Перейти к содержимому

Ссылка для ultralytics/models/sam/modules/tiny_encoder.py

Примечание

Этот файл доступен по адресу https://github.com/ultralytics/ ultralytics/blob/main/ ultralytics/models/ sam/modules/tiny_encoder .py. Если ты обнаружишь проблему, пожалуйста, помоги исправить ее, отправив Pull Request 🛠️. Спасибо 🙏!



ultralytics.models.sam.modules.tiny_encoder.Conv2d_BN

Базы: Sequential

Последовательный контейнер, выполняющий двумерную свертку с последующей пакетной нормализацией.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
class Conv2d_BN(torch.nn.Sequential):
    """A sequential container that performs 2D convolution followed by batch normalization."""

    def __init__(self, a, b, ks=1, stride=1, pad=0, dilation=1, groups=1, bn_weight_init=1):
        """Initializes the MBConv model with given input channels, output channels, expansion ratio, activation, and
        drop path.
        """
        super().__init__()
        self.add_module("c", torch.nn.Conv2d(a, b, ks, stride, pad, dilation, groups, bias=False))
        bn = torch.nn.BatchNorm2d(b)
        torch.nn.init.constant_(bn.weight, bn_weight_init)
        torch.nn.init.constant_(bn.bias, 0)
        self.add_module("bn", bn)

__init__(a, b, ks=1, stride=1, pad=0, dilation=1, groups=1, bn_weight_init=1)

Инициализирует модель MBConv с заданными входными каналами, выходными каналами, коэффициентом расширения, активацией и путь падения.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def __init__(self, a, b, ks=1, stride=1, pad=0, dilation=1, groups=1, bn_weight_init=1):
    """Initializes the MBConv model with given input channels, output channels, expansion ratio, activation, and
    drop path.
    """
    super().__init__()
    self.add_module("c", torch.nn.Conv2d(a, b, ks, stride, pad, dilation, groups, bias=False))
    bn = torch.nn.BatchNorm2d(b)
    torch.nn.init.constant_(bn.weight, bn_weight_init)
    torch.nn.init.constant_(bn.bias, 0)
    self.add_module("bn", bn)



ultralytics.models.sam.modules.tiny_encoder.PatchEmbed

Базы: Module

Встраивай изображения в патчи и проецируй их в заданное измерение встраивания.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
class PatchEmbed(nn.Module):
    """Embeds images into patches and projects them into a specified embedding dimension."""

    def __init__(self, in_chans, embed_dim, resolution, activation):
        """Initialize the PatchMerging class with specified input, output dimensions, resolution and activation
        function.
        """
        super().__init__()
        img_size: Tuple[int, int] = to_2tuple(resolution)
        self.patches_resolution = (img_size[0] // 4, img_size[1] // 4)
        self.num_patches = self.patches_resolution[0] * self.patches_resolution[1]
        self.in_chans = in_chans
        self.embed_dim = embed_dim
        n = embed_dim
        self.seq = nn.Sequential(
            Conv2d_BN(in_chans, n // 2, 3, 2, 1),
            activation(),
            Conv2d_BN(n // 2, n, 3, 2, 1),
        )

    def forward(self, x):
        """Runs input tensor 'x' through the PatchMerging model's sequence of operations."""
        return self.seq(x)

__init__(in_chans, embed_dim, resolution, activation)

Инициализируй класс PatchMerging с указанными размерами входа, выхода, разрешением и функцией активации. Функция.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def __init__(self, in_chans, embed_dim, resolution, activation):
    """Initialize the PatchMerging class with specified input, output dimensions, resolution and activation
    function.
    """
    super().__init__()
    img_size: Tuple[int, int] = to_2tuple(resolution)
    self.patches_resolution = (img_size[0] // 4, img_size[1] // 4)
    self.num_patches = self.patches_resolution[0] * self.patches_resolution[1]
    self.in_chans = in_chans
    self.embed_dim = embed_dim
    n = embed_dim
    self.seq = nn.Sequential(
        Conv2d_BN(in_chans, n // 2, 3, 2, 1),
        activation(),
        Conv2d_BN(n // 2, n, 3, 2, 1),
    )

forward(x)

Прогони входной tensor 'x' через последовательность операций модели PatchMerging.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def forward(self, x):
    """Runs input tensor 'x' through the PatchMerging model's sequence of operations."""
    return self.seq(x)



ultralytics.models.sam.modules.tiny_encoder.MBConv

Базы: Module

Слой Mobile Inverted Bottleneck Conv (MBConv), часть архитектуры EfficientNet.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
class MBConv(nn.Module):
    """Mobile Inverted Bottleneck Conv (MBConv) layer, part of the EfficientNet architecture."""

    def __init__(self, in_chans, out_chans, expand_ratio, activation, drop_path):
        """Initializes a convolutional layer with specified dimensions, input resolution, depth, and activation
        function.
        """
        super().__init__()
        self.in_chans = in_chans
        self.hidden_chans = int(in_chans * expand_ratio)
        self.out_chans = out_chans

        self.conv1 = Conv2d_BN(in_chans, self.hidden_chans, ks=1)
        self.act1 = activation()

        self.conv2 = Conv2d_BN(self.hidden_chans, self.hidden_chans, ks=3, stride=1, pad=1, groups=self.hidden_chans)
        self.act2 = activation()

        self.conv3 = Conv2d_BN(self.hidden_chans, out_chans, ks=1, bn_weight_init=0.0)
        self.act3 = activation()

        # NOTE: `DropPath` is needed only for training.
        # self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
        self.drop_path = nn.Identity()

    def forward(self, x):
        """Implements the forward pass for the model architecture."""
        shortcut = x
        x = self.conv1(x)
        x = self.act1(x)
        x = self.conv2(x)
        x = self.act2(x)
        x = self.conv3(x)
        x = self.drop_path(x)
        x += shortcut
        return self.act3(x)

__init__(in_chans, out_chans, expand_ratio, activation, drop_path)

Инициализируй конволюционный слой с заданными размерами, входным разрешением, глубиной и активационной функцией. Функция.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def __init__(self, in_chans, out_chans, expand_ratio, activation, drop_path):
    """Initializes a convolutional layer with specified dimensions, input resolution, depth, and activation
    function.
    """
    super().__init__()
    self.in_chans = in_chans
    self.hidden_chans = int(in_chans * expand_ratio)
    self.out_chans = out_chans

    self.conv1 = Conv2d_BN(in_chans, self.hidden_chans, ks=1)
    self.act1 = activation()

    self.conv2 = Conv2d_BN(self.hidden_chans, self.hidden_chans, ks=3, stride=1, pad=1, groups=self.hidden_chans)
    self.act2 = activation()

    self.conv3 = Conv2d_BN(self.hidden_chans, out_chans, ks=1, bn_weight_init=0.0)
    self.act3 = activation()

    # NOTE: `DropPath` is needed only for training.
    # self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
    self.drop_path = nn.Identity()

forward(x)

Реализуй прямой проход для архитектуры модели.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def forward(self, x):
    """Implements the forward pass for the model architecture."""
    shortcut = x
    x = self.conv1(x)
    x = self.act1(x)
    x = self.conv2(x)
    x = self.act2(x)
    x = self.conv3(x)
    x = self.drop_path(x)
    x += shortcut
    return self.act3(x)



ultralytics.models.sam.modules.tiny_encoder.PatchMerging

Базы: Module

Объединяет соседние патчи в карте характеристик и проецирует на новое измерение.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
class PatchMerging(nn.Module):
    """Merges neighboring patches in the feature map and projects to a new dimension."""

