siHddlZddlmZmZmZmZmZddlZddlmZm Z ddl m Z ej e ZGdde jZdS)N)CallableDictIterableTupleUnion)Tensornn)SentenceTransformerceZdZdddejfdedededededed ed d ffd Z d e e e e fde d ee ee e fffdZed e fdZxZS) SoftmaxLossTFmodelsentence_embedding_dimension num_labelsconcatenation_sent_repconcatenation_sent_difference!concatenation_sent_multiplicationloss_fctreturnNcztt|||_||_||_||_||_d}|r|dz }|r|dz }|r|dz }t d |tj ||z||j |_||_dS)a\ This loss was used in our SBERT publication (https://arxiv.org/abs/1908.10084) to train the SentenceTransformer model on NLI data. It adds a softmax classifier on top of the output of two transformer networks. :class:`MultipleNegativesRankingLoss` is an alternative loss function that often yields better results, as per https://arxiv.org/abs/2004.09813. Args: model (SentenceTransformer): The SentenceTransformer model. sentence_embedding_dimension (int): The dimension of the sentence embeddings. num_labels (int): The number of different labels. concatenation_sent_rep (bool): Whether to concatenate vectors u,v for the softmax classifier. Defaults to True. concatenation_sent_difference (bool): Whether to add abs(u-v) for the softmax classifier. Defaults to True. concatenation_sent_multiplication (bool): Whether to add u*v for the softmax classifier. Defaults to False. loss_fct (Callable): Custom pytorch loss function. If not set, uses nn.CrossEntropyLoss(). Defaults to nn.CrossEntropyLoss(). References: - Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks: https://arxiv.org/abs/1908.10084 - `Training Examples > Natural Language Inference <../../examples/training/nli/README.html>`_ Requirements: 1. sentence pairs with a class label Inputs: +---------------------------------------+--------+ | Texts | Labels | +=======================================+========+ | (sentence_A, sentence_B) pairs | class | +---------------------------------------+--------+ Example: :: from sentence_transformers import SentenceTransformer, SentenceTransformerTrainer, losses from datasets import Dataset model = SentenceTransformer("microsoft/mpnet-base") train_dataset = Dataset.from_dict({ "sentence1": [ "A person on a horse jumps over a broken down airplane.", "A person on a horse jumps over a broken down airplane.", "A person on a horse jumps over a broken down airplane.", "Children smiling and waving at camera", ], "sentence2": [ "A person is training his horse for a competition.", "A person is at a diner, ordering an omelette.", "A person is outdoors, on a horse.", "There are children present.", ], "label": [1, 2, 0, 0], }) loss = losses.SoftmaxLoss(model, model.get_sentence_embedding_dimension(), num_labels=3) trainer = SentenceTransformerTrainer( model=model, train_dataset=train_dataset, loss=loss, ) trainer.train() rz'Softmax loss: #Vectors concatenated: {})deviceN)superr __init__r rrrrloggerinfoformatr Linearr classifierr) selfr rrrrrrnum_vectors_concatenated __class__s [/var/www/icac/venv/lib/python3.11/site-packages/sentence_transformers/losses/SoftmaxLoss.pyrzSoftmaxLoss.__init__ sN k4  ))+++ $&<#-J*1R.#$ ! * $ ) $ ( * $ ) $ , * $ ) $ =DDE]^^___) $'C CZX]Xd   ! sentence_featureslabelscfd|D}|\}}g}jr*||||jr*|tj||z jr|||ztj|d}|}|+|| d} | S||fS)NcFg|]}|dS)sentence_embedding)r ).0sentence_featurer s r# z'SoftmaxLoss.forward..ks-mmmGW +,,-ABmmmr$r) rappendrtorchabsrcatrrview) r r%r&repsrep_arep_bvectors_concatfeaturesoutputlosss ` r#forwardzSoftmaxLoss.forwardhsnmmm[lmmm u  & )  ! !% ( ( (  ! !% ( ( (  - <  ! !%)EEM":": ; ; ;  1 1  ! !%%- 0 0 09^Q//**  ==R99DK< r$cdS)Na @inproceedings{reimers-2019-sentence-bert, title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks", author = "Reimers, Nils and Gurevych, Iryna", booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing", month = "11", year = "2019", publisher = "Association for Computational Linguistics", url = "https://arxiv.org/abs/1908.10084", } )r s r#citationzSoftmaxLoss.citations   r$)__name__ __module__ __qualname__r CrossEntropyLossr intboolrrrrstrrrrr:propertyr= __classcell__)r"s@r#r r s- (,.2270R022Y!Y!"Y!'*Y! Y! !% Y! (, Y!,0Y!Y! Y!Y!Y!Y!Y!Y!v !)$sF{*;!< FL vuVV^,, -    6 #   X     r$r )loggingtypingrrrrrr/rr )sentence_transformers.SentenceTransformerr getLoggerr>rModuler r<r$r#rLs99999999999999 IIIIII  8 $ $CCCCC")CCCCCr$