Understanding of Generative Adversarial NetworksTensorFlow basics
初心者向け度初回貢献者にどれだけ取り組みやすいかを 1-100 で推定したスコアです。
10
調査方針
Investigate the training loop and model architecture in files like 'train.py' or 'model.py'. Check hyperparameters, loss functions, and data preprocessing for typical DCGAN pitfalls. Compare with known working implementations to identify discrepancies.
Generated samples are very bad · carpedm20/DCGAN-tensorflow#139 | Good First Issue