    def __init__(self, input_resolution, dim, out_dim, activation):
        """Initializes the ConvLayer with specific dimension, input resolution, depth, activation, drop path, and other
        optional parameters.
        """
        super().__init__()

        self.input_resolution = input_resolution
        self.dim = dim
        self.out_dim = out_dim
        self.act = activation()
        self.conv1 = Conv2d_BN(dim, out_dim, 1, 1, 0)
        stride_c = 1 if out_dim in {320, 448, 576} else 2
        self.conv2 = Conv2d_BN(out_dim, out_dim, 3, stride_c, 1, groups=out_dim)
        self.conv3 = Conv2d_BN(out_dim, out_dim, 1, 1, 0)

    def forward(self, x):
        """Applies forward pass on the input utilizing convolution and activation layers, and returns the result."""
        if x.ndim == 3:
            H, W = self.input_resolution
            B = len(x)
            # (B, C, H, W)
            x = x.view(B, H, W, -1).permute(0, 3, 1, 2)

        x = self.conv1(x)
        x = self.act(x)

        x = self.conv2(x)
        x = self.act(x)
        x = self.conv3(x)
        return x.flatten(2).transpose(1, 2)

__init__(input_resolution, dim, out_dim, activation)

Инициализирует ConvLayer с определенной размерностью, входным разрешением, глубиной, активацией, траекторией падения и другими необязательными параметрами.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def __init__(self, input_resolution, dim, out_dim, activation):
    """Initializes the ConvLayer with specific dimension, input resolution, depth, activation, drop path, and other
    optional parameters.
    """
    super().__init__()

    self.input_resolution = input_resolution
    self.dim = dim
    self.out_dim = out_dim
    self.act = activation()
    self.conv1 = Conv2d_BN(dim, out_dim, 1, 1, 0)
    stride_c = 1 if out_dim in {320, 448, 576} else 2
    self.conv2 = Conv2d_BN(out_dim, out_dim, 3, stride_c, 1, groups=out_dim)
    self.conv3 = Conv2d_BN(out_dim, out_dim, 1, 1, 0)

forward(x)

Применяй прямое прохождение на входе, используя слои свертки и активации, и возвращай результат.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def forward(self, x):
    """Applies forward pass on the input utilizing convolution and activation layers, and returns the result."""
    if x.ndim == 3:
        H, W = self.input_resolution
        B = len(x)
        # (B, C, H, W)
        x = x.view(B, H, W, -1).permute(0, 3, 1, 2)

    x = self.conv1(x)
    x = self.act(x)

    x = self.conv2(x)
    x = self.act(x)
    x = self.conv3(x)
    return x.flatten(2).transpose(1, 2)



ultralytics.models.sam.modules.tiny_encoder.ConvLayer

Базы: Module

Конволюционный слой с несколькими инвертированными свертками в стиле MobileNetV3 (MBConv).

Опционально применяй операции понижающей дискретизации к выходу и обеспечивай поддержку градиентной контрольной точки.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
class ConvLayer(nn.Module):
    """
    Convolutional Layer featuring multiple MobileNetV3-style inverted bottleneck convolutions (MBConv).

    Optionally applies downsample operations to the output, and provides support for gradient checkpointing.
    """

    def __init__(
        self,
        dim,
        input_resolution,
        depth,
        activation,
        drop_path=0.0,
        downsample=None,
        use_checkpoint=False,
        out_dim=None,
        conv_expand_ratio=4.0,
    ):
        """
        Initializes the ConvLayer with the given dimensions and settings.

        Args:
            dim (int): The dimensionality of the input and output.
            input_resolution (Tuple[int, int]): The resolution of the input image.
            depth (int): The number of MBConv layers in the block.
            activation (Callable): Activation function applied after each convolution.
            drop_path (Union[float, List[float]]): Drop path rate. Single float or a list of floats for each MBConv.
            downsample (Optional[Callable]): Function for downsampling the output. None to skip downsampling.
            use_checkpoint (bool): Whether to use gradient checkpointing to save memory.
            out_dim (Optional[int]): The dimensionality of the output. None means it will be the same as `dim`.
            conv_expand_ratio (float): Expansion ratio for the MBConv layers.
        """
        super().__init__()
        self.dim = dim
        self.input_resolution = input_resolution
        self.depth = depth
        self.use_checkpoint = use_checkpoint

        # Build blocks
        self.blocks = nn.ModuleList(
            [
                MBConv(
                    dim,
                    dim,
                    conv_expand_ratio,
                    activation,
                    drop_path[i] if isinstance(drop_path, list) else drop_path,
                )
                for i in range(depth)
            ]
        )

        # Patch merging layer
        self.downsample = (
            None
            if downsample is None
            else downsample(input_resolution, dim=dim, out_dim=out_dim, activation=activation)
        )

    def forward(self, x):
        """Processes the input through a series of convolutional layers and returns the activated output."""
        for blk in self.blocks:
            x = checkpoint.checkpoint(blk, x) if self.use_checkpoint else blk(x)
        return x if self.downsample is None else self.downsample(x)

__init__(dim, input_resolution, depth, activation, drop_path=0.0, downsample=None, use_checkpoint=False, out_dim=None, conv_expand_ratio=4.0)

Инициализирует ConvLayer с заданными размерами и настройками.

Параметры:

Имя Тип Описание По умолчанию
dim int

Размерность входных и выходных данных.

требуется
input_resolution Tuple[int, int]

Разрешение входного изображения.

требуется
depth int

Количество слоев MBConv в блоке.

требуется
activation Callable

Функция активации, применяемая после каждой свертки.

требуется
drop_path Union[float, List[float]]

Скорость прохождения траектории падения. Одно плавающее число или список плавающих чисел для каждого MBConv.

0.0
downsample Optional[Callable]

Функция для понижения дискретизации выходного сигнала. None, чтобы пропустить даунсемплинг.

None
use_checkpoint bool

Использовать ли градиентную контрольную точку для экономии памяти.

False
out_dim Optional[int]

Размерность выходных данных. Нет означает, что он будет таким же, как dim.

None
conv_expand_ratio float

Коэффициент расширения для слоев MBConv.

4.0
Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def __init__(
    self,
    dim,
    input_resolution,
    depth,
    activation,
    drop_path=0.0,
    downsample=None,
    use_checkpoint=False,
    out_dim=None,
    conv_expand_ratio=4.0,
):
    """
    Initializes the ConvLayer with the given dimensions and settings.

    Args:
        dim (int): The dimensionality of the input and output.
        input_resolution (Tuple[int, int]): The resolution of the input image.
        depth (int): The number of MBConv layers in the block.
        activation (Callable): Activation function applied after each convolution.
        drop_path (Union[float, List[float]]): Drop path rate. Single float or a list of floats for each MBConv.
        downsample (Optional[Callable]): Function for downsampling the output. None to skip downsampling.
        use_checkpoint (bool): Whether to use gradient checkpointing to save memory.
        out_dim (Optional[int]): The dimensionality of the output. None means it will be the same as `dim`.
        conv_expand_ratio (float): Expansion ratio for the MBConv layers.
    """
    super().__init__()
    self.dim = dim
    self.input_resolution = input_resolution
    self.depth = depth
    self.use_checkpoint = use_checkpoint

    # Build blocks
    self.blocks = nn.ModuleList(
        [
            MBConv(
                dim,
                dim,
                conv_expand_ratio,
                activation,
                drop_path[i] if isinstance(drop_path, list) else drop_path,
            )
            for i in range(depth)
        ]
    )

    # Patch merging layer
    self.downsample = (
        None
        if downsample is None
        else downsample(input_resolution, dim=dim, out_dim=out_dim, activation=activation)
    )

forward(x)

Обрабатывает входные данные через ряд конволюционных слоев и возвращает активированный выход.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def forward(self, x):
    """Processes the input through a series of convolutional layers and returns the activated output."""
    for blk in self.blocks:
        x = checkpoint.checkpoint(blk, x) if self.use_checkpoint else blk(x)
    return x if self.downsample is None else self.downsample(x)



ultralytics.models.sam.modules.tiny_encoder.Mlp

Базы: Module

Многослойный перцептрон (MLP) для трансформаторных архитектур.

Этот слой принимает входной сигнал с in_features, применяет нормализацию слоев и два полностью связанных слоя.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
class Mlp(nn.Module):
    """
    Multi-layer Perceptron (MLP) for transformer architectures.

    This layer takes an input with in_features, applies layer normalization and two fully-connected layers.
    """

    def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.0):
        """Initializes Attention module with the given parameters including dimension, key_dim, number of heads, etc."""
        super().__init__()
        out_features = out_features or in_features
        hidden_features = hidden_features or in_features
        self.norm = nn.LayerNorm(in_features)
        self.fc1 = nn.Linear(in_features, hidden_features)
        self.fc2 = nn.Linear(hidden_features, out_features)
        self.act = act_layer()
        self.drop = nn.Dropout(drop)

    def forward(self, x):
        """Applies operations on input x and returns modified x, runs downsample if not None."""
        x = self.norm(x)
        x = self.fc1(x)
        x = self.act(x)
        x = self.drop(x)
        x = self.fc2(x)
        return self.drop(x)

__init__(in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.0)

Инициализирует модуль Attention с заданными параметрами, включая dimension, key_dim, количество головок и т.д.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.0):
    """Initializes Attention module with the given parameters including dimension, key_dim, number of heads, etc."""
    super().__init__()
    out_features = out_features or in_features
    hidden_features = hidden_features or in_features
    self.norm = nn.LayerNorm(in_features)
    self.fc1 = nn.Linear(in_features, hidden_features)
    self.fc2 = nn.Linear(hidden_features, out_features)
    self.act = act_layer()
    self.drop = nn.Dropout(drop)

forward(x)

Применяет операции над входным x и возвращает измененный x, запускает downsample, если не None.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def forward(self, x):
    """Applies operations on input x and returns modified x, runs downsample if not None."""
    x = self.norm(x)
    x = self.fc1(x)
    x = self.act(x)
    x = self.drop(x)
    x = self.fc2(x)
    return self.drop(x)



ultralytics.models.sam.modules.tiny_encoder.Attention

Базы: Module

Многоголовый модуль внимания с поддержкой пространственного осознания, применяющий предубеждения внимания, основанные на пространственном Разрешение. Реализует обучаемые смещения внимания для каждого уникального смещения между пространственными позициями в сетке разрешения сетки.

Атрибуты:

Имя Тип Описание
ab Tensor

Кэшированные предубеждения внимания для умозаключений, удаленные во время тренировки.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
class Attention(torch.nn.Module):
    """
    Multi-head attention module with support for spatial awareness, applying attention biases based on spatial
    resolution. Implements trainable attention biases for each unique offset between spatial positions in the resolution
    grid.

    Attributes:
        ab (Tensor, optional): Cached attention biases for inference, deleted during training.
    """

    def __init__(
        self,
        dim,
        key_dim,
        num_heads=8,
        attn_ratio=4,
        resolution=(14, 14),
    ):
        """
        Initializes the Attention module.

        Args:
            dim (int): The dimensionality of the input and output.
            key_dim (int): The dimensionality of the keys and queries.
            num_heads (int, optional): Number of attention heads. Default is 8.
            attn_ratio (float, optional): Attention ratio, affecting the dimensions of the value vectors. Default is 4.
            resolution (Tuple[int, int], optional): Spatial resolution of the input feature map. Default is (14, 14).

        Raises:
            AssertionError: If `resolution` is not a tuple of length 2.
        """
        super().__init__()

        assert isinstance(resolution, tuple) and len(resolution) == 2
        self.num_heads = num_heads
        self.scale = key_dim**-0.5
        self.key_dim = key_dim
        self.nh_kd = nh_kd = key_dim * num_heads
        self.d = int(attn_ratio * key_dim)
        self.dh = int(attn_ratio * key_dim) * num_heads
        self.attn_ratio = attn_ratio
        h = self.dh + nh_kd * 2

        self.norm = nn.LayerNorm(dim)
        self.qkv = nn.Linear(dim, h)
        self.proj = nn.Linear(self.dh, dim)

        points = list(itertools.product(range(resolution[0]), range(resolution[1])))
        N = len(points)
        attention_offsets = {}
        idxs = []
        for p1 in points:
            for p2 in points:
                offset = (abs(p1[0] - p2[0]), abs(p1[1] - p2[1]))
                if offset not in attention_offsets:
                    attention_offsets[offset] = len(attention_offsets)
                idxs.append(attention_offsets[offset])
        self.attention_biases = torch.nn.Parameter(torch.zeros(num_heads, len(attention_offsets)))
        self.register_buffer("attention_bias_idxs", torch.LongTensor(idxs).view(N, N), persistent=False)

    @torch.no_grad()
    def train(self, mode=True):
        """Sets the module in training mode and handles attribute 'ab' based on the mode."""
        super().train(mode)
        if mode and hasattr(self, "ab"):
            del self.ab
        else:
            self.ab = self.attention_biases[:, self.attention_bias_idxs]

    def forward(self, x):  # x
        """Performs forward pass over the input tensor 'x' by applying normalization and querying keys/values."""
        B, N, _ = x.shape  # B, N, C

        # Normalization
        x = self.norm(x)

        qkv = self.qkv(x)
        # (B, N, num_heads, d)
        q, k, v = qkv.view(B, N, self.num_heads, -1).split([self.key_dim, self.key_dim, self.d], dim=3)
        # (B, num_heads, N, d)
        q = q.permute(0, 2, 1, 3)
        k = k.permute(0, 2, 1, 3)
        v = v.permute(0, 2, 1, 3)
        self.ab = self.ab.to(self.attention_biases.device)

        attn = (q @ k.transpose(-2, -1)) * self.scale + (
            self.attention_biases[:, self.attention_bias_idxs] if self.training else self.ab
        )
        attn = attn.softmax(dim=-1)
        x = (attn @ v).transpose(1, 2).reshape(B, N, self.dh)
        return self.proj(x)

__init__(dim, key_dim, num_heads=8, attn_ratio=4, resolution=(14, 14))

Инициализирует модуль "Внимание".

Параметры:

Имя Тип Описание По умолчанию
dim int

Размерность входных и выходных данных.

требуется
key_dim int

Размерность ключей и запросов.

требуется
num_heads int

Количество головок внимания. По умолчанию это 8.

8
attn_ratio float

Коэффициент внимания, влияющий на размерность векторов значений. По умолчанию это 4.

4
resolution Tuple[int, int]

Пространственное разрешение входной карты признаков. По умолчанию это (14, 14).

(14, 14)

Поднимает:

Тип Описание
AssertionError

Если resolution не является кортежем длины 2.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def __init__(
    self,
    dim,
    key_dim,
    num_heads=8,
    attn_ratio=4,
    resolution=(14, 14),
):
    """
    Initializes the Attention module.

    Args:
        dim (int): The dimensionality of the input and output.
        key_dim (int): The dimensionality of the keys and queries.
        num_heads (int, optional): Number of attention heads. Default is 8.
        attn_ratio (float, optional): Attention ratio, affecting the dimensions of the value vectors. Default is 4.
        resolution (Tuple[int, int], optional): Spatial resolution of the input feature map. Default is (14, 14).

