Reformatted code using dev/linter.sh

detectron2/modeling/backbone/mvit.py

@@ -2,7 +2,6 @@
 import numpy as np
 import torch
 import torch.nn as nn
-
 from fairscale.nn.checkpoint import checkpoint_wrapper
 from timm.models.layers import DropPath, Mlp, trunc_normal_
 

detectron2/modeling/backbone/swin.py

@@ -17,11 +17,10 @@
 import torch.nn as nn
 import torch.nn.functional as F
 import torch.utils.checkpoint as checkpoint
+from timm.models.layers import DropPath, to_2tuple, trunc_normal_
 
 from detectron2.modeling.backbone.backbone import Backbone
 
-from timm.models.layers import DropPath, to_2tuple, trunc_normal_
-
 
 class Mlp(nn.Module):
     """Multilayer perceptron."""

detectron2/modeling/backbone/vit.py

@@ -3,13 +3,12 @@
 import fvcore.nn.weight_init as weight_init
 import torch
 import torch.nn as nn
+from fairscale.nn.checkpoint import checkpoint_wrapper
+from timm.models.layers import DropPath, Mlp, trunc_normal_
 
 from detectron2.layers import CNNBlockBase, Conv2d, get_norm
 from detectron2.modeling.backbone.fpn import _assert_strides_are_log2_contiguous
 
-from fairscale.nn.checkpoint import checkpoint_wrapper
-from timm.models.layers import DropPath, Mlp, trunc_normal_
-
 from .backbone import Backbone
 from .utils import (
     PatchEmbed,

projects/DensePose/densepose/modeling/cse/utils.py

@@ -32,9 +32,7 @@ def normalize_embeddings(embeddings: torch.Tensor, epsilon: float = 1e-6) -> tor
     Return:
         Normalized embeddings (tensor [N, D]), such that L2 vector norms are all equal to 1.
     """
-    return embeddings / torch.clamp(
-        embeddings.norm(p=None, dim=1, keepdim=True), min=epsilon
-    )
+    return embeddings / torch.clamp(embeddings.norm(p=None, dim=1, keepdim=True), min=epsilon)
 
 
 def get_closest_vertices_mask_from_ES(

projects/DensePose/densepose/modeling/losses/cycle_pix2shape.py

@@ -135,9 +135,7 @@ def forward(
                     pixel_embeddings.reshape((self.embed_size, -1))[:, pixel_indices_flattened].T
                 )
                 # pixel-vertex similarity [M, K]
-                sim_matrix = pixel_embeddings_sampled.mm(
-                    mesh_vertex_embeddings.T
-                )
+                sim_matrix = pixel_embeddings_sampled.mm(mesh_vertex_embeddings.T)
                 c_pix_vertex = F.softmax(sim_matrix / self.temperature_pix_to_vertex, dim=1)
                 c_vertex_pix = F.softmax(sim_matrix.T / self.temperature_vertex_to_pix, dim=1)
                 c_cycle = c_pix_vertex.mm(c_vertex_pix)