add docs

image_classification/main.py

-# Copyright (c) 2019 PaddlePaddle Authors. All Rights Reserved.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -42,7 +41,8 @@ def make_optimizer(step_per_epoch, parameter_list=None):
     weight_decay = FLAGS.weight_decay
 
     if lr_scheduler == 'piecewise':
-        boundaries = [step_per_epoch * e for e in [30, 60, 80]]
+        milestones = FLAGS.milestones
+        boundaries = [step_per_epoch * e for e in milestones]
         values = [base_lr * (0.1**i) for i in range(len(boundaries) + 1)]
         learning_rate = fluid.layers.piecewise_decay(
             boundaries=boundaries, values=values)
@@ -155,6 +155,12 @@ if __name__ == '__main__':
         default='piecewise',
         type=str,
         help="learning rate scheduler")
+    parser.add_argument(
+        "--milestones",
+        nargs='+',
+        type=int,
+        default=[30, 60, 80],
+        help="piecewise decay milestones")
     parser.add_argument(
         "--weight-decay", default=1e-4, type=float, help="weight decay")
     parser.add_argument("--momentum", default=0.9, type=float, help="momentum")

models/mobilenet.py

-# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
-#
-# Licensed under the Apache License, Version 2.0 (the "License");
-# you may not use this file except in compliance with the License.
-# You may obtain a copy of the License at
-#
-#     http://www.apache.org/licenses/LICENSE-2.0
-#
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS,
-# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-# See the License for the specific language governing permissions and
-# limitations under the License.
-
-import os
-import time
-import math
-import sys
-import numpy as np
-import argparse
-import paddle
-import paddle.fluid as fluid
-from paddle.fluid.initializer import MSRA
-from paddle.fluid.param_attr import ParamAttr
-from paddle.fluid.layer_helper import LayerHelper
-from paddle.fluid.dygraph.nn import Conv2D, Pool2D, BatchNorm, Linear
-from paddle.fluid.dygraph.base import to_variable
-from paddle.fluid import framework
-
-from model import Model
-from .download import get_weights_path
-
-__all__ = [
-    'MobileNetV2', 'mobilnetv2_x0_25', 'mobilnetv2_x0_5', 'mobilnetv2_x0_75',
-    'mobilnetv2_x1_0', 'mobilnetv2_x1_25', 'mobilnetv2_x1_5',
-    'mobilnetv2_x1_75', 'mobilnetv2_x2_0'
-]
-
-model_urls = {}
-
-
-class ConvBNLayer(fluid.dygraph.Layer):
-    def __init__(self,
-                 num_channels,
-                 filter_size,
-                 num_filters,
-                 stride,
-                 padding,
-                 channels=None,
-                 num_groups=1,
-                 use_cudnn=True):
-        super(ConvBNLayer, self).__init__()
-
-        tmp_param = ParamAttr(name=self.full_name() + "_weights")
-        self._conv = Conv2D(
-            num_channels=num_channels,
-            num_filters=num_filters,
-            filter_size=filter_size,
-            stride=stride,
-            padding=padding,
-            groups=num_groups,
-            act=None,
-            use_cudnn=use_cudnn,
-            param_attr=tmp_param,
-            bias_attr=False)
-
-        self._batch_norm = BatchNorm(
-            num_filters,
-            param_attr=ParamAttr(name=self.full_name() + "_bn" + "_scale"),
-            bias_attr=ParamAttr(name=self.full_name() + "_bn" + "_offset"),
-            moving_mean_name=self.full_name() + "_bn" + '_mean',
-            moving_variance_name=self.full_name() + "_bn" + '_variance')
-
-    def forward(self, inputs, if_act=True):
-        y = self._conv(inputs)
-        y = self._batch_norm(y)
-        if if_act:
-            y = fluid.layers.relu6(y)
-        return y
-
-
-class InvertedResidualUnit(fluid.dygraph.Layer):
-    def __init__(
-            self,
-            num_channels,
-            num_in_filter,
-            num_filters,
-            stride,
-            filter_size,
-            padding,
-            expansion_factor, ):
-        super(InvertedResidualUnit, self).__init__()
-        num_expfilter = int(round(num_in_filter * expansion_factor))
-        self._expand_conv = ConvBNLayer(
-            num_channels=num_channels,
-            num_filters=num_expfilter,
-            filter_size=1,
-            stride=1,
-            padding=0,
-            num_groups=1)
-
-        self._bottleneck_conv = ConvBNLayer(
-            num_channels=num_expfilter,
-            num_filters=num_expfilter,
-            filter_size=filter_size,
-            stride=stride,
-            padding=padding,
-            num_groups=num_expfilter,
-            use_cudnn=False)
-
-        self._linear_conv = ConvBNLayer(
-            num_channels=num_expfilter,
-            num_filters=num_filters,
-            filter_size=1,
-            stride=1,
-            padding=0,
-            num_groups=1)
-
-    def forward(self, inputs, ifshortcut):
-        y = self._expand_conv(inputs, if_act=True)
