sgl-project / SpecForge
Train speculative decoding models effortlessly and port them smoothly to SGLang serving.
Why this job: matched by what the repo talks about: model, serv, decod, speculative
What it does
SpecForge is an ecosystem project developed by the SGLang team. It is a framework for training speculative decoding models so that you can smoothly port them over to the SGLang serving framework to speed up your inference.
Get started
No single install command. Read the steps in the README (opens in new tab)
- Maintenance
- Active
- Last commit today
- Stars
- 1,280
- +16 this week
- Licence
- MIT
- Allows business use
- Downloads for
- None listed
- No release files
- Latest version
- None yet
- No releases
- Built with
- Python
- 100% of the code
Is it looked after?
- Contributors
- 2
- Last 90 days
- Issues
- 58% closed
- 82 still open
- Pull requests
- 550 merged
- 178 waiting
- Releases
- None
- Never published
- No contributing guide
- No security policy
- No code of conduct
Similar repos
| Repo | Stars | Last commit | Licence | Downloads for |
|---|---|---|---|---|
| eagle-hq / EAGLE-4 | 1,226 | yesterday | Custom | No downloads |
| zengxiao-he / tessera | 557 | 5 months ago | Custom | No downloads |
| bitcloud-ai / Bittune | 246 | 6 days ago | MIT | Linux |
More: languages, topics and project record
Languages
- Python 100%
- Jinja <1%
- Shell <1%
- Dockerfile <1%
Topics
- eagle
- eagle3
- pytorch
- sglang
- fsdp
- training
- llm
Project record
- JobLocal LLMs
- StatusListed
- Commits1,457
- Forks392
- Watching12
- Stars in 28 days+59
- CreatedJun 9, 2025
- First seen by usNov 7, 2026
- CheckedNov 12, 2026