Instructions to use Geraldxm/Math-Qwen3-1.7B-Qwen3-4B-Non-Thinking-RL-Math-Step500 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Geraldxm/Math-Qwen3-1.7B-Qwen3-4B-Non-Thinking-RL-Math-Step500 with Transformers:

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

pipe = pipeline("text-generation", model="Geraldxm/Math-Qwen3-1.7B-Qwen3-4B-Non-Thinking-RL-Math-Step500")
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("Geraldxm/Math-Qwen3-1.7B-Qwen3-4B-Non-Thinking-RL-Math-Step500")
model = AutoModelForCausalLM.from_pretrained("Geraldxm/Math-Qwen3-1.7B-Qwen3-4B-Non-Thinking-RL-Math-Step500", 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 Geraldxm/Math-Qwen3-1.7B-Qwen3-4B-Non-Thinking-RL-Math-Step500 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Geraldxm/Math-Qwen3-1.7B-Qwen3-4B-Non-Thinking-RL-Math-Step500"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Geraldxm/Math-Qwen3-1.7B-Qwen3-4B-Non-Thinking-RL-Math-Step500",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Geraldxm/Math-Qwen3-1.7B-Qwen3-4B-Non-Thinking-RL-Math-Step500

SGLang

How to use Geraldxm/Math-Qwen3-1.7B-Qwen3-4B-Non-Thinking-RL-Math-Step500 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 "Geraldxm/Math-Qwen3-1.7B-Qwen3-4B-Non-Thinking-RL-Math-Step500" \
    --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": "Geraldxm/Math-Qwen3-1.7B-Qwen3-4B-Non-Thinking-RL-Math-Step500",
		"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 "Geraldxm/Math-Qwen3-1.7B-Qwen3-4B-Non-Thinking-RL-Math-Step500" \
        --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": "Geraldxm/Math-Qwen3-1.7B-Qwen3-4B-Non-Thinking-RL-Math-Step500",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Geraldxm/Math-Qwen3-1.7B-Qwen3-4B-Non-Thinking-RL-Math-Step500 with Docker Model Runner:
```
docker model run hf.co/Geraldxm/Math-Qwen3-1.7B-Qwen3-4B-Non-Thinking-RL-Math-Step500
```

Math-Qwen3-1.7B-Qwen3-4B-Non-Thinking-RL-Math-Step500

Paper · GitHub · Project · Collection

This repository contains an on-policy distillation (OPD) student checkpoint released with Towards Understanding On-Policy Distillation through the Lens of Test-Time Scaling. Its base model, teacher, and training data are listed below.

The figure reports paper results across math settings; it is not a scorecard for this checkpoint alone.

Checkpoint details

Base model: Qwen/Qwen3-1.7B
Teacher model: Keven16/Qwen3-4B-Non-Thinking-RL-Math-Step500
Training data: DAPO-Math-17K, using the deduplicated OPD training set prepared in math-vault.

Step500 is part of the teacher's name, not a stated student training step.

More information

See the paper and GitHub repository for training settings, evaluation protocols, and analysis. Upstream model and data terms remain applicable; see third-party notices.

Citation

@misc{ge2026understandingonpolicydistillationlens,
  title={Towards Understanding On-Policy Distillation through the Lens of Test-Time Scaling},
  author={Xinmu Ge and Zizhuo Zhang and Yu Huang and Jianing Zhu and Lin Yuan and Wanli Gu and Weichang Wu and Weiran Huang and Bo Han and Xiaolu Zhang and Jiangchao Yao},
  year={2026},
  eprint={2608.11829},
  archivePrefix={arXiv},
  primaryClass={cs.LG},
  doi={10.48550/arXiv.2608.11829},
  url={https://arxiv.org/abs/2608.11829}
}