sulrash Goto Github PK

ascii-game

Making my first proper game in Godot

biggest-winner-and-loser

Firefox extension that shows the biggest winner and loser for the current trading day

cheatsheet

An attempt at improving facial recognition performance through appending a 'cheatsheet' to an image with one positive sample and multiple negatives during training.

comparing-cnn-mlp

Me researching into the performance differnce between MLPs and CNNs for image classification! Hope this is useful to someone. I had done this back in may as a self exercise that a professor suggested, but just got around to uploading it. I implore you to check out the PDF!

datatrove

Freeing data processing from scripting madness by providing a set of platform-agnostic customizable pipeline processing blocks.

discord-data-reader

Presents the discord data package you can download in a neat and clean way.

easyrogue

A simple rogue like game ripped out from my third year project titled "Perfect Information Versus Imperfect Information in Reinforcement Learning". I had made this simple game to benchmark performance, and I hope other people can get some use out of it too!

engagement-level-prediction

Engagement Intensity Prediction in Real TIme (converted to tf2)

envenc

Repository for environment encoder, an attempt at improving reinforcement learning agents' generalisability through learning how to act on universal multimodal embeddings generated by a vision-language model.

fastertransformer_backend

gradient-based-optimization

An exploration into the different types of gradient based optimizations and the performance discrepancy between each of them. I had done this project in university in an effort to explore more of my passion in machine learning and artificial intelligence under the supervision and guidance of a professor. I implore you to read the results analysis!

graduation-project

My final year honours project that involved developing a rogue like game from scratch and implementing some RL algorithms. The purpose was to compare perfect information vs imperfect information. I'm skeptical about my conclusion that imperfect information is better in certain cases, but enjoy!

huggingface-megatron-lm

Ongoing research training transformer models at scale

inseq

Interpretability for sequence generation models 🐛 🔍

lm-evaluation-harness

Fork of LM harness to add attention mapping

medievalsd

Just a quick little project I'm working on for my dungeons & dragons world, creating a stable diffusion model that can reliably create those medieval sketch drawings. Most probably will LoRA it.

minllmtrain

Minimal yet high performant code for pretraining llms. Attempts to implement some SOTA features. Implements training through: Deepspeed, Megatron-LM, and FSDP. WIP

montecarlo-hex

AI and Games project

numerical-heat-conductance

Numerical Solution of the Heat Conductance Equation. Please read the PDF, it's a lot of maths

python-pathfinding

Implementation of common pathfinding algorithms. My fork just adds naive random walks.

sapien

search-discord-data

Searches the discord messages given to you after you download your data from discord.

sulrash.github.io

My github pages website.

uftrace

Function graph tracer for C/C++/Rust

voice-memo-sharing

Records a voice memo and shares it to your apps conveniently. Made it because I wanted to share voice notes on discord.