SpecForge
✅ Recently maintainedTrain speculative decoding models effortlessly and port them smoothly to SGLang serving.
1,255Stars
383Forks
+147-day stars
Python · MIT · Last push 2026-11-02
Official repository & installation ↗Star history
2025-09-292026-11-02
View historical data
| Date | Stars |
|---|---|
| 2026-10-04 | 1,196 |
| 2026-10-05 | 1,196 |
| 2026-10-06 | 1,197 |
| 2026-10-07 | 1,198 |
| 2026-10-08 | 1,202 |
| 2026-10-09 | 1,204 |
| 2026-10-10 | 1,208 |
| 2026-10-11 | 1,210 |
| 2026-10-12 | 1,214 |
| 2026-10-13 | 1,217 |
| 2026-10-14 | 1,219 |
| 2026-10-15 | 1,221 |
| 2026-10-16 | 1,223 |
| 2026-10-17 | 1,223 |
| 2026-10-18 | 1,225 |
| 2026-10-19 | 1,226 |
| 2026-10-20 | 1,227 |
| 2026-10-21 | 1,228 |
| 2026-10-22 | 1,234 |
| 2026-10-23 | 1,237 |
| 2026-10-24 | 1,239 |
| 2026-10-25 | 1,240 |
| 2026-10-26 | 1,241 |
| 2026-10-27 | 1,243 |
| 2026-10-28 | 1,246 |
| 2026-10-29 | 1,249 |
| 2026-10-30 | 1,251 |
| 2026-10-31 | 1,251 |
| 2026-11-01 | 1,253 |
| 2026-11-02 | 1,255 |
Get the project
git clone https://github.com/sgl-project/SpecForge.git
Related projects & alternative candidates
Related domains do not imply feature parity. Review official documentation and licenses.
LLMs-from-scratch
Step-by-step guide to building a GPT-like LLM from scratch in PyTorch.
★ 107,704vllmHigh-throughput, memory-efficient LLM inference and serving engine that lets developers deploy models faster without building serving infrastructure.
★ 96,161peft🤗 PEFT: State-of-the-art Parameter-Efficient Fine-Tuning.
★ 21,866