SKILL BENCHMARK
Generative AI Models Competency (Intermediate Level)
- 20m
- 20 questions
The Generative AI Models Competency (Intermediate Level) benchmark measures your ability to create an autoencoder model using PyTorch, train an autoencoder and a convolutional autoencoder, and use denoising autoencoders. You will be evaluated on your skills in using variational autoencoders to generate images, recognizing the key concepts of how generative adversarial networks (GANs) work, and creating and viewing the training of a GAN and a deep convolutional GAN. A learner who scores high on this benchmark demonstrates good competency in working with core generative AI models. They are capable of independently creating, training, and evaluating these models using PyTorch with minimal supervision. They can confidently apply their skills to real-world projects involving generative AI techniques.
Topics covered
- create a generator and discriminator
- create a VAE
- describe the architecture of GANs
- describe the denoising autoencoder
- load and explore the CelebFaces dataset
- load and explore the Modified National Institute of Standards and Technology (MNIST) dataset
- provide an overview of DCGANs
- provide an overview of the sparse autoencoder
- recall how GANs work
- reconstruct images with convolutional neural networks (CNNs)
- reconstruct images with denoising autoencoders
- reconstruct images with dense neural networks (DNNs)
- set up a virtual environment and Python notebook for GAN training
- train a convolutional VAE
- train a DCGAN
- train a GAN
- train a VAE
- train a VAE on color images
- train convolutional autoencoders
- train the autoencoder and visualize reconstructions
RECENTLY ADDED COURSES
Rating 5.0 of 2 users
(2)
Rating 5.0 of 2 users
(2)