Revert "add amp support for SE_ResNet50_vd (#509)" (#515)

This reverts commit d5105825.

configs/ResNet/ResNet50_vd_fp16.yaml

-mode: 'train'
-ARCHITECTURE:
-    name: 'ResNet50_vd'
-
-pretrained_model: ""
-model_save_dir: "./output/"
-classes_num: 1000
-total_images: 1281167
-save_interval: 1
-validate: True
-valid_interval: 1
-epochs: 200
-topk: 5
-image_shape: [3, 224, 224]
-
-# mixed precision training
-use_amp: True
-use_pure_fp16: False
-multi_precision: False
-scale_loss: 128.0
-use_dynamic_loss_scaleing: True
-data_format: "NCHW"
-
-use_mix: True
-ls_epsilon: 0.1
-
-LEARNING_RATE:
-    function: 'Cosine'          
-    params:                   
-        lr: 0.1               
-
-OPTIMIZER:
-    function: 'Momentum'
-    params:
-        momentum: 0.9
-    regularizer:
-        function: 'L2'
-        factor: 0.000070
-
-TRAIN:
-    batch_size: 256
-    num_workers: 4
-    file_list: "./dataset/ILSVRC2012/train_list.txt"
-    data_dir: "./dataset/ILSVRC2012/"
-    shuffle_seed: 0
-    transforms:
-        - DecodeImage:
-            to_rgb: True
-            to_np: False
-            channel_first: False
-        - RandCropImage:
-            size: 224
-        - RandFlipImage:
-            flip_code: 1
-        - NormalizeImage:
-            scale: 1./255.
-            mean: [0.485, 0.456, 0.406]
-            std: [0.229, 0.224, 0.225]
-            order: ''
-        - ToCHWImage:
-    mix:                       
-        - MixupOperator:    
-            alpha: 0.2      
-
-VALID:
-    batch_size: 64
-    num_workers: 4
-    file_list: "./dataset/ILSVRC2012/val_list.txt"
-    data_dir: "./dataset/ILSVRC2012/"
-    shuffle_seed: 0
-    transforms:
-        - DecodeImage:
-            to_rgb: True
-            to_np: False
-            channel_first: False
-        - ResizeImage:
-            resize_short: 256
-        - CropImage:
-            size: 224
-        - NormalizeImage:
-            scale: 1.0/255.0
-            mean: [0.485, 0.456, 0.406]
-            std: [0.229, 0.224, 0.225]
-            order: ''
-        - ToCHWImage:

ppcls/modeling/architectures/se_resnet_vd.py

@@ -33,16 +33,16 @@ __all__ = [
 
 
 class ConvBNLayer(nn.Layer):
-    def __init__(self,
-                 num_channels,
-                 num_filters,
-                 filter_size,
-                 stride=1,
-                 groups=1,
-                 is_vd_mode=False,
-                 act=None,
-                 name=None,
-                 data_format="NCHW"):
+    def __init__(
+            self,
+            num_channels,
+            num_filters,
+            filter_size,
+            stride=1,
+            groups=1,
+            is_vd_mode=False,
+            act=None,
+            name=None, ):
         super(ConvBNLayer, self).__init__()
 
         self.is_vd_mode = is_vd_mode
@@ -57,8 +57,7 @@ class ConvBNLayer(nn.Layer):
             padding=(filter_size - 1) // 2,
             groups=groups,
             weight_attr=ParamAttr(name=name + "_weights"),
-            bias_attr=False,
-            data_format=data_format)
+            bias_attr=False)
         if name == "conv1":
             bn_name = "bn_" + name
         else:
@@ -69,8 +68,7 @@ class ConvBNLayer(nn.Layer):
             param_attr=ParamAttr(name=bn_name + '_scale'),
             bias_attr=ParamAttr(bn_name + '_offset'),
             moving_mean_name=bn_name + '_mean',
-            moving_variance_name=bn_name + '_variance',
-            data_format=data_format)
+            moving_variance_name=bn_name + '_variance')
 
     def forward(self, inputs):
         if self.is_vd_mode:
@@ -88,8 +86,7 @@ class BottleneckBlock(nn.Layer):
                  shortcut=True,
                  if_first=False,
                  reduction_ratio=16,
-                 name=None,
-                 data_format="NCHW"):
+                 name=None):
         super(BottleneckBlock, self).__init__()
 
         self.conv0 = ConvBNLayer(
@@ -97,29 +94,25 @@ class BottleneckBlock(nn.Layer):
             num_filters=num_filters,
             filter_size=1,
             act='relu',
-            name=name + "_branch2a",
-            data_format=data_format)
+            name=name + "_branch2a")
         self.conv1 = ConvBNLayer(
             num_channels=num_filters,
             num_filters=num_filters,
             filter_size=3,
             stride=stride,
             act='relu',
-            name=name + "_branch2b",
-            data_format=data_format)
+            name=name + "_branch2b")
         self.conv2 = ConvBNLayer(
             num_channels=num_filters,
             num_filters=num_filters * 4,
             filter_size=1,
             act=None,
-            name=name + "_branch2c",
-            data_format=data_format)
+            name=name + "_branch2c")
         self.scale = SELayer(
             num_channels=num_filters * 4,
             num_filters=num_filters * 4,
             reduction_ratio=reduction_ratio,
-            name='fc_' + name,
-            data_format=data_format)
+            name='fc_' + name)
 
