| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.07248346470961312, | |
| "eval_steps": 50, | |
| "global_step": 100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.007248346470961312, | |
| "grad_norm": 0.07235438376665115, | |
| "learning_rate": 4.998766400914329e-06, | |
| "logits/chosen": -1.9785633087158203, | |
| "logits/rejected": -2.5380234718322754, | |
| "logps/chosen": -0.28137442469596863, | |
| "logps/rejected": -0.37793582677841187, | |
| "loss": 7.3908, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.42206162214279175, | |
| "rewards/margins": 0.14484205842018127, | |
| "rewards/rejected": -0.5669037103652954, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.014496692941922623, | |
| "grad_norm": 0.07455909997224808, | |
| "learning_rate": 4.9950668210706795e-06, | |
| "logits/chosen": -2.0576319694519043, | |
| "logits/rejected": -2.499069929122925, | |
| "logps/chosen": -0.2769497036933899, | |
| "logps/rejected": -0.33560773730278015, | |
| "loss": 7.3753, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.41542449593544006, | |
| "rewards/margins": 0.08798708021640778, | |
| "rewards/rejected": -0.503411591053009, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.021745039412883936, | |
| "grad_norm": 0.09561269730329514, | |
| "learning_rate": 4.9889049115077e-06, | |
| "logits/chosen": -2.1018643379211426, | |
| "logits/rejected": -2.377673625946045, | |
| "logps/chosen": -0.2673099935054779, | |
| "logps/rejected": -0.3057115972042084, | |
| "loss": 7.4516, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -0.40096497535705566, | |
| "rewards/margins": 0.05760239437222481, | |
| "rewards/rejected": -0.45856744050979614, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.028993385883845247, | |
| "grad_norm": 0.08169445395469666, | |
| "learning_rate": 4.980286753286196e-06, | |
| "logits/chosen": -2.1627352237701416, | |
| "logits/rejected": -2.3873391151428223, | |
| "logps/chosen": -0.27636194229125977, | |
| "logps/rejected": -0.3703404664993286, | |
| "loss": 7.3888, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.4145428538322449, | |
| "rewards/margins": 0.14096775650978088, | |
| "rewards/rejected": -0.5555106401443481, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03624173235480656, | |
| "grad_norm": 0.08877147734165192, | |
| "learning_rate": 4.9692208514878445e-06, | |
| "logits/chosen": -2.1400368213653564, | |
| "logits/rejected": -2.44627046585083, | |
| "logps/chosen": -0.25384199619293213, | |
| "logps/rejected": -0.309225857257843, | |
| "loss": 7.518, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.3807629942893982, | |
| "rewards/margins": 0.08307582885026932, | |
| "rewards/rejected": -0.4638388752937317, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03624173235480656, | |
| "eval_logits/chosen": -2.1149837970733643, | |
| "eval_logits/rejected": -2.4759538173675537, | |
| "eval_logps/chosen": -0.28288090229034424, | |
| "eval_logps/rejected": -0.3434317111968994, | |
| "eval_loss": 0.9120966196060181, | |
| "eval_rewards/accuracies": 0.5089285969734192, | |
| "eval_rewards/chosen": -0.42432135343551636, | |
| "eval_rewards/margins": 0.09082622081041336, | |
| "eval_rewards/rejected": -0.5151475667953491, | |
| "eval_runtime": 30.2473, | |
| "eval_samples_per_second": 29.49, | |
| "eval_steps_per_second": 3.703, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04349007882576787, | |
| "grad_norm": 0.11774340271949768, | |
| "learning_rate": 4.9557181268217225e-06, | |
| "logits/chosen": -2.0567967891693115, | |
| "logits/rejected": -2.545668125152588, | |
| "logps/chosen": -0.2460743933916092, | |
| "logps/rejected": -0.32390663027763367, | |
| "loss": 7.3929, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.3691115975379944, | |
| "rewards/margins": 0.1167483925819397, | |
| "rewards/rejected": -0.4858599603176117, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.05073842529672919, | |
| "grad_norm": 0.10012238472700119, | |
| "learning_rate": 4.939791904846869e-06, | |
| "logits/chosen": -1.9286606311798096, | |
| "logits/rejected": -2.4620895385742188, | |
| "logps/chosen": -0.25992274284362793, | |
| "logps/rejected": -0.3665553629398346, | |
| "loss": 7.1863, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.3898841440677643, | |
| "rewards/margins": 0.15994893014431, | |
| "rewards/rejected": -0.5498330593109131, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.057986771767690494, | |
| "grad_norm": 0.07387609034776688, | |
| "learning_rate": 4.921457902821578e-06, | |
| "logits/chosen": -2.0778262615203857, | |
| "logits/rejected": -2.4614813327789307, | |
| "logps/chosen": -0.24521782994270325, | |
| "logps/rejected": -0.3450419306755066, | |
| "loss": 7.3662, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.3678267002105713, | |
| "rewards/margins": 0.1497361809015274, | |
| "rewards/rejected": -0.5175628662109375, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0652351182386518, | |
| "grad_norm": 0.09718403220176697, | |
| "learning_rate": 4.900734214192358e-06, | |
| "logits/chosen": -1.998186707496643, | |
| "logits/rejected": -2.4382071495056152, | |
| "logps/chosen": -0.24080593883991241, | |
| "logps/rejected": -0.3280099332332611, | |
| "loss": 7.268, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.36120888590812683, | |
| "rewards/margins": 0.13080602884292603, | |
| "rewards/rejected": -0.49201488494873047, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.07248346470961312, | |
| "grad_norm": 0.08059138059616089, | |
| "learning_rate": 4.8776412907378845e-06, | |
| "logits/chosen": -1.894728422164917, | |
| "logits/rejected": -2.475407838821411, | |
| "logps/chosen": -0.2019994705915451, | |
| "logps/rejected": -0.29733040928840637, | |
| "loss": 7.2544, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.30299919843673706, | |
| "rewards/margins": 0.14299637079238892, | |
| "rewards/rejected": -0.445995569229126, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07248346470961312, | |
| "eval_logits/chosen": -2.1233773231506348, | |
| "eval_logits/rejected": -2.4810428619384766, | |
| "eval_logps/chosen": -0.24400465190410614, | |
| "eval_logps/rejected": -0.3260345160961151, | |
| "eval_loss": 0.8920583724975586, | |
| "eval_rewards/accuracies": 0.5625, | |
| "eval_rewards/chosen": -0.3660070598125458, | |
| "eval_rewards/margins": 0.12304472178220749, | |
| "eval_rewards/rejected": -0.4890517592430115, | |
| "eval_runtime": 30.28, | |
| "eval_samples_per_second": 29.458, | |
| "eval_steps_per_second": 3.699, | |
| "step": 100 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.7012626709099315e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |