sudomishra Goto Github PK

ag4masses

Making Google Deepmind's AlphaGeometry accessible to the Masses

agents.jl

Agent-based modeling framework in Julia

algowiki

A wiki dedicated to competitive programming

assessment2

FreeSound Kaggle

assessment3

autodiff-puzzles

blog

My Academic blog where I post things related to my research interests.

chaostools.jl

Tools for the exploration of chaos and nonlinear dynamics

cleveralgorithms

Clever Algorithms: Nature-Inspired Programming Recipes

cs433-a2

cuhnsw

CUDA implementation of Hierarchical Navigable Small World Graph algorithm

deduplicate_lsh

Script to find and remove near duplicates in csv file

deepexplain

A unified framework of perturbation and gradient-based attribution methods for Deep Neural Networks interpretability. DeepExplain also includes support for Shapley Values sampling. (ICLR 2018)

dopamine

Dopamine is a research framework for fast prototyping of reinforcement learning algorithms.

dynamicalsystemsbase.jl

Definition of dynamical systems and integrators for DynamicalSystems.jl

fast-autocomplete

Fast Autocomplete: When Elastcsearch suggestions are not fast and flexible enough

flash-attention

Fast and memory-efficient exact attention

flaskr

fullgrad-saliency

Full-gradient saliency maps

github-pages-with-jekyll

googleqachallenge

gpt-fast

Simple and efficient pytorch-native transformer text generation in <1000 LOC of python.

gpu-puzzles

Solve puzzles. Learn CUDA.

graph_comb_opt

Implementation of "Learning Combinatorial Optimization Algorithms over Graphs"

how_to_optimize_in_gpu

This is a series of GPU optimization topics. Here we will introduce how to optimize the CUDA kernel in detail. I will introduce several basic kernel optimizations, including: elementwise, reduce, sgemv, sgemm, etc. The performance of these kernels is basically at or near the theoretical limit.

humanoid

kenlm

KenLM: Faster and Smaller Language Model Queries

llama-recipes

Examples and recipes for Llama 2 model

llama.cpp

LLM inference in C/C++

llamafile

Distribute and run LLMs with a single file.