Skip to content

We propose a new neural network architecture, ModuleFormer, that leverages modularity to improve the efficiency and flexibility of large language models.

active 2023-08-242025-09-13 (UTC)

Partial coverage20,522 / 26,278 hourly files (78%) · 2 absent upstream · 5,753 failed, retryable2023-08-152026-08-13 (UTC)— sampled evenly across the window, so rankings and trends hold; absolute counts scale up.
Events
228
Pushes
10
Pull requests
7
Issues
8
Stars
170
Forks
9

Activity over time

Daily event counts in the loaded window

Line chart, 752 days from 2023-08-24 to 2025-09-13. Pushes: 10 total, peak 3 in a day. Pull requests: 7 total, peak 4 in a day. Issues: 8 total, peak 2 in a day. Comments: 20 total, peak 6 in a day. Stars: 170 total, peak 49 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

ContributorContributionsPushesPRsComments
yikangshen2710610
YixinSong-e9005
LuciferianInk5013
flozi003002
zjwang211000

Recent activity

Latest issues, pull requests and releases

  • Pull request#8yikangshen2024-01-02 22:54
  • Issue comment#8yikangshen2024-01-02 22:54
    Implement transformers activation functions
  • Pull request#8LuciferianInk2024-01-02 10:46
  • Issue comment#6yikangshen2023-11-21 20:01
    torch.dtype is not respected
  • Issue comment#6LuciferianInk2023-11-19 04:26
    torch.dtype is not respected
  • Issue comment#6LuciferianInk2023-11-17 19:31
    torch.dtype is not respected
  • Issue comment#6LuciferianInk2023-11-17 19:21
    torch.dtype is not respected
  • Issue comment#6yikangshen2023-11-17 18:06
    torch.dtype is not respected
  • Issue#6LuciferianInk2023-11-17 05:53
    torch.dtype is not respected
  • Issue#4yikangshen2023-11-16 22:17
    Context length
  • Issue comment#5yikangshen2023-11-16 22:16
    the affects of cumulativing aux loss steps
  • Issue#5zjwang212023-11-14 03:29
    the affects of cumulativing aux loss steps
  • Issue#2YixinSong-e2023-09-26 10:33
    Do we have plans for 30B or larger models?
  • Issue comment#2yikangshen2023-09-23 05:13
    Do we have plans for 30B or larger models?
  • Issue comment#2flozi002023-09-21 15:29
    Do we have plans for 30B or larger models?
  • Issue comment#4yikangshen2023-09-20 20:57
    Context length
  • Issue comment#4flozi002023-09-19 15:31
    Context length
  • Issue comment#4yikangshen2023-09-19 15:25
    Context length
  • Issue#4flozi002023-09-19 13:55
    Context length
  • Issue comment#2YixinSong-e2023-09-19 01:21
    Do we have plans for 30B or larger models?
  • Issue comment#2yikangshen2023-09-18 21:24
    Do we have plans for 30B or larger models?
  • Pull request#3yikangshen2023-09-18 21:20
  • Issue comment#2yikangshen2023-09-15 15:01
    Do we have plans for 30B or larger models?
  • Issue comment#2YixinSong-e2023-09-15 05:57
    Do we have plans for 30B or larger models?
  • Issue#2YixinSong-e2023-09-15 05:55
    Do we have plans for 30B or larger models?

Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 170 stars here means stars gained during the window, not the repo's star count.