o yi @sTddlmZmZmZddlmZmZddlmZm Z ddl m Z Gddde Z dS))AnyListUnion)Tensortensor) _wil_compute _wil_update)MetriccseZdZUdZdZeed<dZeed<dZeed<e ed<e ed<e ed<d e ffd d Z d e e ee fd e e ee fddfddZde fddZZS) WordInfoLostaWord Information Lost (`WIL`_) is a metric of the performance of an automatic speech recognition system. This value indicates the percentage of words that were incorrectly predicted between a set of ground-truth sentences and a set of hypothesis sentences. The lower the value, the better the performance of the ASR system with a WordInfoLost of 0 being a perfect score. Word Information Lost rate can then be computed as: .. math:: wil = 1 - \frac{C}{N} + \frac{C}{P} where: - :math:`C` is the number of correct words, - :math:`N` is the number of words in the reference - :math:`P` is the number of words in the prediction As input to ``forward`` and ``update`` the metric accepts the following input: - ``preds`` (:class:`~List`): Transcription(s) to score as a string or list of strings - ``target`` (:class:`~List`): Reference(s) for each speech input as a string or list of strings As output of ``forward`` and ``compute`` the metric returns the following output: - ``wil`` (:class:`~torch.Tensor`): A tensor with the Word Information Lost score Args: kwargs: Additional keyword arguments, see :ref:`Metric kwargs` for more info. Examples: >>> from torchmetrics import WordInfoLost >>> preds = ["this is the prediction", "there is an other sample"] >>> target = ["this is the reference", "there is another one"] >>> wil = WordInfoLost() >>> wil(preds, target) tensor(0.6528) Fis_differentiablehigher_is_betterfull_state_updateerrors target_total preds_totalkwargsc sRtjdi||jdtddd|jdtddd|jdtddddS)Nrgsum)dist_reduce_fxrr)super__init__ add_stater)selfr __class__rI/home/ubuntu/.local/lib/python3.10/site-packages/torchmetrics/text/wil.pyrBszWordInfoLost.__init__predstargetreturnNcCs>t||\}}}|j|7_|j|7_|j|7_dS)z*Update state with predictions and targets.N)rrrr)rrrrrrrrrupdateKszWordInfoLost.updatecCst|j|j|jS)z$Calculate the Word Information Lost.)rrrr)rrrrcomputeRszWordInfoLost.compute)__name__ __module__ __qualname____doc__r bool__annotations__r r rrrrstrrrr __classcell__rrrrr s  "  . r N) typingrrrtorchrr torchmetrics.functional.text.wilrrtorchmetrics.metricr r rrrrs