Instructions to use arzaan789/gemma-3-4b-uncensored_ with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use arzaan789/gemma-3-4b-uncensored_ with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf arzaan789/gemma-3-4b-uncensored_:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf arzaan789/gemma-3-4b-uncensored_:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf arzaan789/gemma-3-4b-uncensored_:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf arzaan789/gemma-3-4b-uncensored_: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 arzaan789/gemma-3-4b-uncensored_:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf arzaan789/gemma-3-4b-uncensored_: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 arzaan789/gemma-3-4b-uncensored_:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf arzaan789/gemma-3-4b-uncensored_:Q4_K_M

Use Docker

docker model run hf.co/arzaan789/gemma-3-4b-uncensored_:Q4_K_M

LM Studio
Jan

vLLM

How to use arzaan789/gemma-3-4b-uncensored_ with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "arzaan789/gemma-3-4b-uncensored_"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "arzaan789/gemma-3-4b-uncensored_",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/arzaan789/gemma-3-4b-uncensored_:Q4_K_M

Ollama
How to use arzaan789/gemma-3-4b-uncensored_ with Ollama:
```
ollama run hf.co/arzaan789/gemma-3-4b-uncensored_:Q4_K_M
```
Unsloth Desktop
Docker Model Runner
How to use arzaan789/gemma-3-4b-uncensored_ with Docker Model Runner:
```
docker model run hf.co/arzaan789/gemma-3-4b-uncensored_:Q4_K_M
```

Lemonade

How to use arzaan789/gemma-3-4b-uncensored_ with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull arzaan789/gemma-3-4b-uncensored_:Q4_K_M

Run and chat with the model

lemonade run user.gemma-3-4b-uncensored_-Q4_K_M

List all available models

lemonade list

Atomic Chat

gemma-3-4b-uncensored

Uncensored variant of google/gemma-3-4b-it.

Method

Abliteration (strength=0.2) — refusal direction removed from all layers
LoRA fine-tune on Guilherme34/uncensor (2 epochs, r=16, alpha=32)
Re-abliteration (strength=0.35) — stronger pass to remove residual refusals

Eval Results

Split	Refused
Harmful (64 prompts)	0/64
Harmless (64 prompts)	0/64

Usage

llama-cli -m gemma_3_4b_uncensored.Q4_K_M.gguf -p "Your prompt here"

Training Config

Parameter	Value
Base Model	google/gemma-3-4b-it
Fine-tune Dataset	Guilherme34/uncensor
Epochs	2
LoRA r	16
LoRA alpha	32
Learning Rate	0.0002
Abliteration Strength	0.2
Re-abliteration Strength	0.35

Credits

Abliteration technique: andyrdt/refusal_direction
Weight editing: Sumandora/remove-refusals-with-transformers

Downloads last month: 334

GGUF

Model size

4B params

Architecture

gemma2

Hardware compatibility

2-bit

3-bit

4-bit

5-bit

6-bit

8-bit

16-bit

Model tree for arzaan789/gemma-3-4b-uncensored_

Base model

google/gemma-3-4b-pt

Finetuned

google/gemma-3-4b-it

Quantized

(489)

this model