paddlespeech.t2s.modules.positional_encoding module

paddlespeech.t2s.modules.positional_encoding.scaled_position_encoding(num_positions: int, feature_size: int, omega: Tensor, start_pos: int = 0, dtype=None) → Tensor[source]
paddlespeech.t2s.modules.positional_encoding.sinusoid_position_encoding(num_positions: int, feature_size: int, omega: float = 1.0, start_pos: int = 0, dtype=None) → Tensor[source]