基于低秩分解的softmax兼有对分解后的低秩信息的还原
active 2023-08-15 → 2026-03-22 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 951 days from 2023-08-15 to 2026-03-22. Pushes: 9 total, peak 2 in a day. Pull requests: 0 total, peak 0 in a day. Issues: 54 total, peak 3 in a day. Comments: 49 total, peak 5 in a day. Stars: 362 total, peak 8 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 |
|---|---|---|---|---|
| tian-qing001 | 57 | 9 | 0 | 31 |
| magehrig | 4 | 0 | 0 | 1 |
| Asthestarsfalll | 3 | 0 | 0 | 1 |
| feiwushiwo | 3 | 0 | 0 | 2 |
| uestchjw | 3 | 0 | 0 | 2 |
| hankyul2 | 3 | 0 | 0 | 1 |
| nviwch | 3 | 0 | 0 | 1 |
| Rosal-1998 | 2 | 0 | 0 | 0 |
| YuhangGu4 | 2 | 0 | 0 | 1 |
| Zmoonzxz | 2 | 0 | 0 | 1 |
| CodeStarting-design | 2 | 0 | 0 | 0 |
| stella-von | 2 | 0 | 0 | 1 |
| JinchaoChen112 | 1 | 0 | 0 | 0 |
| Sahardastani | 1 | 0 | 0 | 1 |
| jiangsu415 | 1 | 0 | 0 | 0 |
| qescccczmr | 1 | 0 | 0 | 0 |
| mathias-gehrig-skydio | 1 | 0 | 0 | 1 |
| WateverOk | 1 | 0 | 0 | 0 |
| Zhangyuhaoo | 1 | 0 | 0 | 0 |
| thunderbolt-fire | 1 | 0 | 0 | 0 |
Recent activity
Latest issues, pull requests and releases
- Issue comment#32Sahardastani2025-08-29 14:50将softmax注意力替换为线性注意力后，模型训练出现Nan
- Issue#31tian-qing0012025-03-30 03:56Triton implementation
- Issue comment#30tian-qing0012025-03-30 03:56Question on code details
- Issue comment#30Haus2262025-03-30 03:45Question on code details
- Issue#33WateverOk2024-12-30 08:32关于位置编码的疑问
- Issue comment#32tian-qing0012024-12-27 11:05将softmax注意力替换为线性注意力后，模型训练出现Nan
- Issue comment#32jiaobin2024-12-23 08:46将softmax注意力替换为线性注意力后，模型训练出现Nan
- Issue#32wzh3262024-12-22 11:08将softmax注意力替换为线性注意力后，模型训练出现Nan
- Issue comment#31tian-qing0012024-11-25 08:39Triton implementation
- Issue comment#6CHN-0012024-11-22 09:15您好，您的工作非常有意义和参考价值，我像用来嫁接到CNN特征提取网络中，用于P345特征图之后，做一注意力机制，请问您代码支持吗？非常期待你的回信！十分感谢
- Issue#31hp-l332024-11-21 11:11Triton implementation
- Issue#30magehrig2024-11-09 02:53Question on code details
- Issue comment#30tian-qing0012024-11-09 02:27Question on code details
- Issue#30magehrig2024-11-07 13:23Question on code details
- Issue comment#29magehrig2024-11-07 10:20window size swin
- Issue#29magehrig2024-11-07 10:20window size swin
- Issue comment#29mathias-gehrig-skydio2024-11-07 10:19window size swin
- Issue comment#29tian-qing0012024-11-07 03:02window size swin
- Issue#28jiangsu4152024-10-14 13:35model diagram
- Issue#27Senqi-Guan2024-08-22 06:43训练代码的时候，cpu只有一个核心两个线程被调用，这是正常的吗？我的cpu是5950x，显卡是4090。
- Issue#26tian-qing0012024-07-17 11:39There is one place where the code is not well understood.
- Issue comment#26tian-qing0012024-07-17 11:38There is one place where the code is not well understood.
- Issue#26DABINB2024-07-16 02:27There is one place where the code is not well understood.
- Issue#25tian-qing0012024-07-15 10:38Subject: Inquiry About Lightweight Feature Extraction with Your Attention Mechanism
- Issue comment#25tian-qing0012024-07-11 08:43Subject: Inquiry About Lightweight Feature Extraction with Your Attention Mechanism
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 362 stars here means stars gained during the window, not the repo's star count.