ChjSSKJr SSKrSSKJrJr \(aSSKrSSKJ r SSK J r "SS5r g)) annotationsN) TYPE_CHECKINGAny)Tensor)SentenceTransformerc\rSrSrSrSrS S SjjrSSjrSSSjjr\ SSj5r SS jr SS jr S r g)SentenceEvaluator a Base class for all evaluators. Notably, this class introduces the ``greater_is_better`` and ``primary_metric`` attributes. The former is a boolean indicating whether a higher evaluation score is better, which is used for choosing the best checkpoint if ``load_best_model_at_end`` is set to ``True`` in the training arguments. The latter is a string indicating the primary metric for the evaluator. This has to be defined whenever the evaluator returns a dictionary of metrics, and the primary metric is the key pointing to the primary metric, i.e. the one that is used for model selection and/or logging. Extend this class and implement __call__ for custom evaluators. c SUlSUlg)NTgreater_is_betterprimary_metricselfs j/var/www/html/shao/venv/lib/python3.13/site-packages/sentence_transformers/evaluation/SentenceEvaluator.py__init__SentenceEvaluator.__init__s!%"Ncg)a7 This is called during training to evaluate the model. It returns a score for the evaluation with a higher score indicating a better result. Args: model: the model to evaluate output_path: path where predictions and metrics are written to epoch: the epoch where the evaluation takes place. This is used for the file prefixes. If this is -1, then we assume evaluation on test data. steps: the steps in the current epoch at time of the evaluation. This is used for the file prefixes. If this is -1, then we assume evaluation at the end of the epoch. Returns: Either a score for the evaluation with a higher score indicating a better result, or a dictionary with scores. If the latter is chosen, then `evaluator.primary_metric` must be defined N)rmodel output_pathepochstepss r__call__SentenceEvaluator.__call__s2 rcSSjnU(d*UR5VVs0sH upEXC"U5_M snn$UR5VVs0sHupEUS-U-U"U5_M nnn[US5(a:URRUS-5(dUS-UR-UlU$s snnfs snnf)Nc>[U5$![a Us$f=fN)float ValueError)values rmaybe_to_float@SentenceEvaluator.prefix_name_to_metrics..maybe_to_float:s% U|#   s _r)r"rreturnr)itemshasattrr startswith)rmetricsnamer#keyr"s rprefix_name_to_metrics(SentenceEvaluator.prefix_name_to_metrics9s  AHQ:3C..Q QMT]]_]_zs4#:#^E%::_] 4) * *43F3F3Q3QRVY\R\3]3]"&*t/B/B"BD  R]s B4 B:c<URRXX45 gr)model_card_dataset_evaluation_metrics)rrr*rsteps r store_metrics_in_model_card_data2SentenceEvaluator.store_metrics_in_model_card_dataGs 44TEPrcURRnURS5(aSUSS-nURS5nUSUn[ R "SSU5$![a N$f=f)z Returns a human-readable description of the evaluator: BinaryClassificationEvaluator -> Binary Classification 1. Replace "CE" prefix with "CrossEncoder" 2. Remove "Evaluator" from the class name 3. Add a space before every capital letter CE CrossEncoderN Evaluatorz([a-z])([A-Z])z \g<1> \g<2>) __class____name__r)index IndexErrorresub)r class_namer<s r descriptionSentenceEvaluator.descriptionLs^^,,   & &'*QR.8J $$[1E#FU+Jvv'DD   sA$$ A10A1c0$)zk Return a dictionary with all meaningful configuration values of the evaluator to store in the model card. rrs rget_config_dict!SentenceEvaluator.get_config_dictbs  rc (UR"U40UD6$)a- Call the encoder method of the model pass Args: model (SentenceTransformer): Model we are evaluating sentences (str | list[str] | np.ndarray): Text that we are embedding Returns: list[Tensor] | np.ndarray | Tensor | dict[str, Tensor] | list[dict[str, Tensor]]: The associated embedding )encode)rr sentenceskwargss r embed_inputsSentenceEvaluator.embed_inputshs ||I000rr )NrL) rrrz str | NonerintrrMr&zfloat | dict[str, float])r*dict[str, float]r+strr&rN)rr) rrr*dict[str, Any]rrMr2rMr&None)r&rO)r&rP)rrrHzstr | list[str] | np.ndarrayr&zPlist[Tensor] | np.ndarray | Tensor | dict[str, Tensor] | list[dict[str, Tensor]])r; __module__ __qualname____firstlineno____doc__rrr-r3propertyrArDrJ__static_attributes__rrrr r s # ik ( 7A QT be ! 6 `aQ(Q3AQJMQY\Q Q EE* 1"101 Z 1rr ) __future__rr>typingrrnumpynptorchr)sentence_transformers.SentenceTransformerrr rrrr^s&" %Mk1k1r