Import of Machine Learning Extractor folder

Help Activities
, ,
1

Hi Team,

I want to clarify one thing under document understanding process.

  1. I have labelled the data for training purpose, and then run train pipeline for the train dataset.
  2. Same process has been repeated for the test data.
    3 After that I have validated the data and placed the output of Machine Learning Extractor Trainer(which has 3 folders namely documents, metadata and prediction) in ABC folder.

Now I am confused that in Data Manager where to import that zip folder? under trained data Labeling section which is already created. OR I have to create a new instance i.e. ML TRAIN EXTRACTOR import in data labeling section , and after that export it and then import it in the datasets and run training pipeline on it.

@AdiPopa @Lahiru.Fernando
Any help in this regard will be highly appreciated
Thank You

Regards

Enthusiastic

2

Hello @enthusiastic

I’m sorry for the late reply… I read your post yesterday, but got dragged into something else… :slight_smile:

So for your question, you can directly upload the validation data into a DataSet in AI Center. the Data Manager will automatically pick it up from the DataSet and you are not needed to manually upload and create different instances.

This is better explained here in my video… Have a look :slight_smile: