tensorblock/Chinese-Mixtral-8x7B-GGUF

Feedback and support: TensorBlock's Twitter/X, Telegram Group and Discord server

HIT-SCIR/Chinese-Mixtral-8x7B - GGUF

This repo contains GGUF format model files for HIT-SCIR/Chinese-Mixtral-8x7B.

The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b4242.

Filename	Quant type	File Size	Description
Chinese-Mixtral-8x7B-Q2_K.gguf	Q2_K	17.429 GB	smallest, significant quality loss - not recommended for most purposes
Chinese-Mixtral-8x7B-Q3_K_S.gguf	Q3_K_S	20.561 GB	very small, high quality loss
Chinese-Mixtral-8x7B-Q3_K_M.gguf	Q3_K_M	22.675 GB	very small, high quality loss
Chinese-Mixtral-8x7B-Q3_K_L.gguf	Q3_K_L	24.298 GB	small, substantial quality loss
Chinese-Mixtral-8x7B-Q4_0.gguf	Q4_0	26.586 GB	legacy; small, very high quality loss - prefer using Q3_K_M
Chinese-Mixtral-8x7B-Q4_K_S.gguf	Q4_K_S	26.888 GB	small, greater quality loss
Chinese-Mixtral-8x7B-Q4_K_M.gguf	Q4_K_M	28.591 GB	medium, balanced quality - recommended
Chinese-Mixtral-8x7B-Q5_0.gguf	Q5_0	32.386 GB	legacy; medium, balanced quality - prefer using Q4_K_M
Chinese-Mixtral-8x7B-Q5_K_S.gguf	Q5_K_S	32.386 GB	large, low quality loss - recommended
Chinese-Mixtral-8x7B-Q5_K_M.gguf	Q5_K_M	33.385 GB	large, very low quality loss - recommended
Chinese-Mixtral-8x7B-Q6_K.gguf	Q6_K	38.549 GB	very large, extremely low quality loss
Chinese-Mixtral-8x7B-Q8_0.gguf	Q8_0	49.844 GB	very large, extremely low quality loss - not recommended

Firstly, install Huggingface Client

pip install -U "huggingface_hub[cli]"

Then, downoad the individual model file the a local directory

huggingface-cli download tensorblock/Chinese-Mixtral-8x7B-GGUF --include "Chinese-Mixtral-8x7B-Q2_K.gguf" --local-dir MY_LOCAL_DIR

If you wanna download multiple model files with a pattern (e.g., *Q4_K*gguf), you can try:

huggingface-cli download tensorblock/Chinese-Mixtral-8x7B-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'