facebookresearch/flow_matching — GitHub trend

FINDING #8537 · UNIT ID 900093817

facebookresearch/flow_matching

A PyTorch library for implementing flow matching algorithms, featuring continuous and discrete flow matching implementations. It includes practical examples for both text and image modalities.

[ PYTHON ][ ORG ][ VERIFIED ][ GITHUB ↗ ]

SURPRISE SCORE

0.00

Score Breakdown

SURPRISE0.00000368

ENGAGEMENT0.20

FRESHNESS1.00

SCORE = SURPRISE × ENGAGEMENT^0.7 × FRESHNESS × VISIBILITY × CONFIDENCE

SURPRISE = WINDOW STARS / DAYS / (AUDIENCE + FLOOR)

0% OF STARS IN ARCHIVE

Growth Telemetry

VELOCITY /D

3.33

ACCEL

-2.50

RETENTION

21.4%

PEAK 2026-10-09 · FORK-RETENTION 0.0% · 10 STARS / WINDOW

Author Audience

AUDIENCE

906,346

FOLLOWERS

37,963

OWNER ★

718,023

Engagement Signals

FORKS

382

ISSUE AUTH

PR AUTH

UNIQUE STARGAZERS 10 / 10 (DIVERSITY 1.00)

Why This Is A Finding

facebookresearch/flow_matching собрал 10 звёзд за окно, тогда как у автора всего 37,963 подписчиков — эффективная аудитория ≈ 906,346. Это даёт surprise-индекс 0.00000368 (звёзды относительно охвата автора, а не в абсолюте). Удержание форков 0.0% и 0 внешних контрибьюторов отделяют реальный инструмент от разовой вспышки. Акселерация отрицательная — внимание остывает после пика.

Related Findings

RANKS ABOVE 2% OF 8679 FINDINGS

#04Oldcircle/geo-sleuth[ PYTHON ]1.24 #13rehan-remade/universal-modder[ PYTHON ]0.83 #20joshuaswarren/omarchy-apple-dev[ PYTHON ]0.64 #22chengyi-ai/native-subtitle-quote-image[ PYTHON ]0.61 #24spoopyghosty0/frameport[ PYTHON ]0.57

METRICS IN CONTEXT

MEDIAN ACROSS ALL 8679 FINDINGS · Δ vs MEDIAN · PERCENTILE = SHARE RANKED BELOW

METRICVALUEMEDIANΔ MEDPERCENTILE

SCORE0.000.00-0.00ABOVE 2%

VELOCITY3.335.67-2.33ABOVE 18%

RETENTION21.4%42.4%-21.0 PPABOVE 26%

FORKS382334+48ABOVE 53%

SURPRISE0.000.00-0.00ABOVE 1%