Simple and fast low-bit matmul kernels in CUDA / Triton
active 2024-08-15 → 2025-10-21 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 433 days from 2024-08-15 to 2025-10-21. Pushes: 321 total, peak 10 in a day. Pull requests: 24 total, peak 2 in a day. Issues: 40 total, peak 3 in a day. Comments: 140 total, peak 20 in a day. Stars: 322 total, peak 23 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 |
|---|---|---|---|---|
| mobicham | 427 | 317 | 21 | 76 |
| Emily-Ward | 23 | 0 | 0 | 18 |
| TsuWeiSung | 17 | 0 | 0 | 12 |
| hebiao064 | 14 | 0 | 0 | 12 |
| KeremTurgutlu | 11 | 0 | 2 | 5 |
| Leonard-debug | 8 | 0 | 0 | 6 |
| zju-stu-lizheng | 5 | 0 | 0 | 3 |
| MARD1NO | 4 | 0 | 0 | 1 |
| kurzdev | 4 | 0 | 0 | 2 |
| vlejd | 2 | 0 | 0 | 1 |
| jinz2014 | 2 | 0 | 0 | 0 |
| kadirnar | 2 | 0 | 0 | 1 |
| kaizizzzzzz | 2 | 0 | 0 | 1 |
| aur61 | 1 | 0 | 1 | 0 |
| kingflag | 1 | 0 | 0 | 0 |
| FC-Li | 1 | 0 | 0 | 0 |
| hadoop2xu | 1 | 0 | 0 | 0 |
| jerryzh168 | 1 | 0 | 0 | 0 |
| voipmonitor | 1 | 0 | 0 | 1 |
| zhangy659 | 1 | 0 | 0 | 1 |
Recent activity
Latest issues, pull requests and releases
- Issue comment#41voipmonitor2025-10-21 08:35GLM 4.5 4.6 MOE
- Releasemobicham2025-09-22 14:32v0.5.1
- Pull request#40mobicham2025-09-14 19:41
- Pull request#40mobicham2025-09-14 19:06
- Releasemobicham2025-08-18 10:58v0.5.0
- Pull request#37mobicham2025-08-08 08:04
- Issue comment#27mobicham2025-08-06 14:48Does it support the Transformers library?
- Pull request#36mobicham2025-06-14 17:08
- Issue#28mobicham2025-06-06 17:53torchao serialization
- Issue#33mobicham2025-06-06 17:53Strange performance profile on A100
- Releasemobicham2025-06-02 08:01v0.4.7
- Pull request#34mobicham2025-06-02 07:54
- Issue comment#33mobicham2025-05-15 07:40Strange performance profile on A100
- Issue comment#33vlejd2025-05-14 22:33Strange performance profile on A100
- Issue comment#33mobicham2025-05-14 12:50Strange performance profile on A100
- Issue#33vlejd2025-05-14 12:30Strange performance profile on A100
- Issue comment#31kurzdev2025-05-13 11:18`GemLiteLinear.pack()` multi-device storage of Tensor attributes causes issues with `accelerate`
- Issue#31kurzdev2025-05-13 11:18`GemLiteLinear.pack()` multi-device storage of Tensor attributes causes issues with `accelerate`
- Releasemobicham2025-05-13 11:00v0.4.6
- Issue comment#31mobicham2025-05-12 09:40`GemLiteLinear.pack()` multi-device storage of Tensor attributes causes issues with `accelerate`
- Pull request#32mobicham2025-05-12 09:17
- Pull request#32mobicham2025-05-12 09:16
- Issue comment#31kurzdev2025-05-12 08:28`GemLiteLinear.pack()` multi-device storage of Tensor attributes causes issues with `accelerate`
- Issue comment#31mobicham2025-05-12 08:13`GemLiteLinear.pack()` multi-device storage of Tensor attributes causes issues with `accelerate`
- Issue comment#31mobicham2025-05-09 08:33`GemLiteLinear.pack()` causes multi-device storage of Tensor attributes
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 322 stars here means stars gained during the window, not the repo's star count.