Add whisper-style tokens. ([PREV] -> [previous text tokens or hotwords]).optional -- ┌-----------------------------------------------------

Add whisper-style tokens. ([PREV] -> [previous text tokens or hotwords]).optional -- ┌-----------------------------------------------------

Returns: List[List[List[int]]]: nbest result for each utterance

Cross validation on

Compute attention context vector. Args: value (torch.Tensor): Transformed value, size (#batch, ..., n_head, time2

Compute attention context vector. Args: value (torch.Tensor): Transformed value, size (#batch, n_head, time2, d_k

Forward just one chunk Args: xs (torch.Tensor): chunk input, with shape (b=1, time, mel-dim), where `time == (ch

Transform query, key and value. Args: query (torch.Tensor): Query tensor (#batch, ..., time1, size). key (torch.Tenso

Args: input (torch.Tensor): [batch_size, time_step=1] padding (torch.Tensor): [batch_size,1], 1 is padding value

Generate amplitude domain mask randomly in [-100db, 0db] Args: mask_num: the slot number of the mask Returns: A list of tupl

Get a amplitude domain mask from db domain mask Args: db_mask: Optional. A list of tuple. if None, using default value. Returns:

Construct dataset from arguments We have two shuffle stage in the Dataset. The first is global shuffle at shard tar/raw file level.

Args: input_size (int): input dim output_size (int): dimension of attention attention_heads (int): the nu

Apply optional mask for encoder. Args: xs (torch.Tensor): padded input, (B, L, D), L for max length mask (torch.Tensor): mask fo

Add <sos> and <eos> labels. Args: ys_pad (torch.Tensor): batch of padded target sequences (B, Lmax) sos (int): index of <sos>

argmax of frame activations Args: torch.Tensor hs_pad: 3d tensor (B, Tmax, eprojs) Returns: torch.Tensor: arg

Args: input_token_ids: (B, Lmax) ,其词表范围是[0, vocab_size1+vocab_size2) dividing_id: int, 第一个词表的个数 embedding1: nn.Embedding

Export interface for c++ call, return eos symbol id of the model

Forward one step. This is only used for decoding. Args: memory: encoded memory, float32 (batch, maxlen_in, feat)

Forward one step. This is only used for decoding. Args: memory: encoded memory, float32 (batch, maxlen_in, feat)

Generates responses for given prompts using Gemma model.

return: wav_embedding: (b, l, v) wav_mask: (b, l), wav为有效值的位置为true

初始化match_found标志。

Args: text_tokens_padded: (B, Lmax) text_tokens_lens: (B,) speech_tokens_padded: (B, Lmax2) speech_tokens_lens:

If a sequence is finished, all of its branch should be <eos> Args: pred (torch.Tensor): A int array with shape (batch_si

If a sequence is finished, we only allow one alive branch. This function aims to give one branch a zero score and the rest -inf score. A

Reverse padding for the list of tensors. Args: ys_pad (tensor): The padded tensor (B, Tokenmax). ys_lens (tensor): The lens of to

Create mask for subsequent steps (size, size). This mask is used only in decoder which works in an auto-regressive mode. This means the curre

Calculate accuracy. Args: pad_outputs (Tensor): Prediction tensors (B * Lmax, D). pad_targets (LongTensor): Target label tensors

Args: input: [bs, max_time_step, dim] padding: [bs, max_time_step]

Args: input_size (int): input dim output_size (int): dimension of attention attention_heads (int): the nu