Deploy Salesforce/SFR-Embedding-Mistral on HF Inference Endpoint?

#6
by rlsChapters - opened

Hi all. I am trying to deploy the model on a HuggingFace Inference Endpoint and keep getting this error

Server message:Endpoint failed to start. raise NotImplementedError(\"sharded is not supported for AutoModel\")

Has anyone encountered this and how should I go about fixing it? Thanks!

Salesforce org

Hi @rlsChapters ,

We didn't meet this issue on our end. Could you try the following package version:
torch==2.0.0
transformers>=4.31.0

Thanks for the suggestion @yliu279

I added the requirements.txt file with

torch==2.0.0
transformers>=4.31.0

Sign up or log in to comment