Hi, AesSedai here - I've put up a PR to support the text-to-text inference of MiMo V2.5 with llama.cpp (and should also support Pro, will work on those quants after finishing V2.5): https://github.com/ggml-org/llama.cpp/pull/22493 I've also put some quants up on HF (https://huggingface.co/AesSedai/M
