I set out to find out Alpaca/LLama 7B language model, running on my Macbook Pro, can achieve similar performance as chatGPT 3.5. My observation:
- Alpaca/LLaMA 7B: a competent junior high school student
- chatGPT 3.5: a competent and well-rounded college graduate