search

Huage001 / AdaAttN

Officially unofficial PyTorch re-implementation of paper: AdaAttN: Revisit Attention Mechanism in Arbitrary Neural Style Transfer, ICCV 2021.
Apache License 2.0
210 stars 34 forks source link

About the epoch of training #5

Closed 33KU closed 3 years ago

33KU commented 3 years ago

How many epochs have been trained for the model in this paper? Thanks~

Huage001 commented 3 years ago

As configured in train_adaattn.sh, there are 2 epochs with full learning rate and another 3 with learning rate decay.