messages: List[Dict] messages[i] = { "role": "user" | "assistant" | "system", "content": str }

Initialization. INPUT: - in_features: shape of the input - alpha - trainable parameter that controls frequenc

kernel_size should be even number for stylegan3 setup, in this implementation, odd number is also possible.

Load Pytorch pretrained weights and return the loaded model.

Computes the output length of the convolutional layers

Save weights and config.json from a Pytorch model to a local directory.

params: mel: [B, T, num_mels], torch.tensor return: wav: [B, 1, T], torch.tensor

params: wav_data: torch.tensor or numpy array, shape [Batch, T], wav data, should be 24k return: mel: [Batch,

Input is expected to be of size [bsz x seqlen].

Forward pass of DiT. x: (N, T, C) tensor of inputs (latent representations of speech) position_ids: (N, T) tensor of position

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

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

Input shape: Batch x Time x Channel

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

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

r""" Args: input_features (`torch.LongTensor` of shape `(batch_size, feature_size, sequence_length)`): Float value

r""" Args: input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`): Indices of input sequence toke

r""" Returns: Example: ```python >>> import torch >>> from transformers import AutoFeatureExtractor, Whisp

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