-        y = self._bottleneck_conv(y, if_act=True)
-        y = self._linear_conv(y, if_act=False)
-        if ifshortcut:
-            y = fluid.layers.elementwise_add(inputs, y)
-        return y
-
-
-class InvresiBlocks(fluid.dygraph.Layer):
-    def __init__(self, in_c, t, c, n, s):
-        super(InvresiBlocks, self).__init__()
-
-        self._first_block = InvertedResidualUnit(
-            num_channels=in_c,
-            num_in_filter=in_c,
-            num_filters=c,
-            stride=s,
-            filter_size=3,
-            padding=1,
-            expansion_factor=t)
-
-        self._inv_blocks = []
-        for i in range(1, n):
-            tmp = self.add_sublayer(
-                sublayer=InvertedResidualUnit(
-                    num_channels=c,
-                    num_in_filter=c,
-                    num_filters=c,
-                    stride=1,
-                    filter_size=3,
-                    padding=1,
-                    expansion_factor=t),
-                name=self.full_name() + "_" + str(i + 1))
-            self._inv_blocks.append(tmp)
-
-    def forward(self, inputs):
-        y = self._first_block(inputs, ifshortcut=False)
-        for inv_block in self._inv_blocks:
-            y = inv_block(y, ifshortcut=True)
-        return y
-
-
-class MobileNetV2(Model):
-    def __init__(self, class_dim=1000, scale=1.0):
-        super(MobileNetV2, self).__init__()
-        self.scale = scale
-        self.class_dim = class_dim
-
-        bottleneck_params_list = [
-            (1, 16, 1, 1),
-            (6, 24, 2, 2),
-            (6, 32, 3, 2),
-            (6, 64, 4, 2),
-            (6, 96, 3, 1),
-            (6, 160, 3, 2),
-            (6, 320, 1, 1),
-        ]
-
-        #1. conv1 
-        self._conv1 = ConvBNLayer(
-            num_channels=3,
-            num_filters=int(32 * scale),
-            filter_size=3,
-            stride=2,
-            padding=1)
-
-        #2. bottleneck sequences
-        self._invl = []
-        i = 1
-        in_c = int(32 * scale)
-        for layer_setting in bottleneck_params_list:
-            t, c, n, s = layer_setting
-            i += 1
-            tmp = self.add_sublayer(
-                sublayer=InvresiBlocks(
-                    in_c=in_c, t=t, c=int(c * scale), n=n, s=s),
-                name='conv' + str(i))
-            self._invl.append(tmp)
-            in_c = int(c * scale)
-
-        #3. last_conv
-        self._out_c = int(1280 * scale) if scale > 1.0 else 1280
-        self._conv9 = ConvBNLayer(
-            num_channels=in_c,
-            num_filters=self._out_c,
-            filter_size=1,
-            stride=1,
-            padding=0)
-
-        #4. pool
-        self._pool2d_avg = Pool2D(pool_type='avg', global_pooling=True)
-
-        #5. fc
-        tmp_param = ParamAttr(name=self.full_name() + "fc10_weights")
-        self._fc = Linear(
-            self._out_c,
-            class_dim,
-            act='softmax',
-            param_attr=tmp_param,
-            bias_attr=ParamAttr(name="fc10_offset"))
-
-    def forward(self, inputs):
-        y = self._conv1(inputs, if_act=True)
-        for inv in self._invl:
-            y = inv(y)
-        y = self._conv9(y, if_act=True)
-        y = self._pool2d_avg(y)
-        y = fluid.layers.reshape(y, shape=[-1, self._out_c])
-        y = self._fc(y)
-        return y
-
-
-def _mobilenet(arch, pretrained=False, **kwargs):
-    model = MobileNetV2(**kwargs)
-    if pretrained:
-        assert arch in model_urls, "{} model do not have a pretrained model now, you should set pretrained=False".format(
-            arch)
-        weight_path = get_weights_path(model_urls[arch][0],
-                                       model_urls[arch][1])
-        assert weight_path.endswith(
-            '.pdparams'), "suffix of weight must be .pdparams"
-        model.load(weight_path[:-9])
-
-    return model
-
-
-def mobilnetv2_x1_0(pretrained=False):
-    model = _mobilenet('mobilenetv2_1.0', pretrained, scale=1.0)
-    return model
-
-
-def mobilnetv2_x0_25(pretrained=False):
-    model = _mobilenet('mobilenetv2_0.25', pretrained, scale=0.25)
-    return model
-
-
-def mobilnetv2_x0_5(pretrained=False):
-    model = _mobilenet('mobilenetv2_0.5', pretrained, scale=0.5)
-    return model
-
-
-def mobilnetv2_x0_75(pretrained=False):
-    model = _mobilenet('mobilenetv2_0.75', pretrained, scale=0.75)
-    return model
-
-
-def mobilnetv2_x1_25(pretrained=False):
-    model = _mobilenet('mobilenetv2_1.25', pretrained, scale=1.25)
-    return model
-
-
-def mobilnetv2_x1_5(pretrained=False):
-    model = _mobilenet('mobilenetv2_1.5', pretrained, scale=1.5)
-    return model
-
-
-def mobilnetv2_x1_75(pretrained=False):
-    model = _mobilenet('mobilenetv2_1.75', pretrained, scale=1.75)
-    return model
-
-
-def mobilnetv2_x2_0(pretrained=False):
-    model = _mobilenet('mobilenetv2_2.0', pretrained, scale=2.0)
-    return model

