how to load already downloaded models in pvc and i dont want to pass the hf token

[gokulgopinathan]

i have some finetuned llm models stored locally and available via pvc. i dont want to pass the hf_token. is there any way to load these models to production stack

[zerofishnoodles]

@gokulgopinathan I think loading it will be the same as in vllm, the modelURL should be pointed to your model path in your pic