change var_name to property_name

deepmd/pt/infer/deep_eval.py

@@ -204,7 +204,7 @@ def model_type(self) -> type["DeepEvalWrapper"]:
             return DeepGlobalPolar
         elif "wfc" in model_output_type:
             return DeepWFC
-        elif "property" in model_output_type:
+        elif self.dp.model["Default"].get_property_name() in model_output_type:
             return DeepProperty
         else:
             raise RuntimeError("Unknown model type")

deepmd/pt/model/model/property_model.py

@@ -76,7 +76,7 @@ def get_task_dim(self) -> int:
     @torch.jit.export
     def get_intensive(self) -> bool:
         """Get whether the property is intensive."""
-        return self.model_output_def()["property"].intensive
+        return self.model_output_def()[self.get_property_name()].intensive
 
     @torch.jit.export
     def get_property_name(self) -> str:

deepmd/pt/model/task/fitting.py

@@ -368,9 +368,7 @@ def set_case_embd(self, case_idx: int):
         ]
 
     def get_compute_stats_do_not_distinguish_types(self) -> bool:
-        """
-        Get whether the fitting net computes stats which are not distinguished between different types of atoms.
-        """
+        """Get whether the fitting net computes stats which are not distinguished between different types of atoms."""
         return False
 
     def __setitem__(self, key, value) -> None:

deepmd/pt/model/task/property.py

@@ -127,9 +127,7 @@ def output_def(self) -> FittingOutputDef:
         )
 
     def get_compute_stats_do_not_distinguish_types(self) -> bool:
-        """
-        Get whether the fitting net computes stats which are not distinguished between different types of atoms.
-        """
+        """Get whether the fitting net computes stats which are not distinguished between different types of atoms."""
         return True
 
     @classmethod

deepmd/utils/out_stat.py

@@ -175,13 +175,17 @@ def compute_stats_do_not_distinguish_types(
     assert natoms.ndim == 2
     assert output_redu.shape[0] == natoms.shape[0]  # [nf,1]
 
-    computed_output_bias = np.repeat(
-        np.mean(output_redu, axis=0)[np.newaxis, :], ntypes, axis=0
-    )
-    output_std = np.std(output_redu, axis=0)
-
-    computed_output_bias = computed_output_bias.reshape([natoms.shape[1]] + var_shape)  # noqa: RUF005
-    output_std = output_std.reshape(var_shape)
-    output_std = np.tile(output_std, (computed_output_bias.shape[0], 1))
+    if intensive:
+        computed_output_bias = np.repeat(
+            np.mean(output_redu, axis=0)[np.newaxis, :], ntypes, axis=0
+        )
+        output_std = np.std(output_redu, axis=0)
+
+        computed_output_bias = computed_output_bias.reshape([natoms.shape[1]] + var_shape)  # noqa: RUF005
+        output_std = output_std.reshape(var_shape)
+        output_std = np.tile(output_std, (computed_output_bias.shape[0], 1))
+    else:
+        # TODO: implement extensive
+        pass
 
     return computed_output_bias, output_std