Little cleanup in _tensornet.py

Signed-off-by: Tsz Wai Ko <47970742+kenko911@users.noreply.github.com>

src/matgl/models/_tensornet.py

@@ -234,12 +234,10 @@ def forward(self, g: dgl.DGLGraph, state_attr: torch.Tensor | None = None, **kwa
         g.edata["bond_vec"] = bond_vec.to(g.device)
         g.edata["bond_dist"] = bond_dist.to(g.device)
 
-        # This asserts convinces TorchScript that edge_vec is a Tensor and not an Optional[Tensor]
-
         # Expand distances with radial basis functions
         edge_attr = self.bond_expansion(g.edata["bond_dist"])
         g.edata["edge_attr"] = edge_attr
-        # Embedding from edge-wise tensors to node-wise tensors
+        # Embedding layer
         X, edge_feat, state_feat = self.tensor_embedding(g, state_attr)
         # Interaction layers
         for layer in self.layers: