Add model card and metadata

README.md

@@ -1 +1,30 @@
-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 draft model checkpoint introduced in the paper [TAPS: Task Aware Proposal Distributions for Speculative Sampling](https://huggingface.co/papers/2603.27027).
+
+## Introduction
+TAPS studies how the training distribution of draft models shapes the quality and efficiency of speculative decoding. By utilizing task-specific training (e.g., on MathInstruct or ShareGPT), draft models can achieve significant acceleration on matched downstream workloads. This repository provides specialized drafters (such as HASS or EAGLE-2 variants) designed for use with larger verifier models like Meta-Llama-3-8B-Instruct.
+
+## Resources
+- **Paper:** [TAPS: Task Aware Proposal Distributions for Speculative Sampling](https://arxiv.org/abs/2603.27027)
+- **GitHub Repository:** [Moe-Zbeeb/TAPS](https://github.com/Moe-Zbeeb/TAPS)
+
+## Model Details
+This checkpoint is a lightweight, single-layer Llama-style drafter (~0.8B parameters). It is intended to be used in a speculative decoding pipeline to propose future tokens, which are then verified in parallel by a larger target model.
+
+## Citation
+If you find this work useful, please cite:
+```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}
+}
+```