| { | |
| "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": 510.96875, | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 1.453367829322815, | |
| "kl": 0.0006581695124623366, | |
| "learning_rate": 5e-07, | |
| "loss": 0.0, | |
| "reward": 0.7741625420749187, | |
| "reward_std": 0.9262352548539639, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.3125, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.6100688222795725, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.015625, | |
| "rewards/xmlcount_reward_func": -0.16403124295175076, | |
| "step": 2 | |
| }, | |
| { | |
| "completion_length": 421.10714285714283, | |
| "epoch": 1.3636363636363638, | |
| "grad_norm": 0.7830181121826172, | |
| "kl": 0.000890114635694772, | |
| "learning_rate": 4.864543104251586e-07, | |
| "loss": 0.0, | |
| "reward": 0.9196308766092572, | |
| "reward_std": 0.9299162881714957, | |
| "rewards/concensus_correctness_reward_func": 0.008928571428571428, | |
| "rewards/consensus_reward_func": 0.5714285714285714, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.4515237451664039, | |
| "rewards/soft_format_reward_func": 0.017857142857142856, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": -0.13010714309556143, | |
| "step": 4 | |
| }, | |
| { | |
| "completion_length": 456.5, | |
| "epoch": 2.0, | |
| "grad_norm": 4.7355804443359375, | |
| "kl": 0.0007553862607372659, | |
| "learning_rate": 4.472851273490984e-07, | |
| "loss": 0.0, | |
| "reward": 0.9059551358222961, | |
| "reward_std": 0.9832929755960192, | |
| "rewards/concensus_correctness_reward_func": 0.018785715103149414, | |
| "rewards/consensus_reward_func": 0.42857142857142855, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.5872765609196254, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": -0.1286785728963358, | |
| "step": 6 | |
| }, | |
| { | |
| "completion_length": 484.59375, | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 1.9618520736694336, | |
| "kl": 0.0008095701414276846, | |
| "learning_rate": 3.867370395306068e-07, | |
| "loss": 0.0, | |
| "reward": 1.1015731655061245, | |
| "reward_std": 1.1763431653380394, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.5625, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.5851044002920389, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.015625, | |
| "rewards/xmlcount_reward_func": -0.061656259931623936, | |
| "step": 8 | |
| }, | |
| { | |
| "completion_length": 410.0357142857143, | |
| "epoch": 3.3636363636363638, | |
| "grad_norm": 0.5187426805496216, | |
| "kl": 0.0008327829002934907, | |
| "learning_rate": 3.1137137178519977e-07, | |
| "loss": 0.0, | |
| "reward": 1.1646665313414164, | |
| "reward_std": 1.260940168585096, | |
| "rewards/concensus_correctness_reward_func": 0.03757143020629883, | |
| "rewards/consensus_reward_func": 0.35714285714285715, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.4700236969760486, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.017857142857142856, | |
| "rewards/xmlcount_reward_func": 0.28207142438207355, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 464.9642857142857, | |
| "epoch": 4.0, | |
| "grad_norm": 0.388344943523407, | |
| "kl": 0.0007031915218768907, | |
| "learning_rate": 2.2935516363191693e-07, | |
| "loss": 0.0, | |
| "reward": 0.9917694830468723, | |
| "reward_std": 0.8420192471572331, | |
| "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": 0.48387659234660013, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.1507500016263553, | |
| "step": 12 | |
| }, | |
| { | |
| "completion_length": 443.6875, | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.33633482456207275, | |
| "kl": 0.0007299767530639656, | |
| "learning_rate": 1.4957614383675767e-07, | |
| "loss": 0.0, | |
| "reward": 1.0675941202789545, | |
| "reward_std": 1.0126370824873447, | |
| "rewards/concensus_correctness_reward_func": 0.016437500715255737, | |
| "rewards/consensus_reward_func": 0.3125, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.5887503828853369, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.14990626275539398, | |
| "step": 14 | |
| }, | |
| { | |
| "completion_length": 520.8214285714286, | |
| "epoch": 5.363636363636363, | |
| "grad_norm": 0.3216298222541809, | |
| "kl": 0.0007666656913767968, | |
| "learning_rate": 8.067960709356478e-08, | |
| "loss": 0.0, | |
| "reward": 1.5415238652910506, | |
| "reward_std": 1.5021810361317225, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.7857142857142857, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.6235595515796116, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.017857142857142856, | |
| "rewards/xmlcount_reward_func": 0.11439285959516253, | |
| "step": 16 | |
| }, | |
| { | |
| "completion_length": 468.7857142857143, | |
| "epoch": 6.0, | |
| "grad_norm": 0.36451953649520874, | |
| "kl": 0.0007882461733450847, | |
| "learning_rate": 3.013156219837776e-08, | |
| "loss": 0.0, | |
| "reward": 1.0603240898677282, | |
| "reward_std": 0.9460750988551548, | |
| "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": 0.639431217951434, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.05357142857142857, | |
| "rewards/xmlcount_reward_func": 0.01017856172152928, | |
| "step": 18 | |
| }, | |
| { | |
| "completion_length": 503.0, | |
| "epoch": 6.7272727272727275, | |
| "grad_norm": 0.4202514886856079, | |
| "kl": 0.0007476642495021224, | |
| "learning_rate": 3.4096741493194193e-09, | |
| "loss": 0.0, | |
| "reward": 1.4001894146203995, | |
| "reward_std": 1.0581434704363346, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.5, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.7372518964111805, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.015625, | |
| "rewards/xmlcount_reward_func": 0.1473125098273158, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 6.7272727272727275, | |
| "step": 20, | |
| "total_flos": 0.0, | |
| "train_loss": 7.141880530525668e-07, | |
| "train_runtime": 1692.5748, | |
| "train_samples_per_second": 0.189, | |
| "train_steps_per_second": 0.012 | |
| } | |
| ], | |
| "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 | |
| } | |