This method forwards all its arguments to LlamaTokenizerFast's [`~PreTrainedTokenizer.decode`]. Please refer to the docstring of this

Append a new message.

This is the equivalent of torch.repeat_interleave(x, dim=1, repeats=n_rep). The hidden states go from (batch, num_key_value_heads, seqlen, he

Decode discrete audio codes to mel spectrograms. Borrowed from `https://github.com/2noise/ChatTTS/blob/main/ChatTTS/core.py` Args:

Get a conversation template.

Prefill a chunk of new text tokens in streaming setting. Specifically speaking, update `past_key_values` using new text tokens, then the model

Get the prompt for generation.

Register a new conversation template.

Applies Rotary Position Embedding to the query and key tensors. Args: q (`torch.Tensor`): The query tensor. k (`torch.Tensor`): T

This method forwards all its arguments to LlamaTokenizerFast's [`~PreTrainedTokenizer.batch_decode`]. Please refer to the docstring o

Args: data: input_embeddings: chunk_length: whisper use full attention or chunk attention str

Compute all visual embeddings, and set into llm embeddings. Args: data: Dict tgt_sizes: image size after

Computes the output length of the convolutional layers and the output length of the audio encoder

Merge two audio waveforms with smooth in streaming audio generation. Borrowed some codes from `https://github.com/huggingface/transfo

Applies Rotary Position Embedding to the query and key tensors.

embed_dim: output dimension for each position pos: a list of positions to be encoded: size (H, W) out: (H, W, D)

In streaming audio generation, determine which `text` positions the TTS model can attend to when generating each chunk of `audio` tokens. Th

This is the equivalent of torch.repeat_interleave(x, dim=1, repeats=n_rep). The hidden states go from (batch, num_key_value_heads, seqlen, he

Rotates half the hidden dims of the input.

Rotates half the hidden dims of the input.

r""" Sets the attention processor to use to compute attention. Parameters: processor (`dict` of `AttentionProcessor` or o