Course Promo Video
Who's this course for?
This course is for anyone who wants to start learning Tensorflow. You ideally have good knowledge of Python and also of basic concepts in deep learning.
Skills you'll gain
You'll learn how to setup your machine to train deep learning models in Tensorflow.
You'll learn how to build neural networks from scratch using Tensorflow layers.
You'll learn to build a neural network in 3 different approaches supported by Tensorflow : Sequential, functional and by inheriting from the Model class.
You'll build neural networks to solve 2 real world cases : recognizing digits and recognizing German traffic signs.
You'll learn how clean and prepare a dataset for training and evaluation.
You'll learn how to setup a standalone script that can make use of your trained deep learning model to make predictions on new random images.
Deep Learning project to add to your portfolio
You will build 2 projects : one for recognizing handwritten digits and one for recognizing German traffic signs. These projects can be a great addition to your ML portfolio.
Course Curriculum
- We will be using an IDE and not notebooks (0:59)
- Visual Studio Code (how to download and install it) (3:24)
- Miniconda - how to install it (2:33)
- Miniconda - why we need it (4:00)
- How are we going to use conda virtual environments in VS Code? (3:56)
- Installing Tensorflow 2 (CPU version) (8:36)
- Installing Tensorflow 2 (GPU version) (13:38)
- What do we want to achieve? (1:51)
- Exploring MNIST dataset (20:28)
- Tensorflow layers (3:49)
- Building a neural network the sequential way (17:38)
- Compiling the model and fitting the data (33:29)
- Building a neural network the functional way (7:41)
- Building a neural network the Model Class way (5:57)
- Things we should add (3:57)
- Restructuring our code for better readability (4:42)
- Summary (1:00)
- What we want to achieve (1:11)
- Downloading and exploring the dataset (8:57)
- Preparing train and validation sets (19:16)
- Preparing the test set (16:40)
- Building a neural network the functional way (11:55)
- Creating data generators (9:26)
- Instantiating the generators (3:58)
- Compiling the model and fitting the data (4:57)
- Adding callbacks (11:33)
- Evaluating the model (5:56)
- Potential improvements (10:45)
- Running prediction on single images (14:15)
- Summary (0:51)