Hugging Face's logo

tensorblock
/
gpt-neox-20b-GGUF

GGUF
PyTorch
English
causal-lm
TensorBlock
GGUF
Inference Endpoints
Model card Files Community
gpt-neox-20b-GGUF / README.md
morriszms's picture
morriszms
Update README.md
66587a1 verified
preview code
|
raw
history blame contribute delete
4.52 kB
metadata
language:
  - en
tags:
  - pytorch
  - causal-lm
  - TensorBlock
  - GGUF
license: apache-2.0
datasets:
  - EleutherAI/pile
base_model: EleutherAI/gpt-neox-20b
TensorBlock

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

EleutherAI/gpt-neox-20b - GGUF

This repo contains GGUF format model files for EleutherAI/gpt-neox-20b.

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
gpt-neox-20b-Q2_K.gguf Q2_K 7.223 GB smallest, significant quality loss - not recommended for most purposes
gpt-neox-20b-Q3_K_S.gguf Q3_K_S 8.352 GB very small, high quality loss
gpt-neox-20b-Q3_K_M.gguf Q3_K_M 10.030 GB very small, high quality loss
gpt-neox-20b-Q3_K_L.gguf Q3_K_L 10.962 GB small, substantial quality loss
gpt-neox-20b-Q4_0.gguf Q4_0 10.855 GB legacy; small, very high quality loss - prefer using Q3_K_M
gpt-neox-20b-Q4_K_S.gguf Q4_K_S 10.943 GB small, greater quality loss
gpt-neox-20b-Q4_K_M.gguf Q4_K_M 12.233 GB medium, balanced quality - recommended
gpt-neox-20b-Q5_0.gguf Q5_0 13.212 GB legacy; medium, balanced quality - prefer using Q4_K_M
gpt-neox-20b-Q5_K_S.gguf Q5_K_S 13.212 GB large, low quality loss - recommended
gpt-neox-20b-Q5_K_M.gguf Q5_K_M 14.239 GB large, very low quality loss - recommended
gpt-neox-20b-Q6_K.gguf Q6_K 15.715 GB very large, extremely low quality loss
gpt-neox-20b-Q8_0.gguf Q8_0 20.351 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/gpt-neox-20b-GGUF --include "gpt-neox-20b-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/gpt-neox-20b-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'