Deep Learning Inference benchmark. Supports OpenVINO™ toolkit, TensorFlow, TensorFlow Lite, ONNX Runtime, OpenCV DNN, MXNet, PyTorch, Apache TVM, ncnn, PaddlePaddle, etc.

This repository contains the code for training YOLOv5, YOLOv7, and YOLOv8 models for object detection using Python 3. Optimization of models with OpenVINO and Apache TVM, and tracking with ByteTrack and yolov8-pose.

compilation with optimization and code generation that can take neural network inference requests and divide the load between classical hardware and novel photonic hardware.

Optimizing RetinaFace inference using TVM.

Experiments using Machine Learning Compilers to improve Seismic Attributes computation

Machine Learning Compilers