Stack multiple Model checkpoints to log steps, train loss, and log every step for train_loss

[msakarvadia]

https://devblog.pytorchlightning.ai/introducing-multiple-modelcheckpoint-callbacks-e4bc13f9c185#:~:text=You%20can%20configure%20as%20many,to%20the%20Trainer%20callbacks%20list.

Right now we only use one model checkpoint at a time, however configure script to use multiple at the same time.