New model gguf request. MiniMax-M2(llama.cpp now supports it.)
New model gguf request. MiniMax-M2
I request GGUF quantization.
https://huggingface.co/MiniMaxAI/MiniMax-M2
PR is update.(Feature Request: MiniMax M2 support)
https://github.com/ggml-org/llama.cpp/issues/16798
Let's be precise. The PR is still open and work in progress. Only the feature request got closed. Let's follow https://github.com/ggml-org/llama.cpp/pull/16831 and properly do this model once it is merged and so works with latest llama.cpp. Casual users don't know how to build a specific llama.cpp branch from source and would be confused if we already upload quants before a model is officially supported. Beside that this PR is still subject to change and there are no guarantees that there will not be any compatibility breaking changes. If you really want to try the model now build the "pwilkin:minimax-m2" branch of llama.cpp yourself and use the experimental quants from: