Add pipeline tag and project links (#1)

- Add pipeline tag and project links (26276c9ac00dfd238408f2492e4b449cb2240307)


Co-authored-by: Niels Rogge <nielsr@users.noreply.huggingface.co>

README.md

@@ -1,26 +1,32 @@
 ---
-license: mit
 base_model:
 - z-lab/Qwen3-4B-DFlash-b16
+license: mit
+pipeline_tag: text-generation
 ---
+
 # Qwen3-4B-Ins-Draft-OPD
 
 This repository contains **Qwen3-4B-Ins-Draft-OPD**, a draft model for speculative decoding.
 
-The model is post-trained from [`z-lab/Qwen3-4B-DFlash-b16`](https://huggingface.co/z-lab/Qwen3-4B-DFlash-b16). It keeps the overall architecture and inference interface consistent with the original DFlash draft model, while further adapting the draft model through our post-training method.
+The model is post-trained from [`z-lab/Qwen3-4B-DFlash-b16`](https://huggingface.co/z-lab/Qwen3-4B-DFlash-b16). It keeps the overall architecture and inference interface consistent with the original DFlash draft model, while further adapting the draft model through the Draft-OPD post-training method.
+
+- **Paper:** [Draft-OPD: On-Policy Distillation for Speculative Draft Models](https://huggingface.co/papers/2605.29343)
+- **Project Page:** [https://www.haodilei.top/draft-opd/](https://www.haodilei.top/draft-opd/)
+- **Code:** [https://github.com/bingyang-lei/Draft-OPD](https://github.com/bingyang-lei/Draft-OPD)
 
 ## Model Details
 
 - **Base draft model:** [`Qwen3-4B(enable_thinking=true)`](https://huggingface.co/Qwen/Qwen3-4B)
 - **Model type:** Draft model for speculative decoding
 - **Architecture:** Same as the original DFlash draft model
-- **Post-training method:** Draft-OPD
+- **Post-training method:** Draft-OPD (On-Policy Distillation)
 
-## Performance and Training Method
+## Method Summary
 
-For detailed training procedures, evaluation settings, and performance results, please refer to our paper:
+Draft-OPD trains speculative draft models with on-policy target feedback. Instead of only learning from fixed target-generated trajectories (SFT), the drafter is supervised on draft-induced states exposed during speculative verification. This allows the drafter to learn from target feedback on both accepted and rejected proposals, focusing training on the draft-induced errors that limit speculative acceptance.
 
-**Paper:** [https://arxiv.org/abs/2605.29343]
+For detailed training procedures, evaluation settings, and performance results, please refer to the paper.
 
 ## Citation
 
@@ -35,4 +41,5 @@ If you find our work useful, please consider citing our paper:
       archivePrefix={arXiv},
       primaryClass={cs.CL},
       url={https://arxiv.org/abs/2605.29343}, 
-}
+}
+```