Predicting value of external data from GCNN model developed

[pramodkp]

Hi ,
I am curious to know that how to predict the external dataset for a trained model.

For example , I did train and evaluate the GCNN model for logP data and model is giving test set R-squared score of 0.90.
Now I want some 10 molecules logP prediction then how to do it ?

thanks
Pramod