{ "effective_tokens_per_sec": 3555.7280013351283, "epoch": 4.0, "total_flos": 3.554237146892075e+18, "train_loss": 0.8450854239206619, "train_runtime": 19534.9003, "train_samples_per_second": 2.906, "train_steps_per_second": 0.097 }