         if not shortcut:
             self.short = ConvBNLayer(
@@ -128,8 +121,7 @@ class BottleneckBlock(nn.Layer):
                 filter_size=1,
                 stride=1,
                 is_vd_mode=False if if_first else True,
-                name=name + "_branch1",
-                data_format=data_format)
+                name=name + "_branch1")
 
         self.shortcut = shortcut
 
@@ -156,8 +148,7 @@ class BasicBlock(nn.Layer):
                  shortcut=True,
                  if_first=False,
                  reduction_ratio=16,
-                 name=None,
-                 data_format="NCHW"):
+                 name=None):
         super(BasicBlock, self).__init__()
         self.stride = stride
         self.conv0 = ConvBNLayer(
@@ -166,22 +157,19 @@ class BasicBlock(nn.Layer):
             filter_size=3,
             stride=stride,
             act='relu',
-            name=name + "_branch2a",
-            data_format=data_format)
+            name=name + "_branch2a")
         self.conv1 = ConvBNLayer(
             num_channels=num_filters,
             num_filters=num_filters,
             filter_size=3,
             act=None,
-            name=name + "_branch2b",
-            data_format=data_format)
+            name=name + "_branch2b")
 
         self.scale = SELayer(
             num_channels=num_filters,
             num_filters=num_filters,
             reduction_ratio=reduction_ratio,
-            name='fc_' + name,
-            data_format=data_format)
+            name='fc_' + name)
 
         if not shortcut:
             self.short = ConvBNLayer(
@@ -190,8 +178,7 @@ class BasicBlock(nn.Layer):
                 filter_size=1,
                 stride=1,
                 is_vd_mode=False if if_first else True,
-                name=name + "_branch1",
-                data_format=data_format)
+                name=name + "_branch1")
 
         self.shortcut = shortcut
 
@@ -210,12 +197,7 @@ class BasicBlock(nn.Layer):
 
 
 class SELayer(nn.Layer):
-    def __init__(self,
-                 num_channels,
-                 num_filters,
-                 reduction_ratio,
-                 name=None,
-                 data_format="NCHW"):
+    def __init__(self, num_channels, num_filters, reduction_ratio, name=None):
         super(SELayer, self).__init__()
 
         self.pool2d_gap = AdaptiveAvgPool2D(1)
@@ -252,16 +234,10 @@ class SELayer(nn.Layer):
 
 
 class SE_ResNet_vd(nn.Layer):
-    def __init__(self,
-                 layers=50,
-                 class_dim=1000,
-                 input_image_channel=3,
-                 data_format="NCHW"):
+    def __init__(self, layers=50, class_dim=1000):
         super(SE_ResNet_vd, self).__init__()
 
         self.layers = layers
-        self.data_format = data_format
-        self.input_image_channel = input_image_channel
         supported_layers = [18, 34, 50, 101, 152, 200]
         assert layers in supported_layers, \
             "supported layers are {} but input layer is {}".format(
@@ -282,31 +258,27 @@ class SE_ResNet_vd(nn.Layer):
         num_filters = [64, 128, 256, 512]
 
         self.conv1_1 = ConvBNLayer(
-            num_channels=self.input_image_channel,
+            num_channels=3,
             num_filters=32,
             filter_size=3,
             stride=2,
             act='relu',
-            name="conv1_1",
-            data_format=self.data_format)
+            name="conv1_1")
         self.conv1_2 = ConvBNLayer(
             num_channels=32,
             num_filters=32,
             filter_size=3,
             stride=1,
             act='relu',
-            name="conv1_2",
-            data_format=self.data_format)
+            name="conv1_2")
         self.conv1_3 = ConvBNLayer(
             num_channels=32,
             num_filters=64,
             filter_size=3,
             stride=1,
             act='relu',
-            name="conv1_3",
-            data_format=self.data_format)
-        self.pool2d_max = MaxPool2D(
-            kernel_size=3, stride=2, padding=1, data_format=self.data_format)
+            name="conv1_3")
+        self.pool2d_max = MaxPool2D(kernel_size=3, stride=2, padding=1)
 
         self.block_list = []
         if layers >= 50:
@@ -329,8 +301,7 @@ class SE_ResNet_vd(nn.Layer):
                             stride=2 if i == 0 and block != 0 else 1,
                             shortcut=shortcut,
                             if_first=block == i == 0,
-                            name=conv_name,
-                            data_format=self.data_format))
+                            name=conv_name))
                     self.block_list.append(bottleneck_block)
                     shortcut = True
         else: