Module audiocraft.modules.lstm

Classes

class StreamableLSTM (dimension: int, num_layers: int = 2, skip: bool = True)

Expand source code

class StreamableLSTM(nn.Module):
    """LSTM without worrying about the hidden state, nor the layout of the data.
    Expects input as convolutional layout.
    """
    def __init__(self, dimension: int, num_layers: int = 2, skip: bool = True):
        super().__init__()
        self.skip = skip
        self.lstm = nn.LSTM(dimension, dimension, num_layers)

    def forward(self, x):
        x = x.permute(2, 0, 1)
        y, _ = self.lstm(x)
        if self.skip:
            y = y + x
        y = y.permute(1, 2, 0)
        return y

LSTM without worrying about the hidden state, nor the layout of the data. Expects input as convolutional layout.

Initializes internal Module state, shared by both nn.Module and ScriptModule.

Ancestors

• torch.nn.modules.module.Module

Class variables

var call_super_init : bool

var dump_patches : bool

var training : bool

Methods

def forward(self, x) ‑> Callable[..., Any]

Expand source code

def forward(self, x):
    x = x.permute(2, 0, 1)
    y, _ = self.lstm(x)
    if self.skip:
        y = y + x
    y = y.permute(1, 2, 0)
    return y

Defines the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the :class:Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.