torchaudio.models.hubert_pretrain_xlarge¶
- torchaudio.models.hubert_pretrain_xlarge(encoder_projection_dropout: float = 0.0, encoder_attention_dropout: float = 0.0, encoder_ff_interm_dropout: float = 0.0, encoder_dropout: float = 0.0, encoder_layer_drop: float = 0.0, mask_prob: float = 0.8, mask_channel_prob: float = 0.0, mask_channel_length: int = 10, feature_grad_mult: Optional[float] = None) HuBERTPretrainModel [source]¶
构建 HuBERT [Hsu 等人,2021] 的“超大型”
HuBERTPretrainModel
,用于预训练。- 参数:
encoder_projection_dropout (float) – 请参见
hubert_pretrain_model()
。encoder_attention_dropout (float) – 请参见
hubert_pretrain_model()
。encoder_ff_interm_dropout (float) – 请参见
hubert_pretrain_model()
。encoder_dropout (float) – 请参见
hubert_pretrain_model()
。encoder_layer_drop (float) – 请参见
hubert_pretrain_model()
。mask_prob (float) – 请参见
hubert_pretrain_model()
。mask_channel_prob (float) – 请参见
hubert_pretrain_model()
。mask_channel_length (int) – 请参见
hubert_pretrain_model()
。feature_grad_mult (浮点数 或 None) – 请参阅
hubert_pretrain_model()
。
- 返回:
生成的模型。
- 返回类型: