| { | |
| "epoch": 3.0, | |
| "eval_logits/chosen": -2.3620400428771973, | |
| "eval_logits/rejected": -2.259032964706421, | |
| "eval_logps/chosen": -321.4129333496094, | |
| "eval_logps/rejected": -315.8341064453125, | |
| "eval_loss": 0.7384642958641052, | |
| "eval_rewards/accuracies": 0.7559999823570251, | |
| "eval_rewards/chosen": -4.756579875946045, | |
| "eval_rewards/margins": 3.8600614070892334, | |
| "eval_rewards/rejected": -8.6166410446167, | |
| "eval_runtime": 330.136, | |
| "eval_samples": 2000, | |
| "eval_samples_per_second": 6.058, | |
| "eval_steps_per_second": 0.379, | |
| "train_loss": 0.20145473461829064, | |
| "train_runtime": 66730.9037, | |
| "train_samples": 61966, | |
| "train_samples_per_second": 2.786, | |
| "train_steps_per_second": 0.044 | |
| } |