- cv
- AI
- Computer Vision
•
•
-
LLM from scratch - 1.3 Multi Head Self Attention
Why Multi Head Attention is important, mathematical formulations, and implementation.
-
LLM from scratch - 1.2 Single Head Self Attention
Implementation details of single head self attention and causal masks.
-
LLM from scratch - 1.1 Positional Encoding
Positional encoding explanations and code implementations, including learned and sinusoidal encoding.
-
Deep Dive into MicroGPT by Karpathy
A detailed walkthrough of Karpathy's MicroGPT, covering dataset preparation, character-level tokenization, a minimal autograd engine (the Value class), Python special methods, and backpropagation via topological sort.
-
[CS231n] Assignment 1 - Q2. Implement a Softmax Classifier
A walkthrough of implementing a Softmax Classifier for CS231n Assignment 1, including preprocessing, loss function derivation, and gradient computation.