    Raises:
        AssertionError: If `resolution` is not a tuple of length 2.
    """
    super().__init__()

    assert isinstance(resolution, tuple) and len(resolution) == 2
    self.num_heads = num_heads
    self.scale = key_dim**-0.5
    self.key_dim = key_dim
    self.nh_kd = nh_kd = key_dim * num_heads
    self.d = int(attn_ratio * key_dim)
    self.dh = int(attn_ratio * key_dim) * num_heads
    self.attn_ratio = attn_ratio
    h = self.dh + nh_kd * 2

    self.norm = nn.LayerNorm(dim)
    self.qkv = nn.Linear(dim, h)
    self.proj = nn.Linear(self.dh, dim)

    points = list(itertools.product(range(resolution[0]), range(resolution[1])))
    N = len(points)
    attention_offsets = {}
    idxs = []
    for p1 in points:
        for p2 in points:
            offset = (abs(p1[0] - p2[0]), abs(p1[1] - p2[1]))
            if offset not in attention_offsets:
                attention_offsets[offset] = len(attention_offsets)
            idxs.append(attention_offsets[offset])
    self.attention_biases = torch.nn.Parameter(torch.zeros(num_heads, len(attention_offsets)))
    self.register_buffer("attention_bias_idxs", torch.LongTensor(idxs).view(N, N), persistent=False)

forward(x)

Выполняет прямой проход по входному tensor 'x', применяя нормализацию и запрашивая ключи/значения.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def forward(self, x):  # x
    """Performs forward pass over the input tensor 'x' by applying normalization and querying keys/values."""
    B, N, _ = x.shape  # B, N, C

    # Normalization
    x = self.norm(x)

    qkv = self.qkv(x)
    # (B, N, num_heads, d)
    q, k, v = qkv.view(B, N, self.num_heads, -1).split([self.key_dim, self.key_dim, self.d], dim=3)
    # (B, num_heads, N, d)
    q = q.permute(0, 2, 1, 3)
    k = k.permute(0, 2, 1, 3)
    v = v.permute(0, 2, 1, 3)
    self.ab = self.ab.to(self.attention_biases.device)

    attn = (q @ k.transpose(-2, -1)) * self.scale + (
        self.attention_biases[:, self.attention_bias_idxs] if self.training else self.ab
    )
    attn = attn.softmax(dim=-1)
    x = (attn @ v).transpose(1, 2).reshape(B, N, self.dh)
    return self.proj(x)

train(mode=True)

Переводит модуль в режим тренировки и обрабатывает атрибут 'ab' в зависимости от режима.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
@torch.no_grad()
def train(self, mode=True):
    """Sets the module in training mode and handles attribute 'ab' based on the mode."""
    super().train(mode)
    if mode and hasattr(self, "ab"):
        del self.ab
    else:
        self.ab = self.attention_biases[:, self.attention_bias_idxs]



ultralytics.models.sam.modules.tiny_encoder.TinyViTBlock

Базы: Module

Блок TinyViT, который применяет самовнимание и локальную свертку к входным данным.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
class TinyViTBlock(nn.Module):
    """TinyViT Block that applies self-attention and a local convolution to the input."""

    def __init__(
        self,
        dim,
        input_resolution,
        num_heads,
        window_size=7,
        mlp_ratio=4.0,
        drop=0.0,
        drop_path=0.0,
        local_conv_size=3,
        activation=nn.GELU,
    ):
        """
        Initializes the TinyViTBlock.

        Args:
            dim (int): The dimensionality of the input and output.
            input_resolution (Tuple[int, int]): Spatial resolution of the input feature map.
            num_heads (int): Number of attention heads.
            window_size (int, optional): Window size for attention. Default is 7.
            mlp_ratio (float, optional): Ratio of mlp hidden dim to embedding dim. Default is 4.
            drop (float, optional): Dropout rate. Default is 0.
            drop_path (float, optional): Stochastic depth rate. Default is 0.
            local_conv_size (int, optional): The kernel size of the local convolution. Default is 3.
            activation (torch.nn, optional): Activation function for MLP. Default is nn.GELU.

        Raises:
            AssertionError: If `window_size` is not greater than 0.
            AssertionError: If `dim` is not divisible by `num_heads`.
        """
        super().__init__()
        self.dim = dim
        self.input_resolution = input_resolution
        self.num_heads = num_heads
        assert window_size > 0, "window_size must be greater than 0"
        self.window_size = window_size
        self.mlp_ratio = mlp_ratio

        # NOTE: `DropPath` is needed only for training.
        # self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
        self.drop_path = nn.Identity()

        assert dim % num_heads == 0, "dim must be divisible by num_heads"
        head_dim = dim // num_heads

        window_resolution = (window_size, window_size)
        self.attn = Attention(dim, head_dim, num_heads, attn_ratio=1, resolution=window_resolution)

        mlp_hidden_dim = int(dim * mlp_ratio)
        mlp_activation = activation
        self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=mlp_activation, drop=drop)

        pad = local_conv_size // 2
        self.local_conv = Conv2d_BN(dim, dim, ks=local_conv_size, stride=1, pad=pad, groups=dim)

    def forward(self, x):
        """Applies attention-based transformation or padding to input 'x' before passing it through a local
        convolution.
        """
        h, w = self.input_resolution
        b, hw, c = x.shape  # batch, height*width, channels
        assert hw == h * w, "input feature has wrong size"
        res_x = x
        if h == self.window_size and w == self.window_size:
            x = self.attn(x)
        else:
            x = x.view(b, h, w, c)
            pad_b = (self.window_size - h % self.window_size) % self.window_size
            pad_r = (self.window_size - w % self.window_size) % self.window_size
            padding = pad_b > 0 or pad_r > 0
            if padding:
                x = F.pad(x, (0, 0, 0, pad_r, 0, pad_b))

            pH, pW = h + pad_b, w + pad_r
            nH = pH // self.window_size
            nW = pW // self.window_size

            # Window partition
            x = (
                x.view(b, nH, self.window_size, nW, self.window_size, c)
                .transpose(2, 3)
                .reshape(b * nH * nW, self.window_size * self.window_size, c)
            )
            x = self.attn(x)

            # Window reverse
            x = x.view(b, nH, nW, self.window_size, self.window_size, c).transpose(2, 3).reshape(b, pH, pW, c)
            if padding:
                x = x[:, :h, :w].contiguous()

            x = x.view(b, hw, c)

        x = res_x + self.drop_path(x)
        x = x.transpose(1, 2).reshape(b, c, h, w)
        x = self.local_conv(x)
        x = x.view(b, c, hw).transpose(1, 2)

        return x + self.drop_path(self.mlp(x))

    def extra_repr(self) -> str:
        """Returns a formatted string representing the TinyViTBlock's parameters: dimension, input resolution, number of
        attentions heads, window size, and MLP ratio.
        """
        return (
            f"dim={self.dim}, input_resolution={self.input_resolution}, num_heads={self.num_heads}, "
            f"window_size={self.window_size}, mlp_ratio={self.mlp_ratio}"
        )

__init__(dim, input_resolution, num_heads, window_size=7, mlp_ratio=4.0, drop=0.0, drop_path=0.0, local_conv_size=3, activation=nn.GELU)

Инициализирует TinyViTBlock.

Параметры:

Имя Тип Описание По умолчанию
dim int

Размерность входных и выходных данных.

требуется
input_resolution Tuple[int, int]

Пространственное разрешение входной карты признаков.

требуется
num_heads int

Количество головок внимания.

требуется
window_size int

Размер окна для привлечения внимания. По умолчанию это 7.

7
mlp_ratio float

Отношение mlp hidden dim к embedding dim. По умолчанию - 4.

4.0
drop float

Уровень отсева. По умолчанию - 0.

0.0
drop_path float

Стохастический показатель глубины. По умолчанию - 0.

0.0
local_conv_size int

Размер ядра локальной свертки. По умолчанию это 3.

3
activation nn

Функция активации для MLP. По умолчанию это nn.GELU.

GELU

Поднимает:

Тип Описание
AssertionError

Если window_size не больше 0.

AssertionError

Если dim не делится на num_heads.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def __init__(
    self,
    dim,
    input_resolution,
    num_heads,
    window_size=7,
    mlp_ratio=4.0,
    drop=0.0,
    drop_path=0.0,
    local_conv_size=3,
    activation=nn.GELU,
):
    """
    Initializes the TinyViTBlock.

    Args:
        dim (int): The dimensionality of the input and output.
        input_resolution (Tuple[int, int]): Spatial resolution of the input feature map.
        num_heads (int): Number of attention heads.
        window_size (int, optional): Window size for attention. Default is 7.
        mlp_ratio (float, optional): Ratio of mlp hidden dim to embedding dim. Default is 4.
        drop (float, optional): Dropout rate. Default is 0.
        drop_path (float, optional): Stochastic depth rate. Default is 0.
        local_conv_size (int, optional): The kernel size of the local convolution. Default is 3.
        activation (torch.nn, optional): Activation function for MLP. Default is nn.GELU.

    Raises:
        AssertionError: If `window_size` is not greater than 0.
        AssertionError: If `dim` is not divisible by `num_heads`.
    """
    super().__init__()
    self.dim = dim
    self.input_resolution = input_resolution
    self.num_heads = num_heads
    assert window_size > 0, "window_size must be greater than 0"
    self.window_size = window_size
    self.mlp_ratio = mlp_ratio

    # NOTE: `DropPath` is needed only for training.
    # self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
    self.drop_path = nn.Identity()

    assert dim % num_heads == 0, "dim must be divisible by num_heads"
    head_dim = dim // num_heads

    window_resolution = (window_size, window_size)
    self.attn = Attention(dim, head_dim, num_heads, attn_ratio=1, resolution=window_resolution)

    mlp_hidden_dim = int(dim * mlp_ratio)
    mlp_activation = activation
    self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=mlp_activation, drop=drop)

    pad = local_conv_size // 2
    self.local_conv = Conv2d_BN(dim, dim, ks=local_conv_size, stride=1, pad=pad, groups=dim)

extra_repr()

Возвращает форматированную строку, представляющую параметры TinyViTBlock: размерность, разрешение входа, количество головок внимания, размер окна и коэффициент MLP.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def extra_repr(self) -> str:
    """Returns a formatted string representing the TinyViTBlock's parameters: dimension, input resolution, number of
    attentions heads, window size, and MLP ratio.
    """
    return (
        f"dim={self.dim}, input_resolution={self.input_resolution}, num_heads={self.num_heads}, "
        f"window_size={self.window_size}, mlp_ratio={self.mlp_ratio}"
    )

forward(x)

Примени преобразование или набивку на основе внимания к входному сигналу 'x', прежде чем пропустить его через локальную свертку.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def forward(self, x):
    """Applies attention-based transformation or padding to input 'x' before passing it through a local
    convolution.
    """
    h, w = self.input_resolution
    b, hw, c = x.shape  # batch, height*width, channels
    assert hw == h * w, "input feature has wrong size"
    res_x = x
    if h == self.window_size and w == self.window_size:
        x = self.attn(x)
    else:
        x = x.view(b, h, w, c)
        pad_b = (self.window_size - h % self.window_size) % self.window_size
        pad_r = (self.window_size - w % self.window_size) % self.window_size
        padding = pad_b > 0 or pad_r > 0
        if padding:
            x = F.pad(x, (0, 0, 0, pad_r, 0, pad_b))

        pH, pW = h + pad_b, w + pad_r
        nH = pH // self.window_size
        nW = pW // self.window_size

        # Window partition
        x = (
            x.view(b, nH, self.window_size, nW, self.window_size, c)
            .transpose(2, 3)
            .reshape(b * nH * nW, self.window_size * self.window_size, c)
        )
        x = self.attn(x)

        # Window reverse
        x = x.view(b, nH, nW, self.window_size, self.window_size, c).transpose(2, 3).reshape(b, pH, pW, c)
        if padding:
            x = x[:, :h, :w].contiguous()

        x = x.view(b, hw, c)

    x = res_x + self.drop_path(x)
    x = x.transpose(1, 2).reshape(b, c, h, w)
    x = self.local_conv(x)
    x = x.view(b, c, hw).transpose(1, 2)

    return x + self.drop_path(self.mlp(x))



ultralytics.models.sam.modules.tiny_encoder.BasicLayer

Базы: Module

Базовый слой TinyViT для одного этапа в архитектуре TinyViT.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
class BasicLayer(nn.Module):
    """A basic TinyViT layer for one stage in a TinyViT architecture."""

    def __init__(
        self,
        dim,
        input_resolution,
        depth,
        num_heads,
        window_size,
        mlp_ratio=4.0,
        drop=0.0,
        drop_path=0.0,
        downsample=None,
        use_checkpoint=False,
        local_conv_size=3,
        activation=nn.GELU,
        out_dim=None,
    ):
        """
        Initializes the BasicLayer.

        Args:
            dim (int): The dimensionality of the input and output.
            input_resolution (Tuple[int, int]): Spatial resolution of the input feature map.
            depth (int): Number of TinyViT blocks.
            num_heads (int): Number of attention heads.
            window_size (int): Local window size.
            mlp_ratio (float, optional): Ratio of mlp hidden dim to embedding dim. Default is 4.
            drop (float, optional): Dropout rate. Default is 0.
            drop_path (float | tuple[float], optional): Stochastic depth rate. Default is 0.
            downsample (nn.Module | None, optional): Downsample layer at the end of the layer. Default is None.
            use_checkpoint (bool, optional): Whether to use checkpointing to save memory. Default is False.
            local_conv_size (int, optional): Kernel size of the local convolution. Default is 3.
            activation (torch.nn, optional): Activation function for MLP. Default is nn.GELU.
            out_dim (int | None, optional): The output dimension of the layer. Default is None.

        Raises:
            ValueError: If `drop_path` is a list of float but its length doesn't match `depth`.
        """
        super().__init__()
        self.dim = dim
        self.input_resolution = input_resolution
        self.depth = depth
        self.use_checkpoint = use_checkpoint

        # Build blocks
        self.blocks = nn.ModuleList(
            [
                TinyViTBlock(
                    dim=dim,
                    input_resolution=input_resolution,
                    num_heads=num_heads,
                    window_size=window_size,
                    mlp_ratio=mlp_ratio,
                    drop=drop,
                    drop_path=drop_path[i] if isinstance(drop_path, list) else drop_path,
                    local_conv_size=local_conv_size,
                    activation=activation,
                )
                for i in range(depth)
            ]
        )

        # Patch merging layer
        self.downsample = (
            None
            if downsample is None
            else downsample(input_resolution, dim=dim, out_dim=out_dim, activation=activation)
        )

    def forward(self, x):
        """Performs forward propagation on the input tensor and returns a normalized tensor."""
        for blk in self.blocks:
            x = checkpoint.checkpoint(blk, x) if self.use_checkpoint else blk(x)
        return x if self.downsample is None else self.downsample(x)

    def extra_repr(self) -> str:
        """Returns a string representation of the extra_repr function with the layer's parameters."""
        return f"dim={self.dim}, input_resolution={self.input_resolution}, depth={self.depth}"

__init__(dim, input_resolution, depth, num_heads, window_size, mlp_ratio=4.0, drop=0.0, drop_path=0.0, downsample=None, use_checkpoint=False, local_conv_size=3, activation=nn.GELU, out_dim=None)

Инициализирует базовый слой (BasicLayer).

Параметры:

Имя Тип Описание По умолчанию
dim int

Размерность входных и выходных данных.

требуется
input_resolution Tuple[int, int]

Пространственное разрешение входной карты признаков.

требуется
depth int

Количество блоков TinyViT.

требуется
num_heads int

Количество головок внимания.

требуется
window_size int

Размер локального окна.

требуется
mlp_ratio float

Отношение mlp hidden dim к embedding dim. По умолчанию - 4.

4.0
drop float

Уровень отсева. По умолчанию - 0.

0.0
drop_path float | tuple[float]

Стохастический показатель глубины. По умолчанию - 0.

0.0
downsample Module | None

Уменьши слой в конце слоя. По умолчанию - None.

None
use_checkpoint bool

Использовать ли контрольную точку для экономии памяти. По умолчанию - False.

