| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9182209469153515, | |
| "eval_steps": 500, | |
| "global_step": 10, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09182209469153516, | |
| "grad_norm": 0.48370492458343506, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": 0.16483092308044434, | |
| "logits/rejected": 0.3757104277610779, | |
| "logps/chosen": -129.87265014648438, | |
| "logps/ref_chosen": -130.11143493652344, | |
| "logps/ref_rejected": -170.61505126953125, | |
| "logps/rejected": -170.22250366210938, | |
| "loss": 0.6939, | |
| "rewards/accuracies": 0.3359375, | |
| "rewards/chosen": 0.0023880640510469675, | |
| "rewards/margins": -0.0015374040231108665, | |
| "rewards/rejected": 0.003925468306988478, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.1836441893830703, | |
| "grad_norm": 0.2921633720397949, | |
| "learning_rate": 4.849231551964771e-07, | |
| "logits/chosen": 0.21723544597625732, | |
| "logits/rejected": 0.33153384923934937, | |
| "logps/chosen": -136.4618377685547, | |
| "logps/ref_chosen": -136.66879272460938, | |
| "logps/ref_rejected": -170.20431518554688, | |
| "logps/rejected": -169.85755920410156, | |
| "loss": 0.6939, | |
| "rewards/accuracies": 0.390625, | |
| "rewards/chosen": 0.0020694590639322996, | |
| "rewards/margins": -0.0013981539523229003, | |
| "rewards/rejected": 0.003467612899839878, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.27546628407460544, | |
| "grad_norm": 0.3820483684539795, | |
| "learning_rate": 4.415111107797445e-07, | |
| "logits/chosen": 0.278810977935791, | |
| "logits/rejected": 0.3778008222579956, | |
| "logps/chosen": -140.81326293945312, | |
| "logps/ref_chosen": -141.2205047607422, | |
| "logps/ref_rejected": -163.39132690429688, | |
| "logps/rejected": -162.9018096923828, | |
| "loss": 0.6936, | |
| "rewards/accuracies": 0.453125, | |
| "rewards/chosen": 0.004072446841746569, | |
| "rewards/margins": -0.000822849222458899, | |
| "rewards/rejected": 0.004895296413451433, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.3672883787661406, | |
| "grad_norm": 0.34646955132484436, | |
| "learning_rate": 3.75e-07, | |
| "logits/chosen": 0.14010563492774963, | |
| "logits/rejected": 0.28288155794143677, | |
| "logps/chosen": -140.85638427734375, | |
| "logps/ref_chosen": -141.0355987548828, | |
| "logps/ref_rejected": -166.15609741210938, | |
| "logps/rejected": -165.95912170410156, | |
| "loss": 0.6932, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.0017920829122886062, | |
| "rewards/margins": -0.00017783755902200937, | |
| "rewards/rejected": 0.001969920238479972, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.45911047345767575, | |
| "grad_norm": 0.3133969306945801, | |
| "learning_rate": 2.934120444167326e-07, | |
| "logits/chosen": 0.20951783657073975, | |
| "logits/rejected": 0.33099910616874695, | |
| "logps/chosen": -136.3818817138672, | |
| "logps/ref_chosen": -136.40235900878906, | |
| "logps/ref_rejected": -158.99447631835938, | |
| "logps/rejected": -158.85215759277344, | |
| "loss": 0.6938, | |
| "rewards/accuracies": 0.359375, | |
| "rewards/chosen": 0.00020450774172786623, | |
| "rewards/margins": -0.0012185449013486505, | |
| "rewards/rejected": 0.0014230526285246015, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.5509325681492109, | |
| "grad_norm": 0.413125604391098, | |
| "learning_rate": 2.065879555832674e-07, | |
| "logits/chosen": 0.1270689070224762, | |
| "logits/rejected": 0.35084813833236694, | |
| "logps/chosen": -134.93136596679688, | |
| "logps/ref_chosen": -134.92478942871094, | |
| "logps/ref_rejected": -171.18968200683594, | |
| "logps/rejected": -171.1276092529297, | |
| "loss": 0.6935, | |
| "rewards/accuracies": 0.4765625, | |
| "rewards/chosen": -6.584226503036916e-05, | |
| "rewards/margins": -0.0006865662289783359, | |
| "rewards/rejected": 0.0006207239348441362, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.642754662840746, | |
| "grad_norm": 0.3633699417114258, | |
| "learning_rate": 1.2500000000000005e-07, | |
| "logits/chosen": 0.160421222448349, | |
| "logits/rejected": 0.31845855712890625, | |
| "logps/chosen": -131.51528930664062, | |
| "logps/ref_chosen": -131.45591735839844, | |
| "logps/ref_rejected": -161.3601531982422, | |
| "logps/rejected": -161.34597778320312, | |
| "loss": 0.6935, | |
| "rewards/accuracies": 0.390625, | |
| "rewards/chosen": -0.0005936581292189658, | |
| "rewards/margins": -0.0007353991386480629, | |
| "rewards/rejected": 0.00014174105308484286, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.7345767575322812, | |
| "grad_norm": 0.31316861510276794, | |
| "learning_rate": 5.848888922025552e-08, | |
| "logits/chosen": 0.20960983633995056, | |
| "logits/rejected": 0.41680708527565, | |
| "logps/chosen": -136.77139282226562, | |
| "logps/ref_chosen": -136.7902069091797, | |
| "logps/ref_rejected": -181.39559936523438, | |
| "logps/rejected": -181.3594512939453, | |
| "loss": 0.6932, | |
| "rewards/accuracies": 0.4609375, | |
| "rewards/chosen": 0.00018815101066138595, | |
| "rewards/margins": -0.0001733488024910912, | |
| "rewards/rejected": 0.0003614998422563076, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.8263988522238164, | |
| "grad_norm": 0.33593055605888367, | |
| "learning_rate": 1.507684480352292e-08, | |
| "logits/chosen": 0.21831227838993073, | |
| "logits/rejected": 0.32834041118621826, | |
| "logps/chosen": -137.8075408935547, | |
| "logps/ref_chosen": -137.73573303222656, | |
| "logps/ref_rejected": -171.2981719970703, | |
| "logps/rejected": -171.36036682128906, | |
| "loss": 0.6932, | |
| "rewards/accuracies": 0.484375, | |
| "rewards/chosen": -0.0007180434186011553, | |
| "rewards/margins": -9.611125278752297e-05, | |
| "rewards/rejected": -0.0006219320930540562, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.9182209469153515, | |
| "grad_norm": 0.3731705844402313, | |
| "learning_rate": 0.0, | |
| "logits/chosen": 0.15047258138656616, | |
| "logits/rejected": 0.27864834666252136, | |
| "logps/chosen": -146.82630920410156, | |
| "logps/ref_chosen": -146.76913452148438, | |
| "logps/ref_rejected": -181.7820281982422, | |
| "logps/rejected": -181.79539489746094, | |
| "loss": 0.6934, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.000571864191442728, | |
| "rewards/margins": -0.0004380465252324939, | |
| "rewards/rejected": -0.000133817782625556, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.9182209469153515, | |
| "step": 10, | |
| "total_flos": 0.0, | |
| "train_loss": 0.6935156106948852, | |
| "train_runtime": 611.1413, | |
| "train_samples_per_second": 2.279, | |
| "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 | |
| } | |