Development Environment Setup

Get your machine ready for deep learning development. Follow these steps to set up your environment.

Canvas Students

Complete this setup before Module 1. If you encounter issues, post in the Canvas discussion forum or attend office hours for help.

Local Development
Install everything on your own machine for full control and offline access
Full control over environment
Works offline
Use your own GPU
Follow Local Setup Guide
Google Colab
Use Google's free cloud environment with GPU access
No installation required
Free GPU access
Works on any device
Open Google Colab

Local Setup Guide

1
Python Environment Setup
Install Python and essential packages
Beginner
15-20 minutes

Steps:

  1. 1Install Python 3.8+ from python.org or use Anaconda
  2. 2Create a virtual environment: `python -m venv dl-course`
  3. 3Activate the environment: `source dl-course/bin/activate` (Linux/Mac) or `dl-course\Scripts\activate` (Windows)
  4. 4Install required packages: `pip install -r requirements.txt`
2
Deep Learning Frameworks
Install PyTorch and TensorFlow
Beginner
10-15 minutes

Steps:

  1. 1Install PyTorch: Visit pytorch.org and select your configuration
  2. 2Install TensorFlow: `pip install tensorflow`
  3. 3Verify installations by importing both libraries
  4. 4Test GPU support (if available): `torch.cuda.is_available()`
3
Jupyter Notebook Setup
Configure Jupyter for interactive development
Beginner
10 minutes

Steps:

  1. 1Install Jupyter: `pip install jupyter`
  2. 2Install JupyterLab (optional): `pip install jupyterlab`
  3. 3Start Jupyter: `jupyter notebook` or `jupyter lab`
  4. 4Install useful extensions: `pip install jupyter_contrib_nbextensions`
4
Google Colab Setup
Alternative cloud-based environment
Beginner
5 minutes

Steps:

  1. 1Go to colab.research.google.com
  2. 2Sign in with your Google account
  3. 3Test GPU access: Runtime → Change runtime type → GPU
  4. 4Mount Google Drive for file storage (optional)
5
Git and GitHub
Version control for your projects
Intermediate
20 minutes

Steps:

  1. 1Install Git from git-scm.com
  2. 2Create a GitHub account at github.com
  3. 3Configure Git: `git config --global user.name 'Your Name'`
  4. 4Clone the course repository: `git clone https://github.com/course/materials.git`
Requirements File
Download the requirements.txt file with all necessary packages
torch>=1.12.0
torchvision>=0.13.0
tensorflow>=2.9.0
numpy>=1.21.0
pandas>=1.4.0
matplotlib>=3.5.0
seaborn>=0.11.0
scikit-learn>=1.1.0
jupyter>=1.0.0
tqdm>=4.64.0
Download requirements.txt

Common Issues & Solutions

ImportError: No module named 'torch'

Installation

PyTorch not installed correctly. Try reinstalling with the correct command from pytorch.org

CUDA out of memory

GPU

Reduce batch size or use a smaller model. Consider using Google Colab for more GPU memory.

Jupyter kernel keeps dying

Jupyter

Usually a memory issue. Restart the kernel and reduce data size or model complexity.

Package conflicts

Dependencies

Create a fresh virtual environment and install packages one by one to identify conflicts.

Need Help?

If you're stuck with setup, don't worry! Here are ways to get help:

Canvas Discussion ForumEmail InstructorOffice Hours Schedule