tensorblock/CodeLlama-7b-hf-GGUF

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

codellama/CodeLlama-7b-hf - GGUF

This repo contains GGUF format model files for codellama/CodeLlama-7b-hf.

The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b4011.

Prompt template

Model file specification

Filename	Quant type	File Size	Description
CodeLlama-7b-hf-Q2_K.gguf	Q2_K	2.359 GB	smallest, significant quality loss - not recommended for most purposes
CodeLlama-7b-hf-Q3_K_S.gguf	Q3_K_S	2.746 GB	very small, high quality loss
CodeLlama-7b-hf-Q3_K_M.gguf	Q3_K_M	3.072 GB	very small, high quality loss
CodeLlama-7b-hf-Q3_K_L.gguf	Q3_K_L	3.350 GB	small, substantial quality loss
CodeLlama-7b-hf-Q4_0.gguf	Q4_0	3.563 GB	legacy; small, very high quality loss - prefer using Q3_K_M
CodeLlama-7b-hf-Q4_K_S.gguf	Q4_K_S	3.592 GB	small, greater quality loss
CodeLlama-7b-hf-Q4_K_M.gguf	Q4_K_M	3.801 GB	medium, balanced quality - recommended
CodeLlama-7b-hf-Q5_0.gguf	Q5_0	4.332 GB	legacy; medium, balanced quality - prefer using Q4_K_M
CodeLlama-7b-hf-Q5_K_S.gguf	Q5_K_S	4.332 GB	large, low quality loss - recommended
CodeLlama-7b-hf-Q5_K_M.gguf	Q5_K_M	4.455 GB	large, very low quality loss - recommended
CodeLlama-7b-hf-Q6_K.gguf	Q6_K	5.150 GB	very large, extremely low quality loss
CodeLlama-7b-hf-Q8_0.gguf	Q8_0	6.669 GB	very large, extremely low quality loss - not recommended

Downloading instruction

Command line

Firstly, install Huggingface Client

pip install -U "huggingface_hub[cli]"

Then, downoad the individual model file the a local directory

huggingface-cli download tensorblock/CodeLlama-7b-hf-GGUF --include "CodeLlama-7b-hf-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/CodeLlama-7b-hf-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'