Sania G.

We provide incremental feature as a part of Zingg Enterprise - you can check it out - https://www.zingg.ai/product/features/incremental-flow

We provide incremental feature as a part of Zingg Enterprise - you can check it out - https://www.zingg.ai/product/features/incremental-flow

Yes this is the approach

No two should be the same

Each cluster has a different z_cluster?

Could you please check whether each cluster in the training data you are adding has a distinctive z_cluster in case of a match or not a match? Also, could you cross-verify the pair counts are accurate for each cluster in your training data with n*(n-1)/2 -> A cluster with 3 records should have 3 pairs and a cluster with 4 records should have 4*3/2 = 6 pairs

Hi What are your requirements?

https://docs.zingg.ai/latest/stepbystep/hardwaresizing

It is likely the error is because of changing the match type - retraining your model will solve this

Ruchi S.