Pokémon GAN
Unconditional EMA generator for 64x64 Pokémon-style images.
This repository contains generator weights only, as required by the assignment. The model uses a 100-dimensional standard-normal latent vector. See model.py for the exact architecture and training-metrics.json for the complete loss history.
Training
- Dataset:
huggan/pokemon - Epochs: 400
- Batch size: 64
- Seed: 42
- Fixed-sample pixel standard deviation: 0.4797
Weights are stored in safetensors format. The model generates synthetic images and can produce malformed or ambiguous samples; it is intended for coursework and experimentation.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support