Instructions to use xinlai/DeepSeekMath-RL-Step-DPO with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use xinlai/DeepSeekMath-RL-Step-DPO with Transformers:

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

pipe = pipeline("text-generation", model="xinlai/DeepSeekMath-RL-Step-DPO")
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("xinlai/DeepSeekMath-RL-Step-DPO")
model = AutoModelForCausalLM.from_pretrained("xinlai/DeepSeekMath-RL-Step-DPO", 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 xinlai/DeepSeekMath-RL-Step-DPO with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/xinlai/DeepSeekMath-RL-Step-DPO

SGLang

How to use xinlai/DeepSeekMath-RL-Step-DPO 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 "xinlai/DeepSeekMath-RL-Step-DPO" \
    --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": "xinlai/DeepSeekMath-RL-Step-DPO",
		"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 "xinlai/DeepSeekMath-RL-Step-DPO" \
        --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": "xinlai/DeepSeekMath-RL-Step-DPO",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use xinlai/DeepSeekMath-RL-Step-DPO with Docker Model Runner:
```
docker model run hf.co/xinlai/DeepSeekMath-RL-Step-DPO
```

xinlai commited on Jun 28, 2024

Commit

0134ded

verified ·

1 Parent(s): 6d4f1cf

Update README.md

Browse files

Files changed (1) hide show

README.md +13 -3

README.md CHANGED Viewed

@@ -1,3 +1,13 @@
----
-license: apache-2.0
----

+---
+license: apache-2.0
+---
+# Step-DPO: Step-wise Preference Optimization for Long-chain Reasoning of LLMs
+🖥️[Code](https://github.com/dvlab-research/Step-DPO) | 🤗[Data](https://huggingface.co/datasets/xinlai/Math-Step-DPO-10K) | 📄[Paper](https://arxiv.org/pdf/2406.18629)
+This repo contains the **DeepSeekMath-RL-Step-DPO** model for our paper **Step-DPO: Step-wise Preference Optimization for Long-chain Reasoning of LLMs**, **Step-DPO** is a simple, effective, and data-efficient method for boosting the mathematical reasoning ability of LLMs. Notably, Step-DPO, when applied to Qwen2-72B-Instruct, achieves scores of **70.8%** and **94.0%** on the test sets of **MATH** and **GSM8K** without bells and wistles, respectively, surpassing a series of closed-source models, including GPT-4-1106, Claude-3-Opus, and Gemini-1.5-Pro.
+## Contact
+Please submit an issue [here](https://github.com/dvlab-research/Step-DPO) or send me an email [here](mailto:xinlai@cse.cuhk.edu.hk).