| { | |
| "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": 0.3625384569168091, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": 0.12041429430246353, | |
| "logits/rejected": 0.2929443418979645, | |
| "logps/chosen": -134.8220672607422, | |
| "logps/ref_chosen": -135.22447204589844, | |
| "logps/ref_rejected": -175.8645782470703, | |
| "logps/rejected": -175.39471435546875, | |
| "loss": 0.6935, | |
| "rewards/accuracies": 0.4609375, | |
| "rewards/chosen": 0.004024160094559193, | |
| "rewards/margins": -0.000674244191031903, | |
| "rewards/rejected": 0.0046984050422906876, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.1844380403458213, | |
| "grad_norm": 0.43253499269485474, | |
| "learning_rate": 4.849231551964771e-07, | |
| "logits/chosen": 0.18826183676719666, | |
| "logits/rejected": 0.31612664461135864, | |
| "logps/chosen": -134.79527282714844, | |
| "logps/ref_chosen": -135.17901611328125, | |
| "logps/ref_rejected": -167.55538940429688, | |
| "logps/rejected": -167.03347778320312, | |
| "loss": 0.6938, | |
| "rewards/accuracies": 0.3671875, | |
| "rewards/chosen": 0.0038375072181224823, | |
| "rewards/margins": -0.0013817482395097613, | |
| "rewards/rejected": 0.0052192555740475655, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.276657060518732, | |
| "grad_norm": 0.4659630358219147, | |
| "learning_rate": 4.415111107797445e-07, | |
| "logits/chosen": 0.08171764016151428, | |
| "logits/rejected": 0.19102837145328522, | |
| "logps/chosen": -130.4796600341797, | |
| "logps/ref_chosen": -130.83258056640625, | |
| "logps/ref_rejected": -162.6024627685547, | |
| "logps/rejected": -162.10250854492188, | |
| "loss": 0.6939, | |
| "rewards/accuracies": 0.390625, | |
| "rewards/chosen": 0.00352896424010396, | |
| "rewards/margins": -0.001470521092414856, | |
| "rewards/rejected": 0.004999485332518816, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.3688760806916426, | |
| "grad_norm": 0.29478588700294495, | |
| "learning_rate": 3.75e-07, | |
| "logits/chosen": 0.08501344174146652, | |
| "logits/rejected": 0.2843908369541168, | |
| "logps/chosen": -131.01138305664062, | |
| "logps/ref_chosen": -131.26690673828125, | |
| "logps/ref_rejected": -169.983154296875, | |
| "logps/rejected": -169.63084411621094, | |
| "loss": 0.6936, | |
| "rewards/accuracies": 0.4453125, | |
| "rewards/chosen": 0.002555293031036854, | |
| "rewards/margins": -0.0009679353097453713, | |
| "rewards/rejected": 0.0035232282243669033, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.4610951008645533, | |
| "grad_norm": 0.38472220301628113, | |
| "learning_rate": 2.934120444167326e-07, | |
| "logits/chosen": 0.17568331956863403, | |
| "logits/rejected": 0.3496350049972534, | |
| "logps/chosen": -132.62481689453125, | |
| "logps/ref_chosen": -132.73873901367188, | |
| "logps/ref_rejected": -169.95010375976562, | |
| "logps/rejected": -169.81109619140625, | |
| "loss": 0.6933, | |
| "rewards/accuracies": 0.453125, | |
| "rewards/chosen": 0.0011391888838261366, | |
| "rewards/margins": -0.00025093427393585443, | |
| "rewards/rejected": 0.0013901233905926347, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.553314121037464, | |
| "grad_norm": 0.3507808446884155, | |
| "learning_rate": 2.065879555832674e-07, | |
| "logits/chosen": 0.09320589154958725, | |
| "logits/rejected": 0.29253625869750977, | |
| "logps/chosen": -133.10009765625, | |
| "logps/ref_chosen": -133.45578002929688, | |
| "logps/ref_rejected": -173.9835205078125, | |
| "logps/rejected": -173.6336212158203, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.5078125, | |
| "rewards/chosen": 0.003556814743205905, | |
| "rewards/margins": 5.76310558244586e-05, | |
| "rewards/rejected": 0.003499183803796768, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.6455331412103746, | |
| "grad_norm": 0.2825755178928375, | |
| "learning_rate": 1.2500000000000005e-07, | |
| "logits/chosen": 0.16687946021556854, | |
| "logits/rejected": 0.29598817229270935, | |
| "logps/chosen": -139.96351623535156, | |
| "logps/ref_chosen": -140.12921142578125, | |
| "logps/ref_rejected": -171.02523803710938, | |
| "logps/rejected": -170.8452911376953, | |
| "loss": 0.6932, | |
| "rewards/accuracies": 0.4921875, | |
| "rewards/chosen": 0.001656890381127596, | |
| "rewards/margins": -0.00014275184366852045, | |
| "rewards/rejected": 0.0017996423412114382, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.7377521613832853, | |
| "grad_norm": 0.3293827474117279, | |
| "learning_rate": 5.848888922025552e-08, | |
| "logits/chosen": 0.12655451893806458, | |
| "logits/rejected": 0.3025429844856262, | |
| "logps/chosen": -131.79420471191406, | |
| "logps/ref_chosen": -131.82289123535156, | |
| "logps/ref_rejected": -170.08551025390625, | |
| "logps/rejected": -170.10638427734375, | |
| "loss": 0.6929, | |
| "rewards/accuracies": 0.5234375, | |
| "rewards/chosen": 0.00028696414665319026, | |
| "rewards/margins": 0.0004956752527505159, | |
| "rewards/rejected": -0.00020871106244158, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.829971181556196, | |
| "grad_norm": 0.36158400774002075, | |
| "learning_rate": 1.507684480352292e-08, | |
| "logits/chosen": 0.09881015121936798, | |
| "logits/rejected": 0.2709089517593384, | |
| "logps/chosen": -133.0918426513672, | |
| "logps/ref_chosen": -133.32278442382812, | |
| "logps/ref_rejected": -168.0214080810547, | |
| "logps/rejected": -167.9148406982422, | |
| "loss": 0.6925, | |
| "rewards/accuracies": 0.5859375, | |
| "rewards/chosen": 0.002309603150933981, | |
| "rewards/margins": 0.0012440752470865846, | |
| "rewards/rejected": 0.0010655277874320745, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.9221902017291066, | |
| "grad_norm": 0.3515172600746155, | |
| "learning_rate": 0.0, | |
| "logits/chosen": 0.17717128992080688, | |
| "logits/rejected": 0.3308084011077881, | |
| "logps/chosen": -132.75823974609375, | |
| "logps/ref_chosen": -132.620849609375, | |
| "logps/ref_rejected": -157.32113647460938, | |
| "logps/rejected": -157.39456176757812, | |
| "loss": 0.6935, | |
| "rewards/accuracies": 0.3984375, | |
| "rewards/chosen": -0.0013737636618316174, | |
| "rewards/margins": -0.0006395292584784329, | |
| "rewards/rejected": -0.0007342344033531845, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.9221902017291066, | |
| "step": 10, | |
| "total_flos": 0.0, | |
| "train_loss": 0.6933379590511322, | |
| "train_runtime": 621.366, | |
| "train_samples_per_second": 2.232, | |
| "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 | |
| } | |