Converts a frequency from the Hertz scale to the Mel scale. Parameters: - pitch: float or ndarray Frequency in Hertz. Retur

Inverse of `codes_to_indices`.

Initialize the model. Args: args: Dictionary containing model configuration

Initialize the model. Args: args: Dictionary containing model configuration

Initialize the model. Args: args: Dictionary containing model configuration

Converts a JSONL file to a CSV file. This function reads a JSONL file, determines all unique keys present in the file, and writes th

Turn special token of pitch std level.

Turn special token of pitch_std value.

Reads a JSONL file and returns a list of dictionaries. Args: file_path : Path The path to the JSONL file to be read. Return

Read metadata from a file. Args: filename (str): The file from which to read the metadata. Returns: list of dic

Remove silence from both ends of an audio signal. Args: wav: Input audio signal array sample_rate: Audio sample rate in Hz

Resolves the absolute path of a symbolic link. Args: symbolic_link_path (Path): The path to the symbolic link. Returns:

Helper function to run the blocking sync streaming call.

Save metadata to a file. Args: data (list of dict): Metadata to be saved. filename (str): Name of the file to save the m

Turn special token of speed level.

Turn special token of speed value.

Perform STFT and convert to magnitude spectrogram. Args: x (Tensor): Input signal tensor (B, T). fft_size (int): FFT size.

tokenize the input tensor

tokenize the batch of audio Args: batch: wavs (List[np.ndarray]): batch of audio ref_wavs (torch.

Gradio callback to clone voice using text and optional prompt speech. - text: The input text to be synthesised. - prompt_text

Gradio callback to create a synthetic voice with adjustable parameters. - text: The input text for synthesis. - gender: 'male

Writes a list of dictionaries to a JSONL file. Args: metadata : List[dict] A list of dictionaries, each representing a piece of meta.