CBABHI/gan-artwork-generation

GAN Artwork Generation Models (DCGAN, cGAN, CAN)

Trained PyTorch generative models capable of synthesizing multi-style historical and contemporary artwork across 27 artistic movements.

Model Details

• Architecture: Deep Convolutional GAN (DCGAN), Conditional GAN (cGAN), Creative Adversarial Network (CAN)
• Trained Epochs: 75 Epochs (0-74) on NVIDIA RTX 4050 GPU
• Resolution: 64x64x3 RGB
• Latent Dim: 100
• Weight Checkpoint: checkpoint_latest.pth (56.2 MB)

Live Web UI

Visit the deployed web interface: web-ui-tau-six.vercel.app

Quick Load (PyTorch)

`python import torch

checkpoint = torch.load('checkpoint_latest.pth', map_location='cpu')

Generator state dict is inside checkpoint['generator_state_dict']

print('Loaded checkpoint from epoch:', checkpoint.get('epoch', 75)) `