GitLab

* using least-squares to compute atomic self energies from the dataset

* self atomic energy calculation in the example training file

* Change unidiomatic type calls to isinstance and remove deprecated h5py operator

* Add whitespace after , to make flake8 happy

Being 1 does not work well

@farhadrgh Please review and merge

Hi @farhadrgh, I am adding you as a code owner of `/torchani/data` since you have experience on this. Later I will request your review if there are PR changing that part.

Add AdamW optimizer, Add docs of nnp training to train a real ANI-1x model, remove h5py and ignite from dependencies, rename MAEMetric -> MaxAEMetric, use torch.utils.tensorboard to replace tensorboardX (#224)

* implement stress

* fix

* more

* Add analytical stress support