torch_frame.nn.encoding.PositionalEncoding

class PositionalEncoding(out_size: int)

Bases: Module

Positional encoding introduced in “Attention Is All You Need” paper. Given an input tensor of shape (*, ), this encoding expands it into an output tensor of shape (*, out_size).

Parameters:

out_size (int) – The output dimension size.

forward(input_tensor: Tensor) → Tensor

Define 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 Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.