Model

flamingo

1 entry · 29 April 2022

Appears alongside

Tracks

  • Models & capabilities 1
  • Benchmarks & progress 1

  1. DeepMind's Flamingo tackles multimodal few-shot learning

    Flamingo, an 80B-parameter vision-language model, sets few-shot state of the art on image and video benchmarks with minimal task examples.

    Google DeepMindModels & capabilities · Benchmarks & progress

In the commentary

Pieces from around the web that discuss flamingo. External links.