GPU implementation of a fast generalized ANS (asymmetric numeral system) entropy encoder and decoder, with extensions for lossless compression of numerical and other data types in HPC/ML applications.

LLM training in simple, raw C/CUDA

Segmentation models with pretrained backbones. PyTorch.

Inference Llama 2 in one file of pure Rust 🦀

Inference Llama 2 in one file of pure C

Inference Llama 2 in one file of pure C & one file with CUDA

Efficient Deep Learning Systems course materials (HSE, YSDA)

C++ Cheat Sheet for ACM ICPC

Zero Bubble Pipeline Parallelism

Incredibly fast Whisper-large-v3

Inference Llama 2 in one file of pure C

Llama 2 Everywhere (L2E)

Inference Llama 2 in C++ to use Eigen

Implementation of the paper "DeepLSD: Line Segment Detection and Refinement with Deep Image Gradients"

Data processing and simulation for serial crystallography

Locally run an Instruction-Tuned Chat-Style LLM (Android/Linux/Windows/Mac)

bfloat16 dtype for numpy

This is an official implementation for "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows" on Semantic Segmentation.

See what's happening on GitHub in real time

Some experiments on Imagenette/Imagewoof datasets

torch-optimizer -- collection of optimizers for Pytorch

$ git remote <graduation> yearbook

The Incredible PyTorch: a curated list of tutorials, papers, projects, communities and more relating to PyTorch.

Tensors and Dynamic neural networks in Python with strong GPU acceleration