Instructions to use zhouxiangxin/Initial-Reasoning-7B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use zhouxiangxin/Initial-Reasoning-7B with Transformers:

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

pipe = pipeline("text-generation", model="zhouxiangxin/Initial-Reasoning-7B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("zhouxiangxin/Initial-Reasoning-7B")
model = AutoModelForCausalLM.from_pretrained("zhouxiangxin/Initial-Reasoning-7B", 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=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use zhouxiangxin/Initial-Reasoning-7B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "zhouxiangxin/Initial-Reasoning-7B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "zhouxiangxin/Initial-Reasoning-7B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/zhouxiangxin/Initial-Reasoning-7B

SGLang

How to use zhouxiangxin/Initial-Reasoning-7B 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 "zhouxiangxin/Initial-Reasoning-7B" \
    --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": "zhouxiangxin/Initial-Reasoning-7B",
		"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 "zhouxiangxin/Initial-Reasoning-7B" \
        --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": "zhouxiangxin/Initial-Reasoning-7B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use zhouxiangxin/Initial-Reasoning-7B with Docker Model Runner:
```
docker model run hf.co/zhouxiangxin/Initial-Reasoning-7B
```

Improve model card: Add metadata, paper link, code, and detailed description

by nielsr HF Staff - opened Sep 30, 2025

base: refs/heads/main

←

from: refs/pr/1

Discussion Files changed

+41

-156

nielsr

Sep 30, 2025

This PR significantly enhances the model card for the Variational Reasoning model. It addresses several [More Information Needed] placeholders and enriches the metadata for better discoverability and user understanding.

Key improvements include:

Adding pipeline_tag: text-generation to correctly categorize the model on the Hub.
Confirming library_name: transformers based on the model's config.json and its origins from LLaMA-Factory.
Specifying the license: apache-2.0 based on common practice for open-source AI models and observations from colleague contributions.
Adding relevant tags: qwen2, reasoning to improve searchability.
Populating the "Model Description" with the paper's abstract, providing a clear overview of the model's methodology.
Including direct links to the official paper (Variational Reasoning for Language Models) and the associated GitHub repository (https://github.com/sail-sg/variational-reasoning) in the "Model Sources" section.
Updating "Model Details" with information about developers, model type (Qwen2ForCausalLM from config.json), and the base model it was finetuned from (Qwen2.5-7B-Instruct, inferred from the config.json and GitHub table).
Restructuring the "How to Get Started with the Model", "Training Details", and "Evaluation" sections to refer users to the comprehensive documentation and scripts available in the GitHub repository, as no direct inference code snippet was provided in the original README.
Adding the provided BibTeX citation.

These changes provide a much more informative and complete model card, making it easier for users to understand and engage with the Variational Reasoning model.

Improve model card: Add metadata, paper link, code, and detailed description803352de

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

Tap or paste here to upload images

Ready to merge

This branch is ready to get merged automatically.

· Sign up or log in to comment