Method and apparatus for distributed training of artificial intelligence model in channel-sharing network environment
AssigneeElectronics and Telecommunications Research Institute
InventorsKi-Dong KANG, Hong-Yeon KIM, Baik-Song AN, Myung-Hoon CHA
Life of this patent
Estimating the term…
Abstract
Disclosed herein is a method for distributed training of an AI model in a channel-sharing network environment. The method includes determining whether data parallel processing is applied, calculating a computation time and a communication time when input data is evenly distributed across multiple computation devices, and unevenly distributing the input data across the multiple computation devices based on the computation time and the communication time.
Claims14
A method for distributed training of an Artificial Intelligence (AI) model in a channel-sharing network environment including multiple computation devices, comprising:
- determining whether data parallel processing is applied;
- calculating a computation time and a communication time when input data is evenly distributed across the multiple computation devices;
- and
- unevenly distributing the input data across the multiple computation devices based on the computation time and the communication time.
An apparatus for distributed training of an Artificial Intelligence (AI) model in a channel-sharing network environment including multiple computation devices, comprising:
- a parallelism identification unit for determining whether data parallel processing is applied;
- a profiling unit for calculating a computation time and a communication time when input data is evenly distributed across the multiple computation devices;
- and
- a data distribution unit for unevenly distributing the input data across the multiple computation devices based on the computation time and the communication time.
Dependencies are inferred from the claim text.