Hi Guys,
I have a scenario where I’m using “OS Packages > Language Analysis > EnglishTextClassification” ML Package in my process. It is a retrainable ML package.
I need to retrain the ML package from the Action Center after completing the Action by the human.
I didn’t find the proper document to understand how to implement this case.
Kindly please help with a appropriate scenario to achieve this goal.
1 Like
Hey @chethankumar.rpa
Yes this is possible.
You need to capture the action center task user response in the form of training data excel.
The same Excel should be uploaded to the AI center for retraining the model.
Also, automated pipeline should be setup in AI Center for the same to retrain the model with the file once uploaded.
Reference
Hope this helps.
Thanks
#nK
1 Like
Dear Nithin,
Thank for your response. I will check the document.
1 Like
Hi Nithin,
Is it possible to train the ML Package in incremental process ?
Example :
In first phase I have trained the ML Package with 100 records.
As the project run in Production, every day I will get
5 new records to train in addition to the previous train record.
Is it possible to train 5 records and merger to the already trained 100 records or we need to train all the 105 records ?
Kindly help with your valuable inputs.
1 Like
Hey @chethankumar.rpa
Of course yes, but since you have already trained with the initial records then you no need to train with same records as it will be just redundant.
Thanks
#nK