Add model card for TAPS draft model

README.md

@@ -1 +1,32 @@
-arxiv.org/abs/2603.27027
+---
+license: apache-2.0
+library_name: transformers
+pipeline_tag: text-generation
+---
+
+# TAPS: Task-Aware Proposal Distributions for Speculative Sampling
+
+This repository contains a lightweight draft model introduced in the paper [TAPS: Task Aware Proposal Distributions for Speculative Sampling](https://huggingface.co/papers/2603.27027).
+
+## Overview
+Speculative decoding accelerates autoregressive generation by letting a lightweight draft model propose future tokens that a larger target model then verifies in parallel. TAPS (Task-Aware Proposal Distributions) investigates how the draft training distribution shapes speculative decoding quality.
+
+The research demonstrates that task-specific training yields significant specialization:
+- **Task-aware specialization:** Drafts trained on specific datasets (like MathInstruct or ShareGPT) perform significantly better on domain-matched workloads.
+- **Composition strategies:** The paper explores combining specialized drafters at inference time via confidence-based routing and merged-tree verification, which outperform naive weight averaging.
+
+This model is a lightweight (~0.8B parameters) Llama-style drafter designed to be used in conjunction with a larger verifier model, such as Meta-Llama-3-8B-Instruct.
+
+## Resources
+- **Paper:** [https://arxiv.org/abs/2603.27027](https://arxiv.org/abs/2603.27027)
+- **GitHub Repository:** [https://github.com/Moe-Zbeeb/TAPS](https://github.com/Moe-Zbeeb/TAPS)
+
+## Citation
+```bibtex
+@article{zbib2026taps,
+  title={TAPS: Task Aware Proposal Distributions for Speculative Sampling},
+  author={Zbib, Mohamad and Bazzi, Mohamad and Mohanna, Ammar and Ghanem, Bernard and Hammoud, Hasan Abed Al Kader},
+  year={2026},
+  note={Technical report}
+}
+```