| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 1.0, |
| "eval_steps": 500, |
| "global_step": 114, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.08839779005524862, |
| "grad_norm": 13.0625, |
| "learning_rate": 4.605263157894737e-06, |
| "logits/chosen": -3.266218900680542, |
| "logits/rejected": -3.173306703567505, |
| "logps/chosen": -169.16183471679688, |
| "logps/rejected": -190.1395263671875, |
| "loss": 0.6924, |
| "rewards/accuracies": 0.48124998807907104, |
| "rewards/chosen": 0.0008526706951670349, |
| "rewards/margins": 0.0017210483783856034, |
| "rewards/rejected": -0.0008683774503879249, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.17679558011049723, |
| "grad_norm": 12.625, |
| "learning_rate": 4.166666666666667e-06, |
| "logits/chosen": -3.263390302658081, |
| "logits/rejected": -3.2149460315704346, |
| "logps/chosen": -162.94313049316406, |
| "logps/rejected": -179.63369750976562, |
| "loss": 0.6946, |
| "rewards/accuracies": 0.4625000059604645, |
| "rewards/chosen": -0.0010867499513551593, |
| "rewards/margins": -0.002635319484397769, |
| "rewards/rejected": 0.0015485694166272879, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.26519337016574585, |
| "grad_norm": 13.625, |
| "learning_rate": 3.728070175438597e-06, |
| "logits/chosen": -3.299316883087158, |
| "logits/rejected": -3.243040084838867, |
| "logps/chosen": -168.03802490234375, |
| "logps/rejected": -191.59165954589844, |
| "loss": 0.6908, |
| "rewards/accuracies": 0.5562499761581421, |
| "rewards/chosen": 0.005441132001578808, |
| "rewards/margins": 0.004950051195919514, |
| "rewards/rejected": 0.0004910803982056677, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.35359116022099446, |
| "grad_norm": 12.8125, |
| "learning_rate": 3.289473684210527e-06, |
| "logits/chosen": -3.3205394744873047, |
| "logits/rejected": -3.2248268127441406, |
| "logps/chosen": -163.4181365966797, |
| "logps/rejected": -190.28494262695312, |
| "loss": 0.692, |
| "rewards/accuracies": 0.4937500059604645, |
| "rewards/chosen": 0.0033660412300378084, |
| "rewards/margins": 0.0026486157439649105, |
| "rewards/rejected": 0.0007174253696575761, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.4419889502762431, |
| "grad_norm": 13.4375, |
| "learning_rate": 2.8508771929824565e-06, |
| "logits/chosen": -3.266845226287842, |
| "logits/rejected": -3.239501953125, |
| "logps/chosen": -165.8727569580078, |
| "logps/rejected": -180.970703125, |
| "loss": 0.69, |
| "rewards/accuracies": 0.5562499761581421, |
| "rewards/chosen": 0.002517760032787919, |
| "rewards/margins": 0.006597781088203192, |
| "rewards/rejected": -0.004080021288245916, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.5303867403314917, |
| "grad_norm": 13.0, |
| "learning_rate": 2.412280701754386e-06, |
| "logits/chosen": -3.300370454788208, |
| "logits/rejected": -3.2336509227752686, |
| "logps/chosen": -166.43919372558594, |
| "logps/rejected": -177.72369384765625, |
| "loss": 0.6912, |
| "rewards/accuracies": 0.5249999761581421, |
| "rewards/chosen": 0.0010731505462899804, |
| "rewards/margins": 0.004260644782334566, |
| "rewards/rejected": -0.0031874938867986202, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.6187845303867403, |
| "grad_norm": 13.1875, |
| "learning_rate": 1.973684210526316e-06, |
| "logits/chosen": -3.292241334915161, |
| "logits/rejected": -3.263917922973633, |
| "logps/chosen": -166.0662841796875, |
| "logps/rejected": -179.9429168701172, |
| "loss": 0.6886, |
| "rewards/accuracies": 0.550000011920929, |
| "rewards/chosen": 0.004152910318225622, |
| "rewards/margins": 0.009452776983380318, |
| "rewards/rejected": -0.005299866199493408, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.7071823204419889, |
| "grad_norm": 13.1875, |
| "learning_rate": 1.5350877192982458e-06, |
| "logits/chosen": -3.2774970531463623, |
| "logits/rejected": -3.2063546180725098, |
| "logps/chosen": -166.04795837402344, |
| "logps/rejected": -189.38204956054688, |
| "loss": 0.6917, |
| "rewards/accuracies": 0.574999988079071, |
| "rewards/chosen": 0.002573251724243164, |
| "rewards/margins": 0.00316311651840806, |
| "rewards/rejected": -0.0005898644449189305, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.7955801104972375, |
| "grad_norm": 11.75, |
| "learning_rate": 1.0964912280701756e-06, |
| "logits/chosen": -3.3311314582824707, |
| "logits/rejected": -3.2796638011932373, |
| "logps/chosen": -157.93948364257812, |
| "logps/rejected": -183.2423095703125, |
| "loss": 0.6909, |
| "rewards/accuracies": 0.5874999761581421, |
| "rewards/chosen": 0.002239528112113476, |
| "rewards/margins": 0.00483693415299058, |
| "rewards/rejected": -0.002597406040877104, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.8839779005524862, |
| "grad_norm": 14.75, |
| "learning_rate": 6.578947368421053e-07, |
| "logits/chosen": -3.2761390209198, |
| "logits/rejected": -3.2539830207824707, |
| "logps/chosen": -178.083984375, |
| "logps/rejected": -185.00401306152344, |
| "loss": 0.6913, |
| "rewards/accuracies": 0.5625, |
| "rewards/chosen": 0.001332092098891735, |
| "rewards/margins": 0.004107826389372349, |
| "rewards/rejected": -0.0027757335919886827, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.9723756906077348, |
| "grad_norm": 13.3125, |
| "learning_rate": 2.192982456140351e-07, |
| "logits/chosen": -3.2679781913757324, |
| "logits/rejected": -3.1474175453186035, |
| "logps/chosen": -170.65786743164062, |
| "logps/rejected": -183.07583618164062, |
| "loss": 0.692, |
| "rewards/accuracies": 0.550000011920929, |
| "rewards/chosen": 0.001837458461523056, |
| "rewards/margins": 0.0027389838360249996, |
| "rewards/rejected": -0.0009015247924253345, |
| "step": 110 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 114, |
| "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": 2, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
