Inference model optimization
    4.
    发明授权

    公开(公告)号:US12008487B2

    公开(公告)日:2024-06-11

    申请号:US17122774

    申请日:2020-12-15

    CPC classification number: G06N5/048 G06F16/54

    Abstract: An approach to optimize performance for large scale inference models. Data in the form of images is received from sensors such as cameras. The data is processed to generate data tags associated with the context of the image and portion the images. Model tags are generated based on data characteristics or user input. The tags and their associated data are added to a time-based queue for delivery to the appropriate inference models. Based on the embedded delivery time and frequency, the portioned images are delivered to the appropriate inference models.

    Container image migration service

    公开(公告)号:US11556499B2

    公开(公告)日:2023-01-17

    申请号:US17154254

    申请日:2021-01-21

    Abstract: A method, system and computer program product for container image migration service is provided. The method comprises identifying a latest version of a first customer container image stored in a container image repository. The method further comprises determining the latest version of the first customer container image is a migration image from a last version of the first customer container image; determining a set of commands in the Docker file of the last version of the first customer container image that have migrated to a corresponding set of commands in the Docker file of the migration image; identifying a latest version of a second customer container image having at least one Docker file command in common with at least one command in the set of commands; and recommending imminent migration of the second customer container image to include migration of the at least one Docker file command.

    CONTAINER IMAGE MIGRATION SERVICE

    公开(公告)号:US20220229804A1

    公开(公告)日:2022-07-21

    申请号:US17154254

    申请日:2021-01-21

    Abstract: A method, system and computer program product for container image migration service is provided. The method comprises identifying a latest version of a first customer container image stored in a container image repository. The method further comprises determining the latest version of the first customer container image is a migration image from a last version of the first customer container image; determining a set of commands in the Docker file of the last version of the first customer container image that have migrated to a corresponding set of commands in the Docker file of the migration image; identifying a latest version of a second customer container image having at least one Docker file command in common with at least one command in the set of commands; and recommending imminent migration of the second customer container image to include migration of the at least one Docker file command.

    INFERENCE MODEL OPTIMIZATION
    8.
    发明申请

    公开(公告)号:US20220188676A1

    公开(公告)日:2022-06-16

    申请号:US17122774

    申请日:2020-12-15

    Abstract: An approach to optimize performance for large scale inference models. Data in the form of images is received from sensors such as cameras. The data is processed to generate data tags associated with the context of the image and portion the images. Model tags are generated based on data characteristics or user input. The tags and their associated data are added to a time-based queue for delivery to the appropriate inference models. Based on the embedded delivery time and frequency, the portioned images are delivered to the appropriate inference models.

Patent Agency Ranking