8000 Getting capped at 90 samples per label when creating dataset · Issue #2098 · NVIDIA/DIGITS · GitHub
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content
8000
This repository was archived by the owner on Jan 7, 2025. It is now read-only.
This repository was archived by the owner on Jan 7, 2025. It is now read-only.
Getting capped at 90 samples per label when creating dataset #2098
Open
@chrishadiono

Description

@chrishadiono

screenshot from 2018-08-13 12 57 36
Hi,
I am making a dataset where I have hundreds of images per label. However, when I create a dataset with this data in DIGITS it only includes 90 of them maximum per label. I have attached an image of what I mean here. I tried using the mnist data and did not experience this cap using the exact same settings. Originally I thought it was because my images are each around 30kb each and maybe Digits cannot handle large sizes, but I also tried reducing the jpeg quality of the images very low so they were each 3-4kb each (like the mnist example) and instead was capped at 100. Any help would be greatly appreciated!

Note: I am using the master versions of DIGITS and using the Use S3 tab

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions

      0