Versatile Agentic Library and Workflow Orchestrator

Massive Multimodal Open RAG & Extraction A scalable multimodal pipeline for processing, indexing, and querying multimodal documents Ever needed to take 8000 PDFs, 2000 videos, and 500 spreadsheets …

Scalable and Performant Data Loading

The simplest but fast implementation of matrix multiplication in CUDA.

Benchmarking framework to measure performance of data storage and ingestion for deep learning

VF-NeRF code. We learn to densely reconstruct indoor from multi-view images by representing the surface with Vector Fields (VF). We develop a novel density-VF relationship and a training scheme tha…

A tool to configure, launch and manage your machine learning experiments.

Efficient Triton Kernels for LLM Training

This repo contains annotated research papers that I found really good and useful

YaFSDP: Yet another Fully Sharded Data Parallel

DISCO is a code-free and installation-free browser platform that allows any non-technical user to collaboratively train machine learning models without sharing any private data.

Supercharge Your Model Training

Tile primitives for speedy kernels

A PyTorch Native LLM Training Framework

Lighteval is your all-in-one toolkit for evaluating LLMs across multiple backends

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

Minimalistic large language model 3D-parallelism training

GPU programming related news and material links

code for KalmanNet

Everything we actually know about the Apple Neural Engine (ANE)

PyTorch Lightning + Hydra. A very user-friendly template for ML experimentation. ⚡🔥⚡

Latex code for making neural networks diagrams

Examples of how to create colorful, annotated equations in Latex using Tikz.