This project trains a CycleGAN to convert real-world photos into Monet-style paintings using the Kaggle "GAN Getting Started" dataset.

1. Data Preprocessing – Load, normalize, and batch images.
2. Model Architecture – CycleGAN with:
   • Generator: Photo → Monet
   • Discriminator: Monet vs. Fake Monet
3. Training – Adversarial & cycle-consistency losses.
4. Results – Generate Monet-style images from real photos.

• Run cycleGAN_monet_project_fixed.ipynb in Jupyter Notebook.
• Generates Monet-style images and saves them in /images/.

• Longer training for better details.
• Improved hyperparameter tuning.
• Perceptual loss for texture enhancement.