[Fix lint]

Author: Joey Tsai

backends/qualcomm/quantizer/quantizer.py

@@ -26,9 +26,9 @@
     get_16a4w_qnn_ptq_config,
     get_16a8w_qnn_ptq_config,
     get_default_16bit_qnn_ptq_config,
+    get_default_8bit_qat_proto,
     get_default_8bit_qnn_ptq_config,
     get_ptq_per_channel_quant_config,
-    get_default_8bit_qat_proto,
     OP_ANNOTATOR,
     QuantizationConfig,
 )

backends/qualcomm/quantizer/utils.py

@@ -25,9 +25,9 @@
     FixedQParamsObserver,
     MinMaxObserver,
     MovingAverageMinMaxObserver,
+    MovingAveragePerChannelMinMaxObserver,
     PerChannelMinMaxObserver,
     UniformQuantizationObserverBase,
-    MovingAveragePerChannelMinMaxObserver,
 )
 
 from torch.ao.quantization.quantizer import (
@@ -203,7 +203,9 @@ def get_default_8bit_qat_proto(act_symmetric: bool = False) -> QuantizationConfi
         quant_max=torch.iinfo(torch.int8).max,
         qscheme=torch.per_tensor_symmetric,
         ch_axis=0,
-        observer_or_fake_quant_ctr=FusedMovingAvgObsFakeQuantize.with_args(observer=MovingAverageMinMaxObserver),
+        observer_or_fake_quant_ctr=FusedMovingAvgObsFakeQuantize.with_args(
+            observer=MovingAverageMinMaxObserver
+        ),
     )
 
     bias_quantization_spec = QuantizationSpec(

backends/qualcomm/tests/test_qnn_delegate.py

@@ -1054,7 +1054,7 @@ def test_qnn_backend_linear_qat(self):
 
         module = self.get_prepared_qat_module(module, sample_input)
 
-        optimizer = torch.optim.SGD(module.parameters(), lr = 0.1)
+        optimizer = torch.optim.SGD(module.parameters(), lr=0.1)
         criterion = torch.nn.CrossEntropyLoss()
         output = module(*sample_input)
         loss = criterion(output, module(*sample_input))

backends/qualcomm/tests/utils.py

@@ -45,7 +45,11 @@
 from executorch.exir.pass_base import ExportPass
 from executorch.exir.passes.memory_planning_pass import MemoryPlanningPass
 from executorch.exir.program import ExecutorchProgram, ExecutorchProgramManager
-from torch.ao.quantization.quantize_pt2e import convert_pt2e, prepare_pt2e, prepare_qat_pt2e
+from torch.ao.quantization.quantize_pt2e import (
+    convert_pt2e,
+    prepare_pt2e,
+    prepare_qat_pt2e,
+)
 
 
 def generate_context_binary(