{ "effective_tokens_per_sec": 6045.912252555761, "epoch": 3.0, "total_flos": 2.578606960937009e+18, "train_loss": 0.8228938954362648, "train_runtime": 8271.5959, "train_samples_per_second": 4.753, "train_steps_per_second": 0.297 }