Goat: Fine-Tuned LLaMA Outperforms GPT-4 on Arithmetic Tasksarxiv.org2 points·malshe··1 commentOpen articleSaveView on HN