·
4 commits
to merge-mtp-turboquant
since this release
How to Run
Please refer to the README for detailed instructions.
Weights
You can download the weight files from the Hugging Face Hub:
froggeric/Qwen3.6-27B-MTP-GGUF
Release
The newest
llama-cpp-turboquant-mtp-vision.7z build by 2080 ，support turbo2 turbo3 turbo4 ，vision is ok，cuda 12.3
llama-cpp-turboquant-mtp_cuda13.2_vision.7z build by 4090 ，support turbo2 turbo3 turbo4 ，vision is ok，cuda 13.2
llama-cpp-turboquant-mtp_cuda13.2_vision_30.7z build by 30 series ，support turbo2 turbo3 turbo4 ，vision is ok，cuda 13.2
older
llama-cpp-turboquant-mtp.zip cuda 12.3 build by 2080 ，support turbo2 turbo3 turbo4
llama-cpp-turboquant-mtp_cuda13.2.zip cuda 13.2 build by 4090