Ahmadzei's picture
added 3 more tables for large emb model
5fa1a76
The
authors of SegFormer first pre-trained the Transformer encoder on ImageNet-1k to classify images.