This is a Deeply Convolutional Generative Adversarial Network (DCGAN) that generates 64x64 anime images out of random noise. Colab Notebook works with the GPU Hardware Accelerator but may crash mid training.
The model is trained on a local runtime.
Datset Used: https://www.kaggle.com/datasets/soumikrakshit/anime-faces
g_model is the generative model
d_model is the discriminator model
Images genererated:
