Yacine Mahdid Courses

How to Read Math in Deep Learning Papers - A 5-Step Process

Learn a 5-step method to decipher mathematical formulas in deep learning papers, enhancing your understanding and intuition of complex AI concepts.

• YouTube
• 27 minutes
• Self-Paced
• Free Video

Log Softmax Explained with Python

Explore log softmax implementation in Python, enhancing numerical stability for machine learning. Gain insights into softmax limitations and practical coding solutions.

• YouTube
• 12 minutes
• Self-Paced
• Free Video

QHAdam Optimizer - Theory and Implementation in PyTorch

Dive into the theory and implementation of QHAdam optimizer, exploring its formulas, performance benefits, and practical PyTorch implementation through detailed code examples and mathematical breakdowns.

• YouTube
• 54 minutes
• Self-Paced
• Free Video

Guide to Scientific Research in Undergraduate Studies - From Lab Selection to Success

Discover a proven step-by-step approach to securing undergraduate research positions in labs, from identifying your interests and time commitment to successfully landing and maximizing research opportunities.

• YouTube
• 14 minutes
• Self-Paced
• Free Video

Segment Anything Model (SAM) - Deep Dive into Meta's Image Segmentation Foundation Model

Dive into Meta's Segment Anything Model (SAM), exploring its architecture, implementation, and zero-shot capabilities for advanced image segmentation tasks through detailed code analysis and practical demonstrations.

• YouTube
• 45 minutes
• Self-Paced
• Free Video

A Systematic Approach to Reading Deep Learning Codebases

Master a systematic approach to understanding complex deep learning codebases, from initial paper review to mapping structure and components, with practical examples using SAM1.

• YouTube
• 15 minutes
• Self-Paced
• Free Video

Intersection over Union (IoU) Explained with PyTorch

Learn to implement Intersection over Union (IoU) for object recognition using PyTorch. Covers theory, formulas, and practical code for bounding boxes and segmentation masks.

• YouTube
• 18 minutes
• Self-Paced
• Free Video

Backpropagation with Automatic Differentiation from Scratch in Python

Master automatic differentiation and backpropagation by building PyTorch's autograd system from scratch in Python with hands-on implementation of core operations.

• YouTube
• 27 minutes
• Self-Paced
• Free Video

AdamW Optimizer from Scratch in Python

Discover how to implement the state-of-the-art AdamW optimizer from scratch, understanding its unique weight decay regularization that outperforms traditional Adam + L2.

• YouTube
• 20 minutes
• Self-Paced
• Free Video

How to Build RuneScape for RL Agent - Neural MMO v3 with PufferLib 2.0

Dive into building MMO environments for reinforcement learning agents with Dr. Joseph Suarez, exploring PufferLib 2.0 and Neural MMO v3.0 development insights.

• YouTube
• 1 hour 29 minutes
• Self-Paced
• Free Video

How to Contribute to Open Source AI with Zero Experience

Discover how to contribute to open source AI projects with zero experience using PufferLib as an example, featuring insights from contributor DanAdvantage's journey.

• YouTube
• 15 minutes
• Self-Paced
• Free Video

What Are Multi-Agent Systems? - Theory and Code Overview with MetaGPT

Explore multi-agent systems through theoretical foundations and practical implementation with MetaGPT, including a hands-on software project demonstration and code overview.

• YouTube
• 29 minutes
• Self-Paced
• Free Video

MiniMax-01 Theory Overview - Lightning Attention + MoE + FlashAttention Optimization

Explore MiniMax-01's architecture featuring Lightning Attention, MoE, and FlashAttention optimizations that enable a 4M token context window and 456B parameters, outperforming Llama 3.1 and challenging Claude in benchmarks.

• YouTube
• 47 minutes
• Self-Paced
• Free Video

Exploring "Logical Calculus of Nervous Activity" by McCulloch and Pitts - 1943

Dive into the foundational 1943 McCulloch-Pitts paper that predates computers yet laid groundwork for modern deep learning and neural networks.

• YouTube
• 1 hour 35 minutes
• Self-Paced
• Free Video

Joint Self-Supervised Compression Method for ARC-AGI Problem Solving

Explore a controversial compression method achieving 27.5% on ARC-AGI v1 for just $2, featuring an in-depth interview with author Mithil Vakde on joint self-supervised learning.

• YouTube
• 1 hour 12 minutes
• Self-Paced
• Free Video