Skip to content

Enabling Persistence with what-if tensorboard plugin #219

@etsyamitra

Description

@etsyamitra

I have been using this tool through tensorboard and find there is does not seem to be an backend persistence into GCS with this tool unlike other third party tensorboard plugins like profiler.
I would love if there could be a way we can save the tool outputs/stats somewhere, such that we do not loose the data/configured contexts on browser refresh etc.

Curious if there is some expected practices for this tool, using the REST endpoints as exposed here ? would be ideal if can pass the arguments to the endpoints and get a http link to tensorboard instance with a uid/session id etc which would have all the entries baked up and we can just keep visulalizing over the amazing UI. i.e. configure inputs like (inference address, signature, data_path etc) through REST and rest over UI!

I am pretty new to this tool, so please direct me to recommended resources. Thank you in advance!

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions