Скачать с ютуб Torch-MLIR 101 в хорошем качестве

Torch-MLIR 101

Abstract: In this meetup, we will be discussing the basics of MLIR and deep learning computations and dive into torch-mlir. This presentation will give the basic details of different passes a torchscript graph goes through and finally output TOSA, Linalg or MHLO backend (dialects) in torch-mlir. This presentation is meant for someone who wants to contribute to torch-mlir Speaker Bio: Prashant Kumar is a Deep learning compiler engineer at NOD labs. Prior to Nod labs he was working as a Machine learning Compiler Engineer at Qualcomm. He holds a Masters degree in Computer Science from IIT Delhi. He is an active contributor to the Torch-MLIR project and FPL library Github: https://github.com/pashu123 LinkedIn:   / prashantkumar25   Slides: https://drive.google.com/file/d/1ZSlk...