增量预训练：Progressive LLaMA with Block Expansion.
active 2024-01-04 → 2025-09-22 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 628 days from 2024-01-04 to 2025-09-22. Pushes: 25 total, peak 4 in a day. Pull requests: 0 total, peak 0 in a day. Issues: 41 total, peak 2 in a day. Comments: 92 total, peak 5 in a day. Stars: 504 total, peak 38 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 |
|---|---|---|---|---|
| hills-code | 63 | 23 | 0 | 40 |
| bestpredicts | 7 | 0 | 0 | 3 |
| hzgdeerHo | 6 | 0 | 0 | 5 |
| denghj3 | 4 | 0 | 0 | 3 |
| tammypi | 4 | 0 | 0 | 4 |
| Abolfazl-kr | 3 | 0 | 0 | 1 |
| raghavgarg97 | 3 | 0 | 0 | 3 |
| shamanez | 3 | 0 | 0 | 1 |
| FanXiang | 3 | 0 | 0 | 3 |
| GOOD-N-LCM | 3 | 0 | 0 | 2 |
| zhuxiaobin | 3 | 0 | 0 | 2 |
| RuipingWang1986 | 3 | 0 | 0 | 1 |
| yumingfan-0219 | 3 | 0 | 0 | 1 |
| yhyu13 | 3 | 0 | 0 | 2 |
| win10ogod | 2 | 0 | 0 | 1 |
| icoderzqliu | 2 | 0 | 0 | 2 |
| HuXinjing | 2 | 0 | 0 | 1 |
| CharlinChen | 2 | 0 | 0 | 0 |
| HelloWorldLTY | 2 | 0 | 0 | 1 |
| ChrisXULC | 2 | 0 | 0 | 0 |
Recent activity
Latest issues, pull requests and releases
- Issue#34shutttttdown2025-01-05 07:40Pretrain data source request
- Issue comment#27season1blue2024-10-29 09:11finetune_cosmopedia.sh如何训练出来8B模型
- Issue comment#31tammypi2024-10-21 07:16关于论文中通用能力榜单几乎没有下降，部分反而有提升
- Issue comment#20tammypi2024-10-17 11:43请教下训练的显存需求
- Issue comment#20tammypi2024-10-17 11:11请教下训练的显存需求
- Issue comment#20tammypi2024-10-17 08:15请教下训练的显存需求
- Issue#33lucasjinreal2024-08-29 11:50这个方法可以扩展到vit类的视觉encode上吗？
- Issue#32jqtian1232024-08-23 03:41请教大佬可以训练qewn2-7b吗
- Issue comment#30ymourenya2024-08-07 15:13关于运行流程
- Issue comment#28JiwenJ2024-07-26 05:27关于零初始化和扩展层的位置
- Issue#31bestpredicts2024-06-28 08:47关于论文中通用能力榜单几乎没有下降，部分反而有提升
- Issue#29bestpredicts2024-06-28 08:47训练到10B tokens 时loss就收敛了 无法下降
- Issue comment#31bestpredicts2024-06-27 23:52关于论文中通用能力榜单几乎没有下降
- Issue#31bestpredicts2024-06-27 23:50关于论文中通用能力榜单几乎没有下降
- Issue comment#30GOOD-N-LCM2024-06-26 09:15关于运行流程
- Issue comment#30hills-code2024-06-26 08:24关于运行流程
- Issue comment#30GOOD-N-LCM2024-06-26 08:20关于运行流程
- Issue comment#30hills-code2024-06-26 06:49关于运行流程
- Issue#30GOOD-N-LCM2024-06-26 06:24关于运行流程
- Issue comment#29bestpredicts2024-06-13 01:08训练到10B tokens 时loss就收敛了 无法下降
- Issue comment#28bestpredicts2024-06-13 01:06关于零初始化和扩展层的位置
- Issue comment#28hills-code2024-06-08 03:47关于零初始化和扩展层的位置
- Issue#29bestpredicts2024-06-07 23:36训练到10B tokens 时loss就收敛了 无法下降
- Issue comment#28ouyanxi11252024-06-07 10:42关于零初始化和扩展层的位置
- Issue comment#28hills-code2024-06-07 10:00关于零初始化和扩展层的位置
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 504 stars here means stars gained during the window, not the repo's star count.