{ "effective_tokens_per_sec": 602.4478364775058, "epoch": 3.0, "total_flos": 2.906371548002648e+18, "train_loss": 2.8588710447346606, "train_runtime": 9577.8089, "train_samples_per_second": 6.135, "train_steps_per_second": 0.096 }