Instructions to use WithinUsAI/Qwen3-Space.Agent.Claude.Uncensored-4B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use WithinUsAI/Qwen3-Space.Agent.Claude.Uncensored-4B with Transformers:

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

pipe = pipeline("text-generation", model="WithinUsAI/Qwen3-Space.Agent.Claude.Uncensored-4B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("WithinUsAI/Qwen3-Space.Agent.Claude.Uncensored-4B")
model = AutoModelForCausalLM.from_pretrained("WithinUsAI/Qwen3-Space.Agent.Claude.Uncensored-4B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.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(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use WithinUsAI/Qwen3-Space.Agent.Claude.Uncensored-4B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "WithinUsAI/Qwen3-Space.Agent.Claude.Uncensored-4B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "WithinUsAI/Qwen3-Space.Agent.Claude.Uncensored-4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/WithinUsAI/Qwen3-Space.Agent.Claude.Uncensored-4B

SGLang

How to use WithinUsAI/Qwen3-Space.Agent.Claude.Uncensored-4B 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 "WithinUsAI/Qwen3-Space.Agent.Claude.Uncensored-4B" \
    --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": "WithinUsAI/Qwen3-Space.Agent.Claude.Uncensored-4B",
		"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 "WithinUsAI/Qwen3-Space.Agent.Claude.Uncensored-4B" \
        --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": "WithinUsAI/Qwen3-Space.Agent.Claude.Uncensored-4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use WithinUsAI/Qwen3-Space.Agent.Claude.Uncensored-4B with Docker Model Runner:
```
docker model run hf.co/WithinUsAI/Qwen3-Space.Agent.Claude.Uncensored-4B
```

Performance on mlx

by nightmedia - opened 28 days ago

Discussion

nightmedia

28 days ago

•

edited 28 days ago

Great model, love the mix :)

I took a bit of time to make a quant and get some metrics:

           arc   arc/e boolq hswag obkqa piqa  wino
Qwen3-Space.Agent.Claude-Uncensored-4B
qx86-hi    0.575,0.768,0.861,0.707,0.420,0.779,0.697

nightmedia/Qwen3-4B-Agent-Claude-Gemini
qx86-hi    0.572,0.763,0.861,0.708,0.414,0.773,0.676

Very nice work

-G

GODsStrongestSoldier

WithIn Us AI org 28 days ago

Thanks, Really appreciate you taking the time to spin up the quant and run those metrics. It’s awesome to see it holding strong on ARC and BoolQ after the merge. Your Qwen3-4B-Agent-Claude-Gemini model was a massive piece of the puzzle here for getting that agentic behavior dialed in, so huge thanks for your work on that base, too. Glad you like the mix

nightmedia

28 days ago

I appreciate how well the merge went, considering the only way to build higher than that is to have an equally strong model, or an orthogonal set that expands a bit the view. It really worked well, I'd usually be happy if it "just dips a bit" but maintains arc, in your case the openbookqa stabilized on the 0.420 "safe bet", and also improved winogrande, which is hard. That's why I was curious :)

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment