{ "effective_tokens_per_sec": 2954.1245634821435, "epoch": 4.0, "total_flos": 3.554237146892075e+18, "train_loss": 0.8299213015755204, "train_runtime": 23513.1561, "train_samples_per_second": 2.414, "train_steps_per_second": 0.08 }