Instructions to use Qwen/Qwen2.5-Math-7B-Instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Qwen/Qwen2.5-Math-7B-Instruct with Transformers:

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

pipe = pipeline("text-generation", model="Qwen/Qwen2.5-Math-7B-Instruct")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen2.5-Math-7B-Instruct")
model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-Math-7B-Instruct")
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 Qwen/Qwen2.5-Math-7B-Instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Qwen/Qwen2.5-Math-7B-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": "Qwen/Qwen2.5-Math-7B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Qwen/Qwen2.5-Math-7B-Instruct

SGLang

How to use Qwen/Qwen2.5-Math-7B-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 "Qwen/Qwen2.5-Math-7B-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": "Qwen/Qwen2.5-Math-7B-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 "Qwen/Qwen2.5-Math-7B-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": "Qwen/Qwen2.5-Math-7B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Qwen/Qwen2.5-Math-7B-Instruct with Docker Model Runner:
```
docker model run hf.co/Qwen/Qwen2.5-Math-7B-Instruct
```

Independent evaluation results

by yaronr - opened Sep 26, 2024

Discussion

yaronr

Sep 26, 2024

Dear Qwen Math team,

I'm pleased to share our independent evaluation of the model using our implementation of the MMLU-Pro benchmark.

It's puzzling why Qwen 2.5 Math 7B Instruct didn't outperform the general 7B Instruct model in the Math category on our benchmark. We would love to hear your thoughts on the subject.

We know that MMLU-Pro is probably not the best benchmark for math, but this is our starting point for all models, and we decided to share this with you nonetheless, hoping you may find this useful.

JustinLin610

Sep 27, 2024

Any more info about reproduction?

yaronr

Sep 27, 2024

@JustinLin610 We are open. Let's discuss over zoom, we can show you our platform and discuss if there is interest in open sourcing it.

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