{ "effective_tokens_per_sec": 7586.702082571366, "epoch": 4.0, "total_flos": 3.519329211045118e+18, "train_loss": 0.8785166950206239, "train_runtime": 8976.4943, "train_samples_per_second": 6.711, "train_steps_per_second": 0.112 }