{ "effective_tokens_per_sec": 6763.601819652164, "epoch": 3.0, "total_flos": 4.2988160857187287e+18, "train_loss": 0.7857236749096433, "train_runtime": 6229.2125, "train_samples_per_second": 9.154, "train_steps_per_second": 0.572 }