label_lengths (Tensor): The length for each label sequence, it should have shape [batch_size] and dtype int64.
label_lengths (Tensor): The length for each label sequence, it should have shape [batch_size] and dtype int64.
blank (int, optional): The blank label index of Connectionist Temporal Classification (CTC) loss, which is in the half-opened interval [0, num_classes + 1). The data type must be int32. Default is 0.
blank (int, optional): The blank label index of Connectionist Temporal Classification (CTC) loss, which is in the half-opened interval [0, num_classes + 1). The data type must be int32. Default is 0.
reduction (string, optional): Indicate how to average the loss, the candicates are ``'none'`` | ``'mean'`` | ``'sum'``. If :attr:`reduction` is ``'mean'``, the output loss will be divided by the label_lengths, and then return the mean of quotient; If :attr:`reduction` is ``'sum'``, return the sum of loss; If :attr:`reduction` is ``'none'``, no reduction will be applied. Default is ``'mean'``.
reduction (string, optional): Indicate how to average the loss, the candicates are ``'none'`` | ``'mean'`` | ``'sum'``. If :attr:`reduction` is ``'mean'``, the output loss will be divided by the label_lengths, and then return the mean of quotient; If :attr:`reduction` is ``'sum'``, return the sum of loss; If :attr:`reduction` is ``'none'``, no reduction will be applied. Default is ``'mean'``.
norm_by_times (bool, default False) – Whether to normalize the gradients by the number of time-step, which is also the sequence’s length. There is no need to normalize the gradients if reduction mode is 'mean'.
norm_by_times (bool, default False): Whether to normalize the gradients by the number of time-step, which is also the sequence’s length. There is no need to normalize the gradients if reduction mode is 'mean'.
Returns:
Returns:
Tensor, The Connectionist Temporal Classification (CTC) loss between ``log_probs`` and ``labels``. If attr:`reduction` is ``'none'``, the shape of loss is [batch_size], otherwise, the shape of loss is [1]. Data type is the same as ``log_probs``.
Tensor, The Connectionist Temporal Classification (CTC) loss between ``log_probs`` and ``labels``. If attr:`reduction` is ``'none'``, the shape of loss is [batch_size], otherwise, the shape of loss is [1]. Data type is the same as ``log_probs``.
...
@@ -1895,6 +1895,130 @@ def ctc_loss(
...
@@ -1895,6 +1895,130 @@ def ctc_loss(
returnloss_out
returnloss_out
defrnnt_loss(
input,
label,
input_lengths,
label_lengths,
blank=0,
fastemit_lambda=0.001,
reduction='mean',
name=None,
):
"""
An operator integrating the open source Warp-Transducer library (https://github.com/b-flo/warp-transducer.git)
to compute Sequence Transduction with Recurrent Neural Networks (RNN-T) loss.
Parameters:
input (Tensor): The logprobs sequence with padding, which is a 4-D Tensor. The tensor shape is [B, Tmax， Umax, D], where Tmax， is the longest length of input logit sequence. The data type should be float32 or float64.
label (Tensor): The ground truth sequence with padding, which must be a 2-D Tensor. The tensor shape is [B, Umax], where Umax is the longest length of label sequence. The data type must be int32.
input_lengths (Tensor): The length for each input sequence, it should have shape [batch_size] and dtype int64.
label_lengths (Tensor): The length for each label sequence, it should have shape [batch_size] and dtype int64.
blank (int, optional): The blank label index of RNN-T loss, which is in the half-opened interval [0, B). The data type must be int32. Default is 0.
fastemit_lambda (float, default 0.001): Regularization parameter for FastEmit (https://arxiv.org/pdf/2010.11148.pdf)
reduction (string, optional): Indicate how to average the loss, the candicates are ``'none'`` | ``'mean'`` | ``'sum'``. If :attr:`reduction` is ``'mean'``, the output will be sum of loss and be divided by the batch_size; If :attr:`reduction` is ``'sum'``, return the sum of loss; If :attr:`reduction` is ``'none'``, no reduction will be applied. Default is ``'mean'``.
name (str, optional): Name for the operation (optional, default is None). For more information, please refer to :ref:`api_guide_Name`.
Returns:
Tensor, The RNN-T loss between ``logprobs`` and ``labels``. If attr:`reduction` is ``'none'``, the shape of loss is [batch_size], otherwise, the shape of loss is [1]. Data type is the same as ``logprobs``.
fastemit_lambda (float, optional): Regularization parameter for FastEmit (https://arxiv.org/pdf/2010.11148.pdf)
reduction (string, optional): Specifies the reduction to apply to the output:
'none' | 'mean' | 'sum'. 'none': no reduction will be applied,
'mean': the output losses will be divided by the target lengths and
then the mean over the batch is taken. Default: 'mean'
Shape:
input: logprob Tensor of (batch x seqLength x labelLength x outputDim) containing output from network
label: 2 dimensional (batch, labelLength) Tensor containing all the targets of the batch with zero padded
input_lengths: Tensor of size (batch) containing size of each output sequence from the network
label_lengths: Tensor of (batch) containing label length of each example
Returns:
Tensor, The RNN-T loss between ``logprobs`` and ``labels``. If attr:`reduction` is ``'none'``, the shape of loss is [batch_size], otherwise, the shape of loss is [1]. Data type is the same as ``logprobs``.
