{ "effective_tokens_per_sec": 2759.3348239749225, "epoch": 3.0, "total_flos": 2.5683697677065257e+18, "train_loss": 1.0660493817018426, "train_runtime": 18433.9416, "train_samples_per_second": 2.396, "train_steps_per_second": 0.037 }