{ "effective_tokens_per_sec": 1153.798380212981, "epoch": 4.0, "total_flos": 3.545203061907456e+17, "train_loss": 0.08566030774240586, "train_runtime": 13933.4985, "train_samples_per_second": 3.772, "train_steps_per_second": 0.472 }