NVIDIA FastGen: Fast Generation from Diffusion Models
active 2026-01-27 → 2026-07-24 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 179 days from 2026-01-27 to 2026-07-24. Pushes: 6 total, peak 1 in a day. Pull requests: 3 total, peak 1 in a day. Issues: 9 total, peak 5 in a day. Comments: 12 total, peak 2 in a day. Stars: 236 total, peak 82 in a day.
- Pushes
- Pull requests
- Issues
- Comments
- Stars
Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.
Top contributors
Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity
| Contributor | Contributions | Pushes | PRs | Comments |
|---|---|---|---|---|
| juliusberner | 19 | 5 | 2 | 5 |
| cxlcl | 3 | 0 | 0 | 0 |
| Enderfga | 2 | 0 | 0 | 2 |
| greptile-apps[bot] | 2 | 0 | 0 | 2 |
| nom | 1 | 0 | 0 | 1 |
| weilinie | 1 | 1 | 0 | 0 |
| xilongzhou | 1 | 0 | 0 | 0 |
| danieltudosiu | 1 | 0 | 0 | 1 |
| linnanwang | 1 | 0 | 1 | 0 |
| meng-ustc | 1 | 0 | 0 | 0 |
| wlaud1001 | 1 | 0 | 0 | 1 |
| zzkk-spec | 1 | 0 | 0 | 0 |
Recent activity
Latest issues, pull requests and releases
- Issue comment#26greptile-apps[bot]2026-07-04 08:53Fix: per-frame timestep allocation for video2world (image2world) mode in CosmosPredict2
- Issue comment#26juliusberner2026-06-12 00:18Fix: per-frame timestep allocation for video2world (image2world) mode in CosmosPredict2
- Issue comment#25Enderfga2026-05-16 05:22Add AnyFlow algorithm (any-step video diffusion via flow maps)
- Issue comment#25Enderfga2026-05-16 04:40Add AnyFlow algorithm (any-step video diffusion via flow maps)
- Issue comment#20juliusberner2026-04-11 00:45OOM when training AR wan-2.1-14b model
- Issue comment#20juliusberner2026-04-09 17:50OOM when training AR wan-2.1-14b model
- Issue#20xilongzhou2026-04-08 23:15OOM when training AR wan-2.1-14b model
- Issue comment#17wlaud10012026-03-25 05:57Use fake_score features for DMD2 discriminator losses
- Issue#13juliusberner2026-02-26 22:27support FlashAttention JVP?
- Issue#13juliusberner2026-02-26 22:21support FlashAttention JVP?
- Issue#1juliusberner2026-02-26 22:19Support for Qwen-image series?
- Issue#1juliusberner2026-02-26 22:19Support for Qwen-image series?
- Issue#13zzkk-spec2026-02-26 04:13support FlashAttention JVP?
- Issue comment#1danieltudosiu2026-02-26 01:29Support for Qwen-image series?
- Pull request#12juliusberner2026-02-25 21:51
- Pull request#9juliusberner2026-02-21 00:48
- Pull request#8linnanwang2026-02-20 08:40
- Issue#4juliusberner2026-02-10 23:35Thank you for your excellent work. Do you support the models https://huggingface.co/robbyant/lingbot-world-base-cam, https://huggingface.co/Lightricks/LTX-2, and https://huggingface.co/Phr00t/LTX2-Rapid-Merges? Or do you have plans to support them
- Issue comment#4juliusberner2026-02-10 23:34Thank you for your excellent work. Do you support the models https://huggingface.co/robbyant/lingbot-world-base-cam, https://huggingface.co/Lightricks/LTX-2, and https://huggingface.co/Phr00t/LTX2-Rapid-Merges? Or do you have plans to support them
- Issue#2juliusberner2026-02-04 21:58Higher precision of the estimated jvp term in sCM loss
- Issue#2meng-ustc2026-02-02 22:42Higher precision of the estimated jvp term in sCM loss
- Issue comment#1nom2026-01-28 02:46Support for Qwen-image series?
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 236 stars here means stars gained during the window, not the repo's star count.