google-research/timesfm — GitHub trend

FINDING #2053 · UNIT ID 793784055

google-research/timesfm

TimesFM (Time Series Foundation Model) is a pretrained time-series foundation model developed by Google Research for time-series forecasting.

[ PYTHON ][ ORG ][ GITHUB ↗ ]

SURPRISE SCORE

0.00

Score Breakdown

SURPRISE0.000161

ENGAGEMENT0.66

FRESHNESS1.30

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

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

9% OF STARS IN ARCHIVE

Growth Telemetry

VELOCITY /D

13.91

ACCEL

-0.13

RETENTION

4.7%

PEAK 2026-04-02 · FORK-RETENTION 64.3% · 2,504 STARS / WINDOW

Author Audience

AUDIENCE

86,503

FOLLOWERS

16,757

OWNER ★

264,945

Engagement Signals

FORKS

2,630

ISSUE AUTH

PR AUTH

UNIQUE STARGAZERS 2,498 / 2,504 (DIVERSITY 1.00)

Why This Is A Finding

google-research/timesfm собрал 2,504 звёзд за окно, тогда как у автора всего 16,757 подписчиков — эффективная аудитория ≈ 86,503. Это даёт surprise-индекс 0.000161 (звёзды относительно охвата автора, а не в абсолюте). Удержание форков 64.3% и 20 внешних контрибьюторов отделяют реальный инструмент от разовой вспышки. Акселерация отрицательная — внимание остывает после пика.

Related Findings

RANKS ABOVE 21% OF 2593 FINDINGS

#03databricks-solutions/ai-dev-kit[ PYTHON ]0.10 #10sickn33/agentic-awesome-skills[ PYTHON ]0.0191 #11cft0808/edict[ PYTHON ]0.0183 #16ZhuLinsen/daily_stock_analysis[ PYTHON ]0.016 #26jundot/omlx[ PYTHON ]0.0139

METRICS IN CONTEXT

MEDIAN ACROSS ALL 2593 FINDINGS · Δ vs MEDIAN · PERCENTILE = SHARE RANKED BELOW

METRICVALUEMEDIANΔ MEDPERCENTILE

SCORE0.000.00-0.00ABOVE 21%

VELOCITY13.912.88+11.03ABOVE 92%

RETENTION4.7%6.3%-1.6 PPABOVE 40%

FORKS2,6301,166+1,464ABOVE 69%

SURPRISE0.000.00-0.00ABOVE 21%