1 Deep Learning Examples

This is a collection of deep learning examples, tests and related material. Some of the examples are based on courses and tests I've done along the years, which I'd like to mention here:

• Deep Learning (Coursera)
• Deep Learning: Advanced Computer Vision by Lazy Programmer (Udemy)
• Deep Learning: Financial Engineering and Artificial Intelligence in Python by Lazy Programmer
• Deep Learning: Data Science: Transformers for Natural Language Processing by Lazy Programmer

All of the above are fantastic courses and if you're interested in those subjects, you should consider taking them.

---

2 Folder Structure

2.1 Virtual Environments and Docker Images

Since PyTorch stopped supporting Conda in 2024, using any other virtualization technology is recommended. Docker containers are probably the most portable option.

• Virtualenv
  • Instructions for setting up virtual environments using virtualenv.
• Docker
  • Contains files for creating Docker container for running the examples.
• Conda
  • Contains files for creating Conda virtual environments for running the examples.

2.2 Generic Explanations

• Documentation

2.3 Deep Learning and Related Examples

• TensorFlow
  • Deep learning examples using TensorFlow 2.
• PyTorch
  • Deep learning examples using PyTorch.
• Financial Engineering
  • Financial engineering using Python and Ai.
• Hugging Face
  • Examples of using models from Hugging Face.

---

3 Testing Environment

Testing and executing the examples has been done in a system with the following characteristics:

• OS Ubuntu 20.04
• GPU GeForce GTX 1070

3.1 Pre-commit

This repo uses pre-commit git-hooks to verify the code before it is pushed. After cloning the repo, install the pre-commit package into your Python environment:

pip install pre-commit

and then install the pre-commit checks:

pre-commit install

Pre-commit is run automatically every time code is committed. You can also run pre-commit manually as follows:

pre-commit run --all-files

---

4 Running the Examples

You can either create a Conda virtual environment, or a Docker container, for running the examples. In order to do so, take a look at the instructions in conda and docker directories.

---

5 Additional Documentation

Additional documentation regarding the notation, neural network model etc. can be found here

---