models/mobilenetv1.py

@@ -12,29 +12,17 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import os
-import time
-import sys
-import math
 import numpy as np
-import argparse
 import paddle
 import paddle.fluid as fluid
 from paddle.fluid.initializer import MSRA
 from paddle.fluid.param_attr import ParamAttr
-from paddle.fluid.layer_helper import LayerHelper
 from paddle.fluid.dygraph.nn import Conv2D, Pool2D, BatchNorm, Linear
-from paddle.fluid.dygraph.base import to_variable
-from paddle.fluid import framework
 
 from model import Model
 from .download import get_weights_path
 
-__all__ = [
-    'MobileNetV1', 'mobilnetv1_x0_25', 'mobilnetv1_x0_5', 'mobilnetv1_x0_75',
-    'mobilnetv1_x1_0', 'mobilnetv1_x1_25', 'mobilnetv1_x1_5',
-    'mobilnetv1_x1_75', 'mobilnetv1_x2_0'
-]
+__all__ = ['MobileNetV1', 'mobilenet_v1']
 
 model_urls = {}
 
@@ -114,6 +102,14 @@ class DepthwiseSeparable(fluid.dygraph.Layer):
 
 
 class MobileNetV1(Model):
+    """MobileNetV1 model from
+    `"MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications" <https://arxiv.org/abs/1704.04861>`_.
+
+    Args:
+        scale (float): scale of channels in each layer. Default: 1.0.
+        class_dim (int): output dim of last fc layer. Default: 1000.
+    """
+
     def __init__(self, scale=1.0, class_dim=1000):
         super(MobileNetV1, self).__init__()
         self.scale = scale
@@ -261,41 +257,6 @@ def _mobilenet(arch, pretrained=False, **kwargs):
     return model
 
 
-def mobilnetv1_x1_0(pretrained=False):
-    model = _mobilenet('mobilenetv1_1.0', pretrained, scale=1.0)
-    return model
-
-
-def mobilnetv1_x0_25(pretrained=False):
-    model = _mobilenet('mobilenetv1_0.25', pretrained, scale=0.25)
-    return model
-
-
-def mobilnetv1_x0_5(pretrained=False):
-    model = _mobilenet('mobilenetv1_0.5', pretrained, scale=0.5)
-    return model
-
-
-def mobilnetv1_x0_75(pretrained=False):
-    model = _mobilenet('mobilenetv1_0.75', pretrained, scale=0.75)
-    return model
-
-
-def mobilnetv1_x1_25(pretrained=False):
-    model = _mobilenet('mobilenetv1_1.25', pretrained, scale=1.25)
-    return model
-
-
-def mobilnetv1_x1_5(pretrained=False):
-    model = _mobilenet('mobilenetv1_1.5', pretrained, scale=1.5)
-    return model
-
-
-def mobilnetv1_x1_75(pretrained=False):
-    model = _mobilenet('mobilenetv1_1.75', pretrained, scale=1.75)
-    return model
-
-
-def mobilnetv1_x2_0(pretrained=False):
-    model = _mobilenet('mobilenetv1_2.0', pretrained, scale=2.0)
+def mobilenet_v1(pretrained=False, scale=1.0):
+    model = _mobilenet('mobilenetv1_' + str(scale), pretrained, scale=scale)
     return model

