> You will neither have compute nor data to do anything original unless it fall under transfer learning and somebody is kind enough to share pretrained model.
Lots of research recently goes into doing something useful with less data and compute. E.g. there have been a lot of zero/few-shot learning on the last CVPR.
Lots of research recently goes into doing something useful with less data and compute. E.g. there have been a lot of zero/few-shot learning on the last CVPR.