You should benchmark GPT-3 Curie (7B) for comparison since it is the same size as llama-7B and StableLM-7B. That will give us some indication of how much bette

You should benchmark GPT-3 Curie (7B) for comparison since it is the same size as llama-7B and StableLM-7B. That will give us some indication of how much better these models are than GPT-3 at the same size.