Hugging Face's logo

tensorblock
/
granite-3.1-3b-a800m-instruct-GGUF

Text Generation
Transformers
GGUF
language
granite-3.1
TensorBlock
GGUF
conversational
Model card Files Community
granite-3.1-3b-a800m-instruct-GGUF / README.md
morriszms's picture
morriszms
Upload folder using huggingface_hub
ead42b3 verified
preview code
|
raw
history blame contribute delete
5.44 kB
metadata
pipeline_tag: text-generation
inference: false
license: apache-2.0
library_name: transformers
tags:
  - language
  - granite-3.1
  - TensorBlock
  - GGUF
base_model: ibm-granite/granite-3.1-3b-a800m-instruct
TensorBlock

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

ibm-granite/granite-3.1-3b-a800m-instruct - GGUF

This repo contains GGUF format model files for ibm-granite/granite-3.1-3b-a800m-instruct.

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

Prompt template 

<|start_of_role|>system<|end_of_role|>{system_prompt}<|end_of_text|>
<|start_of_role|>user<|end_of_role|>{prompt}<|end_of_text|>
<|start_of_role|>assistant<|end_of_role|>

Model file specification

Filename Quant type File Size Description
granite-3.1-3b-a800m-instruct-Q2_K.gguf Q2_K 1.241 GB smallest, significant quality loss - not recommended for most purposes
granite-3.1-3b-a800m-instruct-Q3_K_S.gguf Q3_K_S 1.456 GB very small, high quality loss
granite-3.1-3b-a800m-instruct-Q3_K_M.gguf Q3_K_M 1.611 GB very small, high quality loss
granite-3.1-3b-a800m-instruct-Q3_K_L.gguf Q3_K_L 1.742 GB small, substantial quality loss
granite-3.1-3b-a800m-instruct-Q4_0.gguf Q4_0 1.884 GB legacy; small, very high quality loss - prefer using Q3_K_M
granite-3.1-3b-a800m-instruct-Q4_K_S.gguf Q4_K_S 1.900 GB small, greater quality loss
granite-3.1-3b-a800m-instruct-Q4_K_M.gguf Q4_K_M 2.017 GB medium, balanced quality - recommended
granite-3.1-3b-a800m-instruct-Q5_0.gguf Q5_0 2.287 GB legacy; medium, balanced quality - prefer using Q4_K_M
granite-3.1-3b-a800m-instruct-Q5_K_S.gguf Q5_K_S 2.287 GB large, low quality loss - recommended
granite-3.1-3b-a800m-instruct-Q5_K_M.gguf Q5_K_M 2.355 GB large, very low quality loss - recommended
granite-3.1-3b-a800m-instruct-Q6_K.gguf Q6_K 2.714 GB very large, extremely low quality loss
granite-3.1-3b-a800m-instruct-Q8_0.gguf Q8_0 3.513 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/granite-3.1-3b-a800m-instruct-GGUF --include "granite-3.1-3b-a800m-instruct-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/granite-3.1-3b-a800m-instruct-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'