yes - we have already published a quantized version here: https://huggingface.co/katanemo/Arch-Router-1.5B.gguf https://huggingface.co/katanemo/Arch-Router-1.5B

yes - we have already published a quantized version here: https://huggingface.co/katanemo/Arch-Router-1.5B.gguf https://huggingface.co/katanemo/Arch-Router-1.5B.gguf. The performance difference with a quant version is negligible. I'll run another analysis and update the thread shortly