False
local_conv_size int

Размер ядра локальной свертки. По умолчанию это 3.

3
activation nn

Функция активации для MLP. По умолчанию это nn.GELU.

GELU
out_dim int | None

Выходное измерение слоя. По умолчанию - None.

None

Поднимает:

Тип Описание
ValueError

Если drop_path это список float, но его длина не соответствует depth.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def __init__(
    self,
    dim,
    input_resolution,
    depth,
    num_heads,
    window_size,
    mlp_ratio=4.0,
    drop=0.0,
    drop_path=0.0,
    downsample=None,
    use_checkpoint=False,
    local_conv_size=3,
    activation=nn.GELU,
    out_dim=None,
):
    """
    Initializes the BasicLayer.

    Args:
        dim (int): The dimensionality of the input and output.
        input_resolution (Tuple[int, int]): Spatial resolution of the input feature map.
        depth (int): Number of TinyViT blocks.
        num_heads (int): Number of attention heads.
        window_size (int): Local window size.
        mlp_ratio (float, optional): Ratio of mlp hidden dim to embedding dim. Default is 4.
        drop (float, optional): Dropout rate. Default is 0.
        drop_path (float | tuple[float], optional): Stochastic depth rate. Default is 0.
        downsample (nn.Module | None, optional): Downsample layer at the end of the layer. Default is None.
        use_checkpoint (bool, optional): Whether to use checkpointing to save memory. Default is False.
        local_conv_size (int, optional): Kernel size of the local convolution. Default is 3.
        activation (torch.nn, optional): Activation function for MLP. Default is nn.GELU.
        out_dim (int | None, optional): The output dimension of the layer. Default is None.

    Raises:
        ValueError: If `drop_path` is a list of float but its length doesn't match `depth`.
    """
    super().__init__()
    self.dim = dim
    self.input_resolution = input_resolution
    self.depth = depth
    self.use_checkpoint = use_checkpoint

    # Build blocks
    self.blocks = nn.ModuleList(
        [
            TinyViTBlock(
                dim=dim,
                input_resolution=input_resolution,
                num_heads=num_heads,
                window_size=window_size,
                mlp_ratio=mlp_ratio,
                drop=drop,
                drop_path=drop_path[i] if isinstance(drop_path, list) else drop_path,
                local_conv_size=local_conv_size,
                activation=activation,
            )
            for i in range(depth)
        ]
    )

    # Patch merging layer
    self.downsample = (
        None
        if downsample is None
        else downsample(input_resolution, dim=dim, out_dim=out_dim, activation=activation)
    )

extra_repr()

Возвращает строковое представление функции extra_repr с параметрами слоя.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def extra_repr(self) -> str:
    """Returns a string representation of the extra_repr function with the layer's parameters."""
    return f"dim={self.dim}, input_resolution={self.input_resolution}, depth={self.depth}"

forward(x)

Выполняет прямое распространение на входе tensor и возвращает нормализованный tensor.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def forward(self, x):
    """Performs forward propagation on the input tensor and returns a normalized tensor."""
    for blk in self.blocks:
        x = checkpoint.checkpoint(blk, x) if self.use_checkpoint else blk(x)
    return x if self.downsample is None else self.downsample(x)



ultralytics.models.sam.modules.tiny_encoder.LayerNorm2d

Базы: Module

Реализация нормализации слоев в 2D на PyTorch .

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
class LayerNorm2d(nn.Module):
    """A PyTorch implementation of Layer Normalization in 2D."""

    def __init__(self, num_channels: int, eps: float = 1e-6) -> None:
        """Initialize LayerNorm2d with the number of channels and an optional epsilon."""
        super().__init__()
        self.weight = nn.Parameter(torch.ones(num_channels))
        self.bias = nn.Parameter(torch.zeros(num_channels))
        self.eps = eps

    def forward(self, x: torch.Tensor) -> torch.Tensor:
        """Perform a forward pass, normalizing the input tensor."""
        u = x.mean(1, keepdim=True)
        s = (x - u).pow(2).mean(1, keepdim=True)
        x = (x - u) / torch.sqrt(s + self.eps)
        return self.weight[:, None, None] * x + self.bias[:, None, None]

__init__(num_channels, eps=1e-06)

Инициализируй LayerNorm2d с количеством каналов и необязательным эпсилоном.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def __init__(self, num_channels: int, eps: float = 1e-6) -> None:
    """Initialize LayerNorm2d with the number of channels and an optional epsilon."""
    super().__init__()
    self.weight = nn.Parameter(torch.ones(num_channels))
    self.bias = nn.Parameter(torch.zeros(num_channels))
    self.eps = eps

forward(x)

Выполни прямое прохождение, нормализуя вход tensor.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def forward(self, x: torch.Tensor) -> torch.Tensor:
    """Perform a forward pass, normalizing the input tensor."""
    u = x.mean(1, keepdim=True)
    s = (x - u).pow(2).mean(1, keepdim=True)
    x = (x - u) / torch.sqrt(s + self.eps)
    return self.weight[:, None, None] * x + self.bias[:, None, None]



ultralytics.models.sam.modules.tiny_encoder.TinyViT

Базы: Module

Архитектура TinyViT для задач зрения.

Атрибуты:

Имя Тип Описание
img_size int

Размер входного изображения.

in_chans int

Количество входных каналов.

num_classes int

Количество классификационных классов.

embed_dims List[int]

Список размеров вкраплений для каждого слоя.

depths List[int]

Список глубин для каждого слоя.

num_heads List[int]

Список количества головок внимания для каждого слоя.

window_sizes List[int]

Список размеров окон для каждого слоя.

mlp_ratio float

Отношение скрытой размерности MLP к размерности встраивания.

drop_rate float

Показатель отсева для дроп-слоев.

drop_path_rate float

Скорость прохождения капли для стохастической глубины.

use_checkpoint bool

Используй контрольные точки для эффективного использования памяти.

mbconv_expand_ratio float

Коэффициент расширения для слоя MBConv.

local_conv_size int

Размер ядра локальной свертки.

layer_lr_decay float

Затухание скорости обучения в слоях.

Примечание

Эта реализация обобщена и может принимать список глубин, голов внимания, размеров встраивания и размеров окна, что позволяет тебе создавать "стопку" моделей TinyViT различной конфигурации.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
class TinyViT(nn.Module):
    """
    The TinyViT architecture for vision tasks.

    Attributes:
        img_size (int): Input image size.
        in_chans (int): Number of input channels.
        num_classes (int): Number of classification classes.
        embed_dims (List[int]): List of embedding dimensions for each layer.
        depths (List[int]): List of depths for each layer.
        num_heads (List[int]): List of number of attention heads for each layer.
        window_sizes (List[int]): List of window sizes for each layer.
        mlp_ratio (float): Ratio of MLP hidden dimension to embedding dimension.
        drop_rate (float): Dropout rate for drop layers.
        drop_path_rate (float): Drop path rate for stochastic depth.
        use_checkpoint (bool): Use checkpointing for efficient memory usage.
        mbconv_expand_ratio (float): Expansion ratio for MBConv layer.
        local_conv_size (int): Local convolution kernel size.
        layer_lr_decay (float): Layer-wise learning rate decay.

    Note:
        This implementation is generalized to accept a list of depths, attention heads,
        embedding dimensions and window sizes, which allows you to create a
        "stack" of TinyViT models of varying configurations.
    """

    def __init__(
        self,
        img_size=224,
        in_chans=3,
        num_classes=1000,
        embed_dims=(96, 192, 384, 768),
        depths=(2, 2, 6, 2),
        num_heads=(3, 6, 12, 24),
        window_sizes=(7, 7, 14, 7),
        mlp_ratio=4.0,
        drop_rate=0.0,
        drop_path_rate=0.1,
        use_checkpoint=False,
        mbconv_expand_ratio=4.0,
        local_conv_size=3,
        layer_lr_decay=1.0,
    ):
        """
        Initializes the TinyViT model.

