[codemod][usort] apply import merging for fbcode (3 of 11)

Summary:
Applies new import merging and sorting from µsort v1.0.

When merging imports, µsort will make a best-effort to move associated
comments to match merged elements, but there are known limitations due to
the diynamic nature of Python and developer tooling. These changes should
not produce any dangerous runtime changes, but may require touch-ups to
satisfy linters and other tooling.

Note that µsort uses case-insensitive, lexicographical sorting, which
results in a different ordering compared to isort. This provides a more
consistent sorting order, matching the case-insensitive order used when
sorting import statements by module name, and ensures that "frog", "FROG",
and "Frog" always sort next to each other.

For details on µsort's sorting and merging semantics, see the user guide:
https://usort.readthedocs.io/en/stable/guide.html#sorting

Reviewed By: lisroach

Differential Revision: D36402148

fbshipit-source-id: 8eab032f38c0af6c83b38e18cfb69506e32fa53e

Author: amyreese

fx/converter_registry.py

@@ -1,4 +1,4 @@
-from typing import Callable, Dict, Any
+from typing import Any, Callable, Dict
 
 from torch.fx.node import Target
 

fx/converters/acc_ops_converters.py

@@ -2,7 +2,7 @@
 import math
 import operator
 import warnings
-from typing import Dict, Tuple, Sequence, cast, Optional, Union
+from typing import cast, Dict, Optional, Sequence, Tuple, Union
 
 import fx2trt_oss.tracer.acc_tracer.acc_ops as acc_ops
 import numpy as np
@@ -13,12 +13,12 @@
 from fx2trt_oss.fx.converter_registry import tensorrt_converter
 from fx2trt_oss.fx.types import *  # noqa: F403
 from fx2trt_oss.fx.utils import (
+    get_dynamic_dims,
     torch_dtype_from_trt,
     torch_dtype_to_trt,
-    get_dynamic_dims,
 )
 from torch.fx.immutable_collections import immutable_list
-from torch.fx.node import Target, Argument
+from torch.fx.node import Argument, Target
 
 from .converter_utils import *  # noqa: F403
 

fx/converters/adaptive_avgpool.py

@@ -3,10 +3,7 @@
 import torch
 from fx2trt_oss.fx.converter_registry import tensorrt_converter
 
-from .converter_utils import (
-    mark_as_int8_layer,
-    extend_mod_attr_to_tuple,
-)
+from .converter_utils import extend_mod_attr_to_tuple, mark_as_int8_layer
 
 
 @tensorrt_converter(torch.nn.modules.pooling.AdaptiveAvgPool2d)

fx/converters/batchnorm.py

@@ -5,11 +5,7 @@
 import torch
 from fx2trt_oss.fx.converter_registry import tensorrt_converter
 
-from .converter_utils import (
-    mark_as_int8_layer,
-    to_numpy,
-    get_dyn_range,
-)
+from .converter_utils import get_dyn_range, mark_as_int8_layer, to_numpy
 
 
 def common_batchnorm(network, mod, input_val, layer_name, is_quantized):

fx/converters/converter_utils.py

@@ -1,24 +1,24 @@
 import operator
 import warnings
-from typing import Any, Tuple, Sequence, Union, List, Optional, Dict, Callable
+from typing import Any, Callable, Dict, List, Optional, Sequence, Tuple, Union
 
 import numpy as np
 
 # @manual=//deeplearning/trt/python:py_tensorrt
 import tensorrt as trt
 import torch
 from fx2trt_oss.fx.types import (
-    TRTNetwork,
-    TRTTensor,
-    TRTLayer,
-    TRTPluginFieldCollection,
-    TRTPlugin,
+    Shape,
     TRTDataType,
     TRTElementWiseOp,
-    Shape,
+    TRTLayer,
+    TRTNetwork,
+    TRTPlugin,
+    TRTPluginFieldCollection,
+    TRTTensor,
 )
 from fx2trt_oss.fx.utils import torch_dtype_from_trt
-from torch.fx.node import Target, Argument
+from torch.fx.node import Argument, Target
 
 
 def get_trt_plugin(

fx/converters/convolution.py

@@ -6,9 +6,9 @@
 
 from .converter_utils import (
     extend_mod_attr_to_tuple,
+    get_dyn_range,
     mark_as_int8_layer,
     to_numpy,
-    get_dyn_range,
 )
 
 

fx/converters/linear.py

@@ -3,11 +3,7 @@
 import torch
 from fx2trt_oss.fx.converter_registry import tensorrt_converter
 
-from .converter_utils import (
-    mark_as_int8_layer,
-    to_numpy,
-    get_dyn_range,
-)
+from .converter_utils import get_dyn_range, mark_as_int8_layer, to_numpy
 
 
 def common_linear(network, mod, input_val, layer_name, is_quantized):

fx/converters/maxpool.py

@@ -3,10 +3,7 @@
 import torch
 from fx2trt_oss.fx.converter_registry import tensorrt_converter
 
-from .converter_utils import (
-    mark_as_int8_layer,
-    extend_mod_attr_to_tuple,
-)
+from .converter_utils import extend_mod_attr_to_tuple, mark_as_int8_layer
 
 
 def common_maxpool(network, mod, dimension, input_val, layer_name):

fx/converters/quantization.py

@@ -3,10 +3,7 @@
 import torch
 from fx2trt_oss.fx.converter_registry import tensorrt_converter
 
-from .converter_utils import (
-    get_dyn_range,
-    get_inputs_from_args_and_kwargs,
-)
+from .converter_utils import get_dyn_range, get_inputs_from_args_and_kwargs
 
 
 quantize_per_tensor_inputs = ["input", "scale", "zero_point", "dtype"]

fx/example/fx2trt_example.py

@@ -4,7 +4,7 @@
 import torch
 import torch.fx
 import torch.nn as nn
-from fx2trt_oss.fx import TRTInterpreter, InputTensorSpec, TRTModule
+from fx2trt_oss.fx import InputTensorSpec, TRTInterpreter, TRTModule
 from fx2trt_oss.fx.tools.trt_splitter import TRTSplitter