GitHub repositories

You can find all of my public GitHub content here: gperdrizet.

Unit 2: Data science fundamentals

Lesson	Topic	Description	Repo
Lesson 16	Feature engineering	Instructor solution for feature engineering challenge	gperdrizet/featurely

Unit 3: Machine learning

Lesson	Topic	Description	Repo
Lesson 19	Supervised learning, regression	Deployed web app for predicting concrete compressive strength from composition and age	gperdrizet/concrete-strength
Lesson 20	Supervised learning, classification	Kaggle student health risk competition submission	gperdrizet/student-health-risk
Lesson 23	Recommendation systems, collaborative, content and hybrid filtering	Deployed hybrid filtering movie recommendation app using the Movie Lens 100k dataset	gperdrizet/movie-lens-recommendations

Unit 4: Deeplearning

Lesson	Topic	Description	Repo
Lesson 28-33	PyTorch	Demo & activity notebooks for PyTorch regression and image classification	gperdrizet/pytorch-demo
Lesson 29-33	Image classification	RSNA knee abnormality Kaggle competition solution: multiclass image classification	gperdrizet/RSNA-knee-abnormality
Lesson 29-33	Image classification	CIFAR10 model training and optimization, image classification tools package	gperdrizet/CIFAR10
Lesson 34	Object detection	Live object detection with YOLO26 and streamlit	gperdrizet/YOLO26
Lesson 34	Sequence learning with RNNs	Simple RNN models for sequence learning and NLP sentiment analysis	gperdrizet/RNNs
Lesson 36	Autoencoders	Demos for image compression, denoising and anomaly detection with convolutional autoencoders	gperdrizet/autoencoders

Unit 5: MLOps

Lesson	Topic	Description	Repo
Lesson 38	Introduction to MLOps	CI/CD with GitHub actions, training experiment tracking and model versioning with MLFlow	gperdrizet/fsa-mlops

Unit 6: Natural language processing

Lesson	Topic	Description	Repo
Lesson 46,47,48	NLP with sequence models	Machine translation, encoder decoder architecture, recurrent models, attention and transformers	gperdrizet/language-models

Unit 7: Essentials and applications of generative AI

Lesson	Topic	Description	Repo
Lesson 50-56	LLM applications	Prompting, LangChain, Agents, Fine-tuning and evaluation	gperdrizet/llms-demo
Image generation	Open weight image generation model hardware benchmark	Stable diffusion, FLUX, Kandinsky, PixArt	gperdrizet/open-weight-image-generation
Unit end project(s)	OpenAI compatible inference APIs, Gradio UI	OpenAI Python client, LangChain ChatOpenAI module, credential management, Gradio	gperdrizet/unit7-cep

Unit 8: Agentic AI frameworks with model context and tooling protocols

Lesson	Topic	Description	Repo
Lesson 59	Intro to LangGraph	Nodes, edges, sequential and parallel edges, decision making	gperdrizet/fsa-agentic-ai

DevOps/other

Purpose	Description	Repo
Containerized development environments	General purpose data science	gperdrizet/datascience-devcontainer
	Kaggle competitions	gperdrizet/kaggle-devcontainer
	Deeplearning & neural network development	gperdrizet/deeplearning-devcontainer
	LLM hosting & application development	gperdrizet/llms-devcontainer
	Master repo for base Docker images	gperdrizet/docker-images
Learning tools & course resources	This GitHub pages site source	gperdrizet/fullstack-2605
	Bug Hunter basic Python practice app	gperdrizet/bug-hunter
	Kaggle style leaderboard for feature engineering competition	gperdrizet/leaderboard
PostgreSQL + PGvector server setup	gperdrizet/postgreSQL-server

How to use GitHub repos

First, make a fork of the repo: click the 'fork' button in the upper right on the repository main page. Set yourself as the owner and click create. This gives you a copy of the original repository that you can make changes to. Then, to run the code or make changes you can work in a GitHub codespace, or clone your fork to work locally.

GitHub codespaces

From your fork's main page on GitHub, click the green 'Code' dropdown button at the upper right of the file list, switch to the 'Codespaces' tab of the menu and click 'Create codespace on main'. This will spin up a containerized development environment and show you a clone of the repo via VS code running on a GiHub server.

Codespaces are best when the repo is configured with a .devcontainer/devcontainer.json file, meaning the developer set things to up to work well in VS Code via Microsoft's devcontainers extension (what codespaces use...), but it is possible to interact with any repo this way.

Clone

You can also clone the repository locally or to another cloud development environment provider (like Vocareum). After you have made your fork, do the following:

1. Click the green 'Code' button at the upper right above the file list on the repo's GitHub page
2. Make sure you are on the 'Local' tab
3. Copy the URL shown under 'Clone'
4. Then clone the repo on whatever machine you are using with:

git clone git@github.com:USERNAME/REPONAME.git

Now you should have a local directory called REPONAME/ with the repository's contents.

It is possible to clone a public GitHub repository that you don't own, but you will not be able to push changes back to GitHub. It's best to use a fork - even if you want to contribute to a project by adding/editing code to the original repo. This is usually done via pull request, but check the specific project's documentation to see what their preferred collaboration workflow is.