models/mobilenetv2.py

@@ -12,29 +12,16 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import os
-import time
-import math
-import sys
 import numpy as np
-import argparse
 import paddle
 import paddle.fluid as fluid
-from paddle.fluid.initializer import MSRA
 from paddle.fluid.param_attr import ParamAttr
-from paddle.fluid.layer_helper import LayerHelper
 from paddle.fluid.dygraph.nn import Conv2D, Pool2D, BatchNorm, Linear
-from paddle.fluid.dygraph.base import to_variable
-from paddle.fluid import framework
 
 from model import Model
 from .download import get_weights_path
 
-__all__ = [
-    'MobileNetV2', 'mobilnetv2_x0_25', 'mobilnetv2_x0_5', 'mobilnetv2_x0_75',
-    'mobilnetv2_x1_0', 'mobilnetv2_x1_25', 'mobilnetv2_x1_5',
-    'mobilnetv2_x1_75', 'mobilnetv2_x2_0'
-]
+__all__ = ['MobileNetV2', 'mobilenet_v2']
 
 model_urls = {}
 
@@ -160,7 +147,15 @@ class InvresiBlocks(fluid.dygraph.Layer):
 
 
 class MobileNetV2(Model):
-    def __init__(self, class_dim=1000, scale=1.0):
+    """MobileNetV2 model from
+    `"MobileNetV2: Inverted Residuals and Linear Bottlenecks" <https://arxiv.org/abs/1801.04381>`_.
+
+    Args:
+        scale (float): scale of channels in each layer. Default: 1.0.
+        class_dim (int): output dim of last fc layer. Default: 1000.
+    """
+
+    def __init__(self, scale=1.0, class_dim=1000):
         super(MobileNetV2, self).__init__()
         self.scale = scale
         self.class_dim = class_dim
@@ -243,41 +238,11 @@ def _mobilenet(arch, pretrained=False, **kwargs):
     return model
 
 
-def mobilnetv2_x1_0(pretrained=False):
-    model = _mobilenet('mobilenetv2_1.0', pretrained, scale=1.0)
-    return model
-
-
-def mobilnetv2_x0_25(pretrained=False):
-    model = _mobilenet('mobilenetv2_0.25', pretrained, scale=0.25)
-    return model
-
-
-def mobilnetv2_x0_5(pretrained=False):
-    model = _mobilenet('mobilenetv2_0.5', pretrained, scale=0.5)
-    return model
-
-
-def mobilnetv2_x0_75(pretrained=False):
-    model = _mobilenet('mobilenetv2_0.75', pretrained, scale=0.75)
-    return model
-
-
-def mobilnetv2_x1_25(pretrained=False):
-    model = _mobilenet('mobilenetv2_1.25', pretrained, scale=1.25)
-    return model
-
-
-def mobilnetv2_x1_5(pretrained=False):
-    model = _mobilenet('mobilenetv2_1.5', pretrained, scale=1.5)
-    return model
-
-
-def mobilnetv2_x1_75(pretrained=False):
-    model = _mobilenet('mobilenetv2_1.75', pretrained, scale=1.75)
-    return model
-
-
-def mobilnetv2_x2_0(pretrained=False):
-    model = _mobilenet('mobilenetv2_2.0', pretrained, scale=2.0)
+def mobilenet_v2(pretrained=False, scale=1.0):
+    """MobileNetV2
+    
+    Args:
+        pretrained (bool): If True, returns a model pre-trained on ImageNet
+    """
+    model = _mobilenet('mobilenetv2_' + str(scale), pretrained, scale=scale)
     return model

models/resnet.py

@@ -157,6 +157,15 @@ class BottleneckBlock(fluid.dygraph.Layer):
 
 
 class ResNet(Model):
+    """ResNet model from
+    `"Deep Residual Learning for Image Recognition" <https://arxiv.org/pdf/1512.03385.pdf>`_
+
+    Args:
+        Block (BasicBlock|BottleneckBlock): block module of model.
+        depth (int): layers of resnet, default: 50.
+        num_classes (int): output dim of last fc layer, default: 1000.
+    """
+
     def __init__(self, Block, depth=50, num_classes=1000):
         super(ResNet, self).__init__()
 
