Langyuan Mo

icon indicating an email [email protected]

icon company Alibaba Group icon location Guangdong, CN icon location Working on computer vision and multi-modal representation learning.

Results 2 issues of Langyuan Mo

Could you please share the data cleaning code?

2 comment

Hi, in your paper, you use a data cleaning strategy to filter some training images. Can you share this part of code? Thank you!

Issue about cross entropy loss in tensorflow

Hi, The cross entropy loss you used in Tensforflow is ['sparse_softmax_cross_entropy_with_logits()'.](https://github.com/WeitaoVan/L-GM-loss/blob/598582f0631bac876b3eeb8d6c4cd1d780269e03/tensorflow/cifar100_train.py#L180) however, this loss function will perform a 'softmax' on 'logits'. Since your lgm loss doesn't contain a 'softmax' operation,...