Online Course – Certified Professional Internship in Advanced DeepLearning.AI Techniques

Custom models, layers, and loss functions with TensorFlow

Course 1 • 31 hours • 4.9 (1,046 ratings)

Course Details

What you’ll learn

• You will compare functional and continuous APIs, discover new models you can build with the functional API, and build a model that produces multiple outputs including a Siamese network.
• Build customized loss functions (including the contrastive loss function used in a Siamese network) to measure the success of the model and help your neural network learn from the training data.
• You will build on existing standard layers to create layers tailored to your model, adapt a network layer using an input layer, understand the differences between them, learn what constitutes an adapted layer, and explore activation functions.
• You will build on existing models to add custom functions, learn how to define your own custom class instead of using functional or sequential APIs, build models that can be inherited from the TensorFlow Model class, and build a residual network (ResNet) by defining a custom model class.

DeepLearning.AI’s TensorFlow specialization: Advanced techniques

Introduces TensorFlow features that give learners more control over their model architecture and tools that help them create and train advanced machine learning models.

This internship is intended for

For early and mid-career software and machine learning engineers who have a basic understanding of TensorFlow and want to expand their knowledge and skills by learning advanced TensorFlow features to build powerful models.

Skills you will acquire

• Category: Functional API
• Functional API
• Category: Customized and special models with functional API
• Customized and special models with functional API
• Category: Customized loss functions
• Customized loss functions
• Category: Custom Layers
• Custom layers

Personalized and Distributed Training with TensorFlow

Course 2 • 24 hours • 4.8 (406 ratings)

Course Details

What you’ll learn

• You will learn about tensor objects, the basic building blocks of TensorFlow, understand the difference between “passive” and “graphic” estimators in TensorFlow, and learn how to use TensorFlow tools to compute gradients.
• Build custom training loops using GradientTape and TensorFlow Datasets to gain more flexibility and visibility in training your model.
• You will learn about the benefits of creating code that runs in graphical mode, take a look at what graphical code looks like, and practice automatically migrating this efficient code with TensorFlow tools.
• You will harness the power of distributed training to process more data and train larger models, faster, get an overview of different distributed training strategies, and practice working with a strategy that trains on multiple GPUs, and another that trains on multiple TPUs.

DeepLearning.AI’s TensorFlow specialization: Advanced techniques

Introduces TensorFlow features that give learners more control over their model architecture and tools that help them create and train advanced machine learning models.

This internship is intended for

For early and mid-career software and machine learning engineers who have a basic understanding of TensorFlow and want to expand their knowledge and skills by learning advanced TensorFlow features to build powerful models.

Skills you will acquire

• Category: Distribution Strategies
• Distribution strategies
• Category: GradientTape for optimization
• GradientTape for optimization
• Category: Customized Training Loops
• Customized training loops
• Category: Basic Tensor Functionality
• Basic functionality of tensors

Advanced Computer Vision with TensorFlow

Course 3 • 19 hours • 4.8 (498 ratings)

Course Details

What you’ll learn

• You will explore image classification, image segmentation, object localization, and object detection. You will apply transfer learning to object localization and detection.
• Implement object detection models such as regional-CNN and ResNet-50, adapt existing models, and build your own models to detect, locate, and label your duck images.
• Implement image segmentation using variations of the full convolutional network (FCN) including U-Net and Mask-RCNN to identify and detect numbers, pets, zombies, and more.
• You will identify which parts of the image are used by your model to make its predictions using class activation maps and importance maps, and apply these machine learning interpretation methods to test and improve the design of a famous network, AlexNet.

DeepLearning.AI’s TensorFlow specialization: Advanced techniques

Introduces TensorFlow features that give learners more control over their model architecture and tools that help them create and train advanced machine learning models.

This internship is intended for

For early and mid-career software and machine learning engineers who have a basic understanding of TensorFlow and want to expand their knowledge and skills by learning advanced TensorFlow features to build powerful models.

Skills you will acquire

• Category: Importance
• importance
• Category: Image Segmentation
• Image segmentation
• Category: Model Interpretation
• Interpreting models
• Category: Department Operation Maps
• Department operation maps
• Category: TensorFlow Object Discovery API
• Object Discovery API in TensorFlow

Generative Deep Learning with TensorFlow

Course 4 • 16 hours • 4.9 (279 ratings)

Course Details

What you’ll learn

• You will learn about neural style transfer through transfer learning: you will find the content of an image (e.g., a duck), and the style of a painting (e.g., cubist or impressionist), and combine the content and style into a new image.
• You will build simple AutoEncoders on the well-known MNIST dataset, and more complex deep and convolutional architectures on the Fashion MNIST dataset, understand the difference in results between DNN and CNN models, identify ways to remove noise from noisy images, and build a CNN AutoEncoder using TensorFlow to produce a clean image from a noisy image.
• You will explore Variational AutoEncoders (VAEs) to create entirely new data, and create anime faces to compare to reference images.
• You will learn about GANs; invention, features, architecture, and how they differ from VAEs, understand the role of the generator and discriminator within the model, the concept of 2 training stages and the role of introduced noise, and build your own GAN that can generate faces.

DeepLearning.AI’s TensorFlow specialization: Advanced techniques

Introduces TensorFlow features that give learners more control over their model architecture, providing them with the tools to create and train advanced machine learning models.

This internship is intended for

For early and mid-career software and machine learning engineers who have a basic understanding of TensorFlow and want to expand their knowledge and skills by learning advanced TensorFlow features to build powerful models.

Skills you will acquire

• Category: Auto Encoders
• Auto Encoders
• Category: Adversarial Generative Networks
• Adversarial generative networks
• Category: Neural Style Transfer
• Neuronal style transfer