| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9893390191897654, | |
| "eval_steps": 100, | |
| "global_step": 58, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "completion_length": 606.2089546203613, | |
| "epoch": 0.08528784648187633, | |
| "grad_norm": 0.29036083817481995, | |
| "kl": 0.00022020339965820313, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.0464, | |
| "reward": 0.6397321701049805, | |
| "reward_std": 0.3235082272440195, | |
| "rewards/accuracy_reward": 0.6390625275671482, | |
| "rewards/format_reward": 0.0006696428870782256, | |
| "step": 5 | |
| }, | |
| { | |
| "completion_length": 612.3254737854004, | |
| "epoch": 0.17057569296375266, | |
| "grad_norm": 2.9889135360717773, | |
| "kl": 0.7559915542602539, | |
| "learning_rate": 2.956412726139078e-06, | |
| "loss": 0.111, | |
| "reward": 0.696651816368103, | |
| "reward_std": 0.2815938711166382, | |
| "rewards/accuracy_reward": 0.6962053865194321, | |
| "rewards/format_reward": 0.00044642859138548373, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 588.2087371826171, | |
| "epoch": 0.255863539445629, | |
| "grad_norm": 58.945621490478516, | |
| "kl": 0.013921737670898438, | |
| "learning_rate": 2.7836719084521715e-06, | |
| "loss": 0.0668, | |
| "reward": 0.7671875312924386, | |
| "reward_std": 0.21175616141408682, | |
| "rewards/accuracy_reward": 0.7671875312924386, | |
| "rewards/format_reward": 0.0, | |
| "step": 15 | |
| }, | |
| { | |
| "completion_length": 577.985066986084, | |
| "epoch": 0.3411513859275053, | |
| "grad_norm": 8.070592880249023, | |
| "kl": 23.920117950439455, | |
| "learning_rate": 2.4946839873611927e-06, | |
| "loss": 2.1306, | |
| "reward": 0.7787946745753288, | |
| "reward_std": 0.18835734333842993, | |
| "rewards/accuracy_reward": 0.7787946745753288, | |
| "rewards/format_reward": 0.0, | |
| "step": 20 | |
| }, | |
| { | |
| "completion_length": 589.9038261413574, | |
| "epoch": 0.42643923240938164, | |
| "grad_norm": 12.532156944274902, | |
| "kl": 0.11749801635742188, | |
| "learning_rate": 2.1156192081791355e-06, | |
| "loss": 0.0537, | |
| "reward": 0.7645089641213417, | |
| "reward_std": 0.1823441507294774, | |
| "rewards/accuracy_reward": 0.7645089641213417, | |
| "rewards/format_reward": 0.0, | |
| "step": 25 | |
| }, | |
| { | |
| "completion_length": 597.8422134399414, | |
| "epoch": 0.511727078891258, | |
| "grad_norm": 338.9886169433594, | |
| "kl": 0.22518844604492189, | |
| "learning_rate": 1.6808050203829845e-06, | |
| "loss": 0.0479, | |
| "reward": 0.7513393253087998, | |
| "reward_std": 0.17386628594249487, | |
| "rewards/accuracy_reward": 0.7513393253087998, | |
| "rewards/format_reward": 0.0, | |
| "step": 30 | |
| }, | |
| { | |
| "completion_length": 591.2631973266601, | |
| "epoch": 0.5970149253731343, | |
| "grad_norm": 10.277711868286133, | |
| "kl": 0.11962966918945313, | |
| "learning_rate": 1.2296174432791415e-06, | |
| "loss": 0.0388, | |
| "reward": 0.744419677555561, | |
| "reward_std": 0.16556358551606537, | |
| "rewards/accuracy_reward": 0.744419677555561, | |
| "rewards/format_reward": 0.0, | |
| "step": 35 | |
| }, | |
| { | |
| "completion_length": 582.1995803833008, | |
| "epoch": 0.6823027718550106, | |
| "grad_norm": 2.5935020446777344, | |
| "kl": 0.018338775634765624, | |
| "learning_rate": 8.029152419343472e-07, | |
| "loss": 0.0428, | |
| "reward": 0.7609375298023224, | |
| "reward_std": 0.181213954789564, | |
| "rewards/accuracy_reward": 0.7609375298023224, | |
| "rewards/format_reward": 0.0, | |
| "step": 40 | |
| }, | |
| { | |
| "completion_length": 602.1290435791016, | |
| "epoch": 0.767590618336887, | |
| "grad_norm": 14.890217781066895, | |
| "kl": 5.168456268310547, | |
| "learning_rate": 4.3933982822017883e-07, | |
| "loss": 0.3779, | |
| "reward": 0.7399553924798965, | |
| "reward_std": 0.19358455892652274, | |
| "rewards/accuracy_reward": 0.7399553924798965, | |
| "rewards/format_reward": 0.0, | |
| "step": 45 | |
| }, | |
| { | |
| "completion_length": 598.5703392028809, | |
| "epoch": 0.8528784648187633, | |
| "grad_norm": 5.467597007751465, | |
| "kl": 0.12263717651367187, | |
| "learning_rate": 1.718159615201853e-07, | |
| "loss": 0.0443, | |
| "reward": 0.7386161044239998, | |
| "reward_std": 0.19092278694733977, | |
| "rewards/accuracy_reward": 0.7386161044239998, | |
| "rewards/format_reward": 0.0, | |
| "step": 50 | |
| }, | |
| { | |
| "completion_length": 591.5207893371582, | |
| "epoch": 0.9381663113006397, | |
| "grad_norm": 74.10267639160156, | |
| "kl": 0.07649765014648438, | |
| "learning_rate": 2.4570139579284723e-08, | |
| "loss": 0.0465, | |
| "reward": 0.7707589626312256, | |
| "reward_std": 0.1925399899482727, | |
| "rewards/accuracy_reward": 0.7707589626312256, | |
| "rewards/format_reward": 0.0, | |
| "step": 55 | |
| }, | |
| { | |
| "completion_length": 587.6321029663086, | |
| "epoch": 0.9893390191897654, | |
| "kl": 0.052463531494140625, | |
| "reward": 0.7678571715950966, | |
| "reward_std": 0.17418795668830475, | |
| "rewards/accuracy_reward": 0.7678571715950966, | |
| "rewards/format_reward": 0.0, | |
| "step": 58, | |
| "total_flos": 0.0, | |
| "train_loss": 0.25991881401117506, | |
| "train_runtime": 12262.8912, | |
| "train_samples_per_second": 0.612, | |
| "train_steps_per_second": 0.005 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 58, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |