tensorblock/Ryu-4x7B-MoE-bf16-GGUF

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

Kquant03/Ryu-4x7B-MoE-bf16 - GGUF

This repo contains GGUF format model files for Kquant03/Ryu-4x7B-MoE-bf16.

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
Ryu-4x7B-MoE-bf16-Q2_K.gguf	Q2_K	8.843 GB	smallest, significant quality loss - not recommended for most purposes
Ryu-4x7B-MoE-bf16-Q3_K_S.gguf	Q3_K_S	10.433 GB	very small, high quality loss
Ryu-4x7B-MoE-bf16-Q3_K_M.gguf	Q3_K_M	11.580 GB	very small, high quality loss
Ryu-4x7B-MoE-bf16-Q3_K_L.gguf	Q3_K_L	12.544 GB	small, substantial quality loss
Ryu-4x7B-MoE-bf16-Q4_0.gguf	Q4_0	13.624 GB	legacy; small, very high quality loss - prefer using Q3_K_M
Ryu-4x7B-MoE-bf16-Q4_K_S.gguf	Q4_K_S	13.743 GB	small, greater quality loss
Ryu-4x7B-MoE-bf16-Q4_K_M.gguf	Q4_K_M	14.610 GB	medium, balanced quality - recommended
Ryu-4x7B-MoE-bf16-Q5_0.gguf	Q5_0	16.626 GB	legacy; medium, balanced quality - prefer using Q4_K_M
Ryu-4x7B-MoE-bf16-Q5_K_S.gguf	Q5_K_S	16.626 GB	large, low quality loss - recommended
Ryu-4x7B-MoE-bf16-Q5_K_M.gguf	Q5_K_M	17.134 GB	large, very low quality loss - recommended
Ryu-4x7B-MoE-bf16-Q6_K.gguf	Q6_K	19.817 GB	very large, extremely low quality loss
Ryu-4x7B-MoE-bf16-Q8_0.gguf	Q8_0	25.666 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/Ryu-4x7B-MoE-bf16-GGUF --include "Ryu-4x7B-MoE-bf16-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/Ryu-4x7B-MoE-bf16-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'