Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

fixed aggregation_mode bug for avg pooling #361

Merged
merged 4 commits into from
Nov 10, 2021
Merged

fixed aggregation_mode bug for avg pooling #361

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
34a9cf2
fixed aggregation_mode bug for avg pooling
jnsLs Nov 1, 2021
03aeb89
Merge branch 'atomistic-machine-learning:dev' into dev
jnsLs Nov 9, 2021
a15ff8c
fixed units A-->Ang, deleted deprecated preprocessing in AtomsConverter
jnsLs Nov 10, 2021
dfcb332
fixed units A-->Ang, deleted deprecated preprocessing in AtomsConvert…
jnsLs Nov 10, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 3 additions & 3 deletions src/schnetpack/atomistic/atomwise.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -69,14 +69,14 @@ def forward(self, inputs: Dict[str, torch.Tensor]) -> Dict[str, torch.Tensor]:

# aggregate
if self.aggregation_mode is not None:
if self.aggregation_mode == "avg":
y = y / inputs[properties.n_atoms][:, None]

idx_m = inputs[properties.idx_m]
maxm = int(idx_m[-1]) + 1
y = snn.scatter_add(y, idx_m, dim_size=maxm)
y = torch.squeeze(y, -1)

if self.aggregation_mode == "avg":
y = y / inputs[properties.n_atoms]

inputs[self.output_key] = y
return inputs

Expand Down
5 changes: 1 addition & 4 deletions src/schnetpack/interfaces/ase_interface.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -76,9 +76,6 @@ def __init__(
else:
raise AtomsConverterError(f"Unrecognized precision {dtype}")

for t in self.transforms:
t.preprocessor()

def __call__(self, atoms: Atoms):
"""

Expand Down Expand Up @@ -162,7 +159,7 @@ def __init__(
self.property_units = {
self.energy: convert_units(energy_units, "eV"),
self.forces: convert_units(forces_units, "eV/Angstrom"),
self.stress: convert_units(stress_units, "eV/A/A/A"),
self.stress: convert_units(stress_units, "eV/Ang/Ang/Ang"),
}

def calculate(
Expand Down