| { | |
| "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": 548.0, | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.3813503384590149, | |
| "kl": 0.0005652548934449442, | |
| "learning_rate": 5e-07, | |
| "loss": 0.0, | |
| "reward": 1.268306341022253, | |
| "reward_std": 0.9305344503372908, | |
| "rewards/concensus_correctness_reward_func": -0.013500000350177288, | |
| "rewards/consensus_reward_func": 0.5, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.6313688531517982, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.015625, | |
| "rewards/xmlcount_reward_func": 0.1348125096410513, | |
| "step": 2 | |
| }, | |
| { | |
| "completion_length": 519.0714285714286, | |
| "epoch": 1.3636363636363638, | |
| "grad_norm": 0.4107059836387634, | |
| "kl": 0.0005838624409599495, | |
| "learning_rate": 4.864543104251586e-07, | |
| "loss": 0.0, | |
| "reward": 1.1471687395657812, | |
| "reward_std": 1.296517321041652, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.5, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.14285714285714285, | |
| "rewards/question_recreation_reward_func": 0.5604187846183777, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": -0.05610714427062443, | |
| "step": 4 | |
| }, | |
| { | |
| "completion_length": 493.42857142857144, | |
| "epoch": 2.0, | |
| "grad_norm": 3.2317256927490234, | |
| "kl": 0.0007849977825701769, | |
| "learning_rate": 4.472851273490984e-07, | |
| "loss": 0.0, | |
| "reward": 1.0174174010753632, | |
| "reward_std": 1.1762256452015467, | |
| "rewards/concensus_correctness_reward_func": -0.015428571828774043, | |
| "rewards/consensus_reward_func": 0.42857142857142855, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.21428571428571427, | |
| "rewards/question_recreation_reward_func": 0.49534599695886883, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": -0.10535713844001293, | |
| "step": 6 | |
| }, | |
| { | |
| "completion_length": 424.34375, | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 0.7373853921890259, | |
| "kl": 0.0007588231637782883, | |
| "learning_rate": 3.867370395306068e-07, | |
| "loss": 0.0, | |
| "reward": 1.6647616177797318, | |
| "reward_std": 1.481652483344078, | |
| "rewards/concensus_correctness_reward_func": -0.027000000700354576, | |
| "rewards/consensus_reward_func": 0.875, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0625, | |
| "rewards/question_recreation_reward_func": 0.5569491200149059, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.046875, | |
| "rewards/xmlcount_reward_func": 0.15043749660253525, | |
| "step": 8 | |
| }, | |
| { | |
| "completion_length": 659.6071428571429, | |
| "epoch": 3.3636363636363638, | |
| "grad_norm": 0.2880741059780121, | |
| "kl": 0.0005647774627764843, | |
| "learning_rate": 3.1137137178519977e-07, | |
| "loss": 0.0, | |
| "reward": 1.6125414307628358, | |
| "reward_std": 1.161760653768267, | |
| "rewards/concensus_correctness_reward_func": 0.029214286378451755, | |
| "rewards/consensus_reward_func": 0.7142857142857143, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.07142857142857142, | |
| "rewards/question_recreation_reward_func": 0.4839700332709721, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.31364286158766064, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 432.17857142857144, | |
| "epoch": 4.0, | |
| "grad_norm": 0.40186750888824463, | |
| "kl": 0.0006883981654287449, | |
| "learning_rate": 2.2935516363191693e-07, | |
| "loss": 0.0, | |
| "reward": 1.3254563638142176, | |
| "reward_std": 1.0523893194539207, | |
| "rewards/concensus_correctness_reward_func": -0.04628571442195347, | |
| "rewards/consensus_reward_func": 0.35714285714285715, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.6414920730250222, | |
| "rewards/soft_format_reward_func": 0.017857142857142856, | |
| "rewards/strict_format_reward_func": 0.03571428571428571, | |
| "rewards/xmlcount_reward_func": 0.31953572162560057, | |
| "step": 12 | |
| }, | |
| { | |
| "completion_length": 525.5, | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.3948386013507843, | |
| "kl": 0.0006550081307068467, | |
| "learning_rate": 1.4957614383675767e-07, | |
| "loss": 0.0, | |
| "reward": 0.9727559667080641, | |
| "reward_std": 1.115324780344963, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.375, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0625, | |
| "rewards/question_recreation_reward_func": 0.4061622004956007, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.12909374572336674, | |
| "step": 14 | |
| }, | |
| { | |
| "completion_length": 505.17857142857144, | |
| "epoch": 5.363636363636363, | |
| "grad_norm": 0.40845388174057007, | |
| "kl": 0.0007241717290266284, | |
| "learning_rate": 8.067960709356478e-08, | |
| "loss": 0.0, | |
| "reward": 0.9902927960668292, | |
| "reward_std": 1.2754220792225428, | |
| "rewards/concensus_correctness_reward_func": -0.030857143657548085, | |
| "rewards/consensus_reward_func": 0.5714285714285714, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.5254356435367039, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": -0.07571428694895335, | |
| "step": 16 | |
| }, | |
| { | |
| "completion_length": 498.17857142857144, | |
| "epoch": 6.0, | |
| "grad_norm": 0.45181164145469666, | |
| "kl": 0.000708604252265234, | |
| "learning_rate": 3.013156219837776e-08, | |
| "loss": 0.0, | |
| "reward": 1.1312974180494035, | |
| "reward_std": 1.1744655881609236, | |
| "rewards/concensus_correctness_reward_func": -0.015428571828774043, | |
| "rewards/consensus_reward_func": 0.42857142857142855, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.14285714285714285, | |
| "rewards/question_recreation_reward_func": 0.5619760027953556, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.017857142857142856, | |
| "rewards/xmlcount_reward_func": -0.0045357027224131995, | |
| "step": 18 | |
| }, | |
| { | |
| "completion_length": 618.46875, | |
| "epoch": 6.7272727272727275, | |
| "grad_norm": 0.43022793531417847, | |
| "kl": 0.0007380955321423244, | |
| "learning_rate": 3.4096741493194193e-09, | |
| "loss": 0.0, | |
| "reward": 1.2054723352193832, | |
| "reward_std": 1.4775859341025352, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.5625, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.125, | |
| "rewards/question_recreation_reward_func": 0.5296285934746265, | |
| "rewards/soft_format_reward_func": 0.015625, | |
| "rewards/strict_format_reward_func": 0.03125, | |
| "rewards/xmlcount_reward_func": -0.05853124521672726, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 6.7272727272727275, | |
| "step": 20, | |
| "total_flos": 0.0, | |
| "train_loss": 6.219314457212021e-07, | |
| "train_runtime": 974.0742, | |
| "train_samples_per_second": 0.329, | |
| "train_steps_per_second": 0.021 | |
| } | |
| ], | |
| "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 | |
| } | |