Add metadata and links to paper and code

Hi! I'm Niels from the community science team at Hugging Face.

I'm opening this pull request to improve the model card for **KernelGen-LM-32B**. This PR adds:
- **YAML metadata**: Including `pipeline_tag`, `library_name`, and `license`.
- **References**: Linking the model card to the original research paper and the official GitHub repository for easier access.

These changes help ensure the model is correctly categorized on the Hugging Face Hub and remains discoverable for researchers and developers.

README.md

@@ -1,19 +1,22 @@
 ---
 language:
 - en
-...
+license: apache-2.0
+library_name: transformers
+pipeline_tag: text-generation
 ---
+
 # AscendKernelGen/KernelGen-LM-32B
 
 ![License](https://img.shields.io/badge/License-Apache-yellow)
 [![arXiv](https://img.shields.io/badge/arXiv-2601.07160-b31b1b.svg)](https://arxiv.org/abs/2601.07160)
 
-KernelGen-LM-32B is a state-of-the-art domain-adaptive large language model specialized for low-level NPU kernel generation, specifically for the Huawei Ascend architecture using the AscendC programming language. Built upon the Qwen3-32B backbone, it is trained on the Ascend-CoT dataset and refined via reinforcement learning with execution feedback. It achieves unprecedented success rates in generating complex, functional hardware kernels, improving compilation success on L2 tasks from 0% (baseline) to 96.5% (Pass@10), while functional correctness achieves
-40.5% compared to the baseline’s complete failure.
+KernelGen-LM-32B is a state-of-the-art domain-adaptive large language model specialized for low-level NPU kernel generation, specifically for the Huawei Ascend architecture using the AscendC programming language. Built upon the Qwen3-32B backbone, it is trained on the Ascend-CoT dataset and refined via reinforcement learning with execution feedback. It achieves unprecedented success rates in generating complex, functional hardware kernels, improving compilation success on L2 tasks from 0% (baseline) to 96.5% (Pass@10), while functional correctness achieves 40.5% compared to the baseline’s complete failure.
 
-**Other artifacts:**
-* The **AscendKernelGen Technical Report** is published at https://arxiv.org/abs/2601.07160.
-* The **NPUKernelBench** evaluation framework is published at https://git.openi.org.cn/PCL-Benchmark/NPUKernelBench.
+**Links:**
+* **Paper:** [AscendKernelGen: A Systematic Study of LLM-Based Kernel Generation for Neural Processing Units](https://huggingface.co/papers/2601.07160)
+* **Code:** [GitHub - weich97/NPUKernelBench](https://github.com/weich97/NPUKernelBench)
+* **Evaluation Framework:** [NPUKernelBench (OpenI)](https://git.openi.org.cn/PCL-Benchmark/NPUKernelBench)
 
 ## Introduction
 
@@ -25,6 +28,7 @@ Our framework, **AscendKernelGen (AKGen)**, bridges the gap between general-purp
 * **Performance:** The model demonstrates siginificant improvement on complex Level-2 kernels compared to baselines, and effectively solving tasks where general-purpose models (like Qwen3, Llama3.1) fail completely.
 
 ## Citation
+```bibtex
 @article{cao2026ascendkernelgen,
   title={AscendKernelGen: A Systematic Study of LLM-Based Kernel Generation for Neural Processing Units},
   author={Xinzi Cao and Jianyang Zhai and Pengfei Li and Zhiheng Hu and Cen Yan and Bingxu Mu and Guanghuan Fang and Bin She and Jiayu Li and Yihan Su and Dongyang Tao and Xiansong Huang and Fan Xu and Feidiao Yang and Yao Lu and Chang-Dong Wang and Yutong Lu and Weicheng Xue and Bin Zhou and Yonghong Tian},
@@ -32,3 +36,4 @@ Our framework, **AscendKernelGen (AKGen)**, bridges the gap between general-purp
   year={2026},
   url=https://arxiv.org/abs/2601.07160
 }
+```