Cleanup some junks inside ANIModel introduced due to JIT (#362)

torchani/nn.py

@@ -16,19 +16,11 @@ class ANIModel(torch.nn.Module):
             :attr:`modules`, which means, for example ``modules[i]`` must be
             the module for atom type ``i``. Different atom types can share a
             module by putting the same reference in :attr:`modules`.
-        padding_fill (float): The value to fill output of padding atoms.
-            Padding values will participate in reducing, so this value should
-            be appropriately chosen so that it has no effect on the result. For
-            example, if the reducer is :func:`torch.sum`, then
-            :attr:`padding_fill` should be 0, and if the reducer is
-            :func:`torch.min`, then :attr:`padding_fill` should be
-            :obj:`math.inf`.
     """
 
-    def __init__(self, modules, padding_fill=0):
+    def __init__(self, modules):
         super(ANIModel, self).__init__()
         self.module_list = torch.nn.ModuleList(modules)
-        self.padding_fill = padding_fill
 
     def __getitem__(self, i):
         return self.module_list[i]
@@ -39,8 +31,7 @@ def forward(self, species_aev):
         species_ = species.flatten()
         aev = aev.flatten(0, 1)
 
-        output = torch.full(species_.shape, self.padding_fill,
-                            dtype=aev.dtype, device=species.device)
+        output = aev.new_zeros(species_.shape)
 
         for i, m in enumerate(self.module_list):
             mask = (species_ == i)