Aleksa Gordić - The AI Epiphany Courses

ETA Prediction with Graph Neural Networks in Google Maps - Paper Explained

1 rating

Explore how Google Maps uses graph neural networks to predict estimated arrival times, including graph construction, multi-horizon models, variance reduction, inference, and production engineering.

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

Will LLMs Kill Search? The Future of Information Retrieval

Explore the future of search with Nils Reimers, discussing RAG, multilingual embedding, complex data search, and Cohere-Rerank. Gain insights on LLMs' impact on search technology.

• YouTube
• 1 hour 7 minutes
• Self-Paced
• Free Video

Flash Attention 2.0 with Tri Dao - Discord Server Talks

Explore Flash Attention 2.0 with author Tri Dao, covering its motivation, improvements, and future directions. Gain insights into memory bottlenecks and IO awareness in attention mechanisms for long sequences.

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

Building Web Apps Using Streamlit - Streamlit Crash Course - MLOps Series

Build and deploy a Python Streamlit web app using the Hugging Face Inference API.

• YouTube
• 1 hour 6 minutes
• Self-Paced
• Free Video

Implementing DeepMind's DQN from Scratch - Project Update

A candid code walkthrough of an unfinished Deep Q-Network project, focusing on debugging reinforcement-learning components, logging, and implementation workflow.

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

EfficientNetV2 - Smaller Models and Faster Training - Paper Explained

Explains EfficientNetV2’s neural architecture search, fused-MBConv blocks, progressive training, and regularization for smaller models and faster image-model training.

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

MuZero - Mastering Atari, Go, Chess, and Shogi by Planning with a Learned Model - RL Paper Explained

An explanation of MuZero’s model-based reinforcement learning approach, which learns game dynamics and uses tree search to play Atari, Go, chess, and shogi without rules.

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

OpenAI - Solving Rubik's Cube with a Robot Hand - RL Paper Explained

A paper explainer on training a robotic hand in simulation to solve a Rubik’s Cube using automatic domain randomization for sim-to-real transfer.

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

DeepMind's AlphaGo Zero and AlphaZero - RL Paper Explained

Explains how AlphaGo Zero and AlphaZero master Go, chess, and shogi through self-play reinforcement learning without human game data or domain knowledge.

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

AlphaGo - Mastering the Game of Go with Deep Neural Networks and Tree Search - RL Paper Explained

A detailed walkthrough of AlphaGo’s deep neural networks, reinforcement-learning policies, and Monte Carlo tree search for playing Go.

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

DQN - Playing Atari with Deep Reinforcement Learning - RL Paper Explained

A detailed walkthrough of the DQN paper, from experience replay and target networks to Atari training results.

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

How to Get Started With Graph ML - Blog Walkthrough

A research-oriented guide to getting started with graph machine learning, covering applications, graph embeddings, GNNs, geometric deep learning, resources, and a GAT project.

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

Graph Attention Network Project Walkthrough

Walk through a Graph Attention Network implementation, from Cora graph data and edge indices to neighborhood-aware softmax and aggregation.

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

Temporal Graph Networks - GNN Paper Explained

A visual deep dive into Temporal Graph Networks, explaining how memory, temporal sampling, and graph attention model evolving interactions.

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

OpenAI CLIP - Connecting Text and Images - Paper Explained

Explains CLIP’s contrastive image-text pretraining and its zero-shot transfer, embedding quality, robustness, and limitations.

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