GitHub stars · StarDay

1alibaba / MNNMNN: A blazing-fast, lightweight inference engine battle-tested by Alibaba, powering high-performance on-device LLMs and Edge AI.C++Apache-2.0 · Permissive16,343

2disintegration / imagingImaging is a simple image processing package for GoGoMIT · Permissive5,768

3Jounce / SurgeA Swift library that uses the Accelerate framework to provide high-performance functions for matrix math, digital signal processing, and image manipulation.SwiftMIT · Permissive5,325

4uxlfoundation / oneDNNOpen-source library of optimized deep learning operations (matmul, convolution, attention) for CPUs (x64, AArch64, RISC-V) and Intel GPUs. Used by PyTorch, TensorFlow, OpenVINO, and ONNX Runtime.C++Apache-2.0 · Permissive4,090

5zjhellofss / KuiperInferGood projects for campus recruitment, autumn recruitment, spring recruitment, and internships! It guides you to build a high-performance deep learning inference library from scratch, supporting the inference of models such as llama2, Unet, Yolov5, and Resnet. Implement a high-performance deep learning inference library step by step.C++MIT · Permissive3,520