all changes

Author: ChongWei905

mindcv/models/convnext.py

@@ -10,7 +10,7 @@
 import mindspore.common.initializer as init
 from mindspore import Parameter, Tensor
 from mindspore import dtype as mstype
-from mindspore import nn, ops
+from mindspore import mint, nn, ops
 
 from .helpers import build_model_with_cfg
 from .layers.drop_path import DropPath
@@ -69,15 +69,14 @@ def __init__(self, dim: int):
         super().__init__()
         self.gamma = Parameter(Tensor(np.zeros([1, 1, 1, dim]), mstype.float32))
         self.beta = Parameter(Tensor(np.zeros([1, 1, 1, dim]), mstype.float32))
-        self.norm = ops.LpNorm(axis=[1, 2], p=2, keep_dims=True)
 
     def construct(self, x: Tensor) -> Tensor:
-        gx = self.norm(x)
-        nx = gx / (ops.mean(gx, axis=-1, keep_dims=True) + 1e-6)
+        gx = mint.norm(x, p=2, dim=(1, 2), keepdim=True)
+        nx = gx / (mint.mean(gx, dim=-1, keepdim=True) + 1e-6)
         return self.gamma * (x * nx) + self.beta + x
 
 
-class ConvNextLayerNorm(nn.LayerNorm):
+class ConvNextLayerNorm(mint.nn.LayerNorm):
     """
     LayerNorm for channels_first tensors with 2d spatial dimensions (ie N, C, H, W).
     """
@@ -88,17 +87,17 @@ def __init__(
         epsilon: float,
         norm_axis: int = -1,
     ) -> None:
-        super().__init__(normalized_shape=normalized_shape, epsilon=epsilon)
+        super().__init__(normalized_shape=normalized_shape, eps=epsilon)
         assert norm_axis in (-1, 1), "ConvNextLayerNorm's norm_axis must be 1 or -1."
         self.norm_axis = norm_axis
 
     def construct(self, input_x: Tensor) -> Tensor:
         if self.norm_axis == -1:
-            y, _, _ = self.layer_norm(input_x, self.gamma, self.beta)
+            y = ops.layer_norm(input_x, self.normalized_shape, self.weight, self.bias, self.eps)
         else:
-            input_x = ops.transpose(input_x, (0, 2, 3, 1))
-            y, _, _ = self.layer_norm(input_x, self.gamma, self.beta)
-            y = ops.transpose(y, (0, 3, 1, 2))
+            input_x = mint.permute(input_x, (0, 2, 3, 1))
+            y = ops.layer_norm(input_x, self.normalized_shape, self.weight, self.bias, self.eps)
+            y = mint.permute(y, (0, 3, 1, 2))
         return y
 
 
@@ -124,22 +123,22 @@ def __init__(
         use_grn: bool = False,
     ) -> None:
         super().__init__()
-        self.dwconv = nn.Conv2d(dim, dim, kernel_size=7, group=dim, has_bias=True)  # depthwise conv
+        self.dwconv = mint.nn.Conv2d(dim, dim, kernel_size=7, padding=3, groups=dim, bias=True)  # depthwise conv
         self.norm = ConvNextLayerNorm((dim,), epsilon=1e-6)
-        self.pwconv1 = nn.Dense(dim, 4 * dim)  # pointwise/1x1 convs, implemented with Dense layers
-        self.act = nn.GELU()
+        self.pwconv1 = mint.nn.Linear(dim, 4 * dim)  # pointwise/1x1 convs, implemented with Dense layers
+        self.act = mint.nn.GELU()
         self.use_grn = use_grn
         if use_grn:
             self.grn = GRN(4 * dim)
-        self.pwconv2 = nn.Dense(4 * dim, dim)
+        self.pwconv2 = mint.nn.Linear(4 * dim, dim)
         self.gamma_ = Parameter(Tensor(layer_scale_init_value * np.ones((dim)), dtype=mstype.float32),
                                 requires_grad=True) if layer_scale_init_value > 0 else None
         self.drop_path = DropPath(drop_path) if drop_path > 0.0 else Identity()
 
     def construct(self, x: Tensor) -> Tensor:
         downsample = x
         x = self.dwconv(x)
-        x = ops.transpose(x, (0, 2, 3, 1))
+        x = mint.permute(x, (0, 2, 3, 1))
         x = self.norm(x)
         x = self.pwconv1(x)
         x = self.act(x)
@@ -148,7 +147,7 @@ def construct(self, x: Tensor) -> Tensor:
         x = self.pwconv2(x)
         if self.gamma_ is not None:
             x = self.gamma_ * x
-        x = ops.transpose(x, (0, 3, 1, 2))
+        x = mint.permute(x, (0, 3, 1, 2))
         x = downsample + self.drop_path(x)
         return x
 
