l^(hVddlmZddlmZmZddlmZGddejZy)) annotations)Tensornn) CrossEncodercJeZdZejfdfd ZddZxZS)CrossEntropyLossc t|||_||_t j di||_t|jts8t|jjdt|jdy)a Computes the Cross Entropy Loss for a CrossEncoder model. This loss is used to train a model to predict the correct class label for a given pair of sentences. The number of classes should be equal to the number of model output labels. Args: model (:class:`~sentence_transformers.cross_encoder.CrossEncoder`): A CrossEncoder model to be trained. activation_fn (:class:`~torch.nn.Module`): Activation function applied to the logits before computing the loss. Defaults to :class:`~torch.nn.Identity`. **kwargs: Additional keyword arguments passed to the underlying :class:`torch.nn.CrossEntropyLoss`. References: - :class:`torch.nn.CrossEntropyLoss` - `Cross Encoder > Training Examples > Natural Language Inference <../../../examples/cross_encoder/training/nli/README.html>`_ Requirements: 1. Your model can be initialized with `num_labels > 1` to predict multiple classes. 2. The number of dataset classes should be equal to the number of model output labels (`model.num_labels`). Inputs: +-------------------------------------------------+--------+-------------------------------+ | Texts | Labels | Number of Model Output Labels | +=================================================+========+===============================+ | (sentence_A, sentence_B) pairs | class | `num_classes` | +-------------------------------------------------+--------+-------------------------------+ Example: :: from sentence_transformers.cross_encoder import CrossEncoder, CrossEncoderTrainer, losses from datasets import Dataset model = CrossEncoder("microsoft/mpnet-base", num_labels=2) train_dataset = Dataset.from_dict({ "sentence1": ["How can I be a good geologist?", "What is the capital of France?"], "sentence2": ["What should I do to be a great geologist?", "What is the capital of Germany?"], "label": [1, 0], # 1: duplicate, 0: not duplicate }) loss = losses.CrossEntropyLoss(model) trainer = CrossEncoderTrainer( model=model, train_dataset=train_dataset, loss=loss, ) trainer.train() z? expects a model of type CrossEncoder, but got a model of type .N) super__init__model activation_fnrrce_loss isinstancer ValueError __class____name__type)selfrrkwargsrs y/var/www/skyplay_api_hub/venv/lib/python3.12/site-packages/sentence_transformers/cross_encoder/losses/CrossEntropyLoss.pyr zCrossEntropyLoss.__init__ s^  ***4V4 $**l3>>**+,++/ +;*r8s "ILryyLr