        Args:
            img_size (int, optional): The input image size. Defaults to 224.
            in_chans (int, optional): Number of input channels. Defaults to 3.
            num_classes (int, optional): Number of classification classes. Defaults to 1000.
            embed_dims (List[int], optional): List of embedding dimensions per layer. Defaults to [96, 192, 384, 768].
            depths (List[int], optional): List of depths for each layer. Defaults to [2, 2, 6, 2].
            num_heads (List[int], optional): List of number of attention heads per layer. Defaults to [3, 6, 12, 24].
            window_sizes (List[int], optional): List of window sizes for each layer. Defaults to [7, 7, 14, 7].
            mlp_ratio (float, optional): Ratio of MLP hidden dimension to embedding dimension. Defaults to 4.
            drop_rate (float, optional): Dropout rate. Defaults to 0.
            drop_path_rate (float, optional): Drop path rate for stochastic depth. Defaults to 0.1.
            use_checkpoint (bool, optional): Whether to use checkpointing for efficient memory usage. Defaults to False.
            mbconv_expand_ratio (float, optional): Expansion ratio for MBConv layer. Defaults to 4.0.
            local_conv_size (int, optional): Local convolution kernel size. Defaults to 3.
            layer_lr_decay (float, optional): Layer-wise learning rate decay. Defaults to 1.0.
        """
        super().__init__()
        self.img_size = img_size
        self.num_classes = num_classes
        self.depths = depths
        self.num_layers = len(depths)
        self.mlp_ratio = mlp_ratio

        activation = nn.GELU

        self.patch_embed = PatchEmbed(
            in_chans=in_chans, embed_dim=embed_dims[0], resolution=img_size, activation=activation
        )

        patches_resolution = self.patch_embed.patches_resolution
        self.patches_resolution = patches_resolution

        # Stochastic depth
        dpr = [x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))]  # stochastic depth decay rule

        # Build layers
        self.layers = nn.ModuleList()
        for i_layer in range(self.num_layers):
            kwargs = dict(
                dim=embed_dims[i_layer],
                input_resolution=(
                    patches_resolution[0] // (2 ** (i_layer - 1 if i_layer == 3 else i_layer)),
                    patches_resolution[1] // (2 ** (i_layer - 1 if i_layer == 3 else i_layer)),
                ),
                #   input_resolution=(patches_resolution[0] // (2 ** i_layer),
                #                     patches_resolution[1] // (2 ** i_layer)),
                depth=depths[i_layer],
                drop_path=dpr[sum(depths[:i_layer]) : sum(depths[: i_layer + 1])],
                downsample=PatchMerging if (i_layer < self.num_layers - 1) else None,
                use_checkpoint=use_checkpoint,
                out_dim=embed_dims[min(i_layer + 1, len(embed_dims) - 1)],
                activation=activation,
            )
            if i_layer == 0:
                layer = ConvLayer(conv_expand_ratio=mbconv_expand_ratio, **kwargs)
            else:
                layer = BasicLayer(
                    num_heads=num_heads[i_layer],
                    window_size=window_sizes[i_layer],
                    mlp_ratio=self.mlp_ratio,
                    drop=drop_rate,
                    local_conv_size=local_conv_size,
                    **kwargs,
                )
            self.layers.append(layer)

        # Classifier head
        self.norm_head = nn.LayerNorm(embed_dims[-1])
        self.head = nn.Linear(embed_dims[-1], num_classes) if num_classes > 0 else torch.nn.Identity()

        # Init weights
        self.apply(self._init_weights)
        self.set_layer_lr_decay(layer_lr_decay)
        self.neck = nn.Sequential(
            nn.Conv2d(
                embed_dims[-1],
                256,
                kernel_size=1,
                bias=False,
            ),
            LayerNorm2d(256),
            nn.Conv2d(
                256,
                256,
                kernel_size=3,
                padding=1,
                bias=False,
            ),
            LayerNorm2d(256),
        )

    def set_layer_lr_decay(self, layer_lr_decay):
        """Sets the learning rate decay for each layer in the TinyViT model."""
        decay_rate = layer_lr_decay

        # Layers -> blocks (depth)
        depth = sum(self.depths)
        lr_scales = [decay_rate ** (depth - i - 1) for i in range(depth)]

        def _set_lr_scale(m, scale):
            """Sets the learning rate scale for each layer in the model based on the layer's depth."""
            for p in m.parameters():
                p.lr_scale = scale

        self.patch_embed.apply(lambda x: _set_lr_scale(x, lr_scales[0]))
        i = 0
        for layer in self.layers:
            for block in layer.blocks:
                block.apply(lambda x: _set_lr_scale(x, lr_scales[i]))
                i += 1
            if layer.downsample is not None:
                layer.downsample.apply(lambda x: _set_lr_scale(x, lr_scales[i - 1]))
        assert i == depth
        for m in [self.norm_head, self.head]:
            m.apply(lambda x: _set_lr_scale(x, lr_scales[-1]))

        for k, p in self.named_parameters():
            p.param_name = k

        def _check_lr_scale(m):
            """Checks if the learning rate scale attribute is present in module's parameters."""
            for p in m.parameters():
                assert hasattr(p, "lr_scale"), p.param_name

        self.apply(_check_lr_scale)

    def _init_weights(self, m):
        """Initializes weights for linear layers and layer normalization in the given module."""
        if isinstance(m, nn.Linear):
            # NOTE: This initialization is needed only for training.
            # trunc_normal_(m.weight, std=.02)
            if m.bias is not None:
                nn.init.constant_(m.bias, 0)
        elif isinstance(m, nn.LayerNorm):
            nn.init.constant_(m.bias, 0)
            nn.init.constant_(m.weight, 1.0)

    @torch.jit.ignore
    def no_weight_decay_keywords(self):
        """Returns a dictionary of parameter names where weight decay should not be applied."""
        return {"attention_biases"}

    def forward_features(self, x):
        """Runs the input through the model layers and returns the transformed output."""
        x = self.patch_embed(x)  # x input is (N, C, H, W)

        x = self.layers[0](x)
        start_i = 1

        for i in range(start_i, len(self.layers)):
            layer = self.layers[i]
            x = layer(x)
        batch, _, channel = x.shape
        x = x.view(batch, 64, 64, channel)
        x = x.permute(0, 3, 1, 2)
        return self.neck(x)

    def forward(self, x):
        """Executes a forward pass on the input tensor through the constructed model layers."""
        return self.forward_features(x)

__init__(img_size=224, in_chans=3, num_classes=1000, embed_dims=(96, 192, 384, 768), depths=(2, 2, 6, 2), num_heads=(3, 6, 12, 24), window_sizes=(7, 7, 14, 7), mlp_ratio=4.0, drop_rate=0.0, drop_path_rate=0.1, use_checkpoint=False, mbconv_expand_ratio=4.0, local_conv_size=3, layer_lr_decay=1.0)

Инициализирует модель TinyViT.

Параметры:

Имя Тип Описание По умолчанию
img_size int

Размер входного изображения. По умолчанию равен 224.

224
in_chans int

Количество входных каналов. По умолчанию равно 3.

3
num_classes int

Количество классов классификации. По умолчанию равно 1000.

1000
embed_dims List[int]

Список размеров встраивания для каждого слоя. По умолчанию [96, 192, 384, 768].

(96, 192, 384, 768)
depths List[int]

Список глубин для каждого слоя. По умолчанию [2, 2, 6, 2].

(2, 2, 6, 2)
num_heads List[int]

Список количества головок внимания на слой. По умолчанию [3, 6, 12, 24].

(3, 6, 12, 24)
window_sizes List[int]

Список размеров окон для каждого слоя. По умолчанию [7, 7, 14, 7].

