tensorblock/deepseekmath-7b-sft-GSM8K-GGUF

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

realtreetune/deepseekmath-7b-sft-GSM8K - GGUF

This repo contains GGUF format model files for realtreetune/deepseekmath-7b-sft-GSM8K.

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

Run them on the TensorBlock client using your local machine ↗

Prompt template

Model file specification

Filename	Quant type	File Size	Description
deepseekmath-7b-sft-GSM8K-Q2_K.gguf	Q2_K	2.718 GB	smallest, significant quality loss - not recommended for most purposes
deepseekmath-7b-sft-GSM8K-Q3_K_S.gguf	Q3_K_S	3.138 GB	very small, high quality loss
deepseekmath-7b-sft-GSM8K-Q3_K_M.gguf	Q3_K_M	3.461 GB	very small, high quality loss
deepseekmath-7b-sft-GSM8K-Q3_K_L.gguf	Q3_K_L	3.746 GB	small, substantial quality loss
deepseekmath-7b-sft-GSM8K-Q4_0.gguf	Q4_0	4.000 GB	legacy; small, very high quality loss - prefer using Q3_K_M
deepseekmath-7b-sft-GSM8K-Q4_K_S.gguf	Q4_K_S	4.025 GB	small, greater quality loss
deepseekmath-7b-sft-GSM8K-Q4_K_M.gguf	Q4_K_M	4.223 GB	medium, balanced quality - recommended
deepseekmath-7b-sft-GSM8K-Q5_0.gguf	Q5_0	4.811 GB	legacy; medium, balanced quality - prefer using Q4_K_M
deepseekmath-7b-sft-GSM8K-Q5_K_S.gguf	Q5_K_S	4.811 GB	large, low quality loss - recommended
deepseekmath-7b-sft-GSM8K-Q5_K_M.gguf	Q5_K_M	4.926 GB	large, very low quality loss - recommended
deepseekmath-7b-sft-GSM8K-Q6_K.gguf	Q6_K	5.673 GB	very large, extremely low quality loss
deepseekmath-7b-sft-GSM8K-Q8_0.gguf	Q8_0	7.347 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/deepseekmath-7b-sft-GSM8K-GGUF --include "deepseekmath-7b-sft-GSM8K-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/deepseekmath-7b-sft-GSM8K-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'