GitLab

* cleanup after test_vibrational to avoid .pckl files everywhere

* only dump test_neurochem's runs into torchani/tests

* avoid runs inside tests/

* Add coverage to gitignore

* consts is unnecessary here

* Remove workaround for TorchScript bug on torch.unique_consecutive

On par with https://github.com/pytorch/pytorch/pull/39339



This breaks the compatibility with PyTorch 1.5.0, but hopefully will be compatible with 1.5.1.

* trigger ci

* trigger

* trigger
Co-authored-by: Farhad Ramezanghorbani <farhadrgh@users.noreply.github.com>

* Update model URL

* Update models.py

* Update models.py

* fix

* cleanup

* save

* download resources from ASE_ANI

* update gitignore

* get resources with request

* fix flke8

* rm svn checkout for resources

* update

* fix flake8

* document species_indices

* Improve documentation

* More documentation for data

* More clearly specify behaviour of SpeciesConverter

* improve docs for SpeciesConverter

* flake8

* Add order recommendation to ChemicalSymbolsToInts

* Fix examples where species_to_tensor is used wrong

* Fix argument

* Add possibility of sorting in different order to subtract_self_energies

* make species_to_indices behave the same as subtract_self_energies

* Revert "make species_to_indices behave the same as subtract_self_energies"

This reverts commit a415df29bd6e270f3225f0b6721b3c873aeed40e.

* Fix examples

* make species_to_indices behave the same whithout breaking API

* Improve documentation to reflect correct usage

* Fix typos in docstring

* some more docstring issues

* more docstring

* Fix error in species_to_indices

* Adding ANI-2x parameters

* Moving 2x to seperate branch

* Adding ANI-2x to new branch

* Adding 2x network

* Update models.py

* Update api.rst

* Update __init__.py
Co-authored-by: Christian Devereux <cdever01@ufl.edu>
Co-authored-by: Farhad Ramezanghorbani <farhadrgh@users.noreply.github.com>

* Empty list is a useless function

* trigger tests to make sure test_data works

* change call to 'list' because of tests taking suspiciously long
Co-authored-by: Farhad Ramezanghorbani <farhadrgh@users.noreply.github.com>

* Allow easily loading individual models

* Hopefully make the tests a tiny bit faster by not loading the whole network each time

* Refactor BuiltinNet into BuiltinModel and BuiltinEnsemble

* Change test name

* flake8 whitespace

* Get rid of redundant comparison

* remove unused import

* Fix info file bug

* flake8

* don't use nonlocal on get_resource

* Revert "don't use nonlocal on get_resource"

This reverts commit efe221bb4f7ce4a1ccb8c3c73b886bdbad019ab0.

* Fix get_resource bug

* Add missing init call

* Fix species_to_tensor bug

* Add species attribute

* Add some extra attributes

* Fix bug in getitem

* Always assign species_converter

* Species to tensor is also an entry point

* Lets not make it an entry point for now

* delete duplicated definition

* Add a function to recast all buffer tensors

* flake8

* switch AdamW from torchany to pytorch

* fix __init__
Co-authored-by: Farhad Ramezanghorbani <farhadrgh@users.noreply.github.com>

* cleanup

* move forces to additional_properties

* fix typo

* Reduce user inputs for training custom models

* Update nnp_training.py

ANI1x NNs are updated to be consistent with what we used in the paper

PyTorch has supported type promotion since the last PTDC. We shouldn't manually convert the dtype now.
Co-authored-by: Farhad Ramezanghorbani <farhadrgh@users.noreply.github.com>

* Don't compute shifts when no PBC

* Update aev.py

* Don't pass cell and shifts when not needed (#451)

* Don't pass cell and shifts when not needed

* Update aev.py

* Update aev.py

* Update aev.py
Co-authored-by: Farhad Ramezanghorbani <farhadrgh@users.noreply.github.com>

* Reduce number of kernel calls in AEV

* try

* Revert "try"

This reverts commit 04e56fde671c5168cb7825ab3c33f64a24196d98.

* more

* more

* merge more for angular terms

* neighbor_pairs

* more

* more

* save

* save

* save

* save
Co-authored-by: Farhad Ramezanghorbani <farhadrgh@users.noreply.github.com>

Co-authored-by: Farhad Ramezanghorbani <farhadrgh@users.noreply.github.com>

Co-authored-by: Farhad Ramezanghorbani <farhadrgh@users.noreply.github.com>

* ../aten/src/ATen/native/BinaryOps.cpp:81: UserWarning: Integer division of tensors using div or / is deprecated, and in a future release div will perform true division as in Python 3. Use true_divide or floor_divide (// in Python) instead.

* Update aev.py
Co-authored-by: Farhad Ramezanghorbani <farhadrgh@users.noreply.github.com>

* Clean up unused args

* fix
Co-authored-by: Farhad Ramezanghorbani <farhadrgh@users.noreply.github.com>

* Improve new dataset API

* Improve more on new dataset API

* fix

* fix reentrance

* Allow all intermediate state of transformation to be reentered

* Add length inference

* fix

* split by ratio

* add dataloader example

* add test for data loader

* Remove everything about chunking

* aev.py

* neurochem trainer

* training-benchmark-nsys-profile.py

* fix eval

* training-benchmark.py

* nnp_training.py

* flake8

* nnp_training_force.py

* fix dtype of species

* fix

* flake8

* requires_grad_

* git ignore

* fix

* original

* fix

* fix

* fix

* fix

* save

* save

* save

* save

* save

* save

* save

* save

* save

* collate

* fix

* save

* fix

* save

* save

* fix

* save

* fix

* fix

* no len

* float

* save

* save

* save

* save

* save

* save

* save

* save

* save

* fix

* save

* save

* save

* save

* fix

* fix

* fix

* fix mypy

* don't remove outliers

* save

* save

* save

* fix

* flake8

* save

* fix

* flake8

* docs

* more docs

* fix test_data

* remove test_data_new

* fix

* massive training time improvement

* flake8

* improve validation also

* improve training-benchmark.py

* fix devices

* send num_atoms to GPU

* rerun

* Add an argument for inserting synchronization

* Update training-benchmark.py

* Update training-benchmark.py
Co-authored-by: Farhad Ramezanghorbani <farhadrgh@users.noreply.github.com>

* Add tuple of atomic masses

* Add get_atomic_masses function, to get masses from periodic table indices (znumbers)

* Fix bug that prevented using ASE together with periodic_table_index

* ASE has to know if the model has PTI or not

* Allow for models that don't have a periodic_table_index attribute

* now vibrational analysis can be computed without explicit masses if periodic_table_index=True

* Add atomic masses calculator to docs

* flake8

* flake8

* flake8
Co-authored-by: Gao, Xiang <qasdfgtyuiop@gmail.com>