Error: An error occurred while fetching the blob

[Marie-Yang]

Describe the bug
A clear and concise description of what the bug is.

Screenshots
If applicable, add screenshots to help explain your problem.

Environment
I am on:

• MacOS
• Windows
• Linux

Using the model:

• Local model with python
• Inference API with Huggingface

I have tried both of these two respectively(Local model or Inference API), but they all got similar mistakes.

Is there a problem with the requested website address？

[lucasvanmol]

Interesting, it seems it might be on huggingface's end:

https://discuss.huggingface.co/t/inference-api-stopped-working/150492/35

hopefully it gets resolved, but I haven't found any comment on it from HF's side.

However, I am still able to run the model locally, if I use the python application. Could you share screenshots of the problems you are running into there?

[Nuvole0217]

Hi, the problem persists, my system is manjaro linux. On the thread of hf they said they resolve it but the pb still occurs

[hasanakhiar]

Facing the same issue
Platform: MacOS

[lucasvanmol]

Unfortunately this is due to HuggingFace changing to a new inference API, which has somehow dropped support for image-to-text models, even Huggingface's own provider doesn't support it: https://huggingface.co/docs/inference-providers/index

I'm currently looking into ways around this issue.

[abjil]

First, please accept a thousand thanks for your awesome plugin.
Regarding the issue - I am using Obsidian on a few machines and it is not practical to run a local model on all of them to use LatexOCR. It would make more sense for me to run it on a dedicated machine and for all others to send their requests there.
I hope you could add it as an option, hopefully it is not too much change in the code base.