| { | |
| "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": 277.96875, | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 0.5224992632865906, | |
| "kl": 0.0010273568695993163, | |
| "learning_rate": 5e-07, | |
| "loss": 0.0, | |
| "reward": 1.826040230691433, | |
| "reward_std": 1.124132301658392, | |
| "rewards/concensus_correctness_reward_func": 0.3780624852515757, | |
| "rewards/consensus_reward_func": 0.3125, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.3125, | |
| "rewards/question_recreation_reward_func": 0.4694151869043708, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.03125, | |
| "rewards/xmlcount_reward_func": 0.3223125096410513, | |
| "step": 2 | |
| }, | |
| { | |
| "completion_length": 278.4583333333333, | |
| "epoch": 1.0, | |
| "grad_norm": 0.2708077132701874, | |
| "kl": 0.0008795141960339, | |
| "learning_rate": 4.864543104251586e-07, | |
| "loss": 0.0, | |
| "reward": 3.4186474283536277, | |
| "reward_std": 4.419917240738869, | |
| "rewards/concensus_correctness_reward_func": 1.6840000003576279, | |
| "rewards/consensus_reward_func": 0.25, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.5833333333333334, | |
| "rewards/question_recreation_reward_func": 0.6104390869537989, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.041666666666666664, | |
| "rewards/xmlcount_reward_func": 0.249208340421319, | |
| "step": 4 | |
| }, | |
| { | |
| "completion_length": 289.21875, | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 0.5563597083091736, | |
| "kl": 0.0009632354849600233, | |
| "learning_rate": 4.472851273490984e-07, | |
| "loss": 0.0, | |
| "reward": 1.834355428814888, | |
| "reward_std": 0.7832910157740116, | |
| "rewards/concensus_correctness_reward_func": 0.1694375015795231, | |
| "rewards/consensus_reward_func": 0.25, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.5, | |
| "rewards/question_recreation_reward_func": 0.5292928880080581, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.38562500942498446, | |
| "step": 6 | |
| }, | |
| { | |
| "completion_length": 275.1666666666667, | |
| "epoch": 2.0, | |
| "grad_norm": 0.32892560958862305, | |
| "kl": 0.0008922182799627384, | |
| "learning_rate": 3.867370395306068e-07, | |
| "loss": 0.0, | |
| "reward": 1.53997902572155, | |
| "reward_std": 1.0947819550832112, | |
| "rewards/concensus_correctness_reward_func": 0.028833332161108654, | |
| "rewards/consensus_reward_func": 0.16666666666666666, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.3333333333333333, | |
| "rewards/question_recreation_reward_func": 0.517020749549071, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.020833333333333332, | |
| "rewards/xmlcount_reward_func": 0.47329166034857434, | |
| "step": 8 | |
| }, | |
| { | |
| "completion_length": 246.28125, | |
| "epoch": 2.571428571428571, | |
| "grad_norm": 0.8310402035713196, | |
| "kl": 0.0013346855557756498, | |
| "learning_rate": 3.1137137178519977e-07, | |
| "loss": 0.0, | |
| "reward": 3.223034232854843, | |
| "reward_std": 2.629302829504013, | |
| "rewards/concensus_correctness_reward_func": 1.426312480121851, | |
| "rewards/consensus_reward_func": 0.5, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.4375, | |
| "rewards/question_recreation_reward_func": 0.4841280337423086, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.3750937506556511, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 262.3333333333333, | |
| "epoch": 3.0, | |
| "grad_norm": 0.36352404952049255, | |
| "kl": 0.0007948433534086993, | |
| "learning_rate": 2.2935516363191693e-07, | |
| "loss": 0.0, | |
| "reward": 1.390676478544871, | |
| "reward_std": 1.0079821795225143, | |
| "rewards/concensus_correctness_reward_func": 0.04925000170866648, | |
| "rewards/consensus_reward_func": 0.25, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.3333333333333333, | |
| "rewards/question_recreation_reward_func": 0.4735514372587204, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.28454166650772095, | |
| "step": 12 | |
| }, | |
| { | |
| "completion_length": 291.3125, | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 0.4861568808555603, | |
| "kl": 0.0008321171044372022, | |
| "learning_rate": 1.4957614383675767e-07, | |
| "loss": 0.0, | |
| "reward": 2.122737444937229, | |
| "reward_std": 2.4020534493029118, | |
| "rewards/concensus_correctness_reward_func": 0.7350624809041619, | |
| "rewards/consensus_reward_func": 0.3125, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.3125, | |
| "rewards/question_recreation_reward_func": 0.47873741388320923, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.28393751569092274, | |
| "step": 14 | |
| }, | |
| { | |
| "completion_length": 282.8333333333333, | |
| "epoch": 4.0, | |
| "grad_norm": 0.35824164748191833, | |
| "kl": 0.0008225665078498423, | |
| "learning_rate": 8.067960709356478e-08, | |
| "loss": 0.0, | |
| "reward": 1.6974951724211376, | |
| "reward_std": 1.195303310950597, | |
| "rewards/concensus_correctness_reward_func": 0.2313333327571551, | |
| "rewards/consensus_reward_func": 0.25, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.5, | |
| "rewards/question_recreation_reward_func": 0.3424951570729415, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.37366667141517, | |
| "step": 16 | |
| }, | |
| { | |
| "completion_length": 258.875, | |
| "epoch": 4.571428571428571, | |
| "grad_norm": 0.9892435669898987, | |
| "kl": 0.0009560112812323496, | |
| "learning_rate": 3.013156219837776e-08, | |
| "loss": 0.0, | |
| "reward": 2.7269940972328186, | |
| "reward_std": 2.214615745469928, | |
| "rewards/concensus_correctness_reward_func": 0.67250000173226, | |
| "rewards/consensus_reward_func": 0.5625, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.5625, | |
| "rewards/question_recreation_reward_func": 0.4880565758794546, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.015625, | |
| "rewards/xmlcount_reward_func": 0.42581248842179775, | |
| "step": 18 | |
| }, | |
| { | |
| "completion_length": 275.0416666666667, | |
| "epoch": 5.0, | |
| "grad_norm": 0.5237920880317688, | |
| "kl": 0.0010033420112449676, | |
| "learning_rate": 3.4096741493194193e-09, | |
| "loss": 0.0, | |
| "reward": 2.285256117582321, | |
| "reward_std": 2.5205470621585846, | |
| "rewards/concensus_correctness_reward_func": 1.0016666327913601, | |
| "rewards/consensus_reward_func": 0.25, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.25, | |
| "rewards/question_recreation_reward_func": 0.32992271333932877, | |
| "rewards/soft_format_reward_func": 0.020833333333333332, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.4328333338101705, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 20, | |
| "total_flos": 0.0, | |
| "train_loss": 8.486275987706904e-07, | |
| "train_runtime": 1484.7719, | |
| "train_samples_per_second": 0.216, | |
| "train_steps_per_second": 0.013 | |
| } | |
| ], | |
| "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": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |