Instructions to use marcusmi4n/qwen2.5-3b-qnn-ready with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use marcusmi4n/qwen2.5-3b-qnn-ready with Transformers:

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

pipe = pipeline("text-generation", model="marcusmi4n/qwen2.5-3b-qnn-ready")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("marcusmi4n/qwen2.5-3b-qnn-ready")
model = AutoModelForCausalLM.from_pretrained("marcusmi4n/qwen2.5-3b-qnn-ready")
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]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use marcusmi4n/qwen2.5-3b-qnn-ready with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "marcusmi4n/qwen2.5-3b-qnn-ready"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "marcusmi4n/qwen2.5-3b-qnn-ready",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/marcusmi4n/qwen2.5-3b-qnn-ready

SGLang

How to use marcusmi4n/qwen2.5-3b-qnn-ready 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 "marcusmi4n/qwen2.5-3b-qnn-ready" \
    --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": "marcusmi4n/qwen2.5-3b-qnn-ready",
		"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 "marcusmi4n/qwen2.5-3b-qnn-ready" \
        --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": "marcusmi4n/qwen2.5-3b-qnn-ready",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use marcusmi4n/qwen2.5-3b-qnn-ready with Docker Model Runner:
```
docker model run hf.co/marcusmi4n/qwen2.5-3b-qnn-ready
```

Qwen 2.5 3B - QNN Ready

This repository contains the Qwen 2.5 3B model prepared for QNN deployment.

Model Details

Base Model: Qwen/Qwen2.5-3B
Architecture: Qwen2ForCausalLM
Parameters: ~3B
Languages: English, Chinese, and others
Format: PyTorch (Safetensors)

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("marcusmi4n/qwen2.5-3b-qnn-ready")
tokenizer = AutoTokenizer.from_pretrained("marcusmi4n/qwen2.5-3b-qnn-ready")

# Generate text
inputs = tokenizer("Hello, I am", return_tensors="pt")
outputs = model.generate(**inputs, max_length=50)
print(tokenizer.decode(outputs[0]))

QNN Conversion

This model can be converted to QNN format using the scripts in this repository:

# Quantize the model
python scripts/simple_quantize_abeja.py --model-path marcusmi4n/qwen2.5-3b-qnn-ready

# Convert to ONNX
python scripts/create_mock_onnx.py --model-path marcusmi4n/qwen2.5-3b-qnn-ready

# Compile for QNN
python scripts/mock_qnn_compile.py --model-path marcusmi4n/qwen2.5-3b-qnn-ready

License

Apache 2.0

Citation

@misc{qwen25-3b-qnn-ready,
  title={Qwen 2.5 3B - QNN Ready},
  author={QNN Conversion Pipeline},
  year={2025},
  url={https://huggingface.co/marcusmi4n/qwen2.5-3b-qnn-ready}
}

Downloads last month: 8

Safetensors

Model size

3B params

Tensor type

BF16

Model tree for marcusmi4n/qwen2.5-3b-qnn-ready

Base model

Qwen/Qwen2.5-3B

Finetuned

(460)

this model