remove import pdb (#1897)

Signed-off-by: changwangss <[email protected]>

Author: changwangss

examples/onnxrt/nlp/huggingface_model/token_classification/layoutlmv2/quantization/ptq_dynamic/main.py

@@ -216,8 +216,6 @@ def _process_dataset(self):
         self.label = []
         self.onnx_inputs = []
         for inputs in self.dataset:
-            # import pdb;
-            # pdb.set_trace()
             onnx_inputs = []
             has_labels = all(inputs.get(k) is not None for k in self.label_names)
             if has_labels:
@@ -237,8 +235,6 @@ def _process_dataset(self):
             }
             """
             for key in self.onnx_input_names:
-                # import pdb;
-                # pdb.set_trace()
                 if key in inputs:
                     # onnx_inputs[key] = np.array([inputs[key]])
                     onnx_inputs.append(np.array(inputs[key]))

examples/onnxrt/nlp/huggingface_model/token_classification/layoutlmv2/quantization/ptq_static/main.py

@@ -216,8 +216,6 @@ def _process_dataset(self):
         self.label = []
         self.onnx_inputs = []
         for inputs in self.dataset:
-            # import pdb;
-            # pdb.set_trace()
             onnx_inputs = []
             has_labels = all(inputs.get(k) is not None for k in self.label_names)
             if has_labels:
@@ -237,8 +235,6 @@ def _process_dataset(self):
             }
             """
             for key in self.onnx_input_names:
-                # import pdb;
-                # pdb.set_trace()
                 if key in inputs:
                     # onnx_inputs[key] = np.array([inputs[key]])
                     onnx_inputs.append(np.array(inputs[key]))

examples/pytorch/object_detection/ssd_resnet34/quantization/ptq/fx/python/models/utils.py

@@ -59,7 +59,6 @@ def _compute_padding(self, input, dim):
         return additional_padding, total_padding
 
     def forward(self, input):
-        #import pdb; pdb.set_trace()
         if self.padding == "VALID":
             return F.conv2d(
                 input,
@@ -180,7 +179,6 @@ def decode_boxes(rel_codes, boxes, weights):
     dh = dh / wh
 
     pred_ctr_x = dx * widths + ctr_x
-    #import pdb; pdb.set_trace()
     pred_ctr_y = dy * heights + ctr_y
     pred_w = torch.exp(dw) * widths
     pred_h = torch.exp(dh) * heights
@@ -194,5 +192,4 @@ def decode_boxes(rel_codes, boxes, weights):
         ],
         dim=2,
     )
-    #import pdb; pdb.set_trace()
     return pred_boxes

neural_coder/coders/tensorflow/amp.py

@@ -22,8 +22,6 @@ def __init__(self, file) -> None:
         self.keras_edited_flag = False
 
     def transform(self):
-        # import pdb
-        # pdb.set_trace()
         lines = self.file.split("\n")
         for line in lines:
             if self.is_modify(line):

neural_coder/coders/tensorflow/inc.py

@@ -21,8 +21,6 @@ def __init__(self, file) -> None:
         self.result = []
 
     def transform(self):
-        # import pdb
-        # pdb.set_trace()
         lines = self.file.split("\n")
         for line in lines:
             if self.is_modify(line):

neural_compressor/strategy/strategy.py

@@ -485,7 +485,6 @@ def traverse(self):
                 return self.distributed_traverse()
         self._setup_pre_tuning_algo_scheduler()
         self._prepare_tuning()
-        # import pdb;pdb.set_trace()
         traverse_start_time = time()
         for op_tuning_cfg in self.next_tune_cfg():
             tuning_start_time = time()

neural_compressor/torch/algorithms/habana_fp8/fp8_quant.py

@@ -131,7 +131,6 @@ def input_observer_forward_pre_hook(self, input):
 
     ### Insert input observer into model, only for fp8_e4m3 static quantization ###
     observer_cls = observer_mapping[act_observer]
-    # import pdb;pdb.set_trace()
 
     if isinstance(module, white_list):
         observer_obj = observer_cls(dtype=dtype_mapping[qconfig.act_dtype])

test/pruning_with_pt/pruning_2.x/test_auto_excluding_classifier.py

@@ -24,7 +24,6 @@ def forward(self, x):
 
 class TestPruning(unittest.TestCase):
     def test_pruning_basic(self):
-        # import pdb;pdb.set_trace()
         hidden_size = 32
         model = NaiveMLP(hidden_size)
         # import classifier searching functions

test/pruning_with_pt/pruning_2.x/test_auto_slim.py

@@ -50,7 +50,6 @@ def test_pruning_basic(self):
         # run mha and ffn pruning
         compression_manager = prepare_compression(model=model, confs=configs)
         compression_manager.callbacks.on_train_begin()
-        # import pdb;pdb.set_trace()
         for epoch in range(3):
             model.train()
             compression_manager.callbacks.on_epoch_begin(epoch)