You can deploy this model and start earning money today!

tgallup/
forest-gan

Views

No views yet

Progressive GAN Image Generator for Cities - 128x128 Resolution

This model was trained on a custom city image dataset using a Progressive GAN architecture with techniques to prevent mode collapse. It generates images up to 128x128 resolution.

Training Information

Epochs: 8800/10000
Current Date: 2025-05-12 06:02:11
Current Resolution: 5x5
Alpha: 0.8
Latent Dimension: 100
Batch Size: 8

Latest Training Metrics

Generator Loss: 0.3617730140686035
Discriminator Loss: 0.9162907600402832
Discriminator Accuracy: 0.5

Sample Images

Training Techniques

Progressive Growing from 8x8 to 128x128
Minibatch Discrimination to prevent mode collapse
Instance Noise for training stability
Dynamic batch sizing based on resolution

Usage

python

import torch
from torch import nn

# Define generator architecture
class Generator(nn.Module):
    # ... (architecture code)

# Load the model
generator = Generator(latent_dim=100, channels=3)
generator.load_state_dict(torch.load('model/generator_epoch_8800.pt'))
generator.eval()

# Generate images
z = torch.randn(1, 100)
with torch.no_grad():
    fake_image = generator(z)
    # Convert to range [0, 1] for display
    fake_image = (fake_image + 1) / 2.0

Training Progress

Training is currently 88.0% complete. Current resolution phase: 5 of 5

tgallup/forest-gan