Instructions to use sunil-pathak/gemma-3n-E2B-it-Q5_K_M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use sunil-pathak/gemma-3n-E2B-it-Q5_K_M with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="sunil-pathak/gemma-3n-E2B-it-Q5_K_M",
	filename="gemma-3n-E2B-it-Q5_K_M.gguf",
)

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

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use sunil-pathak/gemma-3n-E2B-it-Q5_K_M with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf sunil-pathak/gemma-3n-E2B-it-Q5_K_M:Q5_K_M
# Run inference directly in the terminal:
llama-cli -hf sunil-pathak/gemma-3n-E2B-it-Q5_K_M:Q5_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf sunil-pathak/gemma-3n-E2B-it-Q5_K_M:Q5_K_M
# Run inference directly in the terminal:
llama-cli -hf sunil-pathak/gemma-3n-E2B-it-Q5_K_M:Q5_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 sunil-pathak/gemma-3n-E2B-it-Q5_K_M:Q5_K_M
# Run inference directly in the terminal:
./llama-cli -hf sunil-pathak/gemma-3n-E2B-it-Q5_K_M:Q5_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 sunil-pathak/gemma-3n-E2B-it-Q5_K_M:Q5_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf sunil-pathak/gemma-3n-E2B-it-Q5_K_M:Q5_K_M

Use Docker

docker model run hf.co/sunil-pathak/gemma-3n-E2B-it-Q5_K_M:Q5_K_M

LM Studio
Jan

vLLM

How to use sunil-pathak/gemma-3n-E2B-it-Q5_K_M with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "sunil-pathak/gemma-3n-E2B-it-Q5_K_M"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "sunil-pathak/gemma-3n-E2B-it-Q5_K_M",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/sunil-pathak/gemma-3n-E2B-it-Q5_K_M:Q5_K_M

Ollama
How to use sunil-pathak/gemma-3n-E2B-it-Q5_K_M with Ollama:
```
ollama run hf.co/sunil-pathak/gemma-3n-E2B-it-Q5_K_M:Q5_K_M
```

Unsloth Studio new

How to use sunil-pathak/gemma-3n-E2B-it-Q5_K_M 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 sunil-pathak/gemma-3n-E2B-it-Q5_K_M 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 sunil-pathak/gemma-3n-E2B-it-Q5_K_M to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for sunil-pathak/gemma-3n-E2B-it-Q5_K_M to start chatting

Docker Model Runner
How to use sunil-pathak/gemma-3n-E2B-it-Q5_K_M with Docker Model Runner:
```
docker model run hf.co/sunil-pathak/gemma-3n-E2B-it-Q5_K_M:Q5_K_M
```

Lemonade

How to use sunil-pathak/gemma-3n-E2B-it-Q5_K_M with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull sunil-pathak/gemma-3n-E2B-it-Q5_K_M:Q5_K_M

Run and chat with the model

lemonade run user.gemma-3n-E2B-it-Q5_K_M-Q5_K_M

List all available models

lemonade list

gemma-3n-E2B-it — GGUF (Q5_K_M)

📊 Performance Metrics

Hardware: Intel(R) Xeon(R) CPU @ 2.20GHz (4 vCPUs)
Size: 3.27 GB
Speed (Generation): 5.67 tokens/sec
Speed (Prompt): 10.31 tokens/sec
KV Cache Usage: 0.0143 GB
Quantization: Q5_K_M

🔷 Model Overview

This repository contains a GGUF quantized version of:

Base Model: gemma-3n-E2B-it
Format: GGUF (optimized for llama.cpp inference)
Precision: Q5_K_M
Efficiency Score: 1.7347 (TPS/GB)

GGUF format provides:

Fast loading via memory mapping
Single-file model distribution
Cross-platform compatibility
Efficient inference with llama.cpp

📦 Files

File	Description
`gemma-3n-E2B-it-Q5_K_M.gguf`	Quantized GGUF model file

⚙️ Technical Details

Parameter	Value
Architecture	gemma-3n-E2B-it
Format	GGUF
Precision	Q5_K_M
Runtime	llama.cpp
Benchmark Hardware	Intel(R) Xeon(R) CPU @ 2.20GHz (4 vCPUs)
Context Latency	30.64s
Memory (KV)	0.0143 GB

⚡ Why GGUF?

GGUF is designed for efficient inference:

Optimized for llama.cpp
Supports CPU and GPU inference
Single-file deployment
Memory-mapped loading for speed
Ideal for edge / local environments

⚠️ License & Usage

This is a converted derivative model.

You must comply with the original model license of gemma-3n-E2B-it
This is not an official release
No additional rights are granted
Original ownership remains with the base model creator

🚀 Quick Start (llama.cpp)

./llama-cli -m gemma-3n-E2B-it-Q5_K_M.gguf -p "Explain AI simply"

Downloads last month: 4

GGUF

Model size

4B params

Architecture

gemma3n

Hardware compatibility

5-bit