| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 6.7272727272727275, | |
| "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": 545.1875, | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.424740731716156, | |
| "kl": 0.0016840042117109988, | |
| "learning_rate": 5e-07, | |
| "loss": 0.0, | |
| "reward": 0.5074135232716799, | |
| "reward_std": 1.0041742585599422, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.1875, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.4731947723776102, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.015625, | |
| "rewards/xmlcount_reward_func": -0.16890624107327312, | |
| "step": 2 | |
| }, | |
| { | |
| "completion_length": 591.0714285714286, | |
| "epoch": 1.3636363636363638, | |
| "grad_norm": 0.34182819724082947, | |
| "kl": 0.0006761621168282415, | |
| "learning_rate": 4.864543104251586e-07, | |
| "loss": 0.0, | |
| "reward": 0.9373667751039777, | |
| "reward_std": 0.8502260233674731, | |
| "rewards/concensus_correctness_reward_func": 0.03392857100282397, | |
| "rewards/consensus_reward_func": 0.07142857142857142, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.5631881867136274, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.2688214283968721, | |
| "step": 4 | |
| }, | |
| { | |
| "completion_length": 746.8571428571429, | |
| "epoch": 2.0, | |
| "grad_norm": 0.23797620832920074, | |
| "kl": 0.0005519500950218312, | |
| "learning_rate": 4.472851273490984e-07, | |
| "loss": 0.0, | |
| "reward": 0.27029475356851307, | |
| "reward_std": 1.1702849609511239, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.07142857142857142, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.3723304644227028, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": -0.17346429824829102, | |
| "step": 6 | |
| }, | |
| { | |
| "completion_length": 547.09375, | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 0.3837575614452362, | |
| "kl": 0.0007064485362207051, | |
| "learning_rate": 3.867370395306068e-07, | |
| "loss": 0.0, | |
| "reward": 0.8859070055186749, | |
| "reward_std": 1.2122889906167984, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.1875, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.4487507492303848, | |
| "rewards/soft_format_reward_func": 0.015625, | |
| "rewards/strict_format_reward_func": 0.03125, | |
| "rewards/xmlcount_reward_func": 0.20278123719617724, | |
| "step": 8 | |
| }, | |
| { | |
| "completion_length": 580.3928571428571, | |
| "epoch": 3.3636363636363638, | |
| "grad_norm": 0.4430127441883087, | |
| "kl": 0.0005681784283037164, | |
| "learning_rate": 3.1137137178519977e-07, | |
| "loss": 0.0, | |
| "reward": 0.7612441182136536, | |
| "reward_std": 0.8055442699364254, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.14285714285714285, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.4517084253685815, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.017857142857142856, | |
| "rewards/xmlcount_reward_func": 0.14882143906184606, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 620.6071428571429, | |
| "epoch": 4.0, | |
| "grad_norm": 0.16699042916297913, | |
| "kl": 0.0006943187077662774, | |
| "learning_rate": 2.2935516363191693e-07, | |
| "loss": 0.0, | |
| "reward": 0.8246711769274303, | |
| "reward_std": 0.8966725638934544, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.14285714285714285, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.6077068788664681, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.017857142857142856, | |
| "rewards/xmlcount_reward_func": 0.05625000170298985, | |
| "step": 12 | |
| }, | |
| { | |
| "completion_length": 615.5, | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 2.4189465045928955, | |
| "kl": 0.0005959382724540774, | |
| "learning_rate": 1.4957614383675767e-07, | |
| "loss": 0.0, | |
| "reward": 0.8793371524661779, | |
| "reward_std": 0.9095463119447231, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.3125, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.5272433646023273, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": 0.03959375061094761, | |
| "step": 14 | |
| }, | |
| { | |
| "completion_length": 695.0714285714286, | |
| "epoch": 5.363636363636363, | |
| "grad_norm": 0.27298060059547424, | |
| "kl": 0.0006642445472867361, | |
| "learning_rate": 8.067960709356478e-08, | |
| "loss": 0.0, | |
| "reward": 0.4072608468788011, | |
| "reward_std": 0.679928679551397, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.07142857142857142, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.39158229316983906, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": -0.05574998472418104, | |
| "step": 16 | |
| }, | |
| { | |
| "completion_length": 617.8571428571429, | |
| "epoch": 6.0, | |
| "grad_norm": 0.31612712144851685, | |
| "kl": 0.0006989725474600814, | |
| "learning_rate": 3.013156219837776e-08, | |
| "loss": 0.0, | |
| "reward": 0.42214020022324156, | |
| "reward_std": 0.9410014876297542, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.07142857142857142, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.43992590797798975, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": -0.0892142823764256, | |
| "step": 18 | |
| }, | |
| { | |
| "completion_length": 536.96875, | |
| "epoch": 6.7272727272727275, | |
| "grad_norm": 3.633192777633667, | |
| "kl": 0.0007375037239398807, | |
| "learning_rate": 3.4096741493194193e-09, | |
| "loss": 0.0, | |
| "reward": 0.49759526643902063, | |
| "reward_std": 0.8468644013628364, | |
| "rewards/concensus_correctness_reward_func": 0.0, | |
| "rewards/consensus_reward_func": 0.1875, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.4367202501744032, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.0, | |
| "rewards/xmlcount_reward_func": -0.12662499537691474, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 6.7272727272727275, | |
| "step": 20, | |
| "total_flos": 0.0, | |
| "train_loss": 7.198314960987773e-07, | |
| "train_runtime": 1874.2425, | |
| "train_samples_per_second": 0.171, | |
| "train_steps_per_second": 0.011 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 20, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "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 | |
| } | |