X-Coder-RL-Qwen2.5-7B

X-Coder-RL-Qwen2.5-7B is a strong code reasoning foundation model trained with RLVR on fully synthetic rl data, achieving strong reasoning performance on competitive programming.

Model Description

Base Model: IIGroup/X-Coder-SFT-Qwen2.5-7B
Training Method: GRPO
Training Data: IIGroup/X-Coder-RL-40k
Parameters: 7B

Training

This model was trained using the X-Coder RLVR framework. For training details and code, please refer to the X-Coder GitHub repository.

Performance

Performance on LiveCodeBench v5.

Recommended Inference Parameters

Parameter	Value
temperature	0.6
top_p	0.95
top_k	20 (or -1 to disable)
max_new_tokens	32768

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "IIGroup/X-Coder-RL-Qwen2.5-7B"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name, device_map="auto")

prompt = "Write a Python function to solve the two sum problem."
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
outputs = model.generate(
    **inputs,
    max_new_tokens=32768,
    temperature=0.6,
    top_p=0.95,
    top_k=20,
    do_sample=True
)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Citation

@misc{wu2026xcoderadvancingcompetitiveprogramming,
      title={X-Coder: Advancing Competitive Programming with Fully Synthetic Tasks, Solutions, and Tests}, 
      author={Jie Wu and Haoling Li and Xin Zhang and Jiani Guo and Jane Luo and Steven Liu and Yangyu Huang and Ruihang Chu and Scarlett Li and Yujiu Yang},
      year={2026},
      eprint={2601.06953},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2601.06953}, 
}