Any Machine learning Extractor which only required one-time training

Help Studio
, ,
1

Hi, I am using ML Extractor, but I have to train the robot every single time and train the robot for each of my 25 pdf files to extract the selected data that I require from them. That is taking a long time. But is there any other way that I can just train it one time, and at the second execution, it will extract the data to the Excel file?

1 Like
2

@Haseeb_Farrukh

Sequence:

-Load Pre-trained Model
-For Each PDF File in Directory
-Extract Data Using ML Extractor
 -Append Extracted Data to a DataTable
 -Write DataTable to Excel File
1 Like
3

1
1577Ă—756 22.9 KB

2
2540Ă—813 35.5 KB

3
3527Ă—820 29.2 KB

From the load-pre-trained model, it did not understand it.
This is the sequence from the image I’ve done, but the problem is that I am using ML Extractor, which is asking to be trained every single time. Will Trained ML Extractor solve my issue? If so, how can I get any help? Thank You.

1 Like
4
  • Make sure to capture a diverse range of document layouts that you expect to encounter.
    -After the initial training, you may want to fine-tune the model using additional samples if needed.
    -Save the trained and fine-tuned model so that you can load it in subsequent executions without having to train the model again.
    -use the “Load Model” activity to load the pre-trained model. Then, use the ML Extractor activity to extract data from your PDF files.
1 Like