Method, apparatus and computer program product for resource scheduling

Patent metadata

Field	Value
Publication number	US-11436050-B2
Application number	US-201916380654-A
Country	US
Kind code	B2
Filing date	Apr 10, 2019
Priority date	Apr 20, 2018
Publication date	Sep 6, 2022
Grant date	Sep 6, 2022

Abstract

Embodiments of the present disclosure provide a method, apparatus and computer program product for resource scheduling. The method comprises obtaining a processing requirement for a deep learning task, the processing requirement being specified by a user and at least including a requirement related to a completion time of the deep learning task. The method further comprises determining, based on the processing requirement, a resource required by the deep learning task such that processing of the deep learning task based on the resource satisfies the processing requirement. Through the embodiments of the present disclosure, the resources can be scheduled reasonably and flexibly to satisfy the user's processing requirement for a particular deep learning task without requiring the user to manually specify the requirement on the resources.

First claim

What is claimed is:

1. A method of resource scheduling, comprising steps of:
obtaining a processing requirement for a deep learning task, the processing requirement being specified by a user and at least comprising a requirement related to a completion time of the deep learning task; and
determining, based on the processing requirement, a resource required by the deep learning task such that processing of the deep learning task based on the resource satisfies the processing requirement;
wherein determining the resource required by the deep learning task comprises:
determining a plurality of sets of candidate resources that satisfy the processing requirement;
determining at least a predicated completion time associated with each set of candidate resources;
presenting the sets of candidate resources and respective predicated completion times to a user interface;
receiving, from the user interface, a user selection of a given set of the sets of candidate resources; and
in response to the user selection of the given set, selecting the resource required by the deep learning task; and

wherein one or more of the predicated completion times associated with respective sets of candidate resources is less than or equal to the completion time specified in the processing requirement; and
wherein the steps are performed by a processor and a memory coupled to the processor and having instructions stored thereon which are executed by the processor.

2. The method of claim 1 , wherein determining the resource required by the deep learning task further comprises:
obtaining representation data and a processing parameter of the deep learning task; and
determining the resource based on the representation data and the processing parameter.

3. The method of claim 1 , wherein the processing requirement further comprise a requirement related to a processing cost of the deep learning task.

4. The method of claim 1 , wherein determining the resource required by the deep learning task comprises determining at least one of:
a dedicated processing resource;
a general processing resource; and
a storage resource.

5. The method of claim 1 , further comprising:
allocating the determined resource from a resource pool for processing the deep learning task.

6. An apparatus for resource scheduling, comprising:
a processor; and
a memory coupled to the processor and having instructions stored thereon which, when executed by the processor, cause the apparatus to perform steps comprising:
obtaining a processing requirement for a deep learning task, the processing requirement being specified by a user and at least comprising a requirement related to a completion time of the deep learning task; and
determining, based on the processing requirement, a resource required by the deep learning task such that processing of the deep learning task based on the resource satisfies the processing requirement;

wherein determining the resource required by the deep learning task comprises:
determining a plurality of sets of candidate resources that satisfy the processing requirement;
determining at least a predicated completion time associated with each set of candidate resources;
presenting the sets of candidate resources and respective predicated completion times to a user interface;
receiving, from the user interface, a user selection of a given set of the sets of candidate resources; and
in response to the user selection of the given set, selecting the resource required by the deep learning task; and

wherein one or more of the predicated completion times associated with respective sets of candidate resources is less than or equal to the completion time specified in the processing requirement.

7. The apparatus of claim 6 , wherein determining the resource required by the deep learning task further comprises:
obtaining representation data and a processing parameter of the deep learning task; and
determining the resource based on the representation data and the processing parameter.

8. The apparatus of claim 6 , wherein the processing requirement further comprises a requirement related to a processing cost of the deep learning task.

9. The apparatus of claim 6 , wherein determining the resource required by the deep learning task comprises determining at least one of:
a dedicated processing resource;
a general processing resource; and
a storage resource.

10. The apparatus of claim 6 , wherein the steps further comprise:
allocating the determined resource from a resource pool, for processing the deep learning task.

11. A computer program product being tangibly stored on a non-transitory computer-readable medium and comprising machine-executable instructions which, when executed, cause a machine to perform steps comprising:
obtaining a processing requirement for a deep learning task, the processing requirement being specified by a user and at least comprising a requirement related to a completion time of the deep learning task; and
determining, based on the processing requirement, a resource required by the deep learning task such that processing of the deep learning task based on the resource satisfies the processing requirement;
wherein determining the resource required by the deep learning task comprises:
determining a plurality of sets of candidate resources that satisfy the processing requirement;
determining at least a predicated completion time associated with each set of candidate resources;
presenting the sets of candidate resources and respective predicated completion times to a user interface;
receiving, from the user interface, a user selection of a given set of the sets of candidate resources; and
in response to the user selection of the given set, selecting the resource required by the deep learning task; and

wherein one or more of the predicated completion times associated with respective sets of candidate resources is less than or equal to the completion time specified in the processing requirement.

12. The computer program product of claim 11 , wherein determining the resource required by the deep learning task further comprises:
obtaining representation data and a processing parameter of the deep learning task; and
determining the resource based on the representation data and the processing parameter.

13. The computer program product of claim 11 , wherein the processing requirement further comprise a requirement related to a processing cost of the deep learning task.

14. The computer program product of claim 11 , wherein determining the resource required by the deep learning task comprises determining at least one of:
a dedicated processing resource;
a general processing resource; and
a storage resource.

15. The method of claim 1 , wherein the user selection is performed via a user input interface.

16. The apparatus of claim 6 , wherein the user selection is performed via a user input interface.

17. The computer program product of claim 11 , wherein the steps further comprise:
allocating the determined resource from a resource pool, for processing the deep learning task.

18. The method of claim 3 , further comprising presenting the processing cost to the user interface.

19. The apparatus of claim 8 , further comprising presenting the processing cost to the user interface.

20. The computer program product of claim 13 , further comprising presenting the processing cost to the user interface.

Assignees

• Emc Ip Holding Co Llc

Inventors

• Peng Layne Lin
• Wang Kun
• Li Sanping

Classifications

• G06N3/08

  Learning methods · CPC title

• G06F9/4887 Primary

  involving deadlines, e.g. rate based, periodic · CPC title

• G06F2209/485

  Resource constraint · CPC title

• G06N20/00

  Machine learning · CPC title

• G06F9/4881

  Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues · CPC title