Instructions to use sillykiwi/Gemma3-4B-CodeCenturion with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use sillykiwi/Gemma3-4B-CodeCenturion with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="sillykiwi/Gemma3-4B-CodeCenturion")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForImageTextToText

processor = AutoProcessor.from_pretrained("sillykiwi/Gemma3-4B-CodeCenturion")
model = AutoModelForImageTextToText.from_pretrained("sillykiwi/Gemma3-4B-CodeCenturion")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use sillykiwi/Gemma3-4B-CodeCenturion with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "sillykiwi/Gemma3-4B-CodeCenturion"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "sillykiwi/Gemma3-4B-CodeCenturion",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/sillykiwi/Gemma3-4B-CodeCenturion

SGLang

How to use sillykiwi/Gemma3-4B-CodeCenturion with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "sillykiwi/Gemma3-4B-CodeCenturion" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "sillykiwi/Gemma3-4B-CodeCenturion",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "sillykiwi/Gemma3-4B-CodeCenturion" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "sillykiwi/Gemma3-4B-CodeCenturion",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use sillykiwi/Gemma3-4B-CodeCenturion with Docker Model Runner:
```
docker model run hf.co/sillykiwi/Gemma3-4B-CodeCenturion
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Made using mergekit-gui.

vanta-research/scout-4b has an absolutely amazing personality and is good at cooperative problem solving (something I find most models are not good at).
GetSoloTech/Gemma3-Code-Reasoning-4B is a strong coder.
coder3101/gemma-3-4b-it-heretic as a base (I am a professional who understands that this model is a tool and that I am responsable for how I use it). I also feel like people don't talk enough about how abliteration reduces hallucinations when asking the bot difficult questions.

The end result is hopefully useful for programming and {cyber,information}-security.

Quantizations courtesy of team mradermacher

GGUF: GGUF
iMatrix: iMatrix-GGUF

name: Gemma3-4B-CodeCenturion
base_model: coder3101/gemma-3-4b-it-heretic
dtype: bfloat16
merge_method: model_stock
models:
  - model: vanta-research/scout-4b
  - model: GetSoloTech/Gemma3-Code-Reasoning-4B
tokenizer_source: coder3101/gemma-3-4b-it-heretic

Downloads last month: 2

Safetensors

Model size

4B params

Tensor type

BF16

Model tree for sillykiwi/Gemma3-4B-CodeCenturion

GetSoloTech/Gemma3-Code-Reasoning-4B

coder3101/gemma-3-4b-it-heretic

vanta-research/scout-4b

Merge model

this model

Quantizations

2 models