The request model: jarvis_vla_qwen2_vl_7b_sft does not exist!

[jackieyung]

Hi @MuyaoLi-jimo , thanks for your excellent work.

When I run the script: CUDA_VISIBLE_DEVICES=0 vllm serve jarvis_vla_qwen2_vl_7b_sft --port 8000, it occurs requests.exceptions.HTTPError: The request model: jarvis_vla_qwen2_vl_7b_sft does not exist! . It seems that the model jarvis_vla_qwen2_vl_7b_sft doesn't exist. Please recheck it.

Thanks a lot.

[MuyaoLi-jimo]

vllm serve is expecting a Hugging Face model_id or a local model_path.
try downloading the model first, and then start the server using the local path, like:

CUDA_VISIBLE_DEVICES=0 vllm serve /the/path/of/JarvisVLA --port 8000