o
    ©ôjÅ*  ã                   @   sÒ   d dl Z d dlmZ d dlmZmZmZmZmZ d dl	Z	d dl	m
Z
 d dlmZ d dlmZmZmZmZmZmZ d dlmZ d dlmZ d d	lmZmZ d d
lmZmZ esYdgZes_ddgZG dd„ deƒZdS )é    N)ÚSequence)ÚAnyÚClassVarÚListÚOptionalÚUnion)ÚTensor)Ú_load_tokenizer_and_model)Ú$_ALLOWED_INFORMATION_MEASURE_LITERALÚ_get_dataloaderÚ_get_special_tokens_mapÚ_infolm_computeÚ_infolm_updateÚ_InformationMeasure)ÚMetric)Údim_zero_cat)Ú_MATPLOTLIB_AVAILABLEÚ_TRANSFORMERS_GREATER_EQUAL_4_4)Ú_AX_TYPEÚ_PLOT_OUT_TYPEúInfoLM.plotÚInfoLMc                       s‚  e Zd ZU dZdZee ed< ee ed< ee ed< ee ed< ddddddddddœ	Ze	ed	< 	
											d,de
eejf dedededee dee dee
eejf  dee dededededeeef ddf‡ fdd„Zedefd d!„ƒZd"e
eee f d#e
eee f ddfd$d%„Zde
eeeef f fd&d'„Z	d-d(ee
eee f  d)ee defd*d+„Z ‡  Z!S ).r   uÆ  Calculate `InfoLM`_.

    InfoLM measures a distance/divergence between predicted and reference sentence discrete distribution using one of
    the following information measures:

        - `KL divergence`_
        - `alpha divergence`_
        - `beta divergence`_
        - `AB divergence`_
        - `RÃ©nyi divergence`_
        - L1 distance
        - L2 distance
        - L-infinity distance
        - `Fisher-Rao distance`_

    `InfoLM`_ is a family of untrained embedding-based metrics which addresses some famous flaws of standard
    string-based metrics thanks to the usage of pre-trained masked language models. This family of metrics is mainly
    designed for summarization and data-to-text tasks.

    The implementation of this metric is fully based HuggingFace ``transformers``' package.

    As input to ``forward`` and ``update`` the metric accepts the following input:

    - ``preds`` (:class:`~Sequence`): An iterable of hypothesis corpus
    - ``target`` (:class:`~Sequence`): An iterable of reference corpus

    As output of ``forward`` and ``compute`` the metric returns the following output:

    -  ``infolm`` (:class:`~torch.Tensor`): If `return_sentence_level_score=True` return a tuple with a tensor
       with the corpus-level InfoLM score and a list of sentence-level InfoLM scores, else return a corpus-level
       InfoLM score

    Args:
        model_name_or_path:
            A name or a model path used to load ``transformers`` pretrained model.
            By default the `"bert-base-uncased"` model is used.
        temperature:
            A temperature for calibrating language modelling. For more information, please reference `InfoLM`_ paper.
        information_measure:
            A name of information measure to be used. Please use one of: ['kl_divergence', 'alpha_divergence',
            'beta_divergence', 'ab_divergence', 'renyi_divergence', 'l1_distance', 'l2_distance', 'l_infinity_distance',
            'fisher_rao_distance']
        idf:
            An indication of whether normalization using inverse document frequencies should be used.
        alpha:
            Alpha parameter of the divergence used for alpha, AB and RÃ©nyi divergence measures.
        beta:
            Beta parameter of the divergence used for beta and AB divergence measures.
        device:
            A device to be used for calculation.
        max_length:
            A maximum length of input sequences. Sequences longer than ``max_length`` are to be trimmed.
        batch_size:
            A batch size used for model processing.
        num_threads:
            A number of threads to use for a dataloader.
        verbose:
            An indication of whether a progress bar to be displayed during the embeddings calculation.
        return_sentence_level_score:
            An indication whether a sentence-level InfoLM score to be returned.

    Example:
        >>> from torchmetrics.text.infolm import InfoLM
        >>> preds = ['he read the book because he was interested in world history']
        >>> target = ['he was interested in world history because he read the book']
        >>> infolm = InfoLM('google/bert_uncased_L-2_H-128_A-2', idf=False)
        >>> infolm(preds, target)
        tensor(-0.1784)

    FÚpreds_input_idsÚpreds_attention_maskÚtarget_input_idsÚtarget_attention_maskT)	Úkl_divergenceÚalpha_divergenceÚbeta_divergenceÚab_divergenceÚrenyi_divergenceÚl1_distanceÚl2_distanceÚl_infinity_distanceÚfisher_rao_distanceÚ%_information_measure_higher_is_betterúbert-base-uncasedç      Ð?r   Né@   r   Úmodel_name_or_pathÚtemperatureÚinformation_measureÚidfÚalphaÚbetaÚdeviceÚ
max_lengthÚ
batch_sizeÚnum_threadsÚverboseÚreturn_sentence_level_scoreÚkwargsÚreturnc                    sÞ   t ƒ jdi |¤Ž || _|| _|| _|| _|| _|| _t 	|p d¡| _
|	| _|
| _|| _|| _t||ƒ\| _| _t|||ƒ| _|pE| jjj| _t| jƒ| _| jdg dd� | jdg dd� | jdg dd� | jdg dd� d S )	NÚcpur   Úcat)Údist_reduce_fxr   r   r   © )ÚsuperÚ__init__r)   r*   r+   r,   r-   r.   Útorchr/   Ú_devicer1   r2   r3   r4   r	   Ú	tokenizerÚmodelr   Úinformation_measure_clsÚconfigr0   r   Úspecial_tokens_mapÚ	add_state)Úselfr)   r*   r+   r,   r-   r.   r/   r0   r1   r2   r3   r4   r5   ©Ú	__class__r:   úk/var/www/html/academy/tools/voices/clone/cosy-venv/lib/python3.10/site-packages/torchmetrics/text/infolm.pyr<   †   s(   zInfoLM.__init__c                 C   s   | j | j S )z¹Returns a bool indicating whether a higher value of the information measure is better.

        Done this way as depends on if the information measure is positive or negative.

        )r%   r+   )rE   r:   r:   rH   Úhigher_is_better­   s   zInfoLM.higher_is_betterÚpredsÚtargetc                 C   sN   t ||| j| jƒ\}}}}| j |¡ | j |¡ | j |¡ | j |¡ dS )z*Update state with predictions and targets.N)r   r?   r0   r   Úappendr   r   r   )rE   rJ   rK   r   r   r   r   r:   r:   rH   Úupdate¶   s   ÿzInfoLM.updatec              	   C   s„   t t| jƒt| jƒ| j| j| jd�}t t| jƒt| jƒ| j| j| jd�}t	| j
||| j| j| j| j| jƒ}| jr>| ¡ |fS | ¡ S )zLCalculate selected information measure using the pre-trained language model.)Ú	input_idsÚattention_maskr,   r1   Únum_workers)r   r   r   r   r,   r1   r2   r   r   r   r@   r*   rA   rC   r3   r4   Úmean)rE   Úpreds_dataloaderÚtarget_dataloaderÚinfo_lm_scorer:   r:   rH   ÚcomputeÀ   s6   ûûøzInfoLM.computeÚvalÚaxc                 C   s   |   ||¡S )a1  Plot a single or multiple values from the metric.

        Args:
            val: Either a single result from calling `metric.forward` or `metric.compute` or a list of these results.
                If no value is provided, will automatically call `metric.compute` and plot that result.
            ax: An matplotlib axis object. If provided will add plot to that axis

        Returns:
            Figure and Axes object

        Raises:
            ModuleNotFoundError:
                If `matplotlib` is not installed

        .. plot::
            :scale: 75

            >>> # Example plotting a single value
            >>> from torchmetrics.text.infolm import InfoLM
            >>> metric = InfoLM('google/bert_uncased_L-2_H-128_A-2', idf=False)
            >>> preds = ['he read the book because he was interested in world history']
            >>> target = ['he was interested in world history because he read the book']
            >>> metric.update(preds, target)
            >>> fig_, ax_ = metric.plot()

        .. plot::
            :scale: 75

            >>> # Example plotting multiple values
            >>> from torchmetrics.text.infolm import InfoLM
            >>> metric = InfoLM('google/bert_uncased_L-2_H-128_A-2', idf=False)
            >>> preds = ["this is the prediction", "there is an other sample"]
            >>> target = ["this is the reference", "there is another one"]
            >>> values = [ ]
            >>> for _ in range(10):
            ...     values.append(metric(preds, target))
            >>> fig_, ax_ = metric.plot(values)

        )Ú_plot)rE   rV   rW   r:   r:   rH   Úplotá   s   *r   )r&   r'   r   TNNNNr(   r   TF)NN)"Ú__name__Ú
__module__Ú__qualname__Ú__doc__Úis_differentiabler   r   Ú__annotations__r%   r   r   ÚstrÚosÚPathLikeÚfloatr
   Úboolr   r=   r/   ÚintÚdictr   r<   ÚpropertyrI   r   rM   ÚtuplerU   r   r   rY   Ú__classcell__r:   r:   rF   rH   r   *   sŒ   
 Gõóþýüûúùø	÷
öõôó
òñ'.
"ÿÿÿþ) ra   Úcollections.abcr   Útypingr   r   r   r   r   r=   r   Ú4torchmetrics.functional.text.helper_embedding_metricr	   Ú#torchmetrics.functional.text.infolmr
   r   r   r   r   r   Útorchmetrics.metricr   Útorchmetrics.utilities.datar   Útorchmetrics.utilities.importsr   r   Útorchmetrics.utilities.plotr   r   Ú__doctest_skip__r   r:   r:   r:   rH   Ú<module>   s     