| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "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": 345.78125, | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 1.91469407081604, | |
| "kl": 0.002231337522971444, | |
| "learning_rate": 4.965903258506806e-07, | |
| "loss": 0.0, | |
| "reward": 0.29195431526750326, | |
| "reward_std": 0.752438448369503, | |
| "rewards/concensus_correctness_reward_func": 0.039124999195337296, | |
| "rewards/consensus_reward_func": 0.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0625, | |
| "rewards/question_recreation_reward_func": 0.34751682076603174, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": -0.15718749910593033, | |
| "step": 2 | |
| }, | |
| { | |
| "completion_length": 349.125, | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 1.6588529348373413, | |
| "kl": 0.002210008809925057, | |
| "learning_rate": 4.698684378016222e-07, | |
| "loss": 0.0, | |
| "reward": 0.7831258047372103, | |
| "reward_std": 1.008952866308391, | |
| "rewards/concensus_correctness_reward_func": 0.12256249785423279, | |
| "rewards/consensus_reward_func": 0.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.375, | |
| "rewards/question_recreation_reward_func": 0.2274695634841919, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.0580937466584146, | |
| "step": 4 | |
| }, | |
| { | |
| "completion_length": 402.57142857142856, | |
| "epoch": 1.2105263157894737, | |
| "grad_norm": 2.6808555126190186, | |
| "kl": 0.003944302243845803, | |
| "learning_rate": 4.193203929064353e-07, | |
| "loss": 0.0, | |
| "reward": 1.028327831200191, | |
| "reward_std": 1.1498553667749678, | |
| "rewards/concensus_correctness_reward_func": 0.14014285802841187, | |
| "rewards/consensus_reward_func": 0.07142857142857142, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.21428571428571427, | |
| "rewards/question_recreation_reward_func": 0.3410063875573022, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.2614642892565046, | |
| "step": 6 | |
| }, | |
| { | |
| "completion_length": 319.4375, | |
| "epoch": 1.631578947368421, | |
| "grad_norm": 4.16013765335083, | |
| "kl": 0.0022768215858377516, | |
| "learning_rate": 3.5042385616324236e-07, | |
| "loss": 0.0, | |
| "reward": 1.5005913376808167, | |
| "reward_std": 2.108041614294052, | |
| "rewards/concensus_correctness_reward_func": 0.7163750007748604, | |
| "rewards/consensus_reward_func": 0.0625, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.125, | |
| "rewards/question_recreation_reward_func": 0.3670601084595546, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.22965625021606684, | |
| "step": 8 | |
| }, | |
| { | |
| "completion_length": 285.0, | |
| "epoch": 2.0, | |
| "grad_norm": 2.294445514678955, | |
| "kl": 0.0026452127578003065, | |
| "learning_rate": 2.706448363680831e-07, | |
| "loss": 0.0, | |
| "reward": 1.4123743091310774, | |
| "reward_std": 1.01517368214471, | |
| "rewards/concensus_correctness_reward_func": 0.13742856894220626, | |
| "rewards/consensus_reward_func": 0.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.7142857142857143, | |
| "rewards/question_recreation_reward_func": 0.3533743140952928, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.2072857107434954, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 369.0, | |
| "epoch": 2.4210526315789473, | |
| "grad_norm": 3.066178560256958, | |
| "kl": 0.002823778530000709, | |
| "learning_rate": 1.886286282148002e-07, | |
| "loss": 0.0, | |
| "reward": 0.9884280562400818, | |
| "reward_std": 1.076359886676073, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.0, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.4375, | |
| "rewards/question_recreation_reward_func": 0.3941780310124159, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.15675000147894025, | |
| "step": 12 | |
| }, | |
| { | |
| "completion_length": 258.8125, | |
| "epoch": 2.8421052631578947, | |
| "grad_norm": 4.03523063659668, | |
| "kl": 0.0028213907207828015, | |
| "learning_rate": 1.1326296046939333e-07, | |
| "loss": 0.0, | |
| "reward": 1.5553401838988066, | |
| "reward_std": 2.0138666601851583, | |
| "rewards/concensus_correctness_reward_func": 0.7452500015497208, | |
| "rewards/consensus_reward_func": 0.0625, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.1875, | |
| "rewards/question_recreation_reward_func": 0.30774645414203405, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.25234374590218067, | |
| "step": 14 | |
| }, | |
| { | |
| "completion_length": 310.35714285714283, | |
| "epoch": 3.2105263157894735, | |
| "grad_norm": 3.178826093673706, | |
| "kl": 0.001912726937527103, | |
| "learning_rate": 5.271487265090163e-08, | |
| "loss": 0.0, | |
| "reward": 0.4004705489746162, | |
| "reward_std": 1.051499856369836, | |
| "rewards/concensus_correctness_reward_func": 0.06871428659984044, | |
| "rewards/consensus_reward_func": 0.07142857142857142, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.07142857142857142, | |
| "rewards/question_recreation_reward_func": 0.21786337984459742, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.017857142857142856, | |
| "rewards/xmlcount_reward_func": -0.04682143724390438, | |
| "step": 16 | |
| }, | |
| { | |
| "completion_length": 413.9375, | |
| "epoch": 3.6315789473684212, | |
| "grad_norm": 6.53667688369751, | |
| "kl": 0.0016312935331370682, | |
| "learning_rate": 1.3545689574841341e-08, | |
| "loss": 0.0, | |
| "reward": 0.5633416045457125, | |
| "reward_std": 0.9396487567573786, | |
| "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": 0.36087287683039904, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.014968743547797203, | |
| "step": 18 | |
| }, | |
| { | |
| "completion_length": 340.4642857142857, | |
| "epoch": 4.0, | |
| "grad_norm": 1.7881169319152832, | |
| "kl": 0.0021190502946930273, | |
| "learning_rate": 0.0, | |
| "loss": 0.0, | |
| "reward": 0.9210910541670663, | |
| "reward_std": 1.4995469110352653, | |
| "rewards/concensus_correctness_reward_func": 0.1758571437426976, | |
| "rewards/consensus_reward_func": 0.07142857142857142, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.2857142857142857, | |
| "rewards/question_recreation_reward_func": 0.30434104667178224, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.083749991442476, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 20, | |
| "total_flos": 0.0, | |
| "train_loss": 2.333011286737019e-06, | |
| "train_runtime": 1452.1013, | |
| "train_samples_per_second": 0.22, | |
| "train_steps_per_second": 0.014 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 20, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "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 | |
| } | |