This repository provides a comprehensive 3-month PyTorch curriculum for learning machine learning, designed for beginners to Python. It aims to equip learners with the skills to build, train, deploy, and maintain modern ML/DL models, culminating in a self-driving car simulation project.

How It Works

The curriculum is structured into weekly assignments, each focusing on a specific ML/DL concept or tool, such as PyTorch, Transformers, or MLOps. It emphasizes a project-based learning approach, integrating various domains like Computer Vision, NLP, and Reinforcement Learning. Assignments are designed to be beginner-friendly, often runnable within web-based environments like Google Colab or HuggingFace Spaces, minimizing local setup dependencies.

Quick Start & Requirements

• Install/Run: Primarily uses web-based tools like Google Colab and HuggingFace Spaces for assignments. Local setup may involve Python, Pip, and various ML libraries.
• Prerequisites: Python, Pip, NumPy, Pandas, PyTorch, HuggingFace, GCP/AWS (for MLOps). Specific assignments may require Jupyter Notebooks, Streamlit, Gradio, or Docker.
• Resources: Links to external learning tools like ExplainPaper, Summari, and CoPilot are provided.

Highlighted Details

• Covers a broad spectrum of ML/DL topics: Computer Vision, NLP, Time Series, Audio, Recommender Systems, and Reinforcement Learning.
• Integrates modern tools and frameworks: PyTorch, HuggingFace Transformers/Diffusers, Airflow, GCP/AWS, Spark, Kafka, Terraform.
• Includes a practical MLOps component covering design, development, production deployment, and data engineering.
• Culminates in a capstone project: a self-driving car simulation integrating multiple ML disciplines.

Maintenance & Community

• Associated with Siraj Raval's "Learn Machine Learning in 3 Months" YouTube series.
• Encourages community learning via a Discord channel for study buddies.

Licensing & Compatibility

• The repository itself appears to be under an unspecified license. Individual components or linked resources may have their own licenses.

Limitations & Caveats

• The curriculum relies heavily on external resources and YouTube videos, which may change or be deprecated.
• While aiming for beginner-friendliness, the breadth of topics and tools covered can be demanding.
• The "3 Months" timeline is ambitious and may require significant dedication.