US 12,333,300 B2
Method and apparatus for managing model file in inference application
Zhilong Yang, Shenzhen (CN)
Assigned to HUAWEI TECHNOLOGIES CO., LTD., Shenzhen (CN)
Filed by HUAWEI TECHNOLOGIES CO., LTD., Guangdong (CN)
Filed on Jan. 6, 2023, as Appl. No. 18/150,967.
Application 18/150,967 is a continuation of application No. PCT/CN2021/102489, filed on Jun. 25, 2021.
Claims priority of application No. 202010669862.9 (CN), filed on Jul. 13, 2020.
Prior Publication US 2023/0153100 A1, May 18, 2023
Int. Cl. G06F 8/71 (2018.01); G06F 8/65 (2018.01)
CPC G06F 8/71 (2013.01) [G06F 8/65 (2013.01)] 10 Claims
OG exemplary drawing
 
1. A method for managing deep-learning model files in a Kubernetes system, comprising:
monitoring, by a network device functioning as a master node in the Kubernetes system and via a Kubernetes interface, a status of a target model management object;
determining, by the network device, a target inference application that is deployed on a worker node in the Kubernetes system and matches the target model management object based on a preset field of the target inference application, wherein the preset field of the target inference application identifies the target model management object being corresponding to a target model file of the target inference application;
based on the status of the target model management object, notifying, by the network device and via the Kubernetes interface, the target inference application to perform an operation on the target model file of the target inference application according to the status of the target model management object; and
performing, by the worker node in the Kubernetes system, the operation on the target model file of the target inference application according to the status of the target model management object.