0.0, | |
| "reward": 1.149305330855506, | |
| "reward_std": 1.067141784088952, | |
| "rewards/concensus_correctness_reward_func": 0.0497857140643256, | |
| "rewards/consensus_reward_func": 0.2857142857142857, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.6258052842957633, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.03571428571428571, | |
| "rewards/xmlcount_reward_func": 0.15228570784841264, | |
| "step": 4 | |
| }, | |
| { | |
| "completion_length": 584.25, | |
| "epoch": 2.0, | |
| "grad_norm": 1.4154448509216309, | |
| "kl": 0.0009216495839479778, | |
| "learning_rate": 4.472851273490984e-07, | |
| "loss": 0.0, | |
| "reward": 0.5566322782209941, | |
| "reward_std": 1.250387300338064, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.14285714285714285, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.07142857142857142, | |
| "rewards/question_recreation_reward_func": 0.48673940343516214, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.017857142857142856, | |
| "rewards/xmlcount_reward_func": -0.16224999300071172, | |
| "step": 6 | |
| }, | |
| { | |
| "completion_length": 540.21875, | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 0.7249035239219666, | |
| "kl": 0.0008380638319067657, | |
| "learning_rate": 3.867370395306068e-07, | |
| "loss": 0.0, | |
| "reward": 1.1796844955533743, | |
| "reward_std": 1.1831015311181545, | |
| "rewards/concensus_correctness_reward_func": 0.03125, | |
| "rewards/consensus_reward_func": 0.5, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.6402782127261162, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.015625, | |
| "rewards/xmlcount_reward_func": -0.007468747906386852, | |
| "step": 8 | |
| }, | |
| { | |
| "completion_length": 617.6428571428571, | |
| "epoch": 3.3636363636363638, | |
| "grad_norm": 0.4157549738883972, | |
| "kl": 0.0008354247339801597, | |
| "learning_rate": 3.1137137178519977e-07, | |
| "loss": 0.0, | |
| "reward": 1.2782049434525626, | |
| "reward_std": 1.1169604999678475, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.42857142857142855, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.597062064068658, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.03571428571428571, | |
| "rewards/xmlcount_reward_func": 0.21685715871197836, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 592.6428571428571, | |
| "epoch": 4.0, | |
| "grad_norm": 3.815882682800293, | |
| "kl": 0.0014564325267981207, | |
| "learning_rate": 2.2935516363191693e-07, | |
| "loss": 0.0, | |
| "reward": 0.8755534759589604, | |
| "reward_std": 1.5114734172821045, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.35714285714285715, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.5599820060389382, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": -0.04157141702515738, | |
| "step": 12 | |
| }, | |
| { | |
| "completion_length": 576.09375, | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.4694238305091858, | |
| "kl": 0.0019940473794122227, | |
| "learning_rate": 1.4957614383675767e-07, | |
| "loss": 0.0, | |
| "reward": 1.0448056980967522, | |
| "reward_std": 1.276182696223259, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.4375, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.6383682060986757, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": -0.03106249589473009, | |
| "step": 14 | |
| }, | |
| { | |
| "completion_length": 582.3214285714286, | |
| "epoch": 5.363636363636363, | |
| "grad_norm": 0.788042426109314, | |
| "kl": 0.0007741098192387394, | |
| "learning_rate": 8.067960709356478e-08, | |
| "loss": 0.0, | |
| "reward": 0.9345985140119281, | |
| "reward_std": 0.7620645059006554, | |
| "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": 0.5092770840440478, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.3538928670542581, | |
| "step": 16 | |
| }, | |
| { | |
| "completion_length": 601.0714285714286, | |
| "epoch": 6.0, | |
| "grad_norm": 0.3955620229244232, | |
| "kl": 0.0008603796637284436, | |
| "learning_rate": 3.013156219837776e-08, | |
| "loss": 0.0, | |
| "reward": 1.498078465461731, | |
| "reward_std": 1.03599237544196, | |
| "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": 0.669364162853786, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.017857142857142856, | |
| "rewards/xmlcount_reward_func": 0.5251428782939911, | |
| "step": 18 | |
| }, | |
| { | |
| "completion_length": 555.21875, | |
| "epoch": 6.7272727272727275, | |
| "grad_norm": 0.38449040055274963, | |
| "kl": 0.000928697467315942, | |
| "learning_rate": 3.4096741493194193e-09, | |
| "loss": 0.0, | |
| "reward": 1.5113182216882706, | |
| "reward_std": 1.1425050217658281, | |
| "rewards/concensus_correctness_reward_func": 0.03125, | |
| "rewards/consensus_reward_func": 0.375, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.6907557286322117, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.03125, | |
| "rewards/xmlcount_reward_func": 0.3830625033006072, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 6.7272727272727275, | |
| "step": 20, | |
| "total_flos": 0.0, | |
| "train_loss": 9.848059619343985e-07, | |
| "train_runtime": 706.4449, | |
| "train_samples_per_second": 0.453, | |
| "train_steps_per_second": 0.028 | |
| } | |
| ], | |
| "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 | |
| } | |