Google · Course
Generative Adversarial Networks
Generative Adversarial Networks is a Google for Developers course on GANs, generative models that pair a generator with a discriminator to produce new data resembling the training data, including loss functions, training challenges and the TF-GAN library.
Who it suits
- Practitioners who know TensorFlow and want to understand how GANs work and fail.
- Learners who have completed Machine Learning Crash Course.
What it covers
- The difference between generative and discriminative models.
- How a generator and a discriminator work against each other to improve realism.
- Common GAN loss functions and the challenges of training.
- Building GANs with the TF-GAN library.
Before you start
- Completion of Machine Learning Crash Course.
- Some TensorFlow programming experience.
Common questions about Generative Adversarial Networks
- What is a GAN?
- Google describes a generative model that creates new data resembling your training data, such as images that look like photographs of faces that do not belong to a real person.
- Which library does it use?
- TF-GAN, which builds on TensorFlow.
- Is it free to reuse?
- Google says the content is licensed under Creative Commons Attribution 4.0, and code samples under Apache 2.0.
More like this
Keep learning on Infrato
- Path: AI for software developers
- Path: AI for data analysts
- More from Google: all Google listings in the directory
- Explore all AI courses
Where this comes from
Written from Google’s own pages, read on 2026-10-14. Content, prices and availability change, so the provider’s page is the source. We are not affiliated with Google and this is not an endorsement.