I want to add a system prompt when I run the Llama3 test on my computer.

[scj0709]

🚀 The feature, motivation and pitch

https://github.com/pytorch/executorch/tree/main/examples/models/llama
I successfully executed the part 'Step 3: Run on your computer to validate' from the UR mentioned above. One thing I would like is to add a system prompt, but I only have an input for the prompt. Is there a way to do that?

Alternatives

No response

Additional context

I successfully executed the command cmake-out/examples/models/llama/llama_main --model_path= --tokenizer_path=<tokenizer.model> --prompt=, but I would like to add a separate system prompt!

RFC (Optional)

No response

[JacobSzwejbka]

Hmm I definitely added this to torchchat, do we have this in the ET folders @larryliu0820 ?