Fix exception chaining

detectron2/config/config.py

@@ -138,8 +138,10 @@ def from_config(cls, cfg):
     def wrapped(self, *args, **kwargs):
         try:
             from_config_func = type(self).from_config
-        except AttributeError:
-            raise AttributeError("Class with @configurable must have a 'from_config' classmethod.")
+        except AttributeError as e:
+            raise AttributeError(
+                "Class with @configurable must have a 'from_config' classmethod."
+            ) from e
         if not inspect.ismethod(from_config_func):
             raise TypeError("Class with @configurable must have a 'from_config' classmethod.")
 

detectron2/data/catalog.py

@@ -49,12 +49,12 @@ def get(self, name):
         """
         try:
             f = self[name]
-        except KeyError:
+        except KeyError as e:
             raise KeyError(
                 "Dataset '{}' is not registered! Available datasets are: {}".format(
                     name, ", ".join(list(self.keys()))
                 )
-            )
+            ) from e
         return f()
 
     def list(self) -> List[str]:

detectron2/data/transforms/augmentation.py

@@ -85,10 +85,10 @@ def _get_aug_input_args(aug, aug_input) -> List[Any]:
     for f in aug.input_args:
         try:
             args.append(getattr(aug_input, f))
-        except AttributeError:
+        except AttributeError as e:
             raise AttributeError(
                 f"Augmentation {aug} needs '{f}', which is not an attribute of {aug_input}!"
-            )
+            ) from e
     return args
 
 

detectron2/engine/hooks.py

@@ -331,11 +331,11 @@ def _do_eval(self):
             for k, v in flattened_results.items():
                 try:
                     v = float(v)
-                except Exception:
+                except Exception as e:
                     raise ValueError(
                         "[EvalHook] eval_function should return a nested dict of float. "
                         "Got '{}: {}' instead.".format(k, v)
-                    )
+                    ) from e
             self.trainer.storage.put_scalars(**flattened_results, smoothing_hint=False)
 
         # Evaluation may take different time among workers.