Instructions to use TMLR-Group-HF/GT-Qwen3-4B-Base-OpenRS with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use TMLR-Group-HF/GT-Qwen3-4B-Base-OpenRS with Transformers:

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

pipe = pipeline("text-generation", model="TMLR-Group-HF/GT-Qwen3-4B-Base-OpenRS")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("TMLR-Group-HF/GT-Qwen3-4B-Base-OpenRS")
model = AutoModelForCausalLM.from_pretrained("TMLR-Group-HF/GT-Qwen3-4B-Base-OpenRS")
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 TMLR-Group-HF/GT-Qwen3-4B-Base-OpenRS with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/TMLR-Group-HF/GT-Qwen3-4B-Base-OpenRS

SGLang

How to use TMLR-Group-HF/GT-Qwen3-4B-Base-OpenRS 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 "TMLR-Group-HF/GT-Qwen3-4B-Base-OpenRS" \
    --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": "TMLR-Group-HF/GT-Qwen3-4B-Base-OpenRS",
		"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 "TMLR-Group-HF/GT-Qwen3-4B-Base-OpenRS" \
        --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": "TMLR-Group-HF/GT-Qwen3-4B-Base-OpenRS",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use TMLR-Group-HF/GT-Qwen3-4B-Base-OpenRS with Docker Model Runner:
```
docker model run hf.co/TMLR-Group-HF/GT-Qwen3-4B-Base-OpenRS
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Co-rewarding: GT-GRPO Qwen3-4B-Base trained on OpenRS

This model is a checkpoint of the Qwen3-4B-Base model, specifically trained using the GT-GRPO (Ground-Truth Guided Policy Optimization) method on the OpenRS training set. It is part of the work presented in the paper Co-rewarding: Stable Self-supervised RL for Eliciting Reasoning in Large Language Models.

Paper Abstract Summary

The paper introduces Co-rewarding, a novel self-supervised reinforcement learning (RL) framework designed to enhance the reasoning abilities of large language models (LLMs). It aims to achieve training stability by leveraging complementary supervision from multiple views. The framework is instantiated in two ways:

Co-rewarding-I: A data-side approach that derives reward signals from contrastive agreement across semantically analogous questions.
Co-rewarding-II: A model-side approach that uses a slowly-updated reference teacher with pseudo labels for self-distillation. These instantiations introduce discrepancies to prevent training collapse on trivial reasoning solutions. Empirically, Co-rewarding demonstrates stable training and superior performance compared to other self-rewarding baselines on various mathematical reasoning benchmarks, notably surpassing ground-truth RLVR in some cases.

GitHub Repository

For comprehensive details on the Co-rewarding framework, installation instructions, training scripts, and additional checkpoints, please visit the official GitHub repository: https://github.com/tmlr-group/Co-rewarding

Citation

If you use this model or any resources from the Co-rewarding project, please cite the following paper:

@article{zhang2025co,
  title={Co-rewarding: Stable Self-supervised RL for Eliciting Reasoning in Large Language Models},
  author={Zhang, Zizhuo and Zhu, Jianing and Ge, Xinmu and Zhao, Zihua and Zhou, Zhanke and Li, Xuan and Feng, Xiao and Yao, Jiangchao and Han, Bo},
  journal={arXiv preprint arXiv:2508.00410},
  year={2025}
}