Currently getting this “could not download model artifacts” error when I am trying to deploy the model using TEI -
It seems like whenever it tries to go download the model weights from HF, it is constantly looking for the ONNX version of the weights, which does not exist. In the current model repo, the model weights only exist in a safetensor format, which is why I suppose I am getting the error above. Anyone encountered the same problem? Thank you!
I am also getting this error when I try and serve a TEI endpoint using docker. I had to switch to a model that had weights in ONNX, and unfortunately could not use gte-Qwen for my embeddings.
I am also having this problem. Anyone know a way to fix this?
Error: Could not download model artifacts
Caused by:
0: request error: HTTP status client error (401 Unauthorized) for url (https://huggingface.co/model=BAAI/bge-large-en-v1.5/resolve/main/config.json)
1: HTTP status client error (401 Unauthorized) for url (https://huggingface.co/model=BAAI/bge-large-en-v1.5/resolve/main/config.json)
That model isn’t gated or anything…
Is it a TEI bug?
@olivierdehaene There seems to be a phenomenon where the HF repo cannot be accessed from TEI.
