Get the dtype of the model parameters.

Emit all remaining token ids at end of stream.

r""" Args: input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length, 1 + n_vq)`): Indices of input seq

Quantized the input tensor using a fixed codebook and returns the corresponding codebook vectors Parameters ----------

Quantized the input tensor using a fixed set of `n` codebooks and returns the corresponding codebook vectors Parameters ------

Model forward pass Parameters ---------- audio_data : Tensor[B x 1 x T] Audio data to encode sample_rate

Args: ids: [batch, seq_len] token ids mask: [batch, seq_len] attention mask (1=valid, 0=pad) Returns:

Given the unquantized latents, reconstruct the continuous representation after quantization. Parameters ---------- la

Load a pipeline from a local diffusers-style dir or a HF hub repo id.