Generate samples using pre-trained model.

williamSYSU / TextGAN-PyTorch

TextGAN is a PyTorch framework for Generative Adversarial Networks (GANs) based text generation models.

MIT License

878 stars 207 forks source link

Generate samples using pre-trained model. #32

Open IbrahimSahibzada opened 4 years ago

IbrahimSahibzada commented 4 years ago

Once I restore a model by directly changing pretrained_gen_path and pretrained_dis_path in config.py. How do I then only generate samples and not have the model train again ?

IbrahimSahibzada commented 4 years ago

During training the generator outputs good samples, but when I load the same model as a pre_trained generator model and use the sample() function, the samples generated are not even close to good. Can you please help me out ?

anujwwt commented 3 years ago

@IbrahimSahibzada - were you able to generate samples on the pre trained model? if yes, please et me know how?

JoseLu201 commented 1 month ago

Hi @IbrahimSahibzada, I am facing the same problem you mention here. https://github.com/williamSYSU/TextGAN-PyTorch/issues/32#issuecomment-658574135 in my case both in pre-trained model and in ADV train saved models, any solution?