Through multi-scale visual prompt and post-pretrain token scaling, Chain-of-Sight achieves 3.7x pre-training speedup without performance sacrifice.
Jul 4, 2024
Temporally-adaptive convolutions and transformers.
Aug 10, 2023