Spaces:

---

ggml-org

/

gguf-my-repo

like 1.99k

Running on A10G

MOE models like Gemma-4-26B or Qwen A3B have good results when MXFP4, But no option here to make MXFP4, Is addition of option possible?