| { | |
| "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": 0.28995242714881897, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": 0.15359075367450714, | |
| "logits/rejected": 0.28793957829475403, | |
| "logps/chosen": -134.6827850341797, | |
| "logps/ref_chosen": -134.9701690673828, | |
| "logps/ref_rejected": -165.28143310546875, | |
| "logps/rejected": -164.89645385742188, | |
| "loss": 0.6936, | |
| "rewards/accuracies": 0.4140625, | |
| "rewards/chosen": 0.0028739594854414463, | |
| "rewards/margins": -0.0009759241947904229, | |
| "rewards/rejected": 0.003849883796647191, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.1844380403458213, | |
| "grad_norm": 0.3918780982494354, | |
| "learning_rate": 4.849231551964771e-07, | |
| "logits/chosen": 0.10343426465988159, | |
| "logits/rejected": 0.3228362500667572, | |
| "logps/chosen": -140.3193817138672, | |
| "logps/ref_chosen": -140.7550506591797, | |
| "logps/ref_rejected": -185.51889038085938, | |
| "logps/rejected": -185.00494384765625, | |
| "loss": 0.6935, | |
| "rewards/accuracies": 0.4453125, | |
| "rewards/chosen": 0.0043566604144871235, | |
| "rewards/margins": -0.0007828354137018323, | |
| "rewards/rejected": 0.005139495711773634, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.276657060518732, | |
| "grad_norm": 0.4100656509399414, | |
| "learning_rate": 4.415111107797445e-07, | |
| "logits/chosen": 0.15804989635944366, | |
| "logits/rejected": 0.3310811519622803, | |
| "logps/chosen": -141.03822326660156, | |
| "logps/ref_chosen": -141.44863891601562, | |
| "logps/ref_rejected": -178.04934692382812, | |
| "logps/rejected": -177.44610595703125, | |
| "loss": 0.6941, | |
| "rewards/accuracies": 0.3828125, | |
| "rewards/chosen": 0.004104274790734053, | |
| "rewards/margins": -0.0019280731212347746, | |
| "rewards/rejected": 0.006032348144799471, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.3688760806916426, | |
| "grad_norm": 0.35813120007514954, | |
| "learning_rate": 3.75e-07, | |
| "logits/chosen": 0.15596161782741547, | |
| "logits/rejected": 0.2932696044445038, | |
| "logps/chosen": -138.91563415527344, | |
| "logps/ref_chosen": -139.12986755371094, | |
| "logps/ref_rejected": -170.19546508789062, | |
| "logps/rejected": -169.88096618652344, | |
| "loss": 0.6937, | |
| "rewards/accuracies": 0.4453125, | |
| "rewards/chosen": 0.0021424877922981977, | |
| "rewards/margins": -0.0010024935472756624, | |
| "rewards/rejected": 0.00314498133957386, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.4610951008645533, | |
| "grad_norm": 0.3679925203323364, | |
| "learning_rate": 2.934120444167326e-07, | |
| "logits/chosen": 0.1751195341348648, | |
| "logits/rejected": 0.2788362205028534, | |
| "logps/chosen": -135.33677673339844, | |
| "logps/ref_chosen": -135.38316345214844, | |
| "logps/ref_rejected": -162.080322265625, | |
| "logps/rejected": -161.86647033691406, | |
| "loss": 0.694, | |
| "rewards/accuracies": 0.3984375, | |
| "rewards/chosen": 0.0004639970720745623, | |
| "rewards/margins": -0.0016744149615988135, | |
| "rewards/rejected": 0.002138411859050393, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.553314121037464, | |
| "grad_norm": 0.33639660477638245, | |
| "learning_rate": 2.065879555832674e-07, | |
| "logits/chosen": 0.13116978108882904, | |
| "logits/rejected": 0.315093994140625, | |
| "logps/chosen": -134.42941284179688, | |
| "logps/ref_chosen": -134.54176330566406, | |
| "logps/ref_rejected": -169.29995727539062, | |
| "logps/rejected": -169.15988159179688, | |
| "loss": 0.6933, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": 0.0011235313722863793, | |
| "rewards/margins": -0.00027740595396608114, | |
| "rewards/rejected": 0.0014009373262524605, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.6455331412103746, | |
| "grad_norm": 0.39756712317466736, | |
| "learning_rate": 1.2500000000000005e-07, | |
| "logits/chosen": 0.06657359004020691, | |
| "logits/rejected": 0.2838346064090729, | |
| "logps/chosen": -139.63577270507812, | |
| "logps/ref_chosen": -139.84368896484375, | |
| "logps/ref_rejected": -191.49578857421875, | |
| "logps/rejected": -191.23190307617188, | |
| "loss": 0.6934, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": 0.0020792591385543346, | |
| "rewards/margins": -0.0005596381961368024, | |
| "rewards/rejected": 0.0026388971600681543, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.7377521613832853, | |
| "grad_norm": 0.3440794348716736, | |
| "learning_rate": 5.848888922025552e-08, | |
| "logits/chosen": 0.22174814343452454, | |
| "logits/rejected": 0.31970879435539246, | |
| "logps/chosen": -138.87844848632812, | |
| "logps/ref_chosen": -139.12380981445312, | |
| "logps/ref_rejected": -156.94317626953125, | |
| "logps/rejected": -156.6990203857422, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.0024535127449780703, | |
| "rewards/margins": 1.1861870007123798e-05, | |
| "rewards/rejected": 0.0024416507221758366, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.829971181556196, | |
| "grad_norm": 0.4615296721458435, | |
| "learning_rate": 1.507684480352292e-08, | |
| "logits/chosen": 0.17079558968544006, | |
| "logits/rejected": 0.3105747401714325, | |
| "logps/chosen": -141.0509490966797, | |
| "logps/ref_chosen": -141.11288452148438, | |
| "logps/ref_rejected": -176.02235412597656, | |
| "logps/rejected": -175.96676635742188, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.0006195049500092864, | |
| "rewards/margins": 6.36618715361692e-05, | |
| "rewards/rejected": 0.0005558431148529053, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.9221902017291066, | |
| "grad_norm": 0.30170878767967224, | |
| "learning_rate": 0.0, | |
| "logits/chosen": 0.11142128705978394, | |
| "logits/rejected": 0.29976457357406616, | |
| "logps/chosen": -134.70315551757812, | |
| "logps/ref_chosen": -134.6975555419922, | |
| "logps/ref_rejected": -171.15438842773438, | |
| "logps/rejected": -171.20152282714844, | |
| "loss": 0.6929, | |
| "rewards/accuracies": 0.515625, | |
| "rewards/chosen": -5.5965225328691304e-05, | |
| "rewards/margins": 0.00041539373341947794, | |
| "rewards/rejected": -0.0004713588859885931, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.9221902017291066, | |
| "step": 10, | |
| "total_flos": 0.0, | |
| "train_loss": 0.6934871196746826, | |
| "train_runtime": 628.3288, | |
| "train_samples_per_second": 2.207, | |
| "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 | |
| } | |