Skip to content

A project to improve skills of large language models

active 2025-04-082025-10-08 (UTC)

Partial coverage11,158 / 11,795 hourly files (95%) · 2 absent upstream · 634 failed, retryable2025-04-062026-08-10 (UTC)— sampled evenly across the window, so rankings and trends hold; absolute counts scale up.
Events
4.3K
Pushes
1.2K
Pull requests
411
Issues
75
Stars
158
Forks
29

Activity over time

Daily event counts in the loaded window

Line chart, 184 days from 2025-04-08 to 2025-10-08. Pushes: 1,217 total, peak 35 in a day. Pull requests: 411 total, peak 11 in a day. Issues: 75 total, peak 10 in a day. Comments: 1,102 total, peak 52 in a day. Stars: 158 total, peak 10 in a day.

  • Pushes
  • Pull requests
  • Issues
  • Comments
  • Stars

Top contributors

Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity

ContributorContributionsPushesPRsComments
Kipok1.3K324202440
shtoshni4832484797
coderabbitai[bot]33000167
gwarmstrong2691011177
activatedgeek169601057
wedu-nvidia154721635
titu199412730754
wasiahmad124801022
smahdavi4116581324
fayejf5218613
ekmb422946
i-vainn382754
darraghdog291774
shuoyangd22549
SeanNaren211053
ludwig-n201631
vmendelev19526
artbataev19636
jiacheng-xu19536
AdamRajfer171113

Recent activity

Latest issues, pull requests and releases

  • Issue#909activatedgeek2025-10-08 14:43
    Add copyright checks
  • Issue comment#908coderabbitai[bot]2025-10-08 08:56
    [DRAFT] BFCLv4 support
  • Issue#904wedu-nvidia2025-10-08 04:59
    Add support for interactive node for GB200 cluster
  • Issue comment#906wedu-nvidia2025-10-07 23:01
    Add qos slurm parameter
  • Pull request#907Kipok2025-10-07 22:49
  • Issue comment#907coderabbitai[bot]2025-10-07 22:36
    Add concurrent semaphore control to llm base class
  • Issue#834wedu-nvidia2025-10-07 21:05
    Add avg number of timeouts as metric when python code execution is used
  • Issue comment#905coderabbitai[bot]2025-10-07 19:28
    Slurm tests for code execution timeouts
  • Pull request#905i-vainn2025-10-07 19:28
  • Issue#903Kipok2025-10-07 18:25
    Remove httpx limits and / or use semaphor in base model to limit concurrent requests
  • Pull request#899Kipok2025-10-07 18:25
  • Issue comment#899smahdavi42025-10-07 18:22
    Control max_concurrent_requests in subclasses with parallel generations
  • Issue comment#899Kipok2025-10-07 18:21
    Control max_concurrent_requests in subclasses with parallel generations
  • Issue comment#899Kipok2025-10-07 18:16
    Control max_concurrent_requests in subclasses with parallel generations
  • Issue#901Kipok2025-10-07 18:09
    Refactor BLEU metrics computation so it reuses parent method functions from pass@k
  • Issue comment#899Kipok2025-10-07 17:55
    Control max_concurrent_requests in subclasses with parallel generations
  • Pull request#900Kipok2025-10-07 17:54
  • Pull request#900i-vainn2025-10-07 08:47
  • Issue comment#889Kipok2025-10-07 04:45
    Add responses api type
  • Issue comment#899coderabbitai[bot]2025-10-07 04:43
    Control max_concurrent_requests in subclasses with parallel generations
  • Issue comment#891wedu-nvidia2025-10-07 03:50
    update nemo rl command
  • Issue#898Kipok2025-10-07 01:34
    Add support for TerminalBench
  • Issue comment#897Kipok2025-10-07 00:25
    Add check for mount paths before running experiment and QoL
  • Issue comment#897coderabbitai[bot]2025-10-06 23:38
    Add check for mount paths before running experiment and QoL
  • Pull request#897titu19942025-10-06 23:38

Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 158 stars here means stars gained during the window, not the repo's star count.