| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.96, | |
| "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": 290.5, | |
| "epoch": 0.32, | |
| "grad_norm": 11.468047142028809, | |
| "kl": 0.0, | |
| "learning_rate": 5e-07, | |
| "loss": 0.0, | |
| "reward": 1.224143806612119, | |
| "reward_std": 0.8351118011050858, | |
| "rewards/concensus_correctness_reward_func": 0.24512500315904617, | |
| "rewards/consensus_reward_func": 0.125, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.1875, | |
| "rewards/question_recreation_reward_func": 0.3117375586880371, | |
| "rewards/soft_format_reward_func": 0.015625, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.33915625140070915, | |
| "step": 2 | |
| }, | |
| { | |
| "completion_length": 401.84375, | |
| "epoch": 0.64, | |
| "grad_norm": 5.856527328491211, | |
| "kl": 0.0010117826022906229, | |
| "learning_rate": 4.864543104251586e-07, | |
| "loss": 0.0, | |
| "reward": 0.36932456074282527, | |
| "reward_std": 0.8935166031587869, | |
| "rewards/concensus_correctness_reward_func": 0.05999999865889549, | |
| "rewards/consensus_reward_func": 0.0625, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.28557456005364656, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": -0.03874998493120074, | |
| "step": 4 | |
| }, | |
| { | |
| "completion_length": 300.375, | |
| "epoch": 0.96, | |
| "grad_norm": 9.89108943939209, | |
| "kl": 0.0013540440049837343, | |
| "learning_rate": 4.472851273490984e-07, | |
| "loss": 0.0, | |
| "reward": 0.9661785025382414, | |
| "reward_std": 1.0016355221450794, | |
| "rewards/concensus_correctness_reward_func": 0.1848750002682209, | |
| "rewards/consensus_reward_func": 0.125, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0625, | |
| "rewards/question_recreation_reward_func": 0.3211159927304834, | |
| "rewards/soft_format_reward_func": 0.015625, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.25706250919029117, | |
| "step": 6 | |
| }, | |
| { | |
| "completion_length": 374.5, | |
| "epoch": 1.16, | |
| "grad_norm": 14.001546859741211, | |
| "kl": 0.0014814753434620799, | |
| "learning_rate": 3.867370395306068e-07, | |
| "loss": 0.0, | |
| "reward": 0.7591779669746757, | |
| "reward_std": 0.7673417308134958, | |
| "rewards/concensus_correctness_reward_func": 0.09559999704360962, | |
| "rewards/consensus_reward_func": 0.1, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.24647796172648667, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.3171000026166439, | |
| "step": 8 | |
| }, | |
| { | |
| "completion_length": 353.21875, | |
| "epoch": 1.48, | |
| "grad_norm": 5.270878791809082, | |
| "kl": 0.004905104771751212, | |
| "learning_rate": 3.1137137178519977e-07, | |
| "loss": 0.0, | |
| "reward": 1.23657629895024, | |
| "reward_std": 0.6223840287420899, | |
| "rewards/concensus_correctness_reward_func": 0.11949999630451202, | |
| "rewards/consensus_reward_func": 0.1875, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.125, | |
| "rewards/question_recreation_reward_func": 0.35770130017772317, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.4468749922234565, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 332.1875, | |
| "epoch": 1.8, | |
| "grad_norm": 15.442935943603516, | |
| "kl": 0.0013477384873112896, | |
| "learning_rate": 2.2935516363191693e-07, | |
| "loss": 0.0, | |
| "reward": 0.7857276191934943, | |
| "reward_std": 0.5130172067438252, | |
| "rewards/concensus_correctness_reward_func": 0.11999999731779099, | |
| "rewards/consensus_reward_func": 0.125, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.125, | |
| "rewards/question_recreation_reward_func": 0.3161651297705248, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.0995625001451117, | |
| "step": 12 | |
| }, | |
| { | |
| "completion_length": 308.8, | |
| "epoch": 2.0, | |
| "grad_norm": 3.213531970977783, | |
| "kl": 0.0013160508882720024, | |
| "learning_rate": 1.4957614383675767e-07, | |
| "loss": 0.0, | |
| "reward": 1.1100439364090562, | |
| "reward_std": 1.5506223607808352, | |
| "rewards/concensus_correctness_reward_func": 0.2959999978542328, | |
| "rewards/consensus_reward_func": 0.2, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.1, | |
| "rewards/question_recreation_reward_func": 0.37719390653073787, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.025, | |
| "rewards/xmlcount_reward_func": 0.11185000836849213, | |
| "step": 14 | |
| }, | |
| { | |
| "completion_length": 371.5, | |
| "epoch": 2.32, | |
| "grad_norm": 99.86141204833984, | |
| "kl": 0.01706639961048495, | |
| "learning_rate": 8.067960709356478e-08, | |
| "loss": 0.0, | |
| "reward": 1.0817621119786054, | |
| "reward_std": 1.0203345250920393, | |
| "rewards/concensus_correctness_reward_func": 0.3076249994337559, | |
| "rewards/consensus_reward_func": 0.1875, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0625, | |
| "rewards/question_recreation_reward_func": 0.32141835533548146, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.20271874638274312, | |
| "step": 16 | |
| }, | |
| { | |
| "completion_length": 356.1875, | |
| "epoch": 2.64, | |
| "grad_norm": 7.011810779571533, | |
| "kl": 0.001219689438585192, | |
| "learning_rate": 3.013156219837776e-08, | |
| "loss": 0.0, | |
| "reward": 1.147089685080573, | |
| "reward_std": 1.0848051158245653, | |
| "rewards/concensus_correctness_reward_func": 0.12025000154972076, | |
| "rewards/consensus_reward_func": 0.1875, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0625, | |
| "rewards/question_recreation_reward_func": 0.39693344407714903, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.015625, | |
| "rewards/xmlcount_reward_func": 0.36428126646205783, | |
| "step": 18 | |
| }, | |
| { | |
| "completion_length": 395.96875, | |
| "epoch": 2.96, | |
| "grad_norm": 5.470472812652588, | |
| "kl": 0.0017344929874525405, | |
| "learning_rate": 3.4096741493194193e-09, | |
| "loss": 0.0, | |
| "reward": 0.647011962486431, | |
| "reward_std": 0.705216835369356, | |
| "rewards/concensus_correctness_reward_func": 0.05999999865889549, | |
| "rewards/consensus_reward_func": 0.0625, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.3264807239174843, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.1980312541127205, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "step": 20, | |
| "total_flos": 0.0, | |
| "train_loss": 3.0236318707466125e-06, | |
| "train_runtime": 1388.2855, | |
| "train_samples_per_second": 0.231, | |
| "train_steps_per_second": 0.014 | |
| } | |
| ], | |
| "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 | |
| } | |