Instructions to use DavidAU/Qwen3-30B-A3B-YOYO-V2-Claude-4.6-Opus-High-INSTRUCT with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use DavidAU/Qwen3-30B-A3B-YOYO-V2-Claude-4.6-Opus-High-INSTRUCT with Transformers:

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

pipe = pipeline("text-generation", model="DavidAU/Qwen3-30B-A3B-YOYO-V2-Claude-4.6-Opus-High-INSTRUCT")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("DavidAU/Qwen3-30B-A3B-YOYO-V2-Claude-4.6-Opus-High-INSTRUCT")
model = AutoModelForCausalLM.from_pretrained("DavidAU/Qwen3-30B-A3B-YOYO-V2-Claude-4.6-Opus-High-INSTRUCT", device_map="auto")
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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use DavidAU/Qwen3-30B-A3B-YOYO-V2-Claude-4.6-Opus-High-INSTRUCT with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "DavidAU/Qwen3-30B-A3B-YOYO-V2-Claude-4.6-Opus-High-INSTRUCT"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "DavidAU/Qwen3-30B-A3B-YOYO-V2-Claude-4.6-Opus-High-INSTRUCT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/DavidAU/Qwen3-30B-A3B-YOYO-V2-Claude-4.6-Opus-High-INSTRUCT

SGLang

How to use DavidAU/Qwen3-30B-A3B-YOYO-V2-Claude-4.6-Opus-High-INSTRUCT 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 "DavidAU/Qwen3-30B-A3B-YOYO-V2-Claude-4.6-Opus-High-INSTRUCT" \
    --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": "DavidAU/Qwen3-30B-A3B-YOYO-V2-Claude-4.6-Opus-High-INSTRUCT",
		"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 "DavidAU/Qwen3-30B-A3B-YOYO-V2-Claude-4.6-Opus-High-INSTRUCT" \
        --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": "DavidAU/Qwen3-30B-A3B-YOYO-V2-Claude-4.6-Opus-High-INSTRUCT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Unsloth Desktop
Docker Model Runner
How to use DavidAU/Qwen3-30B-A3B-YOYO-V2-Claude-4.6-Opus-High-INSTRUCT with Docker Model Runner:
```
docker model run hf.co/DavidAU/Qwen3-30B-A3B-YOYO-V2-Claude-4.6-Opus-High-INSTRUCT
```

Sampling parameters?

by McG-221 - opened Feb 14

Discussion

McG-221

Feb 14

Hi David,
tried the following sampling parameters, they don't seem to be correct for this model?

Thanks for looking into this 🙌

DavidAU

Owner Feb 15

Please clarify what is happening? or not? ...
Parameters look good.

Note: This is a pure instruct model, no thinking TRACES.
Gemini dataset was embedded deep into the model ; but model will not produce "think" blocks.

Thanks
DAVE

McG-221

Feb 15

Thanks a lot for your reply! I’m having trouble with my usual logic puzzles (A knows about B and B about C kind of stuff), but this seems to be the case with all models… seems like LM Studio is buggy at the moment, they introduced batch processing recently (and a few bugs, too, it seems).

Much appreciated, will give it another try, once I got my environment fixed! 👍

DavidAU

Owner Feb 15

You may want to try the Claude and/or Gemini Thinking Qwen 30B-A3B I have tuned:

https://huggingface.co/DavidAU/models?search=30b-a3b

NOTE: Some are still awaiting quants as of this writing.

McG-221

Feb 15

Will give it a try, thanks a lot!

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