README.md

---
license: apache-2.0
library_name: transformers
pipeline_tag: text-generation
---

# JetSpec: Parallel Tree Drafting

JetSpec is an implementation of **parallel tree drafting** for fast LLM speculative decoding inference with up to 10x acceptance length, and 1000+ TPS on coding and math tasks using B200 GPUs. This repository contains the draft head model presented in [JetSpec: Breaking the Scaling Ceiling of Speculative Decoding with Parallel Tree Drafting](https://huggingface.co/papers/2606.18394).

A causal-parallel draft head proposes a token tree, and the frozen target model verifies the whole tree in one forward pass under a tree-causal attention mask. The accepted path is selected in accordance with the target's own logits, so decoding is lossless by construction.

For more details, please refer to the [Project Webpage](https://hao-ai-lab.github.io/JetSpec) and the [GitHub Repository](https://github.com/hao-ai-lab/JetSpec).

## Installation

Create an environment and install the package:

```bash
pip install -e '.[bench,kernel]'
```

## Usage

You can run speculative decoding using the lightweight Hugging Face-based reference implementation:

```python
from jetspec import LLM, SamplingParams

llm = LLM("Qwen/Qwen3-8B", attn_implementation="flash_attention_2")
out = llm.generate(
    "The three primary colors are",
    SamplingParams(temperature=0.0, max_new_tokens=64),
)
print(out["text"])
```

## Citation

```bibtex
@inproceedings{jetspec2026,
  title = {JetSpec: Breaking the Scaling Ceiling of Speculative Decoding with Parallel Tree Drafting},
  author = {Hu, Lanxiang and Feng, Zhaoxiang and Wu, Yulun and Yuan, Haoran and Zhao, Yujie and Qian, Yu-Yang and Wang, Bojun and Zhao, Peng and Jiang, Daxin and Zhu, Yibo and Rosing, Tajana and Zhang, Hao},
  year = {2026},
  url = {https://arxiv.org/abs/2606.18394},
  eprint = {2606.18394},
  note = {Preprint}
}
```