Reference implementation of the Transformer architecture optimized for Apple Neural Engine (ANE)

A playbook for systematically maximizing the performance of deep learning models.

Submanifold sparse convolutional networks

AISystem 主要是指AI系统，包括AI芯片、AI编译器、AI推理和训练框架等AI全栈底层技术

Fast and memory-efficient exact attention

Roadmap to becoming an Artificial Intelligence Expert in 2022

深度学习经典、新论文逐段精读

Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more

List of Computer Science courses with video lectures.

The C++ Core Guidelines are a set of tried-and-true guidelines, rules, and best practices about coding in C++

A Python-level JIT compiler designed to make unmodified PyTorch programs faster.

搞定C++:punch:。C++ Primer 中文版第5版学习仓库，包括笔记和课后练习答案。

RapidStream TAPA compiles task-parallel HLS program into high-frequency FPGA accelerators.

by ex-googlers, for ex-googlers - a lookup table of similar tech & services

Development repository for the Triton language and compiler

程序员延寿指南 | A programmer's guide to live longer

Lingvo

Open deep learning compiler stack for cpu, gpu and specialized accelerators

FlexASR: A Reconfigurable Hardware Accelerator for Attention-based Seq-to-Seq Networks

Vitis HLS LLVM source code and examples

Reinforcement learning environments for compiler and program optimization tasks

Brevitas: neural network quantization in PyTorch

Dataflow compiler for QNN inference on FPGAs

Vitis HLS Library for FINN

Stencil with Optimized Dataflow Architecture Compiler

Training neural models with structured signals.

HeteroCL: A Multi-Paradigm Programming Infrastructure for Software-Defined Heterogeneous Computing

Machine learning on FPGAs using HLS

A toolkit to optimize ML models for deployment for Keras and TensorFlow, including quantization and pruning.