DeepSpeech¶ class torchaudio.models.DeepSpeech(n_feature: int, n_hidden: int = 2048, n_class: int = 40, dropout: float = 0.0)[source]¶ DeepSpeech 架構,出自論文《Deep Speech: Scaling up end-to-end speech recognition》[Hannun 等人,2014]。 引數: n_feature – 輸入特徵數量 n_hidden – 內部隱藏單元大小。 n_class – 輸出類別數量 方法¶ forward¶ DeepSpeech.forward(x: Tensor) → Tensor[source]¶ 引數: x (torch.Tensor) – 維度為 (batch, channel, time, feature) 的張量。 返回: 維度為 (batch, time, class) 的預測張量。 返回型別: Tensor