| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "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": 384.84375, | |
| "epoch": 0.1, | |
| "grad_norm": 45.04591751098633, | |
| "kl": 0.0, | |
| "learning_rate": 4.965903258506806e-07, | |
| "loss": -0.0, | |
| "reward": 3.691977483453229, | |
| "reward_std": 1.1356534652295522, | |
| "rewards/concensus_correctness_reward_func": 0.6882499977946281, | |
| "rewards/consensus_reward_func": 0.6875, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.8125, | |
| "rewards/question_recreation_reward_func": 0.6260087669361383, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.15625, | |
| "rewards/xmlcount_reward_func": 0.7214687555097044, | |
| "step": 2 | |
| }, | |
| { | |
| "completion_length": 290.28125, | |
| "epoch": 0.2, | |
| "grad_norm": 56.04582595825195, | |
| "kl": 0.048623996786773205, | |
| "learning_rate": 4.698684378016222e-07, | |
| "loss": 0.0, | |
| "reward": 5.547513723373413, | |
| "reward_std": 0.9548060596571304, | |
| "rewards/concensus_correctness_reward_func": 1.1509375181049109, | |
| "rewards/consensus_reward_func": 1.0625, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 1.0, | |
| "rewards/question_recreation_reward_func": 0.9249514192342758, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.21875, | |
| "rewards/xmlcount_reward_func": 1.1903750076889992, | |
| "step": 4 | |
| }, | |
| { | |
| "completion_length": 216.34375, | |
| "epoch": 0.3, | |
| "grad_norm": 23.09837532043457, | |
| "kl": 892.624653175706, | |
| "learning_rate": 4.193203929064353e-07, | |
| "loss": 0.8926, | |
| "reward": 4.4300857884809375, | |
| "reward_std": 1.2345747215877054, | |
| "rewards/concensus_correctness_reward_func": 0.9410000052303076, | |
| "rewards/consensus_reward_func": 0.75, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.625, | |
| "rewards/question_recreation_reward_func": 0.7900232495740056, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.25, | |
| "rewards/xmlcount_reward_func": 1.074062503874302, | |
| "step": 6 | |
| }, | |
| { | |
| "completion_length": 257.6875, | |
| "epoch": 0.4, | |
| "grad_norm": 25.814899444580078, | |
| "kl": 1.4642285024747252, | |
| "learning_rate": 3.5042385616324236e-07, | |
| "loss": 0.0015, | |
| "reward": 4.062486097216606, | |
| "reward_std": 1.364403745283198, | |
| "rewards/concensus_correctness_reward_func": 0.5777500020340085, | |
| "rewards/consensus_reward_func": 0.625, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.8125, | |
| "rewards/question_recreation_reward_func": 0.7780798338353634, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.171875, | |
| "rewards/xmlcount_reward_func": 1.0972812492400408, | |
| "step": 8 | |
| }, | |
| { | |
| "completion_length": 286.8125, | |
| "epoch": 0.5, | |
| "grad_norm": 17.246906280517578, | |
| "kl": 2.0072783255018294, | |
| "learning_rate": 2.706448363680831e-07, | |
| "loss": 0.002, | |
| "reward": 6.18870684504509, | |
| "reward_std": 0.7351647131145, | |
| "rewards/concensus_correctness_reward_func": 1.5492499843239784, | |
| "rewards/consensus_reward_func": 1.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 1.25, | |
| "rewards/question_recreation_reward_func": 0.929519459605217, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.21875, | |
| "rewards/xmlcount_reward_func": 1.2411874905228615, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 222.3125, | |
| "epoch": 0.6, | |
| "grad_norm": 38.915470123291016, | |
| "kl": 3.321804000530392, | |
| "learning_rate": 1.886286282148002e-07, | |
| "loss": 0.0033, | |
| "reward": 4.649261876940727, | |
| "reward_std": 1.167486259713769, | |
| "rewards/concensus_correctness_reward_func": 1.1363750249147415, | |
| "rewards/consensus_reward_func": 0.75, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.625, | |
| "rewards/question_recreation_reward_func": 0.793293122202158, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.28125, | |
| "rewards/xmlcount_reward_func": 1.0633437521755695, | |
| "step": 12 | |
| }, | |
| { | |
| "completion_length": 259.875, | |
| "epoch": 0.7, | |
| "grad_norm": 12.302684783935547, | |
| "kl": 1.9860680536367, | |
| "learning_rate": 1.1326296046939333e-07, | |
| "loss": 0.002, | |
| "reward": 4.9309286857023835, | |
| "reward_std": 0.8483433704241179, | |
| "rewards/concensus_correctness_reward_func": 1.1531875003129244, | |
| "rewards/consensus_reward_func": 0.9375, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.8125, | |
| "rewards/question_recreation_reward_func": 0.6953349150717258, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.21875, | |
| "rewards/xmlcount_reward_func": 1.113656248897314, | |
| "step": 14 | |
| }, | |
| { | |
| "completion_length": 260.1875, | |
| "epoch": 0.8, | |
| "grad_norm": 12.547839164733887, | |
| "kl": 0.4333856268785894, | |
| "learning_rate": 5.271487265090163e-08, | |
| "loss": 0.0004, | |
| "reward": 5.027402684092522, | |
| "reward_std": 1.352033179719001, | |
| "rewards/concensus_correctness_reward_func": 1.0091875093057752, | |
| "rewards/consensus_reward_func": 0.9375, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.75, | |
| "rewards/question_recreation_reward_func": 0.8164025899022818, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.265625, | |
| "rewards/xmlcount_reward_func": 1.2486874982714653, | |
| "step": 16 | |
| }, | |
| { | |
| "completion_length": 258.15625, | |
| "epoch": 0.9, | |
| "grad_norm": 14.867289543151855, | |
| "kl": 0.38399916142225266, | |
| "learning_rate": 1.3545689574841341e-08, | |
| "loss": 0.0004, | |
| "reward": 5.083541095256805, | |
| "reward_std": 1.3432283122092485, | |
| "rewards/concensus_correctness_reward_func": 1.1145625049248338, | |
| "rewards/consensus_reward_func": 0.75, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 1.0, | |
| "rewards/question_recreation_reward_func": 0.7862598411738873, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.234375, | |
| "rewards/xmlcount_reward_func": 1.1983437463641167, | |
| "step": 18 | |
| }, | |
| { | |
| "completion_length": 269.21875, | |
| "epoch": 1.0, | |
| "grad_norm": 76.03479766845703, | |
| "kl": 0.8027685410343111, | |
| "learning_rate": 0.0, | |
| "loss": 0.0008, | |
| "reward": 4.308647858910263, | |
| "reward_std": 0.8510435534408316, | |
| "rewards/concensus_correctness_reward_func": 0.8827499886974692, | |
| "rewards/consensus_reward_func": 1.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.375, | |
| "rewards/question_recreation_reward_func": 0.767147877253592, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.203125, | |
| "rewards/xmlcount_reward_func": 1.0806249976158142, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 20, | |
| "total_flos": 0.0, | |
| "train_loss": 0.0903062883833627, | |
| "train_runtime": 775.6737, | |
| "train_samples_per_second": 0.413, | |
| "train_steps_per_second": 0.026 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 20, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "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 | |
| } | |