| { | |
| "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": [ | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 594.6719055175781, | |
| "epoch": 0.017057569296375266, | |
| "grad_norm": 0.38836222887039185, | |
| "kl": 0.0, | |
| "learning_rate": 5e-07, | |
| "loss": -0.0015, | |
| "reward": 0.5937500223517418, | |
| "reward_std": 0.36258383840322495, | |
| "rewards/accuracy_reward": 0.5937500223517418, | |
| "rewards/format_reward": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 611.8069486618042, | |
| "epoch": 0.08528784648187633, | |
| "grad_norm": 0.5523127913475037, | |
| "kl": 0.00023105740547180176, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.022, | |
| "reward": 0.6001674383878708, | |
| "reward_std": 0.35108680790290236, | |
| "rewards/accuracy_reward": 0.6001674383878708, | |
| "rewards/format_reward": 0.0, | |
| "step": 5 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 613.4656517028809, | |
| "epoch": 0.17057569296375266, | |
| "grad_norm": 0.22458800673484802, | |
| "kl": 0.053302574157714847, | |
| "learning_rate": 2.956412726139078e-06, | |
| "loss": 0.0567, | |
| "reward": 0.6870536044239998, | |
| "reward_std": 0.29336604475975037, | |
| "rewards/accuracy_reward": 0.6870536044239998, | |
| "rewards/format_reward": 0.0, | |
| "step": 10 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 601.2154266357422, | |
| "epoch": 0.255863539445629, | |
| "grad_norm": 0.2508162558078766, | |
| "kl": 0.007234764099121094, | |
| "learning_rate": 2.7836719084521715e-06, | |
| "loss": 0.0668, | |
| "reward": 0.7562500417232514, | |
| "reward_std": 0.2151686219498515, | |
| "rewards/accuracy_reward": 0.7562500417232514, | |
| "rewards/format_reward": 0.0, | |
| "step": 15 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 584.3326148986816, | |
| "epoch": 0.3411513859275053, | |
| "grad_norm": 0.4706142842769623, | |
| "kl": 0.12990455627441405, | |
| "learning_rate": 2.4946839873611927e-06, | |
| "loss": 0.0528, | |
| "reward": 0.7720982491970062, | |
| "reward_std": 0.18607007805258036, | |
| "rewards/accuracy_reward": 0.7720982491970062, | |
| "rewards/format_reward": 0.0, | |
| "step": 20 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 597.6643104553223, | |
| "epoch": 0.42643923240938164, | |
| "grad_norm": 0.12214601784944534, | |
| "kl": 0.004122734069824219, | |
| "learning_rate": 2.1156192081791355e-06, | |
| "loss": 0.0421, | |
| "reward": 0.7667411029338836, | |
| "reward_std": 0.17663938011974095, | |
| "rewards/accuracy_reward": 0.7667411029338836, | |
| "rewards/format_reward": 0.0, | |
| "step": 25 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 600.2207824707032, | |
| "epoch": 0.511727078891258, | |
| "grad_norm": 0.131424218416214, | |
| "kl": 0.00368804931640625, | |
| "learning_rate": 1.6808050203829845e-06, | |
| "loss": 0.0483, | |
| "reward": 0.760491107404232, | |
| "reward_std": 0.17339718509465457, | |
| "rewards/accuracy_reward": 0.760491107404232, | |
| "rewards/format_reward": 0.0, | |
| "step": 30 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 586.7152046203613, | |
| "epoch": 0.5970149253731343, | |
| "grad_norm": 0.12775097787380219, | |
| "kl": 0.0038127899169921875, | |
| "learning_rate": 1.2296174432791415e-06, | |
| "loss": 0.0301, | |
| "reward": 0.7609375335276127, | |
| "reward_std": 0.16967349722981453, | |
| "rewards/accuracy_reward": 0.7609375335276127, | |
| "rewards/format_reward": 0.0, | |
| "step": 35 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 577.0308288574219, | |
| "epoch": 0.6823027718550106, | |
| "grad_norm": 0.1416534185409546, | |
| "kl": 0.0044384002685546875, | |
| "learning_rate": 8.029152419343472e-07, | |
| "loss": 0.0357, | |
| "reward": 0.7772321820259094, | |
| "reward_std": 0.17331824395805598, | |
| "rewards/accuracy_reward": 0.7772321820259094, | |
| "rewards/format_reward": 0.0, | |
| "step": 40 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 589.5172180175781, | |
| "epoch": 0.767590618336887, | |
| "grad_norm": 0.10860209167003632, | |
| "kl": 0.0040283203125, | |
| "learning_rate": 4.3933982822017883e-07, | |
| "loss": 0.0321, | |
| "reward": 0.755803607404232, | |
| "reward_std": 0.18665391970425843, | |
| "rewards/accuracy_reward": 0.755803607404232, | |
| "rewards/format_reward": 0.0, | |
| "step": 45 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 598.7453392028808, | |
| "epoch": 0.8528784648187633, | |
| "grad_norm": 0.2108219861984253, | |
| "kl": 0.005596160888671875, | |
| "learning_rate": 1.718159615201853e-07, | |
| "loss": 0.0425, | |
| "reward": 0.7546875357627869, | |
| "reward_std": 0.17801200095564126, | |
| "rewards/accuracy_reward": 0.7546875357627869, | |
| "rewards/format_reward": 0.0, | |
| "step": 50 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 585.473462677002, | |
| "epoch": 0.9381663113006397, | |
| "grad_norm": 0.1318359225988388, | |
| "kl": 0.004902267456054687, | |
| "learning_rate": 2.4570139579284723e-08, | |
| "loss": 0.0298, | |
| "reward": 0.7910714641213417, | |
| "reward_std": 0.1715991634875536, | |
| "rewards/accuracy_reward": 0.7910714641213417, | |
| "rewards/format_reward": 0.0, | |
| "step": 55 | |
| }, | |
| { | |
| "clip_ratio": 0.0, | |
| "completion_length": 581.1535606384277, | |
| "epoch": 0.9893390191897654, | |
| "kl": 0.0061969757080078125, | |
| "reward": 0.7686012287934622, | |
| "reward_std": 0.17060346140836677, | |
| "rewards/accuracy_reward": 0.7686012287934622, | |
| "rewards/format_reward": 0.0, | |
| "step": 58, | |
| "total_flos": 0.0, | |
| "train_loss": 0.040922524543992915, | |
| "train_runtime": 12307.4855, | |
| "train_samples_per_second": 0.609, | |
| "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 | |
| } | |