分发

DiscretePriorDistribution

基类：PriorDistribution

表示离散先验分布的抽象基类。

源代码位于 bionemo/moco/distributions/prior/distribution.py

class DiscretePriorDistribution(PriorDistribution):
    """An abstract base class representing a discrete prior distribution."""

    def __init__(self, num_classes: int, prior_dist: Tensor):
        """Initializes a DiscretePriorDistribution instance.

        Args:
        num_classes (int): The number of classes in the discrete distribution.
        prior_dist (Tensor): The prior distribution over the classes.

        Returns:
        None
        """
        self.num_classes = num_classes
        self.prior_dist = prior_dist

    def get_num_classes(self) -> int:
        """Getter for num_classes."""
        return self.num_classes

    def get_prior_dist(self) -> Tensor:
        """Getter for prior_dist."""
        return self.prior_dist

__init__(num_classes, prior_dist)

初始化 DiscretePriorDistribution 实例。

Args: num_classes (int): 离散分布中的类别数量。prior_dist (Tensor): 类别的先验分布。

返回：None

源代码位于 bionemo/moco/distributions/prior/distribution.py

def __init__(self, num_classes: int, prior_dist: Tensor):
    """Initializes a DiscretePriorDistribution instance.

    Args:
    num_classes (int): The number of classes in the discrete distribution.
    prior_dist (Tensor): The prior distribution over the classes.

    Returns:
    None
    """
    self.num_classes = num_classes
    self.prior_dist = prior_dist

get_num_classes()

num_classes 的 Getter。

源代码位于 bionemo/moco/distributions/prior/distribution.py

def get_num_classes(self) -> int:
    """Getter for num_classes."""
    return self.num_classes

get_prior_dist()

prior_dist 的 Getter。

源代码位于 bionemo/moco/distributions/prior/distribution.py

def get_prior_dist(self) -> Tensor:
    """Getter for prior_dist."""
    return self.prior_dist

PriorDistribution

基类：ABC

表示先验分布的抽象基类。

源代码位于 bionemo/moco/distributions/prior/distribution.py

class PriorDistribution(ABC):
    """An abstract base class representing a prior distribution."""

    @abstractmethod
    def sample(self, shape: Tuple, mask: Optional[Tensor] = None, device: Union[str, torch.device] = "cpu") -> Tensor:
        """Generates a specified number of samples from the time distribution.

        Args:
        shape (Tuple): The shape of the samples to generate.
        mask (Optional[Tensor], optional): A tensor indicating which samples should be masked. Defaults to None.
        device (str, optional): The device on which to generate the samples. Defaults to "cpu".

        Returns:
            Float: A tensor of samples.
        """
        pass

sample(shape, mask=None, device='cpu') abstractmethod

从时间分布中生成指定数量的样本。

Args: shape (Tuple): 要生成的样本的形状。mask (Optional[Tensor], optional): 指示哪些样本应被掩码的张量。默认为 None。device (str, optional): 生成样本的设备。默认为 "cpu"。

返回

名称	类型	描述
浮点数	张量	样本的张量。

源代码位于 bionemo/moco/distributions/prior/distribution.py

@abstractmethod
def sample(self, shape: Tuple, mask: Optional[Tensor] = None, device: Union[str, torch.device] = "cpu") -> Tensor:
    """Generates a specified number of samples from the time distribution.

    Args:
    shape (Tuple): The shape of the samples to generate.
    mask (Optional[Tensor], optional): A tensor indicating which samples should be masked. Defaults to None.
    device (str, optional): The device on which to generate the samples. Defaults to "cpu".

    Returns:
        Float: A tensor of samples.
    """
    pass