modern-fortran/neural-fortran

DaCe - Data Centric Parallel Programming

Automatically Discovering Fast Parallelization Strategies for Distributed Deep Neural Network Training

Deep Learning for humans

Rust based framework to train Neural Network

Tile primitives for speedy kernels

Distributed training framework for TensorFlow, Keras, PyTorch, and Apache MXNet.

Development repository for the Triton language and compiler

Deep Declarative Networks

cuDNN Frontend is NVIDIA's modern, open-source entry point to the cuDNN library and a growing collection of high-performance open-source kernels.