Training Pipeline is getting failed for Custom Named Entity model

Help AI Center
,
1

Hello all, I am currently trying to classify emails for one of the clients and I am using custom NER to retrieve few data from the mail. I used Label Studio to annotate the emails and extracted the output in CONLL format. I was able to train the model but since Label Studio is a third-party application, I tried to label the emails in UiPath’s Action Center. I was able to label the emails and extract the output in JSON format. Now, the pipeline is getting failed each time with the error “Pipeline failed due to ML Package Issue”. Kindly help to resolve this issue. Thank you in advance.

2

Hi Suman,
Please check the Pipeline logs to check who is causing the error. In my opinion this problem is asociated to Data Labeling. Please confirm if all the input documents(trainning and validation set) are labeled correctly. Remember that all the trainning and validation documents are together part of trainning datasets used by trainning pipelines.

Cheers!

2 Likes
3

Hi Jose, thank you for the response. I re-labelled the emails again and now my pipeline got successful. Thanks again.

1 Like
4

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