| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 4500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0, | |
| "eval_logits/chosen": 2.8232438564300537, | |
| "eval_logits/rejected": 3.1432998180389404, | |
| "eval_logps/chosen": -289.45379638671875, | |
| "eval_logps/rejected": -299.86944580078125, | |
| "eval_loss": 0.6931473612785339, | |
| "eval_rewards/accuracies": 0.0, | |
| "eval_rewards/chosen": 0.0, | |
| "eval_rewards/margins": 0.0, | |
| "eval_rewards/rejected": 0.0, | |
| "eval_runtime": 94.7465, | |
| "eval_samples_per_second": 10.554, | |
| "eval_steps_per_second": 5.277, | |
| "step": 0 | |
| }, | |
| { | |
| "epoch": 0.00022222222222222223, | |
| "grad_norm": 52.75, | |
| "learning_rate": 4.998888888888889e-05, | |
| "logits/chosen": 3.920553207397461, | |
| "logits/rejected": 3.5924313068389893, | |
| "logps/chosen": -188.6703338623047, | |
| "logps/rejected": -166.2074432373047, | |
| "loss": 0.6901, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.006243896670639515, | |
| "rewards/margins": 0.0062103271484375, | |
| "rewards/rejected": -0.012454223819077015, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "logits/chosen": 3.2731025218963623, | |
| "logits/rejected": 3.3286705017089844, | |
| "logps/chosen": -581.079833984375, | |
| "logps/rejected": -1166.0023193359375, | |
| "loss": 0.0054, | |
| "rewards/accuracies": 0.9969939589500427, | |
| "rewards/chosen": -29.18913459777832, | |
| "rewards/margins": 56.91911315917969, | |
| "rewards/rejected": -86.10824584960938, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 1.4551915228366852e-10, | |
| "learning_rate": 3.888888888888889e-05, | |
| "logits/chosen": 3.5261189937591553, | |
| "logits/rejected": 3.5355136394500732, | |
| "logps/chosen": -574.294189453125, | |
| "logps/rejected": -1153.108154296875, | |
| "loss": 0.0012, | |
| "rewards/accuracies": 0.9990000128746033, | |
| "rewards/chosen": -29.01220703125, | |
| "rewards/margins": 57.46413040161133, | |
| "rewards/rejected": -86.47632598876953, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "logits/chosen": 4.643353462219238, | |
| "logits/rejected": 4.771726131439209, | |
| "logps/chosen": -768.5440063476562, | |
| "logps/rejected": -2062.18212890625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -47.54058837890625, | |
| "rewards/margins": 128.3414306640625, | |
| "rewards/rejected": -175.8820343017578, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.777777777777778e-05, | |
| "logits/chosen": 5.105159759521484, | |
| "logits/rejected": 5.223153591156006, | |
| "logps/chosen": -579.4481201171875, | |
| "logps/rejected": -1619.2967529296875, | |
| "loss": 0.0023, | |
| "rewards/accuracies": 0.9990000128746033, | |
| "rewards/chosen": -28.543167114257812, | |
| "rewards/margins": 103.26885223388672, | |
| "rewards/rejected": -131.81201171875, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "logits/chosen": 5.657288074493408, | |
| "logits/rejected": 6.1234517097473145, | |
| "logps/chosen": -626.1488037109375, | |
| "logps/rejected": -2221.31640625, | |
| "loss": 0.0008, | |
| "rewards/accuracies": 0.9990000128746033, | |
| "rewards/chosen": -33.11819839477539, | |
| "rewards/margins": 158.8884735107422, | |
| "rewards/rejected": -192.00665283203125, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "logits/chosen": 5.869004726409912, | |
| "logits/rejected": 6.477242946624756, | |
| "logps/chosen": -766.905029296875, | |
| "logps/rejected": -2810.428955078125, | |
| "loss": 0.0002, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -47.25943374633789, | |
| "rewards/margins": 203.467041015625, | |
| "rewards/rejected": -250.72650146484375, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "logits/chosen": 5.754186630249023, | |
| "logits/rejected": 6.341037750244141, | |
| "logps/chosen": -730.4443359375, | |
| "logps/rejected": -2717.199951171875, | |
| "loss": 0.0014, | |
| "rewards/accuracies": 0.9990000128746033, | |
| "rewards/chosen": -44.389244079589844, | |
| "rewards/margins": 197.8465118408203, | |
| "rewards/rejected": -242.23577880859375, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.555555555555556e-06, | |
| "logits/chosen": 5.7864789962768555, | |
| "logits/rejected": 6.368413925170898, | |
| "logps/chosen": -765.2401733398438, | |
| "logps/rejected": -2764.18994140625, | |
| "loss": 0.0004, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -47.166805267333984, | |
| "rewards/margins": 199.2112274169922, | |
| "rewards/rejected": -246.37806701660156, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0, | |
| "logits/chosen": 5.893672943115234, | |
| "logits/rejected": 6.538663864135742, | |
| "logps/chosen": -827.6431884765625, | |
| "logps/rejected": -3085.54248046875, | |
| "loss": 0.0007, | |
| "rewards/accuracies": 0.9990000128746033, | |
| "rewards/chosen": -53.547733306884766, | |
| "rewards/margins": 225.18955993652344, | |
| "rewards/rejected": -278.7373046875, | |
| "step": 4500 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "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 | |
| } | |