tensorblock/bloom-1b7-fp32-GGUF

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

LazarusNLP/bloom-1b7-fp32 - GGUF

This repo contains GGUF format model files for LazarusNLP/bloom-1b7-fp32.

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
bloom-1b7-fp32-Q2_K.gguf	Q2_K	1.053 GB	smallest, significant quality loss - not recommended for most purposes
bloom-1b7-fp32-Q3_K_S.gguf	Q3_K_S	1.177 GB	very small, high quality loss
bloom-1b7-fp32-Q3_K_M.gguf	Q3_K_M	1.286 GB	very small, high quality loss
bloom-1b7-fp32-Q3_K_L.gguf	Q3_K_L	1.346 GB	small, substantial quality loss
bloom-1b7-fp32-Q4_0.gguf	Q4_0	1.405 GB	legacy; small, very high quality loss - prefer using Q3_K_M
bloom-1b7-fp32-Q4_K_S.gguf	Q4_K_S	1.411 GB	small, greater quality loss
bloom-1b7-fp32-Q4_K_M.gguf	Q4_K_M	1.495 GB	medium, balanced quality - recommended
bloom-1b7-fp32-Q5_0.gguf	Q5_0	1.620 GB	legacy; medium, balanced quality - prefer using Q4_K_M
bloom-1b7-fp32-Q5_K_S.gguf	Q5_K_S	1.620 GB	large, low quality loss - recommended
bloom-1b7-fp32-Q5_K_M.gguf	Q5_K_M	1.687 GB	large, very low quality loss - recommended
bloom-1b7-fp32-Q6_K.gguf	Q6_K	1.849 GB	very large, extremely low quality loss
bloom-1b7-fp32-Q8_0.gguf	Q8_0	2.390 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/bloom-1b7-fp32-GGUF --include "bloom-1b7-fp32-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/bloom-1b7-fp32-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'