Args: hidden_states (`torch.FloatTensor`): input to the layer of shape `(batch, seq_len, embed_dim)` attention_mask (

Forward pass of the function. Applies the function to the input elementwise. SnakeBeta ∶= x + 1/b * sin^2 (xa)

r""" labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*): Labels for computing the masked language mo

Args: hidden_states (`torch.FloatTensor`): input to the layer of shape `(batch, seq_len, embed_dim)` attention_mask (

r""" labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*): Labels for computing the masked language mo

Initialize tokenizer with HuggingFace `from_pretrained` style. Args: pretrained_model_name_or_path (str):

Get the decoder upsample rate (waveform samples per code step). Returns: int: Decode upsample rate.

Get the encoder downsample rate (waveform samples per code step). Returns: int: Encode downsample rate.

Get the expected input sample rate for encoding. Returns: int: Input sample rate (Hz).

List supported language names for the current model. This is a convenience wrapper around `model.get_supported_languages()`.

List supported speaker names for the current model. This is a convenience wrapper around `model.get_supported_speakers()`. I

All layer in code2wav should be sliding attention