tensorblock/moe-x33-GGUF

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

senseable/moe-x33 - GGUF

This repo contains GGUF format model files for senseable/moe-x33.

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

Awesome MCP Servers	TensorBlock Studio

A comprehensive collection of Model Context Protocol (MCP) servers.	A lightweight, open, and extensible multi-LLM interaction studio.
👀 See what we built 👀	👀 See what we built 👀

## Prompt template

Filename	Quant type	File Size	Description
moe-x33-Q2_K.gguf	Q2_K	21.625 GB	smallest, significant quality loss - not recommended for most purposes
moe-x33-Q3_K_S.gguf	Q3_K_S	25.425 GB	very small, high quality loss
moe-x33-Q3_K_M.gguf	Q3_K_M	28.280 GB	very small, high quality loss
moe-x33-Q3_K_L.gguf	Q3_K_L	30.763 GB	small, substantial quality loss
moe-x33-Q4_0.gguf	Q4_0	33.222 GB	legacy; small, very high quality loss - prefer using Q3_K_M
moe-x33-Q4_K_S.gguf	Q4_K_S	33.466 GB	small, greater quality loss
moe-x33-Q4_K_M.gguf	Q4_K_M	35.408 GB	medium, balanced quality - recommended
moe-x33-Q5_0.gguf	Q5_0	40.561 GB	legacy; medium, balanced quality - prefer using Q4_K_M
moe-x33-Q5_K_S.gguf	Q5_K_S	40.561 GB	large, low quality loss - recommended
moe-x33-Q5_K_M.gguf	Q5_K_M	41.687 GB	large, very low quality loss - recommended
moe-x33-Q6_K.gguf	Q6_K	48.358 GB	very large, extremely low quality loss
moe-x33-Q8_0	Q8_0	42.254 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/moe-x33-GGUF --include "moe-x33-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/moe-x33-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'