- Unofficial Pytorch Implementation of "A Learned Representation For Artistic Style"
- Reference: https://arxiv.org/abs/1610.07629
- Content: MSCOCO 2014 train
- Style: ./imgs/style (manually collected images)
- The
style_index
starts in order from the top left (0 ~ 15).
- The
- Trained Model Weights: Link
# Model Training.
python main.py --mode 'train' \
--style_path "./imgs/style/" \
--content_path "/path/to/content/dataset/"
# Model Evaluation.
python main.py --model 'eval' \
--content_path "./imgs/gentlecat.png" \
--style_index -1 (0~15 for a specific style, or -1 for all styles)