| { | |
| "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": 439.21875, | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 5.828135013580322, | |
| "kl": 0.0014942500565666705, | |
| "learning_rate": 4.965903258506806e-07, | |
| "loss": 0.0, | |
| "reward": 0.10679153446108103, | |
| "reward_std": 0.6297805532813072, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.20822900370694697, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": -0.10143748670816422, | |
| "step": 2 | |
| }, | |
| { | |
| "completion_length": 345.7083333333333, | |
| "epoch": 1.0, | |
| "grad_norm": 23.521379470825195, | |
| "kl": 0.003981841291533783, | |
| "learning_rate": 4.698684378016222e-07, | |
| "loss": 0.0, | |
| "reward": 1.658778190612793, | |
| "reward_std": 2.896653901785612, | |
| "rewards/concensus_correctness_reward_func": 0.8787500013907751, | |
| "rewards/consensus_reward_func": 0.08333333333333333, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.4166666666666667, | |
| "rewards/question_recreation_reward_func": 0.39598656445741653, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": -0.1159583330154419, | |
| "step": 4 | |
| }, | |
| { | |
| "completion_length": 438.5, | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 1.3683266639709473, | |
| "kl": 0.0017546406452311203, | |
| "learning_rate": 4.193203929064353e-07, | |
| "loss": 0.0, | |
| "reward": 0.36969452537596226, | |
| "reward_std": 0.7297583511099219, | |
| "rewards/concensus_correctness_reward_func": 0.056062500923871994, | |
| "rewards/consensus_reward_func": 0.0625, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.24291324615478516, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.008218752220273018, | |
| "step": 6 | |
| }, | |
| { | |
| "completion_length": 378.9583333333333, | |
| "epoch": 2.0, | |
| "grad_norm": 17.58197021484375, | |
| "kl": 0.0063056297464451445, | |
| "learning_rate": 3.5042385616324236e-07, | |
| "loss": 0.0, | |
| "reward": 1.823668549458186, | |
| "reward_std": 2.0832131765782833, | |
| "rewards/concensus_correctness_reward_func": 0.8333333333333334, | |
| "rewards/consensus_reward_func": 0.08333333333333333, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.4166666666666667, | |
| "rewards/question_recreation_reward_func": 0.3380434938396017, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.15229166112840176, | |
| "step": 8 | |
| }, | |
| { | |
| "completion_length": 319.0, | |
| "epoch": 2.571428571428571, | |
| "grad_norm": 4.946495056152344, | |
| "kl": 0.0023342472850345075, | |
| "learning_rate": 2.706448363680831e-07, | |
| "loss": 0.0, | |
| "reward": 0.6347062969580293, | |
| "reward_std": 0.7983217537403107, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.0625, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.125, | |
| "rewards/question_recreation_reward_func": 0.3073625322431326, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.1398437414318323, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 299.9583333333333, | |
| "epoch": 3.0, | |
| "grad_norm": 32.36935043334961, | |
| "kl": 0.002557387478494396, | |
| "learning_rate": 1.886286282148002e-07, | |
| "loss": 0.0, | |
| "reward": 1.3059624830881755, | |
| "reward_std": 2.1406236762801805, | |
| "rewards/concensus_correctness_reward_func": 0.8333333333333334, | |
| "rewards/consensus_reward_func": 0.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.3333333333333333, | |
| "rewards/question_recreation_reward_func": 0.22862919544180235, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": -0.08933333307504654, | |
| "step": 12 | |
| }, | |
| { | |
| "completion_length": 364.59375, | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 6.134871006011963, | |
| "kl": 0.002294319980137516, | |
| "learning_rate": 1.1326296046939333e-07, | |
| "loss": 0.0, | |
| "reward": 0.9578792434185743, | |
| "reward_std": 0.7915452528977767, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.0625, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.3125, | |
| "rewards/question_recreation_reward_func": 0.3451604973524809, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.015625, | |
| "rewards/xmlcount_reward_func": 0.22209375258535147, | |
| "step": 14 | |
| }, | |
| { | |
| "completion_length": 402.0, | |
| "epoch": 4.0, | |
| "grad_norm": 1.2527885437011719, | |
| "kl": 0.004929080818934987, | |
| "learning_rate": 5.271487265090163e-08, | |
| "loss": 0.0, | |
| "reward": 0.3584599271416664, | |
| "reward_std": 1.0923715432484944, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.16666666666666666, | |
| "rewards/question_recreation_reward_func": 0.2636265928546588, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": -0.07183333983023961, | |
| "step": 16 | |
| }, | |
| { | |
| "completion_length": 376.75, | |
| "epoch": 4.571428571428571, | |
| "grad_norm": 2.3929598331451416, | |
| "kl": 0.002563515809015371, | |
| "learning_rate": 1.3545689574841341e-08, | |
| "loss": 0.0, | |
| "reward": 0.241681374842301, | |
| "reward_std": 1.120352853089571, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.1875, | |
| "rewards/question_recreation_reward_func": 0.23099386505782604, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": -0.1768124857917428, | |
| "step": 18 | |
| }, | |
| { | |
| "completion_length": 389.0, | |
| "epoch": 5.0, | |
| "grad_norm": 1.3994996547698975, | |
| "kl": 0.002270005274719248, | |
| "learning_rate": 0.0, | |
| "loss": 0.0, | |
| "reward": 0.5229276654620966, | |
| "reward_std": 0.7810152769088745, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.16666666666666666, | |
| "rewards/question_recreation_reward_func": 0.30830267692605656, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.047958330561717354, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 20, | |
| "total_flos": 0.0, | |
| "train_loss": 2.5685094556138212e-06, | |
| "train_runtime": 536.6447, | |
| "train_samples_per_second": 0.596, | |
| "train_steps_per_second": 0.037 | |
| } | |
| ], | |
| "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 | |
| } | |