[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

examples/tutorial/sequence_parallel/model/layers/mlp.py

@@ -1,10 +1,10 @@
-import torch
 import torch.nn as nn
 import torch.nn.functional as F
 
-from .linear import Linear
 from colossalai.kernel.jit import bias_gelu_impl
 
+from .linear import Linear
+
 
 class TransformerMLP(nn.Module):
     """MLP.
@@ -18,19 +18,13 @@ class TransformerMLP(nn.Module):
         super(TransformerMLP, self).__init__()
 
         # Project to 4h.
-        self.dense_h_to_4h = Linear(
-            hidden_size,
-            int(hidden_size*mlp_ratio),
-            skip_bias_add=True)
+        self.dense_h_to_4h = Linear(hidden_size, int(hidden_size * mlp_ratio), skip_bias_add=True)
 
         self.bias_gelu_fusion = fuse_gelu
         self.activation_func = F.gelu
 
         # Project back to h.
-        self.dense_4h_to_h = Linear(
-            int(hidden_size*mlp_ratio),
-            hidden_size,
-            skip_bias_add=True)
+        self.dense_4h_to_h = Linear(int(hidden_size * mlp_ratio), hidden_size, skip_bias_add=True)
 
     def forward(self, hidden_states):
         # hidden states should be in the shape of [s, b, h]
@@ -39,11 +33,9 @@ class TransformerMLP(nn.Module):
         intermediate_parallel, bias_parallel = self.dense_h_to_4h(hidden_states)
 
         if self.bias_gelu_fusion:
-            intermediate_parallel = \
-                bias_gelu_impl(intermediate_parallel, bias_parallel)
+            intermediate_parallel = bias_gelu_impl(intermediate_parallel, bias_parallel)
         else:
-            intermediate_parallel = \
-                self.activation_func(intermediate_parallel + bias_parallel)
+            intermediate_parallel = self.activation_func(intermediate_parallel + bias_parallel)
 
         # [s, b, h]
         output, output_bias = self.dense_4h_to_h(intermediate_parallel)