Any plans of implementing more models & releasing pretrained weights?

[jasperhyp]

Hi! I wonder if you have any plans on implementing more state-of-the-art molecular representation learning models such as those in DeepChem. Others that haven't been incorporated into other libraries such as Transformer-M might also be a good addition. Thank you!

[jasperhyp]

Perhaps would also be interesting to include readily available pretrained model as #167 #44 #130 #63 already suggested. Given you folks' succuess in developing pretraining methods, e.g. GraphMVP, I guess that wouldn't be a problem!