{ "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": 174.6875, "epoch": 0.5714285714285714, "grad_norm": 1.9667098172249098e-08, "kl": 0.0, "learning_rate": 5e-07, "loss": 0.0, "reward": 7.934873551130295, "reward_std": 0.0883883461356163, "rewards/concensus_correctness_reward_func": 2.0439999625086784, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 1.3125, "rewards/question_recreation_reward_func": 0.9377485089353286, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.46875, "rewards/xmlcount_reward_func": 1.171875, "step": 2 }, { "completion_length": 183.45833333333334, "epoch": 1.0, "grad_norm": 0.0010795251000672579, "kl": 0.0001044354720534102, "learning_rate": 4.864543104251586e-07, "loss": 0.0, "reward": 8.588333288828531, "reward_std": 0.11785112818082173, "rewards/concensus_correctness_reward_func": 2.088333328564962, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 1.75, "rewards/question_recreation_reward_func": 1.0, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.5, "rewards/xmlcount_reward_func": 1.25, "step": 4 }, { "completion_length": 176.65625, "epoch": 1.5714285714285714, "grad_norm": 1.278555989265442, "kl": 0.001978079258492471, "learning_rate": 4.472851273490984e-07, "loss": 0.0, "reward": 7.873997002840042, "reward_std": 0.1769556701183319, "rewards/concensus_correctness_reward_func": 2.1707499772310257, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 1.125, "rewards/question_recreation_reward_func": 0.9376219906262122, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.46875, "rewards/xmlcount_reward_func": 1.171875, "step": 6 }, { "completion_length": 189.91666666666666, "epoch": 2.0, "grad_norm": 1.5522246360778809, "kl": 0.001163311706022796, "learning_rate": 3.867370395306068e-07, "loss": 0.0, "reward": 8.419333299001059, "reward_std": 0.11785112818082173, "rewards/concensus_correctness_reward_func": 1.9193333089351654, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 1.75, "rewards/question_recreation_reward_func": 1.0, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.5, "rewards/xmlcount_reward_func": 1.25, "step": 8 }, { "completion_length": 183.15625, "epoch": 2.571428571428571, "grad_norm": 0.002604426583275199, "kl": 0.003158855697620311, "learning_rate": 3.1137137178519977e-07, "loss": 0.0, "reward": 7.8738735020160675, "reward_std": 0.0, "rewards/concensus_correctness_reward_func": 2.045499972999096, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 1.25, "rewards/question_recreation_reward_func": 0.9377485089353286, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.46875, "rewards/xmlcount_reward_func": 1.171875, "step": 10 }, { "completion_length": 181.45833333333334, "epoch": 3.0, "grad_norm": 0.00258829933591187, "kl": 0.0018120275231012783, "learning_rate": 2.2935516363191693e-07, "loss": 0.0, "reward": 8.586333354314169, "reward_std": 0.3535534143447876, "rewards/concensus_correctness_reward_func": 2.0863333145777383, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 1.75, "rewards/question_recreation_reward_func": 1.0, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.5, "rewards/xmlcount_reward_func": 1.25, "step": 12 }, { "completion_length": 175.65625, "epoch": 3.571428571428571, "grad_norm": 2.532172441482544, "kl": 0.0023079599758943914, "learning_rate": 1.4957614383675767e-07, "loss": 0.0, "reward": 8.186795592308044, "reward_std": 0.26527525241544936, "rewards/concensus_correctness_reward_func": 2.0459999814629555, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 1.5625, "rewards/question_recreation_reward_func": 0.937670579587575, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.46875, "rewards/xmlcount_reward_func": 1.171875, "step": 14 }, { "completion_length": 197.83333333333334, "epoch": 4.0, "grad_norm": 0.003439662279561162, "kl": 0.004733134313331296, "learning_rate": 8.067960709356478e-08, "loss": 0.0, "reward": 8.231499989827475, "reward_std": 0.3240906298160553, "rewards/concensus_correctness_reward_func": 2.0856666366259256, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 1.4166666666666667, "rewards/question_recreation_reward_func": 1.0, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.4791666666666667, "rewards/xmlcount_reward_func": 1.25, "step": 16 }, { "completion_length": 169.90625, "epoch": 4.571428571428571, "grad_norm": 0.006752040237188339, "kl": 0.003379843590778364, "learning_rate": 3.013156219837776e-08, "loss": 0.0, "reward": 8.248373538255692, "reward_std": 0.1767766922712326, "rewards/concensus_correctness_reward_func": 2.169999971985817, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 1.5, "rewards/question_recreation_reward_func": 0.9377485089353286, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.46875, "rewards/xmlcount_reward_func": 1.171875, "step": 18 }, { "completion_length": 197.125, "epoch": 5.0, "grad_norm": 1.3831743001937866, "kl": 0.0051221498773278045, "learning_rate": 3.4096741493194193e-09, "loss": 0.0, "reward": 8.086999972661337, "reward_std": 0.11785115798314412, "rewards/concensus_correctness_reward_func": 1.920333315928777, "rewards/consensus_reward_func": 2.0, "rewards/cumulative_reward_2": 0.0, "rewards/final_correctness_reward_func": 1.4166666666666667, "rewards/question_recreation_reward_func": 1.0, "rewards/soft_format_reward_func": 0.0, "rewards/strict_format_reward_func": 0.5, "rewards/xmlcount_reward_func": 1.25, "step": 20 }, { "epoch": 5.0, "step": 20, "total_flos": 0.0, "train_loss": 2.0567678589600293e-06, "train_runtime": 5494.5037, "train_samples_per_second": 0.058, "train_steps_per_second": 0.004 } ], "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": 2, "trial_name": null, "trial_params": null }