BUG: Save model parameters without DataParallel module wrapper

ptychonn/_infer/__main__.py

@@ -202,24 +202,25 @@ def __init__(
         model: typing.Optional[torch.nn.Module] = None,
         model_params_path: typing.Optional[pathlib.Path] = None,
     ):
-        self.device = torch.device(
-            "cuda" if torch.cuda.is_available() else "cpu")
-        print(f"Let's use {torch.cuda.device_count()} GPUs!")
+        self.model = ptychonn.model.ReconSmallPhaseModel(
+        ) if model is None else model
 
-        if model is None or model_params_path is None:
-            self.model = ptychonn.model.ReconSmallPhaseModel()
+        if model_params_path is None:
             with importlib.resources.path(
                     'ptychonn._infer',
                     'weights.pth',
             ) as model_params_path:
-                self.model.load_state_dict(
-                    torch.load(model_params_path, map_location=self.device))
+                self.model.load_state_dict(torch.load(model_params_path))
         else:
-            self.model = model
-            self.model.load_state_dict(
-                torch.load(model_params_path, map_location=self.device))
+            self.model.load_state_dict(torch.load(model_params_path))
+
+        self.device = torch.device(
+            "cuda" if torch.cuda.is_available() else "cpu")
+        print(f"Let's use {torch.cuda.device_count()} GPUs!")
+
+        if torch.cuda.device_count() > 1:
+            self.model = torch.nn.DataParallel(self.model)
 
-        self.model = torch.nn.DataParallel(self.model)  #Default all devices
         self.model = self.model.to(self.device)
 
     def setTestData(self, X_test: np.ndarray, batch_size: int):

ptychonn/_train/__main__.py

@@ -279,13 +279,10 @@ def initModel(self, model_params_path: pathlib.Path | None = None):
 
         self.device = torch.device(
             "cuda" if torch.cuda.is_available() else "cpu")
+        print(f"Let's use {torch.cuda.device_count()} GPUs!")
+
         if torch.cuda.device_count() > 1:
-            logger.info("Let's use %d GPUs!", torch.cuda.device_count())
-            # dim = 0 [30, xxx] -> [10, ...], [10, ...], [10, ...] on 3 GPUs
-            self.model = torch.nn.DataParallel(
-                self.model,
-                device_ids=None,  # Default all devices
-            )
+            self.model = torch.nn.DataParallel(self.model)
 
         self.model = self.model.to(self.device)
 
@@ -414,7 +411,13 @@ def updateSavedModel(
         fname = directory / f'best_model{ suffix }.pth'
         logger.info("Saving best model as %s", fname)
         os.makedirs(directory, exist_ok=True)
-        torch.save(model.state_dict(), fname)
+        if isinstance(model, (
+                torch.nn.DataParallel,
+                torch.nn.parallel.DistributedDataParallel,
+        )):
+            torch.save(model.module.state_dict(), fname)
+        else:
+            torch.save(model.state_dict(), fname)
 
     def getSavedModelPath(self) -> pathlib.Path | None:
         """Return the path where `validate` will save the model weights"""