| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9221902017291066, | |
| "eval_steps": 500, | |
| "global_step": 10, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09221902017291066, | |
| "grad_norm": 0.3389199674129486, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": 0.14791223406791687, | |
| "logits/rejected": 0.28315749764442444, | |
| "logps/chosen": -134.61978149414062, | |
| "logps/ref_chosen": -134.61978149414062, | |
| "logps/ref_rejected": -173.30970764160156, | |
| "logps/rejected": -173.30970764160156, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.1844380403458213, | |
| "grad_norm": 0.2829269468784332, | |
| "learning_rate": 4.849231551964771e-07, | |
| "logits/chosen": 0.14062762260437012, | |
| "logits/rejected": 0.2469622939825058, | |
| "logps/chosen": -130.12448120117188, | |
| "logps/ref_chosen": -130.12448120117188, | |
| "logps/ref_rejected": -163.5547637939453, | |
| "logps/rejected": -163.5547637939453, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.276657060518732, | |
| "grad_norm": 0.32888761162757874, | |
| "learning_rate": 4.415111107797445e-07, | |
| "logits/chosen": 0.06239762529730797, | |
| "logits/rejected": 0.24671903252601624, | |
| "logps/chosen": -129.64842224121094, | |
| "logps/ref_chosen": -129.6482391357422, | |
| "logps/ref_rejected": -167.99598693847656, | |
| "logps/rejected": -167.98696899414062, | |
| "loss": 0.6932, | |
| "rewards/accuracies": 0.4921875, | |
| "rewards/chosen": -1.7636848497204483e-06, | |
| "rewards/margins": -9.190676792059094e-05, | |
| "rewards/rejected": 9.01430903468281e-05, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.3688760806916426, | |
| "grad_norm": 0.39236217737197876, | |
| "learning_rate": 3.75e-07, | |
| "logits/chosen": 0.11659017950296402, | |
| "logits/rejected": 0.26041609048843384, | |
| "logps/chosen": -132.6394500732422, | |
| "logps/ref_chosen": -132.77403259277344, | |
| "logps/ref_rejected": -168.05410766601562, | |
| "logps/rejected": -167.96951293945312, | |
| "loss": 0.6929, | |
| "rewards/accuracies": 0.5546875, | |
| "rewards/chosen": 0.0013458358589559793, | |
| "rewards/margins": 0.0004997455980628729, | |
| "rewards/rejected": 0.0008460902608931065, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.4610951008645533, | |
| "grad_norm": 0.41242164373397827, | |
| "learning_rate": 2.934120444167326e-07, | |
| "logits/chosen": 0.13683654367923737, | |
| "logits/rejected": 0.28962600231170654, | |
| "logps/chosen": -135.99209594726562, | |
| "logps/ref_chosen": -135.82839965820312, | |
| "logps/ref_rejected": -161.7675323486328, | |
| "logps/rejected": -162.00277709960938, | |
| "loss": 0.6928, | |
| "rewards/accuracies": 0.5546875, | |
| "rewards/chosen": -0.0016370582161471248, | |
| "rewards/margins": 0.0007155549246817827, | |
| "rewards/rejected": -0.0023526130244135857, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.553314121037464, | |
| "grad_norm": 0.37831488251686096, | |
| "learning_rate": 2.065879555832674e-07, | |
| "logits/chosen": 0.09734513610601425, | |
| "logits/rejected": 0.20619672536849976, | |
| "logps/chosen": -137.687744140625, | |
| "logps/ref_chosen": -137.62083435058594, | |
| "logps/ref_rejected": -165.1855926513672, | |
| "logps/rejected": -165.27906799316406, | |
| "loss": 0.693, | |
| "rewards/accuracies": 0.5390625, | |
| "rewards/chosen": -0.0006691145244985819, | |
| "rewards/margins": 0.0002655756543390453, | |
| "rewards/rejected": -0.0009346901206299663, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.6455331412103746, | |
| "grad_norm": 0.44487157464027405, | |
| "learning_rate": 1.2500000000000005e-07, | |
| "logits/chosen": 0.1292731761932373, | |
| "logits/rejected": 0.26378369331359863, | |
| "logps/chosen": -140.36293029785156, | |
| "logps/ref_chosen": -140.14544677734375, | |
| "logps/ref_rejected": -173.2746124267578, | |
| "logps/rejected": -173.66172790527344, | |
| "loss": 0.6923, | |
| "rewards/accuracies": 0.5859375, | |
| "rewards/chosen": -0.0021748137660324574, | |
| "rewards/margins": 0.0016963660018518567, | |
| "rewards/rejected": -0.0038711796514689922, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.7377521613832853, | |
| "grad_norm": 0.3908422887325287, | |
| "learning_rate": 5.848888922025552e-08, | |
| "logits/chosen": 0.08686134964227676, | |
| "logits/rejected": 0.25962409377098083, | |
| "logps/chosen": -132.61734008789062, | |
| "logps/ref_chosen": -132.29315185546875, | |
| "logps/ref_rejected": -175.901611328125, | |
| "logps/rejected": -176.36581420898438, | |
| "loss": 0.6925, | |
| "rewards/accuracies": 0.6015625, | |
| "rewards/chosen": -0.0032418149057775736, | |
| "rewards/margins": 0.0014001112431287766, | |
| "rewards/rejected": -0.0046419259160757065, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.829971181556196, | |
| "grad_norm": 0.28832051157951355, | |
| "learning_rate": 1.507684480352292e-08, | |
| "logits/chosen": 0.11615337431430817, | |
| "logits/rejected": 0.2154950648546219, | |
| "logps/chosen": -132.73593139648438, | |
| "logps/ref_chosen": -132.26475524902344, | |
| "logps/ref_rejected": -158.7421112060547, | |
| "logps/rejected": -159.4190673828125, | |
| "loss": 0.6921, | |
| "rewards/accuracies": 0.6171875, | |
| "rewards/chosen": -0.004711654037237167, | |
| "rewards/margins": 0.0020579977426677942, | |
| "rewards/rejected": -0.006769651547074318, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.9221902017291066, | |
| "grad_norm": 0.3114199936389923, | |
| "learning_rate": 0.0, | |
| "logits/chosen": 0.13599035143852234, | |
| "logits/rejected": 0.26238346099853516, | |
| "logps/chosen": -138.9654541015625, | |
| "logps/ref_chosen": -138.58920288085938, | |
| "logps/ref_rejected": -173.31874084472656, | |
| "logps/rejected": -173.79449462890625, | |
| "loss": 0.6927, | |
| "rewards/accuracies": 0.5703125, | |
| "rewards/chosen": -0.0037624253891408443, | |
| "rewards/margins": 0.0009951507672667503, | |
| "rewards/rejected": -0.004757576156407595, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.9221902017291066, | |
| "step": 10, | |
| "total_flos": 0.0, | |
| "train_loss": 0.692774099111557, | |
| "train_runtime": 614.9448, | |
| "train_samples_per_second": 2.255, | |
| "train_steps_per_second": 0.016 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 10, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100000, | |
| "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": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |