metadata
base_model:
- mistralai/Mistral-Large-3-675B-Instruct-2512
pipeline_tag: image-text-to-text
Mistral-Large-3-675B-Instruct-2512-GGUF
Please refer to this PR: https://github.com/ggml-org/llama.cpp/pull/17730
The architecture is deepseekv3, which is already supported out-of-the-box in llama.cpp