Skip to content

luketerry0/ssl-data-curation

View on GitHub ↗Related repositories →

PyTorch code for hierarchical k-means -- a data curation method for self-supervised learning

active 2024-09-192025-03-28 (UTC)

Complete coverage26,578 / 26,578 hourly files (100%) · 2 absent upstream2023-08-152026-08-26 (UTC)
Events
27
Pushes
11
Pull requests
2
Issues
9
Stars
3
Forks
0

Activity over time

Daily event counts in the loaded window

Line chart, 191 days from 2024-09-19 to 2025-03-28. Pushes: 11 total, peak 3 in a day. Pull requests: 2 total, peak 2 in a day. Issues: 9 total, peak 5 in a day. Comments: 1 total, peak 1 in a day. Stars: 3 total, peak 1 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
luketerry0221121
JayRothenberger1000

Recent activity

Latest issues, pull requests and releases

  • Pull request#9luketerry02025-03-28 18:13
  • Pull request#9luketerry02025-03-28 18:12
  • Issue#8luketerry02024-10-17 19:47
    Consider measuring inter cluster variation and intra cluster distance
  • Issue#7luketerry02024-10-17 19:45
    Try Jeffrey Divergence as a distance function
  • Issue#6luketerry02024-10-17 19:44
    Try cross entropy similarity as a distance function
  • Issue#5luketerry02024-10-17 19:43
    Test hierarchical embedding with cluster size stepping down logarithmically
  • Issue#4luketerry02024-10-17 19:43
    Normalize embeddings
  • Issue comment#1luketerry02024-10-02 18:05
    Edit SLURM Script preparation to use `torchrun` rather than `srun`
  • Issue#1luketerry02024-10-02 18:05
    Edit SLURM Script preparation to use `torchrun` rather than `srun`
  • Issue#3JayRothenberger2024-09-19 20:00
    Create a script that embeds and shards the data using DINOv2 and your sharding scheme of choice
  • Issue#2luketerry02024-09-19 19:59
    Edit script to specify data paths for each rank rather than for the script as a whole
  • Issue#1luketerry02024-09-19 19:59
    Edit SLURM Script preparation to use `torchrun` rather than `srun`

Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 3 stars here means stars gained during the window, not the repo's star count.