Using Mistral-7B model

[pgp69]

how to use mistral-7b model, it returns a message that it is not supported ;(

This is the model: NickyNicky/Mistral-7B-OpenOrca-oasst_top1_2023-08-25-v2

[PromtEngineer]

Look for the model in gguf format. You should be able to use that.

[pgp69]

Can I convert a mistral model to GGUF. I don't success using RTX3050/4GB of RAM with cuda

I am running exactly the installation instructions for CMAKE_ARGS="-DLLAMA_CUBLAS=on" FORCE_CMAKE=1 pip install llama-cpp-python==0.1.83

but does not work with both Win11 and WSL-2.

Torch.cuda.is_available() -> True

Does it make sense to use an eGPU with RTX-3060/12GB RAM over Thunderbolt-3. Will the performance be very low?

[clearsitedesigns]

This is a great source of models updates frequently;

https://huggingface.co/TheBloke

[pgp69]

I tried this model works very good, I'm looking for a model to fine tuning to work on Bulgarian, for now without much success
Thank You.