Instructions to use SandLogicTechnologies/gemma-2-9b-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use SandLogicTechnologies/gemma-2-9b-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="SandLogicTechnologies/gemma-2-9b-GGUF",
	filename="gemma-2-9b_Q4_k_m.gguf",
)

output = llm(
	"Once upon a time,",
	max_tokens=512,
	echo=True
)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use SandLogicTechnologies/gemma-2-9b-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/gemma-2-9b-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf SandLogicTechnologies/gemma-2-9b-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/gemma-2-9b-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf SandLogicTechnologies/gemma-2-9b-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/gemma-2-9b-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf SandLogicTechnologies/gemma-2-9b-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/gemma-2-9b-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf SandLogicTechnologies/gemma-2-9b-GGUF:Q4_K_M

Use Docker

docker model run hf.co/SandLogicTechnologies/gemma-2-9b-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use SandLogicTechnologies/gemma-2-9b-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "SandLogicTechnologies/gemma-2-9b-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "SandLogicTechnologies/gemma-2-9b-GGUF",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/SandLogicTechnologies/gemma-2-9b-GGUF:Q4_K_M

Ollama
How to use SandLogicTechnologies/gemma-2-9b-GGUF with Ollama:
```
ollama run hf.co/SandLogicTechnologies/gemma-2-9b-GGUF:Q4_K_M
```

Unsloth Studio new

How to use SandLogicTechnologies/gemma-2-9b-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/gemma-2-9b-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/gemma-2-9b-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/gemma-2-9b-GGUF to start chatting

Docker Model Runner
How to use SandLogicTechnologies/gemma-2-9b-GGUF with Docker Model Runner:
```
docker model run hf.co/SandLogicTechnologies/gemma-2-9b-GGUF:Q4_K_M
```

Lemonade

How to use SandLogicTechnologies/gemma-2-9b-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull SandLogicTechnologies/gemma-2-9b-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.gemma-2-9b-GGUF-Q4_K_M

List all available models

lemonade list

Gemma-2-9B

Gemma-2-9B is a pretrained large language model developed by Google as part of the second generation of the Gemma model family. It is designed as a general-purpose foundation model capable of strong language understanding, reasoning, and text generation across a wide range of tasks.

The model is optimized for efficiency relative to its size and can be used as a base for downstream fine-tuning, instruction alignment, or domain-specific adaptation. It provides a strong starting point for building conversational agents, analytical systems, and custom language applications.

Gemma-2 models incorporate architectural improvements aimed at improving performance, scalability, and inference efficiency compared to earlier releases.

Model Overview

Model Name: Gemma-2-9B
Base Model Family: Gemma 2
Architecture: Decoder-only Transformer
Parameter Count: 9 Billion
Context Window: ~8K tokens
Modalities: Text
Primary Language: English
Developer: Google DeepMind
License: Gemma usage license (gated access required)

Quantization Details

Q4_K_M

Approx. ~70% size reduction (5.37 GB)
Significant reduction in model memory usage
Suitable for local CPU or low-VRAM GPU inference
Faster token generation speeds
Slight reduction in numerical precision for complex reasoning

Q5_K_M

Approx. ~66% size reduction (6.19 GB)
Higher fidelity compared to lower-bit quantization
Improved stability for analytical and structured tasks
Better preservation of original model behavior
Recommended for balanced performance and efficiency

Training Overview

Pretraining

Gemma-2-9B is trained on large-scale curated text corpora designed to provide broad language understanding and reasoning capability. Training emphasizes knowledge acquisition, contextual awareness, and long-range dependency modeling.

Smaller Gemma-2 models are produced using distillation and scaling techniques to retain performance while reducing computational requirements.

Gemma-2-9B is built as a scalable and efficient base language model suitable for customization and research.

Primary design goals include:

Strong general-purpose language modeling
Efficient inference relative to parameter count
Reliable long-context processing
High-quality reasoning capability
Flexible foundation for downstream fine-tuning

Core Capabilities

General language modeling
Generates coherent text across diverse topics.
Contextual reasoning
Handles analytical prompts and structured thinking tasks.
Long-context processing
Supports extended prompts and document-level understanding.
Foundation model flexibility
Serves as a base for instruction tuning and domain adaptation.
Text generation and transformation
Supports summarization, explanation, and content generation after tuning.

Example Usage

llama.cpp


./llama-cli 
-m SandlogicTechnologies\gemma-2-9b_Q4_K_M.gguf 
-p "Explain how attention mechanisms work in transformers."

Recommended Use Cases

Base model for custom fine-tuning
Research and experimentation
Domain-specific model development
Text analysis and generation systems
Conversational AI after alignment
Local deployment of foundation language models

Acknowledgments

These quantized models are based on the original work by Google development team.

Special thanks to:

The Google team for developing and releasing the gemma-2-9b model.
Georgi Gerganov and the entire llama.cpp open-source community for enabling efficient model quantization and inference via the GGUF format.

Contact

For any inquiries or support, please contact us at support@sandlogic.com or visit our Website.

Downloads last month: 13

GGUF

Model size

9B params

Architecture

gemma2

Hardware compatibility

4-bit

5-bit

Model tree for SandLogicTechnologies/gemma-2-9b-GGUF

Base model

google/gemma-2-9b

Quantized

(52)

this model