Recommended Materials
Online Courses
Introduction to Python Programming is a free online course offered by Udacity that aims to introduce students to the Python programming language.
Intro to Deep Learning with PyTorch is a free online course offered by Udacity that serves as an excellent introduction to the world of deep learning using the PyTorch framework. The course is designed for individuals who are new to deep learning but have some basic knowledge of Python and machine learning concepts.
Setting Up a New PyTorch Deep Learning Environment
This course will provide many code examples in notebooks, which require PyTorch v0.4 or newer, and torchvision. To install PyTorch and torchvision locally, you can follow the official instructions provided on the PyTorch website. Choose the stable version, your appropriate OS and Python versions, and how you’d like to install it. You’ll also need to install numpy, matplotlib and jupyter notebooks to run these examples. The newest versions of these should work fine. For hassle-free installations, it’s generally best to use the conda package manager:
conda install numpy matplotlib jupyter notebook
If you haven’t used conda before, please read the offcial documentation to learn how to create environments and install packages. For a lightweight option, consider installing Miniconda instead of the entire Anaconda distribution. However, if you are already comfortable with the normal package manager, pip, feel free to use it based on your preference. Both conda and pip are reliable choices for managing your Python packages and dependencies.
Having a GPU on your computer can significantly accelerate network computations, providing optimal performance. Nevertheless, don’t worry if you don’t have access to a GPU, as you can still run the code and complete all the assignments efficiently. PyTorch leverages CUDA, a library designed to accelerate operations using compatible GPUs. If your GPU is supported by CUDA, you can easily install all the necessary libraries by using conda to install PyTorch. This will enable you to harness the full potential of your GPU for faster and more efficient deep learning tasks.
If you can’t use a local GPU, you can use cloud platforms such as AWS, GCP, and FloydHub to train your networks on a GPU.
CUDA Programming in EECS@UTK
If you want to run the code with GPUs, please note that many EECS Linux systems (e.g., Tesla Lab, and Hydra Lab) have NVIDIA CPUs capble of running CUDA applications. Please refer to CUDA Programming in EECS for more information.