公开(公告)号：US11176489B1

公开(公告)日：2021-11-16

申请号：US15954913

申请日：2018-04-17

Applicant: Amazon Technologies, Inc.

Inventor： Alexander Johannes Smola ,  Edo Liberty ,  Mu Li ,  Leyuan Wang

IPC: G06N20/00 ,  G06F9/54 ,  G06K9/62 ,  G06N3/04 ,  G06N3/02

Abstract: Techniques for determining and utilizing optimal aggregation schedules are described are described. A deep machine learning model can be trained using multiple processing elements implemented in one or multiple computing devices and that are interconnected using one or multiple types of links. An optimal aggregation schedule for such arbitrary topologies can be determined automatically. The determination may include solving a linear program on the spanning tree polytope. The optimal aggregation schedule can be utilized by the multiple processing elements to train the deep machine learning model.

公开(公告)号：US11797876B1

公开(公告)日：2023-10-24

申请号：US16453489

申请日：2019-06-26

Applicant: Amazon Technologies, Inc.

Inventor： Leyuan Wang ,  Yida Wang ,  Mu Li ,  Zhi Chen ,  Yizhi Liu ,  Yao Wang

IPC: G06N20/00 ,  G06T7/10 ,  G06T1/20 ,  G06F9/455 ,  G06N3/082 ,  G06F8/41 ,  G06F18/24

CPC classification number: G06N20/00 ,  G06F8/443 ,  G06F8/447 ,  G06F8/451 ,  G06F9/45558 ,  G06F18/24 ,  G06N3/082 ,  G06T1/20 ,  G06T7/10 ,  G06F2009/4557

Abstract: Techniques for optimizing and deploying convolutional neural network (CNN) machine learning models for inference using integrated graphics processing units are described. A model compilation system optimizes CNN models using optimized vision-specific operators as well as both graph-level tuning and tensor-level tuning to explore the optimization space for achieving heightened performance. The model compilation system may also implement a heuristic-based two-stage technique for falling back certain operators of CNN models to use CPUs when needed or otherwise beneficial.

公开(公告)号：US12265920B2

公开(公告)日：2025-04-01

申请号：US16849537

申请日：2020-04-15

Applicant: Amazon Technologies, Inc.

Inventor： Jonas Mueller ,  Tatsuya Arai ,  Abhi Vinayaka Sharma ,  Ryan Matthew Brand ,  Yohei Nakayama ,  Nick Dean Erickson ,  Hang Zhang ,  Mu Li ,  Alexander Johannes Smola

IPC: H04L29/06 ,  G06N5/04 ,  G06N20/00

Abstract: Techniques for code-free automated machine learning (ML) are described. Users can train high-quality ML models and pipelines without necessarily needing to write code by providing a training dataset to a code-free machine learning service. The service may deploy an ML orchestration function and a storage location on behalf of a user. When a modification is made to the storage bucket, such as by the user providing a training dataset, the orchestration function is invoked and can automatically initiate an AutoML process using at least the training data to train multiple ML model variants. The resultant ML model(s) and associated metrics can be provided to the user, deployed behind an endpoint, and/or used to generate inferences.