Flamingo

tags

Transformers, Computer vision, NLP, Chinchilla

paper

(Alayrac et al. 2022)

Architecture

Uses a frozen language model (e.g. Chinchilla) that is conditioned on a visual representation given from a normalizer-free ResNet.

Parameter count

80B

Bibliography

1. Jean-Baptiste Alayrac, Jeff Donahue, Pauline Luc, Antoine Miech, Iain Barr, Yana Hasson, Karel Lenc, et al.. April 29, 2022. "Flamingo: A Visual Language Model for Few-shot Learning". arXiv. http://arxiv.org/abs/2204.14198.