Back To Module Home

Deep Learning with PyTorch

0% completed

Introduction

Visualizing Gradient Descent

Spoilers Gradient Descent Data Generation Basic Steps for Using Gradient Descent (Step 0 and 1)Step 2a - Compute the Loss Step 2b - Computing the Loss Surface Step 3 - Compute the Gradients Step 4 - Update the Parameters Learning Rate Scaling the Dataset Step 5 - Rinse and Repeat!Recap Quiz Challenge 1 - Visualizing Gradient Descent Solution Review - Visualizing Gradient Descent

A Simple Regression Problem

Reviewing the Steps of Gradient Descent

Linear Regression in Numpy

PyTorch Tensors

Loading Data, Devices, and CUDA

Creating Parameters

Updating Parameters

Dynamic Computation Graphs

Putting It All Together

Challenge 2 - A Simple Regression Problem

Solution Review: A Simple Regression Problem

Rethinking the Training Loop

Spoilers Introducing Higher-Order Functions Rethinking the Training Loop Building a Dataset DataLoader Quiz Mini-Batch Inner Loop and Training Split Evaluation TensorBoard SummaryWriter Saving and Loading Models Deploying / Making Predictions Putting It All Together Recap Quiz Challenge 3 - Rethinking the Training Loop Solution Review - Rethinking the Training Loop

Going Classy

Spoilers Going Classy Functions Different Method Types Training Methods Saving and Loading Methods Finalising the Methods needed for going Classy Classy Pipeline Model Training and Predictions Checkpointing Putting It All Together Recap Quiz Challenge 4 - Going Classy Solution Review - Going Classy

A Simple Classification Problem

Spoilers Classification Problems Model for Classification Problems Sigmoid and Logistic Regression Quiz Loss Binary Cross-Entropy Loss in PyTorch Imbalanced Dataset Model Configuration, Training, and Predictions for Classification Decision Boundary Classification Threshold and Confusion Matrix Metrics Trade-offs and Curves Best + Worst Curves and Models Putting it All Together Recap Quiz Challenge 5 - A Simple Classification Problem Solution Review: A Simple Classification Problem

Appendix

Setup Guide Running TensorBoard

Conclusion

...

/

Recap

Recap

A quick overview of what you have learned and testing it on the notebook provided.

General overview

First of all, congratulations are in order! You have successfully implemented a fully functioning model and training loop in PyTorch!

We have covered a lot of ground in this third chapter:

Implementing a linear regression in Numpy using gradient descent.
Creating tensors in PyTorch, sending them to a device, and making parameters out of them.
Understanding PyTorch’s main feature, ...