DeepSpeech¶
- class torchaudio.models.DeepSpeech(n_feature: int, n_hidden: int = 2048, n_class: int = 40, dropout: float = 0.0)[source]¶
DeepSpeech architecture introduced in Deep Speech: Scaling up end-to-end speech recognition [Hannun et al., 2014].
- Parameters:
n_feature – Number of input features
n_hidden – Internal hidden unit size.
n_class – Number of output classes
Methods¶
forward¶
- DeepSpeech.forward(x: Tensor) Tensor[source]¶
- Parameters:
x (torch.Tensor) – Tensor of dimension (batch, channel, time, feature).
- Returns:
Predictor tensor of dimension (batch, time, class).
- Return type:
Tensor
