| { | |
| "best_global_step": null, | |
| "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": 323.75, | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 0.5956937670707703, | |
| "kl": 0.0008181491066352464, | |
| "learning_rate": 5e-07, | |
| "loss": 0.0, | |
| "reward": 2.8557234406471252, | |
| "reward_std": 2.5599035173654556, | |
| "rewards/concensus_correctness_reward_func": 1.3679999978558044, | |
| "rewards/consensus_reward_func": 0.375, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.375, | |
| "rewards/question_recreation_reward_func": 0.5109110018238425, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.22681249864399433, | |
| "step": 2 | |
| }, | |
| { | |
| "completion_length": 265.0416666666667, | |
| "epoch": 1.0, | |
| "grad_norm": 0.4528919756412506, | |
| "kl": 0.0007884688869429132, | |
| "learning_rate": 4.864543104251586e-07, | |
| "loss": 0.0, | |
| "reward": 6.245500286420186, | |
| "reward_std": 6.114396661520004, | |
| "rewards/concensus_correctness_reward_func": 3.6057499752690396, | |
| "rewards/consensus_reward_func": 0.5, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 1.25, | |
| "rewards/question_recreation_reward_func": 0.5267084365089735, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.36304166664679843, | |
| "step": 4 | |
| }, | |
| { | |
| "completion_length": 289.40625, | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 0.428985059261322, | |
| "kl": 0.0007604565325891599, | |
| "learning_rate": 4.472851273490984e-07, | |
| "loss": 0.0, | |
| "reward": 4.009314224123955, | |
| "reward_std": 3.4844984263181686, | |
| "rewards/concensus_correctness_reward_func": 1.93650000038906, | |
| "rewards/consensus_reward_func": 0.5625, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.5, | |
| "rewards/question_recreation_reward_func": 0.49393921345472336, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0625, | |
| "rewards/xmlcount_reward_func": 0.453874995931983, | |
| "step": 6 | |
| }, | |
| { | |
| "completion_length": 303.625, | |
| "epoch": 2.0, | |
| "grad_norm": 0.36086052656173706, | |
| "kl": 0.0008454761118628085, | |
| "learning_rate": 3.867370395306068e-07, | |
| "loss": 0.0, | |
| "reward": 7.036508142948151, | |
| "reward_std": 6.619733400642872, | |
| "rewards/concensus_correctness_reward_func": 4.317250081027548, | |
| "rewards/consensus_reward_func": 0.5833333333333334, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 1.0833333333333333, | |
| "rewards/question_recreation_reward_func": 0.6239248464504877, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.4286666711171468, | |
| "step": 8 | |
| }, | |
| { | |
| "completion_length": 281.1875, | |
| "epoch": 2.571428571428571, | |
| "grad_norm": 0.3991297483444214, | |
| "kl": 0.0007313712703762576, | |
| "learning_rate": 3.1137137178519977e-07, | |
| "loss": 0.0, | |
| "reward": 4.937329009175301, | |
| "reward_std": 5.1635561510920525, | |
| "rewards/concensus_correctness_reward_func": 2.6058125024137553, | |
| "rewards/consensus_reward_func": 0.5625, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.9375, | |
| "rewards/question_recreation_reward_func": 0.5038917660713196, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.32762501388788223, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 262.75, | |
| "epoch": 3.0, | |
| "grad_norm": 0.2998476028442383, | |
| "kl": 0.0008465783515324196, | |
| "learning_rate": 2.2935516363191693e-07, | |
| "loss": 0.0, | |
| "reward": 2.618976910909017, | |
| "reward_std": 2.5861998895804086, | |
| "rewards/concensus_correctness_reward_func": 0.8422500000645717, | |
| "rewards/consensus_reward_func": 0.25, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.9166666666666666, | |
| "rewards/question_recreation_reward_func": 0.3030185053745906, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.3070416674017906, | |
| "step": 12 | |
| }, | |
| { | |
| "completion_length": 283.65625, | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 0.49871328473091125, | |
| "kl": 0.0007523985877924133, | |
| "learning_rate": 1.4957614383675767e-07, | |
| "loss": 0.0, | |
| "reward": 4.023481532931328, | |
| "reward_std": 3.793632350862026, | |
| "rewards/concensus_correctness_reward_func": 1.9771250014819088, | |
| "rewards/consensus_reward_func": 0.3125, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.8125, | |
| "rewards/question_recreation_reward_func": 0.5669192187488079, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.35443749465048313, | |
| "step": 14 | |
| }, | |
| { | |
| "completion_length": 292.7083333333333, | |
| "epoch": 4.0, | |
| "grad_norm": 0.3254660964012146, | |
| "kl": 0.0007524868318190178, | |
| "learning_rate": 8.067960709356478e-08, | |
| "loss": 0.0, | |
| "reward": 4.177254433433215, | |
| "reward_std": 6.047743608554204, | |
| "rewards/concensus_correctness_reward_func": 2.5, | |
| "rewards/consensus_reward_func": 0.4166666666666667, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.75, | |
| "rewards/question_recreation_reward_func": 0.27996282279491425, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.020833333333333332, | |
| "rewards/xmlcount_reward_func": 0.20979167272647223, | |
| "step": 16 | |
| }, | |
| { | |
| "completion_length": 259.46875, | |
| "epoch": 4.571428571428571, | |
| "grad_norm": 0.47842758893966675, | |
| "kl": 0.0007955787295941263, | |
| "learning_rate": 3.013156219837776e-08, | |
| "loss": 0.0, | |
| "reward": 2.009469695389271, | |
| "reward_std": 1.7849842756986618, | |
| "rewards/concensus_correctness_reward_func": 0.3196250008113566, | |
| "rewards/consensus_reward_func": 0.3125, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.5, | |
| "rewards/question_recreation_reward_func": 0.5244384333491325, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.35290625877678394, | |
| "step": 18 | |
| }, | |
| { | |
| "completion_length": 297.0, | |
| "epoch": 5.0, | |
| "grad_norm": 0.36606544256210327, | |
| "kl": 0.0008040873993498584, | |
| "learning_rate": 3.4096741493194193e-09, | |
| "loss": 0.0, | |
| "reward": 5.436734278996785, | |
| "reward_std": 6.2437847057978315, | |
| "rewards/concensus_correctness_reward_func": 3.3333333333333335, | |
| "rewards/consensus_reward_func": 0.3333333333333333, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 1.0833333333333333, | |
| "rewards/question_recreation_reward_func": 0.4109427034854889, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.2757916674017906, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 20, | |
| "total_flos": 0.0, | |
| "train_loss": 6.795619583499502e-07, | |
| "train_runtime": 332.5348, | |
| "train_samples_per_second": 0.962, | |
| "train_steps_per_second": 0.06 | |
| } | |
| ], | |
| "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 | |
| } | |