| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "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": 268.125, | |
| "epoch": 0.1, | |
| "grad_norm": 212.015380859375, | |
| "kl": 0.0, | |
| "learning_rate": 4.965903258506806e-07, | |
| "loss": -0.0, | |
| "reward": 2.971410434693098, | |
| "reward_std": 1.660103004425764, | |
| "rewards/concensus_correctness_reward_func": 0.8277499973773956, | |
| "rewards/consensus_reward_func": 0.875, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.0, | |
| "rewards/question_recreation_reward_func": 0.5752854123711586, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.15625, | |
| "rewards/xmlcount_reward_func": 0.5371250007301569, | |
| "step": 2 | |
| }, | |
| { | |
| "completion_length": 161.84375, | |
| "epoch": 0.2, | |
| "grad_norm": 4049.1455078125, | |
| "kl": 3116631.138454, | |
| "learning_rate": 4.698684378016222e-07, | |
| "loss": 3116.6313, | |
| "reward": 4.441190658137202, | |
| "reward_std": 2.443159078247845, | |
| "rewards/concensus_correctness_reward_func": 1.127562504261732, | |
| "rewards/consensus_reward_func": 1.25, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.125, | |
| "rewards/question_recreation_reward_func": 0.7441594742704183, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.265625, | |
| "rewards/xmlcount_reward_func": 0.9288437515497208, | |
| "step": 4 | |
| }, | |
| { | |
| "completion_length": 173.96875, | |
| "epoch": 0.3, | |
| "grad_norm": 6921.29296875, | |
| "kl": 18184.31745560351, | |
| "learning_rate": 4.193203929064353e-07, | |
| "loss": 18.1843, | |
| "reward": 5.475079163908958, | |
| "reward_std": 1.6933745071146404, | |
| "rewards/concensus_correctness_reward_func": 1.6708124876022339, | |
| "rewards/consensus_reward_func": 1.375, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.3125, | |
| "rewards/question_recreation_reward_func": 0.7829542085528374, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.3125, | |
| "rewards/xmlcount_reward_func": 1.0213124975562096, | |
| "step": 6 | |
| }, | |
| { | |
| "completion_length": 206.09375, | |
| "epoch": 0.4, | |
| "grad_norm": 600.0673217773438, | |
| "kl": 10434.655427431862, | |
| "learning_rate": 3.5042385616324236e-07, | |
| "loss": 10.4347, | |
| "reward": 5.194255158305168, | |
| "reward_std": 1.982654629391618, | |
| "rewards/concensus_correctness_reward_func": 1.5163750085048378, | |
| "rewards/consensus_reward_func": 1.3125, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.25, | |
| "rewards/question_recreation_reward_func": 0.8320676572620869, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.28125, | |
| "rewards/xmlcount_reward_func": 1.0020624957978725, | |
| "step": 8 | |
| }, | |
| { | |
| "completion_length": 154.8125, | |
| "epoch": 0.5, | |
| "grad_norm": 278.44561767578125, | |
| "kl": 26163472.887304425, | |
| "learning_rate": 2.706448363680831e-07, | |
| "loss": 26163.4727, | |
| "reward": 6.384411156177521, | |
| "reward_std": 1.6228036261891248, | |
| "rewards/concensus_correctness_reward_func": 2.0416249968111515, | |
| "rewards/consensus_reward_func": 1.5625, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.625, | |
| "rewards/question_recreation_reward_func": 0.8268798030912876, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.296875, | |
| "rewards/xmlcount_reward_func": 1.0315312538295984, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 231.46875, | |
| "epoch": 0.6, | |
| "grad_norm": 50832.27734375, | |
| "kl": 6071.537940578186, | |
| "learning_rate": 1.886286282148002e-07, | |
| "loss": 6.0715, | |
| "reward": 4.917726576328278, | |
| "reward_std": 1.8517463966272771, | |
| "rewards/concensus_correctness_reward_func": 1.3334374949336052, | |
| "rewards/consensus_reward_func": 1.25, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.1875, | |
| "rewards/question_recreation_reward_func": 0.8517578095197678, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.25, | |
| "rewards/xmlcount_reward_func": 1.0450312420725822, | |
| "step": 12 | |
| }, | |
| { | |
| "completion_length": 159.0625, | |
| "epoch": 0.7, | |
| "grad_norm": 71.38750457763672, | |
| "kl": 4.22442401223816, | |
| "learning_rate": 1.1326296046939333e-07, | |
| "loss": 0.0042, | |
| "reward": 5.902323350310326, | |
| "reward_std": 0.8223230724979658, | |
| "rewards/concensus_correctness_reward_func": 1.6347499899566174, | |
| "rewards/consensus_reward_func": 1.5625, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.25, | |
| "rewards/question_recreation_reward_func": 0.9046983420848846, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.390625, | |
| "rewards/xmlcount_reward_func": 1.1597499996423721, | |
| "step": 14 | |
| }, | |
| { | |
| "completion_length": 237.5, | |
| "epoch": 0.8, | |
| "grad_norm": 86529.1875, | |
| "kl": 4002.5769818275585, | |
| "learning_rate": 5.271487265090163e-08, | |
| "loss": 4.0026, | |
| "reward": 4.748107491061091, | |
| "reward_std": 1.2591828682525374, | |
| "rewards/concensus_correctness_reward_func": 1.2969374898821115, | |
| "rewards/consensus_reward_func": 1.3125, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.125, | |
| "rewards/question_recreation_reward_func": 0.814795003272593, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.265625, | |
| "rewards/xmlcount_reward_func": 0.9332499988377094, | |
| "step": 16 | |
| }, | |
| { | |
| "completion_length": 169.125, | |
| "epoch": 0.9, | |
| "grad_norm": 37287.53515625, | |
| "kl": 417.28647581877885, | |
| "learning_rate": 1.3545689574841341e-08, | |
| "loss": 0.4173, | |
| "reward": 4.91078519821167, | |
| "reward_std": 1.4972448431071825, | |
| "rewards/concensus_correctness_reward_func": 1.3146874904632568, | |
| "rewards/consensus_reward_func": 1.25, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.125, | |
| "rewards/question_recreation_reward_func": 0.84206647798419, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.328125, | |
| "rewards/xmlcount_reward_func": 1.0509062558412552, | |
| "step": 18 | |
| }, | |
| { | |
| "completion_length": 174.0625, | |
| "epoch": 1.0, | |
| "grad_norm": 79.22607421875, | |
| "kl": 3.6632585607585497, | |
| "learning_rate": 0.0, | |
| "loss": 0.0037, | |
| "reward": 6.3031881004571915, | |
| "reward_std": 0.9906455368909519, | |
| "rewards/concensus_correctness_reward_func": 2.099812502041459, | |
| "rewards/consensus_reward_func": 1.6875, | |
| "rewards/cumulative_reward_2": 0.0, | |
| "rewards/final_correctness_reward_func": 0.3125, | |
| "rewards/question_recreation_reward_func": 0.8537818752229214, | |
| "rewards/soft_format_reward_func": 0.0, | |
| "rewards/strict_format_reward_func": 0.296875, | |
| "rewards/xmlcount_reward_func": 1.0527187511324883, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 20, | |
| "total_flos": 0.0, | |
| "train_loss": 2931.9222262827448, | |
| "train_runtime": 158.7363, | |
| "train_samples_per_second": 2.016, | |
| "train_steps_per_second": 0.126 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 20, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "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 | |
| } | |