Closed sayakpaul closed 2 years ago
Sure!
From the README:
CoAtNet is considered the State-of-the-Art model for Image Classification.
There are better models than CoAtNet. It's better to avoid claims like that and include more specific statements like below:
CoAtNet achieves xx.xx% top-1 classification accuracy on the ImageNet-1k dataset (validation split). To know more, please refer to the original paper [link].
Updated
CoAtNet achieves 86.0% ImageNet top-1 accuracy; When pre-trained with 13M images from ImageNet-21K, our CoAtNet achieves 88.56% top-1 accuracy, matching ViT-huge pre-trained with 300M images from JFT-300M while using 23x less data; Notably, when we further scale up CoAtNet with JFT-3B, it achieves 90.88% top-1 accuracy on ImageNet
From the README:
There are better models than CoAtNet. It's better to avoid claims like that and include more specific statements like below: