Add element names to ANIModel (#398)

* Add element names to ANIModel

* nc trainer

tests/test_neurochem.py

@@ -18,7 +18,10 @@ def testNeuroChemTrainer(self):
         # test if loader construct correct model
         self.assertEqual(trainer.aev_computer.aev_length, 384)
         m = trainer.nn
-        H, C, N, O = m  # noqa: E741
+        H = m['H']
+        C = m['C']
+        N = m['N']
+        O = m['O']  # noqa: E741
         self.assertIsInstance(H[0], torch.nn.Linear)
         self.assertListEqual(list(H[0].weight.shape), [160, 384])
         self.assertIsInstance(H[1], torch.nn.CELU)

torchani/neurochem/__init__.py

@@ -15,6 +15,7 @@
 from ..utils import EnergyShifter, ChemicalSymbolsToInts
 from ..aev import AEVComputer
 from ..optim import AdamW
+from collections import OrderedDict
 
 
 class Constants(collections.abc.Mapping):
@@ -240,10 +241,10 @@ def load_model(species, dir_):
             chemical symbols of each supported atom type in correct order.
         dir_ (str): String for directory storing network configurations.
     """
-    models = []
+    models = OrderedDict()
     for i in species:
         filename = os.path.join(dir_, 'ANN-{}.nnf'.format(i))
-        models.append(load_atomic_network(filename))
+        models[i] = load_atomic_network(filename)
     return ANIModel(models)
 
 
@@ -496,8 +497,8 @@ def init_params(m):
             input_size, network_setup = network_setup
             if input_size != self.aev_computer.aev_length:
                 raise ValueError('AEV size and input size does not match')
-            atomic_nets = {}
-            for atom_type in network_setup:
+            atomic_nets = OrderedDict()
+            for atom_type in self.consts.species:
                 layers = network_setup[atom_type]
                 modules = []
                 i = input_size
@@ -537,7 +538,7 @@ def init_params(m):
                             'unrecognized parameter in layer setup')
                     i = o
                 atomic_nets[atom_type] = torch.nn.Sequential(*modules)
-            self.nn = ANIModel([atomic_nets[s] for s in self.consts.species])
+            self.nn = ANIModel(atomic_nets)
 
             # initialize weights and biases
             self.nn.apply(init_params)

torchani/nn.py

@@ -1,4 +1,5 @@
 import torch
+from collections import OrderedDict
 from torch import Tensor
 from typing import Tuple, NamedTuple, Optional
 
@@ -13,7 +14,7 @@ class SpeciesCoordinates(NamedTuple):
     coordinates: Tensor
 
 
-class ANIModel(torch.nn.ModuleList):
+class ANIModel(torch.nn.ModuleDict):
     """ANI model that compute energies from species and AEVs.
 
     Different atom types might have different modules, when computing
@@ -31,6 +32,18 @@ class ANIModel(torch.nn.ModuleList):
             module by putting the same reference in :attr:`modules`.
     """
 
+    @staticmethod
+    def ensureOrderedDict(modules):
+        if isinstance(modules, OrderedDict):
+            return modules
+        od = OrderedDict()
+        for i, m in enumerate(modules):
+            od[str(i)] = m
+        return od
+
+    def __init__(self, modules):
+        super(ANIModel, self).__init__(self.ensureOrderedDict(modules))
+
     def forward(self, species_aev: Tuple[Tensor, Tensor],
                 cell: Optional[Tensor] = None,
                 pbc: Optional[Tensor] = None) -> SpeciesEnergies:
@@ -42,7 +55,7 @@ def forward(self, species_aev: Tuple[Tensor, Tensor],
 
         output = aev.new_zeros(species_.shape)
 
-        for i, m in enumerate(self):
+        for i, (_, m) in enumerate(self.items()):
             mask = (species_ == i)
             midx = mask.nonzero().flatten()
             if midx.shape[0] > 0: