DCGAN โ€” CelebA Face Generator

Standard DCGAN trained on CelebA aligned faces for 100 epochs.

Property Value
Latent dim 100
Output 64ร—64 RGB
Optimizer Adam lr=0.0002 ฮฒ=(0.5, 0.999)
Epochs 100
Dataset CelebA (train split)

Used by the "What I Saw Before The Darkness" Gradio app, which lets you progressively destroy the generator's ability to render faces by ablating (zeroing out) its convolutional channels.

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support