Blogs

Stable Diffusion was one of the first Open-Souce Text to Image model that could generate high quality images. However, the maths behind it is fairly complex and this blog aims to simplify it. Before understanding Diffusion Models, we need to understand a type of generative model called Energy Based Models. Energy Based Models Recommended reading: How to Train Your Energy-Based Models https://www.youtube.com/watch?v=8TcNXi3A5DI Let us say we have an image dataset $X = \{ x_1, x_2, \dots, x_n \}$, where $ x_i \in \mathbb{R}^d $....

The mathematics behind generative models is a complex combination of Probability, Statistics, Thermodynamics, Optimization, Linear Algebra, and Information Theory. Let is try to wrap our heads around it. Let us say we have an image dataset $X = \{ x_1, x_2, \dots, x_n \}$, where $ x_i \in \mathbb{R}^d $. We want to find a process that can “generate” new data points $ x $ that are similar to images in $ X $....

Collection of helpful commands for Debian based systems

Estimating the parameters of a distribution given the data is a fundamental problem in statistics. In this blog, we will look at different methods to estimate the parameters of a distribution given the data. We will also look at the different assumptions we make about the noise and the parameters.

Machine Learning deals with mathematically modelling real-world data. So it is important to have a good understanding of the math behind it.

Notes on LLMs

Notes on Reinforcement Learning

Deep Learning notes and tips that I have gathered over time. This is a living document and will be updated as I learn more. Useful Links Google Tuning Playbook Efficient Training on a Single GPU [http://karpathy.github.io/2019/04/25/recipe/](Deep Learning Recipe by Andrej Karpathy) General points Try to over fit a small batch, a decent model should almost every time able to fit a small batch, if it fails something is wrong with the model or the data....

A quick guide on screen resolutions.

Docker can be used to quickly try out a new software stack with lot of dependencies but one is generally forced to google the essential docker commands and arguments every time a new container is to be launched. This cheat sheet aims to list the important steps to build and launch a docker container and the arguments used with it. Hope it helps!