"HTTP Error" when exporting from Document Manager to AI Center

Hi, I’m currently learning to use the AI Center, and just finished data labeling in the Document Manager.

I’m getting a strange error when trying to export my labeled data from the Document Manager to the AI Center. Every time I try, I get:

“Export failed due to internal errors (type: HTTPError). Please try again later.”

I’ve tried several times, over the course of a couple of days, and always get the same thing. I tried different browsers, restarting the browser, logging out and back in again, and each time, I get the same thing.

Any idea what could be causing this?

I am able to export if I choose the “Download” option instead of “Export to AI Center”, but I’m not sure what to do with the downloaded data. Is there a way to import it back into the AI Center?

@Capicean

Welcome to the community

May I know what is the end goal?

Can you please verify the steps and few conventions here and see if you are following

Cheers

Hi @Capicean

Welcome to UiPath community.

You can import the downloaded labeled dataset to AI center, by going to dataset in the AI center and selecting upload folder option.

cheers

For context, I’m working on the “Train ML Models for Document Understanding” course. So the end goal is to retrain the out-of-the-box purchase order model with the newly labeled purchase order data, and then deploy that as an ML skill, which I will use in a project.

I went through the steps in the UiPath documentation, and got the same result.

I tried downloading the labeled dataset and then uploading the entire folder into the AI center. It looks like it worked, and I can see the labeled data in “Datasets” (including schema.json, split.csv, images and latest.)

Is there a way to check it to make sure it is really labeled?

I tried creating a Data Labeling Session using that Dataset. But when I open it in Document Manager, I don’t see any documents.

I also tried uploading just the “images” folder as a dataset. If I create a Data Labeling Session using the “images” data, I can see the documents, but they are not labeled.

Any idea what I might be doing wrong?

Hello @Capicean

The dataset you have imported will have latest folder which will have JSON files associated with each file that you have used for data labeling in document manager. The JSON files represent the labeled data.

you can use this labeled dataset to create the pipeline for training the model.

Cheers

1 Like

This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.