Closed shubhambagwari closed 2 years ago
Hi @shubhambagwari
For arbitrary style transfer algorithms, for example, the network is typically trained on 82,783 content images from Microsoft COCO dataset and 79,433 style images from WikiArt dataset.
Best,
All people did excellent work. Can you tell me how many input images you used to train the models?