Initial upload of GAN-MNIST model

README.md

@@ -0,0 +1,19 @@
+# GAN-MNIST
+
+A simple DCGAN trained on the MNIST dataset using PyTorch for data augmentation research.
+
+## Model Details
+
+- Epochs: 20  
+- Latent Dim: 100  
+- Discriminator & Generator: 3-layer CNN  
+- FID Score: 24.16
+
+## Usage
+```python
+from torch import load
+from models.gan.model import Generator
+
+model = Generator(latent_dim=100, img_channels=1)
+model.load_state_dict(load("generator.pt"))
+model.eval()