Tensors and Dynamic neural networks in Python with strong GPU acceleration

A PyTorch native platform for training generative AI models

Ongoing research training transformer models at scale

Development repository for the Triton language and compiler

A high-throughput and memory-efficient inference and serving engine for LLMs

TorchBench is a collection of open source benchmarks used to evaluate PyTorch performance.

Fast and memory-efficient exact attention

DeepEP: an efficient expert-parallel communication library

NeMo: a toolkit for conversational AI

Optimized primitives for collective multi-GPU communication

The LLVM Project is a collection of modular and reusable compiler and toolchain technologies.

Triton backend that enables pre-process, post-processing and other logic to be implemented in Python.

This is now the official location of the Kaldi project.

Triton Python, C++ and Java client libraries, and GRPC-generated client examples for go, java and scala.

Moved to https://github.com/k2-fsa/icefall

FSA/FST algorithms, differentiable, with PyTorch compatibility.

Seamless operability between C++11 and Python

SCIP - Solving Constraint Integer Programs

Google's Operations Research tools:

NVIDIA® TensorRT™, an SDK for high-performance deep learning inference, includes a deep learning inference optimizer and runtime that delivers low latency and high throughput for inference applicat…

The RedPajama-Data repository contains code for preparing large datasets for training large language models.

nanobind: tiny and efficient C++/Python bindings

Easily convert common crawl to videos with their subtitles, while honoring licenses.

An open-source static random access memory (SRAM) compiler.

Python bindings for the simdjson project.

Robust Speech Recognition via Large-Scale Weak Supervision