Instructions to use SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF",
	filename="granite-3.3-8b-instruct_Q4_k_m.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF:Q4_K_M

Use Docker

docker model run hf.co/SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF:Q4_K_M

Ollama
How to use SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF with Ollama:
```
ollama run hf.co/SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF:Q4_K_M
```

Unsloth Studio

How to use SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF to start chatting

How to use SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF:Q4_K_M

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat new
Docker Model Runner
How to use SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF with Docker Model Runner:
```
docker model run hf.co/SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF:Q4_K_M
```

Lemonade

How to use SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Granite-3.3-8B-Instruct-GGUF-Q4_K_M

List all available models

lemonade list

Granite-3.3-8B-Instruct (GGUF / Quantized)

Granite-3.3-8B-Instruct is an instruction-tuned large language model designed for strong conversational ability, instruction compliance, and efficient inference. This repository contains quantized GGUF formats of the model, allowing for practical local use even on limited hardware.

Quantized variants significantly lower memory requirements while preserving generation quality, making Granite-3.3-8B-Instruct suitable for research, experimentation, and on-device deployments.

Model Overview

Model Name: Granite-3.3-8B-Instruct
Base Model: ibm-granite/granite-3.3-8b-instruct
Architecture: Decoder-only Transformer
Parameter Count: 8 Billion
Supported Context Length: Extended
Modalities: Text
Developer: IBM Granite
License: Apache 2.0

Quantization Formats

Q4_K_M

Approx. ~~71% size reduction (~~4.60 GB)
Designed for CPU-only inference
Good balance of speed vs quality
Ideal for machines with limited VRAM

Q5_K_M

Approx. ~~66% size reduction (~~5.40 GB)
Higher numeric precision than Q4
Better stability on reasoning tasks
Stronger general output consistency

Training Background (Original Model)

Granite-3.3-8B-Instruct is trained with an emphasis on instruction comprehension and versatile generalization across diverse tasks.

Pretraining

Large pretraining corpus covering diverse domains
Autoregressive language modeling training regime
Focus on robust language representation

Instruction Tuning

Refined using instruction datasets to improve user input compliance
Promotes clearer answers and more predictable outputs
Enhanced for multi-step reasoning and conversational flow

Key Capabilities

Instruction Compliance
Handles a wide range of user directives and produces tailored responses.
Conversational Fluency
Generates natural dialogue with contextual awareness.
Reasoning and Explanation
Performs well on tasks requiring multi-step logic and analysis.
Efficient Local Inference
Quantized variants enable practical usage without cloud dependencies.
Flexible Text Generation
Suitable for summarization, Q&A, and creative language tasks.

Usage Example

Using llama.cpp

./llama-cli \
  -m granite-3.3-8b-instruct_Q4_K_M.gguf \
  -p "Describe what makes efficient LLM inference challenging."

Recommended Applications

Local Assistants Build offline chat tools without relying on remote servers.
Instruction Following Systems Create task automation helpers, interactive agents, and Q&A systems.
Research and Prototyping Evaluate model behavior and prompting strategies.
Data Privacy Projects Run generation fully under your own control without external APIs.

Acknowledgments

This repository is based on the original Granite-3.3-8B-Instruct model released by IBM Granite.

Thanks to:

The IBM Granite team for contributing an open instruction-tuned model
The llama.cpp community for enabling efficient GGUF inference

Contact

For questions, feedback, or support, please reach out atsupport@sandlogic.com or visit https://www.sandlogic.com/

Downloads last month: 5

GGUF

Model size

8B params

Architecture

granite

Hardware compatibility

4-bit

5-bit

Model tree for SandLogicTechnologies/Granite-3.3-8B-Instruct-GGUF

Base model

ibm-granite/granite-3.3-8b-base

Finetuned

ibm-granite/granite-3.3-8b-instruct

Quantized

(29)

this model