Closed asif-hanif closed 2 years ago
Hi, I think there is a minor mistake in section 4.1 of the ViG paper. ImageNet-1k dataset has around 1.2M training samples as opposed to 120M samples. You may want to correct this disparity in arXiv version of the paper.
Thanks for the correction. We'll fix the typo in the next version.
Hi, I think there is a minor mistake in section 4.1 of the ViG paper. ImageNet-1k dataset has around 1.2M training samples as opposed to 120M samples. You may want to correct this disparity in arXiv version of the paper.