Open mdtayebadnan opened 1 week ago
Hello @mdtayebadnan , For unconditional generation one should see decent face like outputs in 100 epochs with batch size of 16. While training for 200 epochs should further improve results but its strange that even after 100 epochs of training you are not getting any face generated image. Could you share the config file once ? And after autoencoder training you were able to get good reconstructions right ? Could you please also share a sample vqvae output and sample generated image as well ?
Thank you for your amazing work. What is the standard number of epochs to train for each model? I have trained an unconditional one for 100 epochs on celebHq, but the model doesn't generate any faces. Can you specify the standard number of epochs I should try for each?