[misc] update pre-commit and run all files (#4752)

* [misc] update pre-commit

* [misc] run pre-commit

* [misc] remove useless configuration files

* [misc] ignore cuda for clang-format

colossalai/interface/__init__.py

@@ -1,4 +1,4 @@
 from .model import AMPModelMixin, ModelWrapper
 from .optimizer import OptimizerWrapper
 
-__all__ = ['OptimizerWrapper', 'ModelWrapper', 'AMPModelMixin']
+__all__ = ["OptimizerWrapper", "ModelWrapper", "AMPModelMixin"]

colossalai/interface/model.py

@@ -26,11 +26,9 @@ class ModelWrapper(nn.Module):
 
 
 class AMPModelMixin:
-    """This mixin class defines the interface for AMP training.
-    """
+    """This mixin class defines the interface for AMP training."""
 
     def update_master_params(self):
         """
         Update the master parameters for AMP training.
         """
-        pass

colossalai/interface/optimizer.py

@@ -22,7 +22,7 @@ class OptimizerWrapper:
         params = []
 
         for group in self.param_groups:
-            params += group['params']
+            params += group["params"]
         return params
 
     @property
@@ -82,12 +82,14 @@ class OptimizerWrapper:
         """
         nn.utils.clip_grad_value_(self.parameters, clip_value, *args, **kwargs)
 
-    def clip_grad_by_norm(self,
-                          max_norm: Union[float, int],
-                          norm_type: Union[float, int] = 2.0,
-                          error_if_nonfinite: bool = False,
-                          *args,
-                          **kwargs) -> Tensor:
+    def clip_grad_by_norm(
+        self,
+        max_norm: Union[float, int],
+        norm_type: Union[float, int] = 2.0,
+        error_if_nonfinite: bool = False,
+        *args,
+        **kwargs,
+    ) -> Tensor:
         """
         Clips gradient norm of an iterable of parameters.
 
@@ -113,7 +115,8 @@ class OptimizerWrapper:
             loss (Tensor): The loss to be scaled.
         """
         raise NotImplementedError(
-            "The method scale_loss is only available for optimizers with mixed precision training")
+            "The method scale_loss is only available for optimizers with mixed precision training"
+        )
 
     def unscale_grad(self):
         """
@@ -122,7 +125,8 @@ class OptimizerWrapper:
         Note: Only available for optimizers with mixed precision training.
         """
         raise NotImplementedError(
-            "The method unscale_grad is only available for optimizers with mixed precision training")
+            "The method unscale_grad is only available for optimizers with mixed precision training"
+        )
 
     def unwrap(self):
         """