165 repositories starred since Apr 2010. Mostly machine learning and data, and lately AI and language models.
I like to train Deep Neural Nets on large datasets.
Over time
Each line is a starred repository at the date it was starred. Height is its GitHub stars. Drag across the graph or pick a year to filter the list below.
- Apr 2010: first stars, tweepy, jqfundamentals
- 2017–2019: 5 stars in 3 years
- 2024: 30 stars, 18 in AI & LLMs
- AI & LLMs49
- ML & data70
- Web & apps13
- Infra & backend4
- Libraries & tools4
- Systems & graphics12
- Learning & other13
Drift
Since Jan 2023, AI and language models went from 6% to 65% of new stars, while machine learning and data fell from 61% to 15%.
- Libraries & tools 12%
- Web & apps 12%
- AI & LLMs 69%
Terrain
Categories are regions. Where stars cluster, the ground rises. The most-starred repositories are marked like towns. Select a region to filter.
Findings
2024 was the busiest year: 30 stars, 18 of them in AI & LLMs.
They star established projects: the median repo here has 2,757 stars, and 33% have fewer than 1,000.
20 of their stars are archived or deprecated, and 5 name a successor. Another 70 have had no commits in two years.
The longest pause was 12 months, from May 2021 to May 2022.
Python accounts for 51% of their stars, JavaScript for 10%.
Games & graphics
- 2012-01-02115
Is this you, @karpathy?
This page is built from public GitHub data. Sign in with GitHub to claim it: pin repositories, write a headline, get a weekly digest of releases and archived repos, connect your coding agent, or hide the page entirely.
Claim with GitHub
