Continuing the discussion from #672 which was closed for some reason although it is still a standing issue. Are there any plans to add a way to update an existing model while adding new training data instead of retraining the entire model from scratch?
So
linker = dedupe.RecordLink(fields)
linker.prepare_training(messy_data, canonical_data, training_file=tf)
Would become
linker = dedupe.RecordLink(fields)
linker.prepare_training(messy_data, canonical_data, settings_file=sf, training_file=tf)
Continuing the discussion from #672 which was closed for some reason although it is still a standing issue. Are there any plans to add a way to update an existing model while adding new training data instead of retraining the entire model from scratch?
So
Would become