{ "effective_tokens_per_sec": 7604.438610834273, "epoch": 6.0, "total_flos": 5.392281114922451e+18, "train_loss": 0.8344338661696338, "train_runtime": 6866.9503, "train_samples_per_second": 12.425, "train_steps_per_second": 0.414 }