@@ -184,14 +183,14 @@ def __init__(
 
         downsample_layers = []  # stem and 3 intermediate down_sampling conv layers
         stem = nn.SequentialCell(
-            nn.Conv2d(in_channels, dims[0], kernel_size=4, stride=4, has_bias=True),
+            mint.nn.Conv2d(in_channels, dims[0], kernel_size=4, stride=4, bias=True),
             ConvNextLayerNorm((dims[0],), epsilon=1e-6, norm_axis=1),
         )
         downsample_layers.append(stem)
         for i in range(3):
             downsample_layer = nn.SequentialCell(
                 ConvNextLayerNorm((dims[i],), epsilon=1e-6, norm_axis=1),
-                nn.Conv2d(dims[i], dims[i + 1], kernel_size=2, stride=2, has_bias=True),
+                mint.nn.Conv2d(dims[i], dims[i + 1], kernel_size=2, stride=2, bias=True),
             )
             downsample_layers.append(downsample_layer)
 
@@ -226,18 +225,18 @@ def __init__(
             stages[3]
         ])
         self.norm = ConvNextLayerNorm((dims[-1],), epsilon=1e-6)  # final norm layer
-        self.classifier = nn.Dense(dims[-1], num_classes)  # classifier
+        self.classifier = mint.nn.Linear(dims[-1], num_classes)  # classifier
         self.head_init_scale = head_init_scale
         self._initialize_weights()
 
     def _initialize_weights(self) -> None:
         """Initialize weights for cells."""
         for _, cell in self.cells_and_names():
-            if isinstance(cell, (nn.Dense, nn.Conv2d)):
+            if isinstance(cell, (mint.nn.Linear, mint.nn.Conv2d)):
                 cell.weight.set_data(
                     init.initializer(init.TruncatedNormal(sigma=0.02), cell.weight.shape, cell.weight.dtype)
                 )
-                if isinstance(cell, nn.Dense) and cell.bias is not None:
+                if isinstance(cell, mint.nn.Linear) and cell.bias is not None:
                     cell.bias.set_data(init.initializer(init.Zero(), cell.bias.shape, cell.bias.dtype))
         self.classifier.weight.set_data(self.classifier.weight * self.head_init_scale)
         self.classifier.bias.set_data(self.classifier.bias * self.head_init_scale)

mindcv/models/densenet.py

@@ -8,10 +8,9 @@
 from typing import Tuple
 
 import mindspore.common.initializer as init
-from mindspore import Tensor, nn, ops
+from mindspore import Tensor, mint, nn
 
 from .helpers import load_pretrained
-from .layers.compatibility import Dropout
 from .layers.pooling import GlobalAvgPooling
 from .registry import register_model
 
@@ -53,16 +52,17 @@ def __init__(
         drop_rate: float,
     ) -> None:
         super().__init__()
-        self.norm1 = nn.BatchNorm2d(num_input_features)
-        self.relu1 = nn.ReLU()
-        self.conv1 = nn.Conv2d(num_input_features, bn_size * growth_rate, kernel_size=1, stride=1)
+        self.norm1 = mint.nn.BatchNorm2d(num_input_features)
+        self.relu1 = mint.nn.ReLU()
+        self.conv1 = mint.nn.Conv2d(num_input_features, bn_size * growth_rate, kernel_size=1, stride=1, bias=False)
 
-        self.norm2 = nn.BatchNorm2d(bn_size * growth_rate)
-        self.relu2 = nn.ReLU()
-        self.conv2 = nn.Conv2d(bn_size * growth_rate, growth_rate, kernel_size=3, stride=1, pad_mode="pad", padding=1)
+        self.norm2 = mint.nn.BatchNorm2d(bn_size * growth_rate)
+        self.relu2 = mint.nn.ReLU()
+        self.conv2 = mint.nn.Conv2d(
+            bn_size * growth_rate, growth_rate, kernel_size=3, stride=1, padding=1, bias=False)
 
         self.drop_rate = drop_rate
-        self.dropout = Dropout(p=self.drop_rate)
+        self.dropout = mint.nn.Dropout(p=self.drop_rate)
 
     def construct(self, features: Tensor) -> Tensor:
         bottleneck = self.conv1(self.relu1(self.norm1(features)))
@@ -98,7 +98,7 @@ def construct(self, init_features: Tensor) -> Tensor:
         features = init_features
         for layer in self.cell_list:
             new_features = layer(features)
-            features = ops.concat((features, new_features), axis=1)
+            features = mint.concat((features, new_features), dim=1)
         return features
 
 
@@ -112,10 +112,10 @@ def __init__(
     ) -> None:
         super().__init__()
         self.features = nn.SequentialCell(OrderedDict([
-            ("norm", nn.BatchNorm2d(num_input_features)),
-            ("relu", nn.ReLU()),
-            ("conv", nn.Conv2d(num_input_features, num_output_features, kernel_size=1, stride=1)),
-            ("pool", nn.AvgPool2d(kernel_size=2, stride=2))
+            ("norm", mint.nn.BatchNorm2d(num_input_features)),
+            ("relu", mint.nn.ReLU()),
+            ("conv", mint.nn.Conv2d(num_input_features, num_output_features, kernel_size=1, stride=1, bias=False)),
+            ("pool", mint.nn.AvgPool2d(kernel_size=2, stride=2))
         ]))
 
     def construct(self, x: Tensor) -> Tensor:
@@ -152,13 +152,11 @@ def __init__(
         layers = OrderedDict()
         # first Conv2d
         num_features = num_init_features
-        layers["conv0"] = nn.Conv2d(in_channels, num_features, kernel_size=7, stride=2, pad_mode="pad", padding=3)
-        layers["norm0"] = nn.BatchNorm2d(num_features)
-        layers["relu0"] = nn.ReLU()
-        layers["pool0"] = nn.SequentialCell([
-            nn.Pad(paddings=((0, 0), (0, 0), (1, 1), (1, 1)), mode="CONSTANT"),
-            nn.MaxPool2d(kernel_size=3, stride=2),
-        ])
+        layers["conv0"] = mint.nn.Conv2d(
+            in_channels, num_features, kernel_size=7, stride=2, padding=3, bias=False)
+        layers["norm0"] = mint.nn.BatchNorm2d(num_features)
+        layers["relu0"] = mint.nn.ReLU()
+        layers["pool0"] = mint.nn.MaxPool2d(kernel_size=3, stride=2, padding=1)
 
         # DenseBlock
         for i, num_layers in enumerate(block_config):
@@ -177,30 +175,30 @@ def __init__(
                 num_features = num_features // 2
 
         # final bn+ReLU
-        layers["norm5"] = nn.BatchNorm2d(num_features)
-        layers["relu5"] = nn.ReLU()
+        layers["norm5"] = mint.nn.BatchNorm2d(num_features)
+        layers["relu5"] = mint.nn.ReLU()
 
         self.num_features = num_features
         self.features = nn.SequentialCell(layers)
         self.pool = GlobalAvgPooling()
-        self.classifier = nn.Dense(self.num_features, num_classes)
+        self.classifier = mint.nn.Linear(self.num_features, num_classes)
         self._initialize_weights()
 
     def _initialize_weights(self) -> None:
         """Initialize weights for cells."""
         for _, cell in self.cells_and_names():
-            if isinstance(cell, nn.Conv2d):
+            if isinstance(cell, mint.nn.Conv2d):
                 cell.weight.set_data(
                     init.initializer(init.HeNormal(math.sqrt(5), mode="fan_out", nonlinearity="relu"),
                                      cell.weight.shape, cell.weight.dtype))
                 if cell.bias is not None:
                     cell.bias.set_data(
                         init.initializer(init.HeUniform(math.sqrt(5), mode="fan_in", nonlinearity="leaky_relu"),
                                          cell.bias.shape, cell.bias.dtype))
-            elif isinstance(cell, nn.BatchNorm2d):
-                cell.gamma.set_data(init.initializer("ones", cell.gamma.shape, cell.gamma.dtype))
-                cell.beta.set_data(init.initializer("zeros", cell.beta.shape, cell.beta.dtype))
-            elif isinstance(cell, nn.Dense):
+            elif isinstance(cell, mint.nn.BatchNorm2d):
+                cell.weight.set_data(init.initializer("ones", cell.weight.shape, cell.weight.dtype))
+                cell.bias.set_data(init.initializer("zeros", cell.bias.shape, cell.bias.dtype))
+            elif isinstance(cell, mint.nn.Linear):
                 cell.weight.set_data(
                     init.initializer(init.HeUniform(math.sqrt(5), mode="fan_in", nonlinearity="leaky_relu"),
                                      cell.weight.shape, cell.weight.dtype))

mindcv/models/googlenet.py

@@ -7,10 +7,10 @@
 from typing import Tuple, Union
 
 import mindspore.common.initializer as init
-from mindspore import Tensor, nn, ops
+from mindspore import Tensor, mint, nn
 
 from .helpers import load_pretrained
-from .layers.compatibility import Dropout
+from .layers.flatten import Flatten
 from .layers.pooling import GlobalAvgPooling
 from .registry import register_model
 
@@ -45,12 +45,12 @@ def __init__(
         kernel_size: int = 1,
         stride: int = 1,
         padding: int = 0,
-        pad_mode: str = "same",
+        pad_mode: str = "zeros",
     ) -> None:
         super().__init__()
-        self.conv = nn.Conv2d(in_channels, out_channels, kernel_size, stride,
-                              padding=padding, pad_mode=pad_mode)
-        self.relu = nn.ReLU()
+        self.conv = mint.nn.Conv2d(
+            in_channels, out_channels, kernel_size, stride, padding=padding, padding_mode=pad_mode, bias=False)
+        self.relu = mint.nn.ReLU()
 
     def construct(self, x: Tensor) -> Tensor:
         x = self.conv(x)
@@ -75,14 +75,14 @@ def __init__(
         self.b1 = BasicConv2d(in_channels, ch1x1, kernel_size=1)
         self.b2 = nn.SequentialCell([
             BasicConv2d(in_channels, ch3x3red, kernel_size=1),
-            BasicConv2d(ch3x3red, ch3x3, kernel_size=3),
+            BasicConv2d(ch3x3red, ch3x3, kernel_size=3, padding=1),
         ])
         self.b3 = nn.SequentialCell([
             BasicConv2d(in_channels, ch5x5red, kernel_size=1),
-            BasicConv2d(ch5x5red, ch5x5, kernel_size=5),
+            BasicConv2d(ch5x5red, ch5x5, kernel_size=5, padding=2),
         ])
         self.b4 = nn.SequentialCell([
-            nn.MaxPool2d(kernel_size=3, stride=1, pad_mode="same"),
+            mint.nn.MaxPool2d(kernel_size=3, stride=1, padding=1, ceil_mode=True),
             BasicConv2d(in_channels, pool_proj, kernel_size=1),
         ])
 
@@ -91,7 +91,7 @@ def construct(self, x: Tensor) -> Tensor:
         branch2 = self.b2(x)
         branch3 = self.b3(x)
         branch4 = self.b4(x)
-        return ops.concat((branch1, branch2, branch3, branch4), axis=1)
+        return mint.concat((branch1, branch2, branch3, branch4), dim=1)
 
 
 class InceptionAux(nn.Cell):
@@ -104,13 +104,13 @@ def __init__(
         drop_rate: float = 0.7,
     ) -> None:
         super().__init__()
-        self.avg_pool = nn.AvgPool2d(kernel_size=5, stride=3)
+        self.avg_pool = mint.nn.AvgPool2d(kernel_size=5, stride=3)
         self.conv = BasicConv2d(in_channels, 128, kernel_size=1)
-        self.fc1 = nn.Dense(2048, 1024)
-        self.fc2 = nn.Dense(1024, num_classes)
-        self.flatten = nn.Flatten()
-        self.relu = nn.ReLU()
-        self.dropout = Dropout(p=drop_rate)
+        self.fc1 = mint.nn.Linear(2048, 1024)
+        self.fc2 = mint.nn.Linear(1024, num_classes)
+        self.flatten = Flatten()
+        self.relu = mint.nn.ReLU()
+        self.dropout = mint.nn.Dropout(p=drop_rate)
 
     def construct(self, x: Tensor) -> Tensor:
         x = self.avg_pool(x)
@@ -145,23 +145,23 @@ def __init__(
     ) -> None:
         super().__init__()
         self.aux_logits = aux_logits
-        self.conv1 = BasicConv2d(in_channels, 64, kernel_size=7, stride=2)
-        self.maxpool1 = nn.MaxPool2d(kernel_size=3, stride=2, pad_mode="same")
+        self.conv1 = BasicConv2d(in_channels, 64, kernel_size=7, stride=2, padding=3)
+        self.maxpool1 = mint.nn.MaxPool2d(kernel_size=3, stride=2, ceil_mode=True)
 
         self.conv2 = BasicConv2d(64, 64, kernel_size=1)
-        self.conv3 = BasicConv2d(64, 192, kernel_size=3)
-        self.maxpool2 = nn.MaxPool2d(kernel_size=3, stride=2, pad_mode="same")
+        self.conv3 = BasicConv2d(64, 192, kernel_size=3, padding=1)
+        self.maxpool2 = mint.nn.MaxPool2d(kernel_size=3, stride=2, ceil_mode=True)
 
         self.inception3a = Inception(192, 64, 96, 128, 16, 32, 32)
         self.inception3b = Inception(256, 128, 128, 192, 32, 96, 64)
-        self.maxpool3 = nn.MaxPool2d(kernel_size=3, stride=2, pad_mode="same")
+        self.maxpool3 = mint.nn.MaxPool2d(kernel_size=3, stride=2, ceil_mode=True)
 
         self.inception4a = Inception(480, 192, 96, 208, 16, 48, 64)
         self.inception4b = Inception(512, 160, 112, 224, 24, 64, 64)
         self.inception4c = Inception(512, 128, 128, 256, 24, 64, 64)
         self.inception4d = Inception(512, 112, 144, 288, 32, 64, 64)
         self.inception4e = Inception(528, 256, 160, 320, 32, 128, 128)
-        self.maxpool4 = nn.MaxPool2d(kernel_size=2, stride=2, pad_mode="same")
+        self.maxpool4 = mint.nn.MaxPool2d(kernel_size=2, stride=2, ceil_mode=True)
 
         self.inception5a = Inception(832, 256, 160, 320, 32, 128, 128)
         self.inception5b = Inception(832, 384, 192, 384, 48, 128, 128)
@@ -171,22 +171,24 @@ def __init__(
             self.aux2 = InceptionAux(528, num_classes, drop_rate=drop_rate_aux)
 
         self.pool = GlobalAvgPooling()
-        self.dropout = Dropout(p=drop_rate)
-        self.classifier = nn.Dense(1024, num_classes)
+        self.dropout = mint.nn.Dropout(p=drop_rate)
+        self.classifier = mint.nn.Linear(1024, num_classes)
         self._initialize_weights()
 
     def _initialize_weights(self):
         for _, cell in self.cells_and_names():
-            if isinstance(cell, nn.Conv2d):
+            if isinstance(cell, mint.nn.Conv2d):
                 cell.weight.set_data(init.initializer(init.HeNormal(0, mode='fan_in', nonlinearity='leaky_relu'),
                                                       cell.weight.shape, cell.weight.dtype))
                 if cell.bias is not None:
                     cell.bias.set_data(init.initializer(init.Constant(0), cell.bias.shape, cell.bias.dtype))
-            elif isinstance(cell, nn.BatchNorm2d) or isinstance(cell, nn.BatchNorm1d):
-                cell.gamma.set_data(init.initializer(init.Constant(1), cell.gamma.shape, cell.gamma.dtype))
-                if cell.beta is not None:
-                    cell.beta.set_data(init.initializer(init.Constant(0), cell.beta.shape, cell.gamma.dtype))
-            elif isinstance(cell, nn.Dense):
+            elif isinstance(cell, mint.nn.BatchNorm2d) or isinstance(cell, mint.nn.BatchNorm1d):
+                cell.weight.set_data(
+                    init.initializer(init.Constant(1), cell.weight.shape, cell.weight.dtype))
+                if cell.bias is not None:
+                    cell.bias.set_data(
+                        init.initializer(init.Constant(0), cell.bias.shape, cell.weight.dtype))
+            elif isinstance(cell, mint.nn.Linear):
                 cell.weight.set_data(
                     init.initializer(init.HeUniform(math.sqrt(5), mode='fan_in', nonlinearity='leaky_relu'),
                                      cell.weight.shape, cell.weight.dtype))