(7, 7, 14, 7)
mlp_ratio float

Отношение скрытой размерности MLP к размерности встраивания. По умолчанию равно 4.

4.0
drop_rate float

Коэффициент отсева. По умолчанию равен 0.

0.0
drop_path_rate float

Скорость прохождения пути падения для стохастической глубины. По умолчанию 0,1.

0.1
use_checkpoint bool

Использовать ли контрольные точки для эффективного использования памяти. По умолчанию установлено значение False.

False
mbconv_expand_ratio float

Коэффициент расширения для слоя MBConv. По умолчанию равен 4.0.

4.0
local_conv_size int

Размер ядра локальной свертки. По умолчанию равен 3.

3
layer_lr_decay float

Затухание скорости обучения по слоям. По умолчанию 1.0.

1.0
Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def __init__(
    self,
    img_size=224,
    in_chans=3,
    num_classes=1000,
    embed_dims=(96, 192, 384, 768),
    depths=(2, 2, 6, 2),
    num_heads=(3, 6, 12, 24),
    window_sizes=(7, 7, 14, 7),
    mlp_ratio=4.0,
    drop_rate=0.0,
    drop_path_rate=0.1,
    use_checkpoint=False,
    mbconv_expand_ratio=4.0,
    local_conv_size=3,
    layer_lr_decay=1.0,
):
    """
    Initializes the TinyViT model.

    Args:
        img_size (int, optional): The input image size. Defaults to 224.
        in_chans (int, optional): Number of input channels. Defaults to 3.
        num_classes (int, optional): Number of classification classes. Defaults to 1000.
        embed_dims (List[int], optional): List of embedding dimensions per layer. Defaults to [96, 192, 384, 768].
        depths (List[int], optional): List of depths for each layer. Defaults to [2, 2, 6, 2].
        num_heads (List[int], optional): List of number of attention heads per layer. Defaults to [3, 6, 12, 24].
        window_sizes (List[int], optional): List of window sizes for each layer. Defaults to [7, 7, 14, 7].
        mlp_ratio (float, optional): Ratio of MLP hidden dimension to embedding dimension. Defaults to 4.
        drop_rate (float, optional): Dropout rate. Defaults to 0.
        drop_path_rate (float, optional): Drop path rate for stochastic depth. Defaults to 0.1.
        use_checkpoint (bool, optional): Whether to use checkpointing for efficient memory usage. Defaults to False.
        mbconv_expand_ratio (float, optional): Expansion ratio for MBConv layer. Defaults to 4.0.
        local_conv_size (int, optional): Local convolution kernel size. Defaults to 3.
        layer_lr_decay (float, optional): Layer-wise learning rate decay. Defaults to 1.0.
    """
    super().__init__()
    self.img_size = img_size
    self.num_classes = num_classes
    self.depths = depths
    self.num_layers = len(depths)
    self.mlp_ratio = mlp_ratio

    activation = nn.GELU

    self.patch_embed = PatchEmbed(
        in_chans=in_chans, embed_dim=embed_dims[0], resolution=img_size, activation=activation
    )

    patches_resolution = self.patch_embed.patches_resolution
    self.patches_resolution = patches_resolution

    # Stochastic depth
    dpr = [x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))]  # stochastic depth decay rule

    # Build layers
    self.layers = nn.ModuleList()
    for i_layer in range(self.num_layers):
        kwargs = dict(
            dim=embed_dims[i_layer],
            input_resolution=(
                patches_resolution[0] // (2 ** (i_layer - 1 if i_layer == 3 else i_layer)),
                patches_resolution[1] // (2 ** (i_layer - 1 if i_layer == 3 else i_layer)),
            ),
            #   input_resolution=(patches_resolution[0] // (2 ** i_layer),
            #                     patches_resolution[1] // (2 ** i_layer)),
            depth=depths[i_layer],
            drop_path=dpr[sum(depths[:i_layer]) : sum(depths[: i_layer + 1])],
            downsample=PatchMerging if (i_layer < self.num_layers - 1) else None,
            use_checkpoint=use_checkpoint,
            out_dim=embed_dims[min(i_layer + 1, len(embed_dims) - 1)],
            activation=activation,
        )
        if i_layer == 0:
            layer = ConvLayer(conv_expand_ratio=mbconv_expand_ratio, **kwargs)
        else:
            layer = BasicLayer(
                num_heads=num_heads[i_layer],
                window_size=window_sizes[i_layer],
                mlp_ratio=self.mlp_ratio,
                drop=drop_rate,
                local_conv_size=local_conv_size,
                **kwargs,
            )
        self.layers.append(layer)

    # Classifier head
    self.norm_head = nn.LayerNorm(embed_dims[-1])
    self.head = nn.Linear(embed_dims[-1], num_classes) if num_classes > 0 else torch.nn.Identity()

    # Init weights
    self.apply(self._init_weights)
    self.set_layer_lr_decay(layer_lr_decay)
    self.neck = nn.Sequential(
        nn.Conv2d(
            embed_dims[-1],
            256,
            kernel_size=1,
            bias=False,
        ),
        LayerNorm2d(256),
        nn.Conv2d(
            256,
            256,
            kernel_size=3,
            padding=1,
            bias=False,
        ),
        LayerNorm2d(256),
    )

forward(x)

Выполняет прямое прохождение входного tensor через построенные слои модели.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def forward(self, x):
    """Executes a forward pass on the input tensor through the constructed model layers."""
    return self.forward_features(x)

forward_features(x)

Прогони входные данные через слои модели и верни преобразованный выход.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def forward_features(self, x):
    """Runs the input through the model layers and returns the transformed output."""
    x = self.patch_embed(x)  # x input is (N, C, H, W)

    x = self.layers[0](x)
    start_i = 1

    for i in range(start_i, len(self.layers)):
        layer = self.layers[i]
        x = layer(x)
    batch, _, channel = x.shape
    x = x.view(batch, 64, 64, channel)
    x = x.permute(0, 3, 1, 2)
    return self.neck(x)

no_weight_decay_keywords()

Возвращает словарь имен параметров, к которым не следует применять затухание веса.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
@torch.jit.ignore
def no_weight_decay_keywords(self):
    """Returns a dictionary of parameter names where weight decay should not be applied."""
    return {"attention_biases"}

set_layer_lr_decay(layer_lr_decay)

Устанавливает затухание скорости обучения для каждого слоя в модели TinyViT.

Исходный код в ultralytics/models/sam/modules/tiny_encoder.py
def set_layer_lr_decay(self, layer_lr_decay):
    """Sets the learning rate decay for each layer in the TinyViT model."""
    decay_rate = layer_lr_decay

    # Layers -> blocks (depth)
    depth = sum(self.depths)
    lr_scales = [decay_rate ** (depth - i - 1) for i in range(depth)]

    def _set_lr_scale(m, scale):
        """Sets the learning rate scale for each layer in the model based on the layer's depth."""
        for p in m.parameters():
            p.lr_scale = scale

    self.patch_embed.apply(lambda x: _set_lr_scale(x, lr_scales[0]))
    i = 0
    for layer in self.layers:
        for block in layer.blocks:
            block.apply(lambda x: _set_lr_scale(x, lr_scales[i]))
            i += 1
        if layer.downsample is not None:
            layer.downsample.apply(lambda x: _set_lr_scale(x, lr_scales[i - 1]))
    assert i == depth
    for m in [self.norm_head, self.head]:
        m.apply(lambda x: _set_lr_scale(x, lr_scales[-1]))

    for k, p in self.named_parameters():
        p.param_name = k

    def _check_lr_scale(m):
        """Checks if the learning rate scale attribute is present in module's parameters."""
        for p in m.parameters():
            assert hasattr(p, "lr_scale"), p.param_name

    self.apply(_check_lr_scale)





Created 2023-11-12, Updated 2024-06-02
Authors: glenn-jocher (5), Burhan-Q (1), Laughing-q (1)