ImageClassification ML Training after dataset update Best Parctice

Help AI Center
, , , ,
1

Hi,

I had to update the datasets used in ImageClassification custom ML.

What is the Best Practice to retrain the Custom ML on the new Dataset and ignore the initial training.
Should I delete the Training and Evaluation Pipelines, and create new ones? Should I create a new ML Skill?

Appreciate your inputs/hints.
Thanks in Advance.

2

@hgaber105

Create a separate dataset with the required data or clean the current dataset and load new data

And train the minor 0 version again…as it is minor 0 old training wpuld not be there and only new data would be trained

Cheers

1 Like