text_style_transfer icon indicating copy to clipboard operation
text_style_transfer copied to clipboard

README

Open bakszero opened this issue 5 years ago • 1 comments

Hi, could you update the README with possible initial experiments to start with, specifically differentiating how to use with different parameters for train and inference times (for both the multi-decoder and style-embedding setup). Thanks!

bakszero avatar Jul 02 '19 06:07 bakszero

All the parameters are shown in the train_nmt.py file.

Some main parameters: senti_num, senti_dim represent the style number and style embedding dimension for the style embedding model. The style_class and style_adv are set to True to perform the adversarial training.

session_style contains the code for the style embedding model.

There are no style related parameters for the inference.

fuzhenxin avatar Jul 03 '19 12:07 fuzhenxin