{ "effective_tokens_per_sec": 7124.293057605591, "epoch": 4.0, "total_flos": 3.5948540766149673e+18, "train_loss": 0.8339844654994046, "train_runtime": 4886.5014, "train_samples_per_second": 11.64, "train_steps_per_second": 0.388 }