@@ -240,20 +249,45 @@ def _resnet(arch, Block, depth, pretrained):
 
 
 def resnet18(pretrained=False):
+    """ResNet 18-layer model
+    
+    Args:
+        pretrained (bool): If True, returns a model pre-trained on ImageNet
+    """
     return _resnet('resnet18', BasicBlock, 18, pretrained)
 
 
 def resnet34(pretrained=False):
+    """ResNet 34-layer model
+    
+    Args:
+        pretrained (bool): If True, returns a model pre-trained on ImageNet
+    """
     return _resnet('resnet34', BasicBlock, 34, pretrained)
 
 
 def resnet50(pretrained=False):
+    """ResNet 50-layer model
+    
+    Args:
+        pretrained (bool): If True, returns a model pre-trained on ImageNet
+    """
     return _resnet('resnet50', BottleneckBlock, 50, pretrained)
 
 
 def resnet101(pretrained=False):
+    """ResNet 101-layer model
+    
+    Args:
+        pretrained (bool): If True, returns a model pre-trained on ImageNet
+    """
     return _resnet('resnet101', BottleneckBlock, 101, pretrained)
 
 
 def resnet152(pretrained=False):
+    """ResNet 152-layer model
+    
+    Args:
+        pretrained (bool): If True, returns a model pre-trained on ImageNet
+    """
     return _resnet('resnet152', BottleneckBlock, 152, pretrained)

models/vgg.py

@@ -12,21 +12,10 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import os
-import time
-import math
-import sys
-import numpy as np
-import argparse
 import paddle
 import paddle.fluid as fluid
-from paddle.fluid.initializer import MSRA
-from paddle.fluid.param_attr import ParamAttr
-from paddle.fluid.layer_helper import LayerHelper
 from paddle.fluid.dygraph.nn import Conv2D, Pool2D, BatchNorm, Linear
 from paddle.fluid.dygraph.container import Sequential
-from paddle.fluid.dygraph.base import to_variable
-from paddle.fluid import framework
 
 from model import Model
 from .download import get_weights_path
@@ -65,7 +54,15 @@ class Classifier(fluid.dygraph.Layer):
 
 
 class VGG(Model):
-    def __init__(self, features, num_classes=1000, init_weights=True):
+    """VGG model from
+    `"Very Deep Convolutional Networks For Large-Scale Image Recognition" <https://arxiv.org/pdf/1409.1556.pdf>`_
+
+    Args:
+        features (fluid.dygraph.Layer): vgg features create by function make_layers.
+        num_classes (int): output dim of last fc layer. Default: 1000.
+    """
+
+    def __init__(self, features, num_classes=1000):
         super(VGG, self).__init__()
         self.features = features
         classifier = Classifier(num_classes)
@@ -74,9 +71,7 @@ class VGG(Model):
 
     def forward(self, x):
         x = self.features(x)
-        # x = fluid.layers.adaptive_pool2d(x, pool_size=(7, 7), pool_type='avg')
-        # x = fluid.layers.flatten(x, 1)
-        x = fluid.layers.reshape(x, [-1, 7 * 7 * 512])
+        x = fluid.layers.flatten(x, 1)
         x = self.classifier(x)
         return x
 
@@ -116,9 +111,8 @@ cfgs = {
 
 
 def _vgg(arch, cfg, batch_norm, pretrained, **kwargs):
-    if pretrained:
-        kwargs['init_weights'] = False
     model = VGG(make_layers(cfgs[cfg], batch_norm=batch_norm), **kwargs)
+
     if pretrained:
         assert arch in model_urls, "{} model do not have a pretrained model now, you should set pretrained=False".format(
             arch)
@@ -127,6 +121,7 @@ def _vgg(arch, cfg, batch_norm, pretrained, **kwargs):
         assert weight_path.endswith(
             '.pdparams'), "suffix of weight must be .pdparams"
         model.load(weight_path[:-9])
+
     return model