Hello. I have a script that uses rasa trainer from “rasa.nlu.model.Trainer” to train rasa model. This script is called using a REST-API. When the user calls this api, the model starts training, but I want to give the user a estimation about the training time or show the progress bar percentage for him/her. How can I get this training duration or any callback about the training progress? any idea?
I want to do something similar. I want to write a callback that reports validation loss at the end of epoch for Hyper parameter tuning. I see that there is a callback function in rasa.utils.tensorflow.callback. Is it possible to provide custom call back functions to trainer.train ?