Custom Model / GPT4all model.

[taranjeet]

Hey @taranjeet I loved embedchain and wanted to use it in one of my projects. I wanted to use my custom model / GPT4All models, do you have any documentation related to it? Current default model is all-MiniLM-L6-v2

• discord message link: https://discord.com/channels/1121119078191480945/1121119078858358846/1127139683953025094

[TomasTomecek]

I second this ask. Unfortunately all-MiniLM-L6-v2 can only work with 256 tokens which is fairly limiting. GPT4All's default embedding model (ggml-all-MiniLM-L6-v2-f16) can do even several k so it would be amazing to use it.

    raise RepositoryNotFoundError(message, response) from e
huggingface_hub.utils._errors.RepositoryNotFoundError: 401 Client Error. (Request ID: ...)

Repository Not Found for url: https://huggingface.co/api/models/sentence-transformers/ggml-all-MiniLM-L6-v2-f16.

[nav9]

Using embedChain with GPT4All is something I'm looking forward to as well.

[Dev-Khant]

Closing this issue as GPT4All functionality is more robust now and can handle different models. Please feel free to open this issue again if there's still a problem.