| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 100, | |
| "global_step": 96, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010416666666666666, | |
| "grad_norm": 6.584957352606943, | |
| "learning_rate": 5e-08, | |
| "logits/chosen": -2.590585231781006, | |
| "logits/rejected": -2.5664222240448, | |
| "logps/chosen": -80.29847717285156, | |
| "logps/rejected": -53.10200881958008, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.10416666666666667, | |
| "grad_norm": 5.891503663607178, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": -2.556152820587158, | |
| "logits/rejected": -2.5382611751556396, | |
| "logps/chosen": -87.86880493164062, | |
| "logps/rejected": -80.99066162109375, | |
| "loss": 0.6932, | |
| "rewards/accuracies": 0.2013888955116272, | |
| "rewards/chosen": 0.0004636119119822979, | |
| "rewards/margins": -1.479367074352922e-05, | |
| "rewards/rejected": 0.0004784055636264384, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.20833333333333334, | |
| "grad_norm": 5.292395011584372, | |
| "learning_rate": 4.83504027183137e-07, | |
| "logits/chosen": -2.603332996368408, | |
| "logits/rejected": -2.5538673400878906, | |
| "logps/chosen": -102.09828186035156, | |
| "logps/rejected": -88.8565902709961, | |
| "loss": 0.6907, | |
| "rewards/accuracies": 0.33125001192092896, | |
| "rewards/chosen": 0.012526802718639374, | |
| "rewards/margins": 0.002249127021059394, | |
| "rewards/rejected": 0.010277675464749336, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3125, | |
| "grad_norm": 7.28461383816043, | |
| "learning_rate": 4.3619304568594546e-07, | |
| "logits/chosen": -2.4793078899383545, | |
| "logits/rejected": -2.494683027267456, | |
| "logps/chosen": -65.36841583251953, | |
| "logps/rejected": -74.10954284667969, | |
| "loss": 0.6873, | |
| "rewards/accuracies": 0.33125001192092896, | |
| "rewards/chosen": 0.020688403397798538, | |
| "rewards/margins": 0.01252506673336029, | |
| "rewards/rejected": 0.008163334801793098, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.4166666666666667, | |
| "grad_norm": 5.351477963315917, | |
| "learning_rate": 3.643105808261596e-07, | |
| "logits/chosen": -2.5246329307556152, | |
| "logits/rejected": -2.511716365814209, | |
| "logps/chosen": -69.93246459960938, | |
| "logps/rejected": -70.63847351074219, | |
| "loss": 0.6825, | |
| "rewards/accuracies": 0.29374998807907104, | |
| "rewards/chosen": 0.022140705958008766, | |
| "rewards/margins": 0.021392690017819405, | |
| "rewards/rejected": 0.000748016289435327, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5208333333333334, | |
| "grad_norm": 10.304987109787724, | |
| "learning_rate": 2.7734280209446865e-07, | |
| "logits/chosen": -2.4275832176208496, | |
| "logits/rejected": -2.4420082569122314, | |
| "logps/chosen": -48.93372344970703, | |
| "logps/rejected": -58.99285888671875, | |
| "loss": 0.6787, | |
| "rewards/accuracies": 0.23125000298023224, | |
| "rewards/chosen": 0.019673839211463928, | |
| "rewards/margins": 0.020511243492364883, | |
| "rewards/rejected": -0.0008374080061912537, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 8.34908660394481, | |
| "learning_rate": 1.8676665440207977e-07, | |
| "logits/chosen": -2.5127832889556885, | |
| "logits/rejected": -2.4938712120056152, | |
| "logps/chosen": -83.14317321777344, | |
| "logps/rejected": -88.23527526855469, | |
| "loss": 0.6766, | |
| "rewards/accuracies": 0.3125, | |
| "rewards/chosen": -0.04074016213417053, | |
| "rewards/margins": 0.048237670212984085, | |
| "rewards/rejected": -0.08897782862186432, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7291666666666666, | |
| "grad_norm": 8.727443948909048, | |
| "learning_rate": 1.0453527111051183e-07, | |
| "logits/chosen": -2.4711642265319824, | |
| "logits/rejected": -2.4562721252441406, | |
| "logps/chosen": -101.56312561035156, | |
| "logps/rejected": -96.13810729980469, | |
| "loss": 0.6702, | |
| "rewards/accuracies": 0.3375000059604645, | |
| "rewards/chosen": -0.052182745188474655, | |
| "rewards/margins": 0.054697006940841675, | |
| "rewards/rejected": -0.10687975585460663, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 10.576517260450066, | |
| "learning_rate": 4.1500545527530544e-08, | |
| "logits/chosen": -2.450218439102173, | |
| "logits/rejected": -2.391376495361328, | |
| "logps/chosen": -83.28327178955078, | |
| "logps/rejected": -86.51213836669922, | |
| "loss": 0.6626, | |
| "rewards/accuracies": 0.38749998807907104, | |
| "rewards/chosen": 0.023618828505277634, | |
| "rewards/margins": 0.09491557627916336, | |
| "rewards/rejected": -0.07129674404859543, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.9375, | |
| "grad_norm": 9.589939679130225, | |
| "learning_rate": 5.981030611018234e-09, | |
| "logits/chosen": -2.396845817565918, | |
| "logits/rejected": -2.3921046257019043, | |
| "logps/chosen": -53.659645080566406, | |
| "logps/rejected": -68.09242248535156, | |
| "loss": 0.6721, | |
| "rewards/accuracies": 0.2750000059604645, | |
| "rewards/chosen": 0.0032018995843827724, | |
| "rewards/margins": 0.04944448918104172, | |
| "rewards/rejected": -0.04624258726835251, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 96, | |
| "total_flos": 0.0, | |
| "train_loss": 0.0, | |
| "train_runtime": 0.0351, | |
| "train_samples_per_second": 173972.916, | |
| "train_steps_per_second": 2732.112 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 96, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |