This is source code of Adversarial Sub-sequence for Text Generation
.
This code is based on texygen and RelGAN
colorama
numpy>=1.12.1
tensorflow>=1.5.0
scipy>=0.19.0
nltk>=3.2.3
tqdm
λ python .\main.py --help
USAGE: .\main.py [flags]
flags:
.\main.py:
--data: Dataset for real Training
(default: 'image_coco')
--gan: <seqgan|leakgan|mle|relgan>: Type of GAN to Training
(default: 'mle')
--gpu: The GPU used for training
(default: '0')
(an integer)
--mode: <real|oracle|cfg>: Type of training mode
(default: 'real')
--model: Experiment name for LeakGan
(default: 'test')
--[no]pretrain: only pretrain, Stop after pretrain!
(default: 'false')
--[no]restore: Restore pretrain models for relgan
(default: 'false')
Try --helpfull to get a list of all flags.