| { | |
| "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": 614.0942222595215, | |
| "epoch": 0.08528784648187633, | |
| "grad_norm": 39.48369216918945, | |
| "kl": 0.002408742904663086, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.0001, | |
| "reward": 0.6303571693599224, | |
| "reward_std": 0.3278109859675169, | |
| "rewards/accuracy_reward": 0.6299107424914837, | |
| "rewards/format_reward": 0.00044642859138548373, | |
| "step": 5 | |
| }, | |
| { | |
| "completion_length": 618.1955627441406, | |
| "epoch": 0.17057569296375266, | |
| "grad_norm": 0.26418083906173706, | |
| "kl": 0.0033366203308105467, | |
| "learning_rate": 2.956412726139078e-06, | |
| "loss": 0.0001, | |
| "reward": 0.6823660999536514, | |
| "reward_std": 0.2967432256788015, | |
| "rewards/accuracy_reward": 0.6821428865194321, | |
| "rewards/format_reward": 0.00022321429569274187, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 615.4616325378418, | |
| "epoch": 0.255863539445629, | |
| "grad_norm": 0.29952046275138855, | |
| "kl": 0.004865837097167969, | |
| "learning_rate": 2.7836719084521715e-06, | |
| "loss": 0.0002, | |
| "reward": 0.7544643208384514, | |
| "reward_std": 0.22131893783807755, | |
| "rewards/accuracy_reward": 0.7544643208384514, | |
| "rewards/format_reward": 0.0, | |
| "step": 15 | |
| }, | |
| { | |
| "completion_length": 594.9377471923829, | |
| "epoch": 0.3411513859275053, | |
| "grad_norm": 0.26904380321502686, | |
| "kl": 0.0036653518676757813, | |
| "learning_rate": 2.4946839873611927e-06, | |
| "loss": 0.0001, | |
| "reward": 0.7558036029338837, | |
| "reward_std": 0.20889290906488894, | |
| "rewards/accuracy_reward": 0.7558036029338837, | |
| "rewards/format_reward": 0.0, | |
| "step": 20 | |
| }, | |
| { | |
| "completion_length": 605.4913223266601, | |
| "epoch": 0.42643923240938164, | |
| "grad_norm": 0.12021861970424652, | |
| "kl": 0.003992271423339844, | |
| "learning_rate": 2.1156192081791355e-06, | |
| "loss": 0.0002, | |
| "reward": 0.7569196805357933, | |
| "reward_std": 0.19887337032705546, | |
| "rewards/accuracy_reward": 0.7569196805357933, | |
| "rewards/format_reward": 0.0, | |
| "step": 25 | |
| }, | |
| { | |
| "completion_length": 609.8823944091797, | |
| "epoch": 0.511727078891258, | |
| "grad_norm": 0.09571157395839691, | |
| "kl": 0.0034656524658203125, | |
| "learning_rate": 1.6808050203829845e-06, | |
| "loss": 0.0001, | |
| "reward": 0.7562500357627868, | |
| "reward_std": 0.1893269034102559, | |
| "rewards/accuracy_reward": 0.7562500357627868, | |
| "rewards/format_reward": 0.0, | |
| "step": 30 | |
| }, | |
| { | |
| "completion_length": 594.9317237854004, | |
| "epoch": 0.5970149253731343, | |
| "grad_norm": 0.17406129837036133, | |
| "kl": 0.01944389343261719, | |
| "learning_rate": 1.2296174432791415e-06, | |
| "loss": 0.0008, | |
| "reward": 0.7457589656114578, | |
| "reward_std": 0.1772445771843195, | |
| "rewards/accuracy_reward": 0.7457589656114578, | |
| "rewards/format_reward": 0.0, | |
| "step": 35 | |
| }, | |
| { | |
| "completion_length": 587.7542655944824, | |
| "epoch": 0.6823027718550106, | |
| "grad_norm": 0.3893604278564453, | |
| "kl": 0.0033367156982421877, | |
| "learning_rate": 8.029152419343472e-07, | |
| "loss": 0.0001, | |
| "reward": 0.7665178909897804, | |
| "reward_std": 0.17746288534253835, | |
| "rewards/accuracy_reward": 0.7665178909897804, | |
| "rewards/format_reward": 0.0, | |
| "step": 40 | |
| }, | |
| { | |
| "completion_length": 608.1580612182618, | |
| "epoch": 0.767590618336887, | |
| "grad_norm": 0.13640980422496796, | |
| "kl": 0.0030500411987304686, | |
| "learning_rate": 4.3933982822017883e-07, | |
| "loss": 0.0001, | |
| "reward": 0.7549107506871223, | |
| "reward_std": 0.18457430368289351, | |
| "rewards/accuracy_reward": 0.7549107506871223, | |
| "rewards/format_reward": 0.0, | |
| "step": 45 | |
| }, | |
| { | |
| "completion_length": 602.2951156616211, | |
| "epoch": 0.8528784648187633, | |
| "grad_norm": 0.217549666762352, | |
| "kl": 0.003106689453125, | |
| "learning_rate": 1.718159615201853e-07, | |
| "loss": 0.0001, | |
| "reward": 0.7488839671015739, | |
| "reward_std": 0.18157500196248294, | |
| "rewards/accuracy_reward": 0.7488839671015739, | |
| "rewards/format_reward": 0.0, | |
| "step": 50 | |
| }, | |
| { | |
| "completion_length": 599.9350708007812, | |
| "epoch": 0.9381663113006397, | |
| "grad_norm": 0.23885449767112732, | |
| "kl": 0.0033361434936523436, | |
| "learning_rate": 2.4570139579284723e-08, | |
| "loss": 0.0001, | |
| "reward": 0.776116107404232, | |
| "reward_std": 0.19774878825992345, | |
| "rewards/accuracy_reward": 0.776116107404232, | |
| "rewards/format_reward": 0.0, | |
| "step": 55 | |
| }, | |
| { | |
| "completion_length": 583.1644630432129, | |
| "epoch": 0.9893390191897654, | |
| "kl": 0.0035022099812825522, | |
| "reward": 0.7760417039195696, | |
| "reward_std": 0.16163485000530878, | |
| "rewards/accuracy_reward": 0.7760417039195696, | |
| "rewards/format_reward": 0.0, | |
| "step": 58, | |
| "total_flos": 0.0, | |
| "train_loss": -0.00014772719968559928, | |
| "train_runtime": 12261.9552, | |
| "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 | |
| } | |