Morgan Giraud

Hi @fayismahmood ,

Short answer: no.
As our model requires a custom architecture, the GGUF file would not be enough to run our model in llama-cli.

You can try quantization with bitsandbytes on the HF model though.