Light Text Classification Training Job Failed Error ' After Pruning No Terms Remain Try A Lower Min DF Or A Higher Max DF'

News Knowledge Base
, , ,
1

Training pipeline failing with ERROR: Training Job failed, error "After pruning, no terms remain. Try a lower min_df or a higher max_df. "

Issue: When running a pipeline to train the LTC (AI Center - Light Text Classification) package, the pipeline fails with the following error -

"Training Job failed, error: After pruning, no terms remain. Try a lower min_df or a higher max_df."


Root Cause: Correct amount of samples for the label is not provided in the dataset.

Resolution: Each label in the dataset should have 15-20 samples to run the pipeline successfully.