torchaudio.models.hubert_pretrain_base¶
- torchaudio.models.hubert_pretrain_base(encoder_projection_dropout: float = 0.1, encoder_attention_dropout: float = 0.1, encoder_ff_interm_dropout: float =0.0, encoder_dropout: float =0.1, encoder_layer_drop: float =0.05, mask_prob: float =0.8, mask_channel_prob: float =0.0, mask_channel_length: int =10, feature_grad_mult: Optional[float] = 0.1, num_classes: int = 100) HuBERTPretrainModel[source]¶
構建“基礎”
HuBERTPretrainModel來自 HuBERT [Hsu et al., 2021] 用於預訓練。- 引數:
encoder_projection_dropout (float) – 參見
hubert_pretrain_model()。encoder_attention_dropout (float) – 參見
hubert_pretrain_model()。encoder_ff_interm_dropout (float) – 參見
hubert_pretrain_model()。encoder_dropout (float) – 參見
hubert_pretrain_model()。encoder_layer_drop (float) – 參見
hubert_pretrain_model()。mask_prob (float) – 參見
hubert_pretrain_model()。mask_channel_prob (float) – 參見
hubert_pretrain_model()。mask_channel_length (int) – 參見
hubert_pretrain_model()。feature_grad_mult (float 或 None) – 參見
hubert_pretrain_model()。num_classes (int, 可選) – 參見
hubert_pretrain_model()。
- 返回:
結果模型。
- 返回型別:
