# BERT[[BERT]] ## 개요[[Overview]] BERT 모델은 Jacob Devlin. Ming-Wei Chang, Kenton Lee, Kristina Touranova가 제안한 논문 [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://huggingface.co/papers/1810.04805)에서 소개되었습니다. BERT는 사전 학습된 양방향 트랜스포머로, Toronto Book Corpus와 Wikipedia로 구성된 대규모 코퍼스에서 마스킹된 언어 모델링과 다음 문장 예측(Next Sentence Prediction) 목표를 결합해 학습되었습니다. 해당 논문의 초록입니다: *우리는 BERT(Bidirectional Encoder Representations from Transformers)라는 새로운 언어 표현 모델을 소개합니다. 최근의 다른 언어 표현 모델들과 달리, BERT는 모든 계층에서 양방향으로 양쪽 문맥을 조건으로 사용하여 비지도 학습된 텍스트에서 깊이 있는 양방향 표현을 사전 학습하도록 설계되었습니다. 그 결과, 사전 학습된 BERT 모델은 추가적인 출력 계층 하나만으로 질문 응답, 언어 추론과 같은 다양한 작업에서 미세 조정될 수 있으므로, 특정 작업을 위해 아키텍처를 수정할 필요가 없습니다.* *BERT는 개념적으로 단순하면서도 실증적으로 강력한 모델입니다. BERT는 11개의 자연어 처리 과제에서 새로운 최고 성능을 달성했으며, GLUE 점수를 80.5% (7.7% 포인트 절대 개선)로, MultiNLI 정확도를 86.7% (4.6% 포인트 절대 개선), SQuAD v1.1 질문 응답 테스트에서 F1 점수를 93.2 (1.5% 포인트 절대 개선)로, SQuAD v2.0에서 F1 점수를 83.1 (5.1% 포인트 절대 개선)로 향상시켰습니다.* 이 모델은 [thomwolf](https://huggingface.co/thomwolf)가 기여하였습니다. 원본 코드는 [여기](https://github.com/google-research/bert)에서 확인할 수 있습니다. ## 사용 팁[[Usage tips]] - BERT는 절대 위치 임베딩을 사용하는 모델이므로 입력을 왼쪽이 아니라 오른쪽에서 패딩하는 것이 일반적으로 권장됩니다. - BERT는 마스킹된 언어 모델(MLM)과 Next Sentence Prediction(NSP) 목표로 학습되었습니다. 이는 마스킹된 토큰 예측과 전반적인 자연어 이해(NLU)에 뛰어나지만, 텍스트 생성에는 최적화되어있지 않습니다. - BERT의 사전 학습 과정에서는 입력 데이터를 무작위로 마스킹하여 일부 토큰을 마스킹합니다. 전체 토큰 중 약 15%가 다음과 같은 방식으로 마스킹됩니다: * 80% 확률로 마스크 토큰으로 대체 * 10% 확률로 임의의 다른 토큰으로 대체 * 10% 확률로 원래 토큰 그대로 유지 - 모델의 주요 목표는 원본 문장을 예측하는 것이지만, 두 번째 목표가 있습니다: 입력으로 문장 A와 B (사이에는 구분 토큰이 있음)가 주어집니다. 이 문장 쌍이 연속될 확률은 50%이며, 나머지 50%는 서로 무관한 문장들입니다. 모델은 이 두 문장이 아닌지를 예측해야 합니다. ### Scaled Dot Product Attention(SDPA) 사용하기 [[Using Scaled Dot Product Attention (SDPA)]] Pytorch는 `torch.nn.functional`의 일부로 Scaled Dot Product Attention(SDPA) 연산자를 기본적으로 제공합니다. 이 함수는 입력과 하드웨어에 따라 여러 구현 방식을 사용할 수 있습니다. 자세한 내용은 [공식 문서](https://pytorch.org/docs/stable/generated/torch.nn.functional.scaled_dot_product_attention.html)나 [GPU Inference](https://huggingface.co/docs/transformers/main/en/perf_infer_gpu_one#pytorch-scaled-dot-product-attention)에서 확인할 수 있습니다. `torch>=2.1.1`에서는 구현이 가능한 경우 SDPA가 기본적으로 사용되지만, `from_pretrained()`함수에서 `attn_implementation="sdpa"`를 설정하여 SDPA를 명시적으로 사용하도록 지정할 수도 있습니다. ``` from transformers import BertModel model = BertModel.from_pretrained("bert-base-uncased", dtype=torch.float16, attn_implementation="sdpa") ... ``` 최적 성능 향상을 위해 모델을 반정밀도(예: `torch.float16` 또는 `torch.bfloat16`)로 불러오는 것을 권장합니다. 로컬 벤치마크 (A100-80GB, CPUx12, RAM 96.6GB, PyTorch 2.2.0, OS Ubuntu 22.04)에서 `float16`을 사용해 학습 및 추론을 수행한 결과, 다음과 같은 속도 향상이 관찰되었습니다. #### 학습 [[Training]] |batch_size|seq_len|Time per batch (eager - s)|Time per batch (sdpa - s)|Speedup (%)|Eager peak mem (MB)|sdpa peak mem (MB)|Mem saving (%)| |----------|-------|--------------------------|-------------------------|-----------|-------------------|------------------|--------------| |4 |256 |0.023 |0.017 |35.472 |939.213 |764.834 |22.800 | |4 |512 |0.023 |0.018 |23.687 |1970.447 |1227.162 |60.569 | |8 |256 |0.023 |0.018 |23.491 |1594.295 |1226.114 |30.028 | |8 |512 |0.035 |0.025 |43.058 |3629.401 |2134.262 |70.054 | |16 |256 |0.030 |0.024 |25.583 |2874.426 |2134.262 |34.680 | |16 |512 |0.064 |0.044 |46.223 |6964.659 |3961.013 |75.830 | #### 추론 [[Inference]] |batch_size|seq_len|Per token latency eager (ms)|Per token latency SDPA (ms)|Speedup (%)|Mem eager (MB)|Mem BT (MB)|Mem saved (%)| |----------|-------|----------------------------|---------------------------|-----------|--------------|-----------|-------------| |1 |128 |5.736 |4.987 |15.022 |282.661 |282.924 |-0.093 | |1 |256 |5.689 |4.945 |15.055 |298.686 |298.948 |-0.088 | |2 |128 |6.154 |4.982 |23.521 |314.523 |314.785 |-0.083 | |2 |256 |6.201 |4.949 |25.303 |347.546 |347.033 |0.148 | |4 |128 |6.049 |4.987 |21.305 |378.895 |379.301 |-0.107 | |4 |256 |6.285 |5.364 |17.166 |443.209 |444.382 |-0.264 | ## 자료[[Resources]] BERT를 시작하는 데 도움이 되는 Hugging Face와 community 자료 목록(🌎로 표시됨) 입니다. 여기에 포함될 자료를 제출하고 싶다면 PR(Pull Request)를 열어주세요. 리뷰 해드리겠습니다! 자료는 기존 자료를 복제하는 대신 새로운 내용을 담고 있어야 합니다. - [BERT 텍스트 분류 (다른 언어로)](https://www.philschmid.de/bert-text-classification-in-a-different-language)에 대한 블로그 포스트. - [다중 레이블 텍스트 분류를 위한 BERT (및 관련 모델) 미세 조정](https://colab.research.google.com/github/NielsRogge/Transformers-Tutorials/blob/master/BERT/Fine_tuning_BERT_(and_friends)_for_multi_label_text_classification.ipynb)에 대한 노트북. - [PyTorch를 이용해 BERT를 다중 레이블 분류를 위해 미세 조정하는 방법](htt기ps://colab.research.google.com/github/abhimishra91/transformers-tutorials/blob/master/transformers_multi_label_classification.ipynb)에 대한 노트북. 🌎 - [BERT로 EncoderDecoder 모델을 warm-start하여 요약하기](https://colab.research.google.com/github/patrickvonplaten/notebooks/blob/master/BERT2BERT_for_CNN_Dailymail.ipynb)에 대한 노트북. - [BertForSequenceClassification](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForSequenceClassification)이 [예제 스크립트](https://github.com/huggingface/transformers/tree/main/examples/pytorch/text-classification)와 [노트북](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/text_classification.ipynb)에서 지원됩니다. - `TFBertForSequenceClassification`이 [예제 스크립트](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/text-classification)와 [노트북](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/text_classification-tf.ipynb)에서 지원됩니다. - `FlaxBertForSequenceClassification`이 [예제 스크립트](https://github.com/huggingface/transformers/tree/main/examples/flax/text-classification)와 [노트북](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/text_classification_flax.ipynb)에서 지원됩니다. - [텍스트 분류 작업 가이드](../tasks/sequence_classification) - [Keras와 함께 Hugging Face Transformers를 사용하여 비영리 BERT를 개체명 인식(NER)용으로 미세 조정하는 방법](https://www.philschmid.de/huggingface-transformers-keras-tf)에 대한 블로그 포스트. - [BERT를 개체명 인식을 위해 미세 조정하기](https://colab.research.google.com/github/NielsRogge/Transformers-Tutorials/blob/master/BERT/Custom_Named_Entity_Recognition_with_BERT_only_first_wordpiece.ipynb)에 대한 노트북. 각 단어의 첫 번째 wordpiece에만 레이블을 지정하여 학습하는 방법을 설명합니다. 모든 wordpiece에 레이블을 전파하는 방법은 [이 버전](https://github.com/NielsRogge/Transformers-Tutorials/blob/master/BERT/Custom_Named_Entity_Recognition_with_BERT.ipynb)에서 확인할 수 있습니다. - [BertForTokenClassification](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForTokenClassification)이 [예제 스크립트](https://github.com/huggingface/transformers/tree/main/examples/pytorch/token-classification)와 [노트북](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/token_classification.ipynb)에서 지원됩니다. - `TFBertForTokenClassification`이 [예제 스크립트](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/token-classification)와 [노트북](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/token_classification-tf.ipynb)에서 지원됩니다. - `FlaxBertForTokenClassification`이 [예제 스크립트](https://github.com/huggingface/transformers/tree/main/examples/flax/token-classification)에서 지원됩니다. - 🤗 Hugging Face 코스의 [토큰 분류 챕터](https://huggingface.co/course/chapter7/2?fw=pt). - [토큰 분류 작업 가이드](../tasks/token_classification) - [BertForMaskedLM](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForMaskedLM)이 [예제 스크립트](https://github.com/huggingface/transformers/tree/main/examples/pytorch/language-modeling#robertabertdistilbert-and-masked-language-modeling)와 [노트북](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/language_modeling.ipynb)에서 지원됩니다. - `TFBertForMaskedLM`이 [예제 스크립트](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/language-modeling#run_mlmpy) 와 [노트북](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/language_modeling-tf.ipynb)에서 지원됩니다. - `FlaxBertForMaskedLM`이 [예제 스크립트](https://github.com/huggingface/transformers/tree/main/examples/flax/language-modeling#masked-language-modeling)와 [노트북](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/masked_language_modeling_flax.ipynb)에서 지원됩니다. - 🤗 Hugging Face 코스의 [마스킹된 언어 모델링 챕터](https://huggingface.co/course/chapter7/3?fw=pt). - [마스킹된 언어 모델링 작업 가이드](../tasks/masked_language_modeling) - [BertForQuestionAnswering](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForQuestionAnswering)이 [예제 스크립트](https://github.com/huggingface/transformers/tree/main/examples/pytorch/question-answering)와 [노트북](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/question_answering.ipynb)에서 지원됩니다. - `TFBertForQuestionAnswering`이 [예제 스크립트](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/question-answering) 와 [노트북](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/question_answering-tf.ipynb)에서 지원됩니다. - `FlaxBertForQuestionAnswering`이 [예제 스크립트](https://github.com/huggingface/transformers/tree/main/examples/flax/question-answering)에서 지원됩니다. - 🤗 Hugging Face 코스의 [질문 답변 챕터](https://huggingface.co/course/chapter7/7?fw=pt). - [질문 답변 작업 가이드](../tasks/question_answering) **다중 선택** - [BertForMultipleChoice](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForMultipleChoice)이 [예제 스크립트](https://github.com/huggingface/transformers/tree/main/examples/pytorch/multiple-choice)와 [노트북](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/multiple_choice.ipynb)에서 지원됩니다. - `TFBertForMultipleChoice`이 [에제 스크립트](https://github.com/huggingface/transformers/tree/main/examples/tensorflow/multiple-choice)와 [노트북](https://colab.research.google.com/github/huggingface/notebooks/blob/main/examples/multiple_choice-tf.ipynb)에서 지원됩니다. - [다중 선택 작업 가이드](../tasks/multiple_choice) ⚡️ **추론** - [Hugging Face Transformers와 AWS Inferentia를 사용하여 BERT 추론을 가속화하는 방법](https://huggingface.co/blog/bert-inferentia-sagemaker)에 대한 블로그 포스트. - [GPU에서 DeepSpeed-Inference로 BERT 추론을 가속화하는 방법](https://www.philschmid.de/bert-deepspeed-inference)에 대한 블로그 포스트. ⚙️ **사전 학습** - [Hugging Face Optimum으로 Transformers를 ONMX로 변환하는 방법](https://www.philschmid.de/pre-training-bert-habana)에 대한 블로그 포스트. 🚀 **배포** - [Hugging Face Optimum으로 Transformers를 ONMX로 변환하는 방법](https://www.philschmid.de/convert-transformers-to-onnx)에 대한 블로그 포스트. - [AWS에서 Hugging Face Transformers를 위한 Habana Gaudi 딥러닝 환경 설정 방법](https://www.philschmid.de/getting-started-habana-gaudi#conclusion)에 대한 블로그 포스트. - [Hugging Face Transformers, Amazon SageMaker 및 Terraform 모듈을 이용한 BERT 자동 확장](https://www.philschmid.de/terraform-huggingface-amazon-sagemaker-advanced)에 대한 블로그 포스트. - [Hugging Face, AWS Lambda, Docker를 활용하여 서버리스 BERT 설정하는 방법](https://www.philschmid.de/serverless-bert-with-huggingface-aws-lambda-docker)에 대한 블로그 포스트. - [Amazon SageMaker와 Training Compiler를 사용하여 Hugging Face Transformers에서 BERT 미세 조정하는 방법](https://www.philschmid.de/huggingface-amazon-sagemaker-training-compiler)에 대한 블로그. - [Amazon SageMaker를 사용한 Transformers와 BERT의 작업별 지식 증류](https://www.philschmid.de/knowledge-distillation-bert-transformers)에 대한 블로그 포스트. ## BertConfig[[transformers.BertConfig]] #### transformers.BertConfig[[transformers.BertConfig]] [Source](https://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/configuration_bert.py#L25) This is the configuration class to store the configuration of a BertModel. It is used to instantiate a Bert model according to the specified arguments, defining the model architecture. Instantiating a configuration with the defaults will yield a similar configuration to that of the [google-bert/bert-base-uncased](https://huggingface.co/google-bert/bert-base-uncased) Configuration objects inherit from [PreTrainedConfig](/docs/transformers/v5.6.0/ko/main_classes/configuration#transformers.PreTrainedConfig) and can be used to control the model outputs. Read the documentation from [PreTrainedConfig](/docs/transformers/v5.6.0/ko/main_classes/configuration#transformers.PreTrainedConfig) for more information. Examples: ```python >>> from transformers import BertConfig, BertModel >>> # Initializing a BERT google-bert/bert-base-uncased style configuration >>> configuration = BertConfig() >>> # Initializing a model (with random weights) from the google-bert/bert-base-uncased style configuration >>> model = BertModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config ``` **Parameters:** vocab_size (`int`, *optional*, defaults to `30522`) : Vocabulary size of the model. Defines the number of different tokens that can be represented by the `input_ids`. hidden_size (`int`, *optional*, defaults to `768`) : Dimension of the hidden representations. num_hidden_layers (`int`, *optional*, defaults to `12`) : Number of hidden layers in the Transformer decoder. num_attention_heads (`int`, *optional*, defaults to `12`) : Number of attention heads for each attention layer in the Transformer decoder. intermediate_size (`int`, *optional*, defaults to `3072`) : Dimension of the MLP representations. hidden_act (`str`, *optional*, defaults to `gelu`) : The non-linear activation function (function or string) in the decoder. For example, `"gelu"`, `"relu"`, `"silu"`, etc. hidden_dropout_prob (`Union[float, int]`, *optional*, defaults to `0.1`) : The dropout probability for all fully connected layers in the embeddings, encoder, and pooler. attention_probs_dropout_prob (`Union[float, int]`, *optional*, defaults to `0.1`) : The dropout ratio for the attention probabilities. max_position_embeddings (`int`, *optional*, defaults to `512`) : The maximum sequence length that this model might ever be used with. type_vocab_size (`int`, *optional*, defaults to `2`) : The vocabulary size of the `token_type_ids`. initializer_range (`float`, *optional*, defaults to `0.02`) : The standard deviation of the truncated_normal_initializer for initializing all weight matrices. layer_norm_eps (`float`, *optional*, defaults to `1e-12`) : The epsilon used by the layer normalization layers. pad_token_id (`int`, *optional*, defaults to `0`) : Token id used for padding in the vocabulary. use_cache (`bool`, *optional*, defaults to `True`) : Whether or not the model should return the last key/values attentions (not used by all models). Only relevant if `config.is_decoder=True` or when the model is a decoder-only generative model. classifier_dropout (`Union[float, int]`, *optional*) : The dropout ratio for classifier. is_decoder (`bool`, *optional*, defaults to `False`) : Whether the model is used as a decoder or not. If `False`, the model is used as an encoder. add_cross_attention (`bool`, *optional*, defaults to `False`) : Whether cross-attention layers should be added to the model. bos_token_id (`int`, *optional*) : Token id used for beginning-of-stream in the vocabulary. eos_token_id (`Union[int, list[int]]`, *optional*) : Token id used for end-of-stream in the vocabulary. tie_word_embeddings (`bool`, *optional*, defaults to `True`) : Whether to tie weight embeddings according to model's `tied_weights_keys` mapping. ## BertTokenizer[[transformers.BertTokenizer]] #### transformers.BertTokenizer[[transformers.BertTokenizer]] [Source](https://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/tokenization_bert.py#L41) Construct a BERT tokenizer (backed by HuggingFace's tokenizers library). Based on WordPiece. This tokenizer inherits from [TokenizersBackend](/docs/transformers/v5.6.0/ko/main_classes/tokenizer#transformers.TokenizersBackend) which contains most of the main methods. Users should refer to this superclass for more information regarding those methods. get_special_tokens_masktransformers.BertTokenizer.get_special_tokens_maskhttps://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/tokenization_utils_base.py#L1315[{"name": "token_ids_0", "val": ": list[int]"}, {"name": "token_ids_1", "val": ": list[int] | None = None"}, {"name": "already_has_special_tokens", "val": ": bool = False"}]- **token_ids_0** -- List of IDs for the (possibly already formatted) sequence. - **token_ids_1** -- Unused when `already_has_special_tokens=True`. Must be None in that case. - **already_has_special_tokens** -- Whether the sequence is already formatted with special tokens.0A list of integers in the range [0, 1]1 for a special token, 0 for a sequence token. Retrieve sequence ids from a token list that has no special tokens added. For fast tokenizers, data collators call this with `already_has_special_tokens=True` to build a mask over an already-formatted sequence. In that case, we compute the mask by checking membership in `all_special_ids`. **Parameters:** vocab (`str` or `dict[str, int]`, *optional*) : Custom vocabulary dictionary. If not provided, vocabulary is loaded from `vocab_file`. do_lower_case (`bool`, *optional*, defaults to `True`) : Whether or not to lowercase the input when tokenizing. unk_token (`str`, *optional*, defaults to `"[UNK]"`) : The unknown token. A token that is not in the vocabulary cannot be converted to an ID and is set to be this token instead. sep_token (`str`, *optional*, defaults to `"[SEP]"`) : The separator token, which is used when building a sequence from multiple sequences, e.g. two sequences for sequence classification or for a text and a question for question answering. It is also used as the last token of a sequence built with special tokens. pad_token (`str`, *optional*, defaults to `"[PAD]"`) : The token used for padding, for example when batching sequences of different lengths. cls_token (`str`, *optional*, defaults to `"[CLS]"`) : The classifier token which is used when doing sequence classification (classification of the whole sequence instead of per-token classification). It is the first token of the sequence when built with special tokens. mask_token (`str`, *optional*, defaults to `"[MASK]"`) : The token used for masking values. This is the token used when training this model with masked language modeling. This is the token which the model will try to predict. tokenize_chinese_chars (`bool`, *optional*, defaults to `True`) : Whether or not to tokenize Chinese characters. strip_accents (`bool`, *optional*) : Whether or not to strip all accents. If this option is not specified, then it will be determined by the value for `lowercase` (as in the original BERT). **Returns:** `A list of integers in the range [0, 1]` 1 for a special token, 0 for a sequence token. #### save_vocabulary[[transformers.BertTokenizer.save_vocabulary]] [Source](https://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/tokenization_utils_tokenizers.py#L509) ## BertTokenizerLegacy[[transformers.BertTokenizerLegacy]] #### transformers.BertTokenizerLegacy[[transformers.BertTokenizerLegacy]] [Source](https://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/tokenization_bert_legacy.py#L49) Construct a BERT tokenizer. Based on WordPiece. This tokenizer inherits from [PreTrainedTokenizer](/docs/transformers/v5.6.0/ko/main_classes/tokenizer#transformers.PythonBackend) which contains most of the main methods. Users should refer to this superclass for more information regarding those methods. build_inputs_with_special_tokenstransformers.BertTokenizerLegacy.build_inputs_with_special_tokenshttps://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/tokenization_bert_legacy.py#L184[{"name": "token_ids_0", "val": ": list"}, {"name": "token_ids_1", "val": ": list[int] | None = None"}]- **token_ids_0** (`List[int]`) -- List of IDs to which the special tokens will be added. - **token_ids_1** (`List[int]`, *optional*) -- Optional second list of IDs for sequence pairs.0`List[int]`List of [input IDs](../glossary#input-ids) with the appropriate special tokens. Build model inputs from a sequence or a pair of sequence for sequence classification tasks by concatenating and adding special tokens. A BERT sequence has the following format: - single sequence: `[CLS] X [SEP]` - pair of sequences: `[CLS] A [SEP] B [SEP]` **Parameters:** vocab_file (`str`) : File containing the vocabulary. do_lower_case (`bool`, *optional*, defaults to `True`) : Whether or not to lowercase the input when tokenizing. do_basic_tokenize (`bool`, *optional*, defaults to `True`) : Whether or not to do basic tokenization before WordPiece. never_split (`Iterable`, *optional*) : Collection of tokens which will never be split during tokenization. Only has an effect when `do_basic_tokenize=True` unk_token (`str`, *optional*, defaults to `"[UNK]"`) : The unknown token. A token that is not in the vocabulary cannot be converted to an ID and is set to be this token instead. sep_token (`str`, *optional*, defaults to `"[SEP]"`) : The separator token, which is used when building a sequence from multiple sequences, e.g. two sequences for sequence classification or for a text and a question for question answering. It is also used as the last token of a sequence built with special tokens. pad_token (`str`, *optional*, defaults to `"[PAD]"`) : The token used for padding, for example when batching sequences of different lengths. cls_token (`str`, *optional*, defaults to `"[CLS]"`) : The classifier token which is used when doing sequence classification (classification of the whole sequence instead of per-token classification). It is the first token of the sequence when built with special tokens. mask_token (`str`, *optional*, defaults to `"[MASK]"`) : The token used for masking values. This is the token used when training this model with masked language modeling. This is the token which the model will try to predict. tokenize_chinese_chars (`bool`, *optional*, defaults to `True`) : Whether or not to tokenize Chinese characters. This should likely be deactivated for Japanese (see this [issue](https://github.com/huggingface/transformers/issues/328)). strip_accents (`bool`, *optional*) : Whether or not to strip all accents. If this option is not specified, then it will be determined by the value for `lowercase` (as in the original BERT). clean_up_tokenization_spaces (`bool`, *optional*, defaults to `True`) : Whether or not to cleanup spaces after decoding, cleanup consists in removing potential artifacts like extra spaces. **Returns:** ``List[int]`` List of [input IDs](../glossary#input-ids) with the appropriate special tokens. #### convert_tokens_to_string[[transformers.BertTokenizerLegacy.convert_tokens_to_string]] [Source](https://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/tokenization_bert_legacy.py#L179) Converts a sequence of tokens (string) in a single string. #### get_special_tokens_mask[[transformers.BertTokenizerLegacy.get_special_tokens_mask]] [Source](https://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/tokenization_bert_legacy.py#L209) Retrieve sequence ids from a token list that has no special tokens added. This method is called when adding special tokens using the tokenizer `prepare_for_model` method. **Parameters:** token_ids_0 (`List[int]`) : List of IDs. token_ids_1 (`List[int]`, *optional*) : Optional second list of IDs for sequence pairs. already_has_special_tokens (`bool`, *optional*, defaults to `False`) : Whether or not the token list is already formatted with special tokens for the model. **Returns:** ``List[int]`` A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token. ## BertTokenizerFast[[transformers.BertTokenizer]] #### transformers.BertTokenizer[[transformers.BertTokenizer]] [Source](https://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/tokenization_bert.py#L41) Construct a BERT tokenizer (backed by HuggingFace's tokenizers library). Based on WordPiece. This tokenizer inherits from [TokenizersBackend](/docs/transformers/v5.6.0/ko/main_classes/tokenizer#transformers.TokenizersBackend) which contains most of the main methods. Users should refer to this superclass for more information regarding those methods. **Parameters:** vocab (`str` or `dict[str, int]`, *optional*) : Custom vocabulary dictionary. If not provided, vocabulary is loaded from `vocab_file`. do_lower_case (`bool`, *optional*, defaults to `True`) : Whether or not to lowercase the input when tokenizing. unk_token (`str`, *optional*, defaults to `"[UNK]"`) : The unknown token. A token that is not in the vocabulary cannot be converted to an ID and is set to be this token instead. sep_token (`str`, *optional*, defaults to `"[SEP]"`) : The separator token, which is used when building a sequence from multiple sequences, e.g. two sequences for sequence classification or for a text and a question for question answering. It is also used as the last token of a sequence built with special tokens. pad_token (`str`, *optional*, defaults to `"[PAD]"`) : The token used for padding, for example when batching sequences of different lengths. cls_token (`str`, *optional*, defaults to `"[CLS]"`) : The classifier token which is used when doing sequence classification (classification of the whole sequence instead of per-token classification). It is the first token of the sequence when built with special tokens. mask_token (`str`, *optional*, defaults to `"[MASK]"`) : The token used for masking values. This is the token used when training this model with masked language modeling. This is the token which the model will try to predict. tokenize_chinese_chars (`bool`, *optional*, defaults to `True`) : Whether or not to tokenize Chinese characters. strip_accents (`bool`, *optional*) : Whether or not to strip all accents. If this option is not specified, then it will be determined by the value for `lowercase` (as in the original BERT). ## Bert specific outputs[[transformers.models.bert.modeling_bert.BertForPreTrainingOutput]] #### transformers.models.bert.modeling_bert.BertForPreTrainingOutput[[transformers.models.bert.modeling_bert.BertForPreTrainingOutput]] [Source](https://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/modeling_bert.py#L567) Output type of [BertForPreTraining](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForPreTraining). **Parameters:** loss (`*optional*`, returned when `labels` is provided, `torch.FloatTensor` of shape `(1,)`) : Total loss as the sum of the masked language modeling loss and the next sequence prediction (classification) loss. prediction_logits (`torch.FloatTensor` of shape `(batch_size, sequence_length, config.vocab_size)`) : Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax). seq_relationship_logits (`torch.FloatTensor` of shape `(batch_size, 2)`) : Prediction scores of the next sequence prediction (classification) head (scores of True/False continuation before SoftMax). hidden_states (`tuple[torch.FloatTensor]`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`) : Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`. Hidden-states of the model at the output of each layer plus the optional initial embedding outputs. attentions (`tuple[torch.FloatTensor]`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`) : Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, sequence_length)`. Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads. ## BertModel[[transformers.BertModel]] #### transformers.BertModel[[transformers.BertModel]] [Source](https://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/modeling_bert.py#L598) The model can behave as an encoder (with only self-attention) as well as a decoder, in which case a layer of cross-attention is added between the self-attention layers, following the architecture described in [Attention is all you need](https://huggingface.co/papers/1706.03762) by Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser and Illia Polosukhin. To behave as an decoder the model needs to be initialized with the `is_decoder` argument of the configuration set to `True`. To be used in a Seq2Seq model, the model needs to initialized with both `is_decoder` argument and `add_cross_attention` set to `True`; an `encoder_hidden_states` is then expected as an input to the forward pass. This model inherits from [PreTrainedModel](/docs/transformers/v5.6.0/ko/main_classes/model#transformers.PreTrainedModel). Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.) This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior. forwardtransformers.BertModel.forwardhttps://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/modeling_bert.py#L624[{"name": "input_ids", "val": ": torch.Tensor | None = None"}, {"name": "attention_mask", "val": ": torch.Tensor | None = None"}, {"name": "token_type_ids", "val": ": torch.Tensor | None = None"}, {"name": "position_ids", "val": ": torch.Tensor | None = None"}, {"name": "inputs_embeds", "val": ": torch.Tensor | None = None"}, {"name": "encoder_hidden_states", "val": ": torch.Tensor | None = None"}, {"name": "encoder_attention_mask", "val": ": torch.Tensor | None = None"}, {"name": "past_key_values", "val": ": transformers.cache_utils.Cache | None = None"}, {"name": "use_cache", "val": ": bool | None = None"}, {"name": "**kwargs", "val": ": typing_extensions.Unpack[transformers.utils.generic.TransformersKwargs]"}]- **input_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Indices of input sequence tokens in the vocabulary. Padding will be ignored by default. Indices can be obtained using [AutoTokenizer](/docs/transformers/v5.6.0/ko/model_doc/auto#transformers.AutoTokenizer). See [PreTrainedTokenizer.encode()](/docs/transformers/v5.6.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.encode) and [PreTrainedTokenizer.__call__()](/docs/transformers/v5.6.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.__call__) for details. [What are input IDs?](../glossary#input-ids) - **attention_mask** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`: - 1 for tokens that are **not masked**, - 0 for tokens that are **masked**. [What are attention masks?](../glossary#attention-mask) - **token_type_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Segment token indices to indicate first and second portions of the inputs. Indices are selected in `[0, 1]`: - 0 corresponds to a *sentence A* token, - 1 corresponds to a *sentence B* token. [What are token type IDs?](../glossary#token-type-ids) - **position_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0, config.n_positions - 1]`. [What are position IDs?](../glossary#position-ids) - **inputs_embeds** (`torch.Tensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*) -- Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert `input_ids` indices into associated vectors than the model's internal embedding lookup matrix. - **encoder_hidden_states** (`torch.Tensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*) -- Sequence of hidden-states at the output of the last layer of the encoder. Used in the cross-attention if the model is configured as a decoder. - **encoder_attention_mask** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Mask to avoid performing attention on the padding token indices of the encoder input. This mask is used in the cross-attention if the model is configured as a decoder. Mask values selected in `[0, 1]`: - 1 for tokens that are **not masked**, - 0 for tokens that are **masked**. - **past_key_values** (`~cache_utils.Cache`, *optional*) -- Pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention blocks) that can be used to speed up sequential decoding. This typically consists in the `past_key_values` returned by the model at a previous stage of decoding, when `use_cache=True` or `config.use_cache=True`. Only [Cache](/docs/transformers/v5.6.0/ko/internal/generation_utils#transformers.Cache) instance is allowed as input, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache). If no `past_key_values` are passed, [DynamicCache](/docs/transformers/v5.6.0/ko/internal/generation_utils#transformers.DynamicCache) will be initialized by default. The model will output the same cache format that is fed as input. If `past_key_values` are used, the user is expected to input only unprocessed `input_ids` (those that don't have their past key value states given to this model) of shape `(batch_size, unprocessed_length)` instead of all `input_ids` of shape `(batch_size, sequence_length)`. - **use_cache** (`bool`, *optional*) -- If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see `past_key_values`).0[BaseModelOutputWithPoolingAndCrossAttentions](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.BaseModelOutputWithPoolingAndCrossAttentions) or `tuple(torch.FloatTensor)`A [BaseModelOutputWithPoolingAndCrossAttentions](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.BaseModelOutputWithPoolingAndCrossAttentions) or a tuple of `torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various elements depending on the configuration ([BertConfig](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertConfig)) and inputs. The [BertModel](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertModel) forward method, overrides the `__call__` special method. Although the recipe for forward pass needs to be defined within this function, one should call the `Module` instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them. - **last_hidden_state** (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`) -- Sequence of hidden-states at the output of the last layer of the model. - **pooler_output** (`torch.FloatTensor` of shape `(batch_size, hidden_size)`) -- Last layer hidden-state of the first token of the sequence (classification token) after further processing through the layers used for the auxiliary pretraining task. E.g. for BERT-family of models, this returns the classification token after processing through a linear layer and a tanh activation function. The linear layer weights are trained from the next sentence prediction (classification) objective during pretraining. - **hidden_states** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`) -- Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`. Hidden-states of the model at the output of each layer plus the optional initial embedding outputs. - **attentions** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`) -- Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, sequence_length)`. Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads. - **cross_attentions** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` and `config.add_cross_attention=True` is passed or when `config.output_attentions=True`) -- Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, sequence_length)`. Attentions weights of the decoder's cross-attention layer, after the attention softmax, used to compute the weighted average in the cross-attention heads. - **past_key_values** (`Cache`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`) -- It is a [Cache](/docs/transformers/v5.6.0/ko/internal/generation_utils#transformers.Cache) instance. For more details, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache). Contains pre-computed hidden-states (key and values in the self-attention blocks and optionally if `config.is_encoder_decoder=True` in the cross-attention blocks) that can be used (see `past_key_values` input) to speed up sequential decoding. **Parameters:** config ([BertModel](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertModel)) : Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the [from_pretrained()](/docs/transformers/v5.6.0/ko/main_classes/model#transformers.PreTrainedModel.from_pretrained) method to load the model weights. add_pooling_layer (`bool`, *optional*, defaults to `True`) : Whether to add a pooling layer **Returns:** `[BaseModelOutputWithPoolingAndCrossAttentions](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.BaseModelOutputWithPoolingAndCrossAttentions) or `tuple(torch.FloatTensor)`` A [BaseModelOutputWithPoolingAndCrossAttentions](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.BaseModelOutputWithPoolingAndCrossAttentions) or a tuple of `torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various elements depending on the configuration ([BertConfig](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertConfig)) and inputs. ## BertForPreTraining[[transformers.BertForPreTraining]] #### transformers.BertForPreTraining[[transformers.BertForPreTraining]] [Source](https://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/modeling_bert.py#L731) Bert Model with two heads on top as done during the pretraining: a `masked language modeling` head and a `next sentence prediction (classification)` head. This model inherits from [PreTrainedModel](/docs/transformers/v5.6.0/ko/main_classes/model#transformers.PreTrainedModel). Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.) This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior. forwardtransformers.BertForPreTraining.forwardhttps://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/modeling_bert.py#L753[{"name": "input_ids", "val": ": torch.Tensor | None = None"}, {"name": "attention_mask", "val": ": torch.Tensor | None = None"}, {"name": "token_type_ids", "val": ": torch.Tensor | None = None"}, {"name": "position_ids", "val": ": torch.Tensor | None = None"}, {"name": "inputs_embeds", "val": ": torch.Tensor | None = None"}, {"name": "labels", "val": ": torch.Tensor | None = None"}, {"name": "next_sentence_label", "val": ": torch.Tensor | None = None"}, {"name": "**kwargs", "val": ": typing_extensions.Unpack[transformers.utils.generic.TransformersKwargs]"}]- **input_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Indices of input sequence tokens in the vocabulary. Padding will be ignored by default. Indices can be obtained using [AutoTokenizer](/docs/transformers/v5.6.0/ko/model_doc/auto#transformers.AutoTokenizer). See [PreTrainedTokenizer.encode()](/docs/transformers/v5.6.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.encode) and [PreTrainedTokenizer.__call__()](/docs/transformers/v5.6.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.__call__) for details. [What are input IDs?](../glossary#input-ids) - **attention_mask** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`: - 1 for tokens that are **not masked**, - 0 for tokens that are **masked**. [What are attention masks?](../glossary#attention-mask) - **token_type_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Segment token indices to indicate first and second portions of the inputs. Indices are selected in `[0, 1]`: - 0 corresponds to a *sentence A* token, - 1 corresponds to a *sentence B* token. [What are token type IDs?](../glossary#token-type-ids) - **position_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0, config.n_positions - 1]`. [What are position IDs?](../glossary#position-ids) - **inputs_embeds** (`torch.Tensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*) -- Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert `input_ids` indices into associated vectors than the model's internal embedding lookup matrix. - **labels** (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*) -- Labels for computing the masked language modeling loss. Indices should be in `[-100, 0, ..., config.vocab_size]` (see `input_ids` docstring) Tokens with indices set to `-100` are ignored (masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]` - **next_sentence_label** (`torch.LongTensor` of shape `(batch_size,)`, *optional*) -- Labels for computing the next sequence prediction (classification) loss. Input should be a sequence pair (see `input_ids` docstring) Indices should be in `[0, 1]`: - 0 indicates sequence B is a continuation of sequence A, - 1 indicates sequence B is a random sequence.0[BertForPreTrainingOutput](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.models.bert.modeling_bert.BertForPreTrainingOutput) or `tuple(torch.FloatTensor)`A [BertForPreTrainingOutput](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.models.bert.modeling_bert.BertForPreTrainingOutput) or a tuple of `torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various elements depending on the configuration ([BertConfig](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertConfig)) and inputs. The [BertForPreTraining](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForPreTraining) forward method, overrides the `__call__` special method. Although the recipe for forward pass needs to be defined within this function, one should call the `Module` instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them. - **loss** (`*optional*`, returned when `labels` is provided, `torch.FloatTensor` of shape `(1,)`) -- Total loss as the sum of the masked language modeling loss and the next sequence prediction (classification) loss. - **prediction_logits** (`torch.FloatTensor` of shape `(batch_size, sequence_length, config.vocab_size)`) -- Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax). - **seq_relationship_logits** (`torch.FloatTensor` of shape `(batch_size, 2)`) -- Prediction scores of the next sequence prediction (classification) head (scores of True/False continuation before SoftMax). - **hidden_states** (`tuple[torch.FloatTensor]`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`) -- Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`. Hidden-states of the model at the output of each layer plus the optional initial embedding outputs. - **attentions** (`tuple[torch.FloatTensor]`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`) -- Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, sequence_length)`. Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads. Example: ```python >>> from transformers import AutoTokenizer, BertForPreTraining >>> import torch >>> tokenizer = AutoTokenizer.from_pretrained("google-bert/bert-base-uncased") >>> model = BertForPreTraining.from_pretrained("google-bert/bert-base-uncased") >>> inputs = tokenizer("Hello, my dog is cute", return_tensors="pt") >>> outputs = model(**inputs) >>> prediction_logits = outputs.prediction_logits >>> seq_relationship_logits = outputs.seq_relationship_logits ``` **Parameters:** config ([BertForPreTraining](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForPreTraining)) : Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the [from_pretrained()](/docs/transformers/v5.6.0/ko/main_classes/model#transformers.PreTrainedModel.from_pretrained) method to load the model weights. **Returns:** `[BertForPreTrainingOutput](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.models.bert.modeling_bert.BertForPreTrainingOutput) or `tuple(torch.FloatTensor)`` A [BertForPreTrainingOutput](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.models.bert.modeling_bert.BertForPreTrainingOutput) or a tuple of `torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various elements depending on the configuration ([BertConfig](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertConfig)) and inputs. ## BertLMHeadModel[[transformers.BertLMHeadModel]] #### transformers.BertLMHeadModel[[transformers.BertLMHeadModel]] [Source](https://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/modeling_bert.py#L828) Bert Model with a `language modeling` head on top for CLM fine-tuning. This model inherits from [PreTrainedModel](/docs/transformers/v5.6.0/ko/main_classes/model#transformers.PreTrainedModel). Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.) This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior. forwardtransformers.BertLMHeadModel.forwardhttps://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/modeling_bert.py#L853[{"name": "input_ids", "val": ": torch.Tensor | None = None"}, {"name": "attention_mask", "val": ": torch.Tensor | None = None"}, {"name": "token_type_ids", "val": ": torch.Tensor | None = None"}, {"name": "position_ids", "val": ": torch.Tensor | None = None"}, {"name": "inputs_embeds", "val": ": torch.Tensor | None = None"}, {"name": "encoder_hidden_states", "val": ": torch.Tensor | None = None"}, {"name": "encoder_attention_mask", "val": ": torch.Tensor | None = None"}, {"name": "labels", "val": ": torch.Tensor | None = None"}, {"name": "past_key_values", "val": ": transformers.cache_utils.Cache | None = None"}, {"name": "use_cache", "val": ": bool | None = None"}, {"name": "logits_to_keep", "val": ": int | torch.Tensor = 0"}, {"name": "**kwargs", "val": ": typing_extensions.Unpack[transformers.utils.generic.TransformersKwargs]"}]- **input_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Indices of input sequence tokens in the vocabulary. Padding will be ignored by default. Indices can be obtained using [AutoTokenizer](/docs/transformers/v5.6.0/ko/model_doc/auto#transformers.AutoTokenizer). See [PreTrainedTokenizer.encode()](/docs/transformers/v5.6.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.encode) and [PreTrainedTokenizer.__call__()](/docs/transformers/v5.6.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.__call__) for details. [What are input IDs?](../glossary#input-ids) - **attention_mask** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`: - 1 for tokens that are **not masked**, - 0 for tokens that are **masked**. [What are attention masks?](../glossary#attention-mask) - **token_type_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Segment token indices to indicate first and second portions of the inputs. Indices are selected in `[0, 1]`: - 0 corresponds to a *sentence A* token, - 1 corresponds to a *sentence B* token. [What are token type IDs?](../glossary#token-type-ids) - **position_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0, config.n_positions - 1]`. [What are position IDs?](../glossary#position-ids) - **inputs_embeds** (`torch.Tensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*) -- Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert `input_ids` indices into associated vectors than the model's internal embedding lookup matrix. - **encoder_hidden_states** (`torch.Tensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*) -- Sequence of hidden-states at the output of the last layer of the encoder. Used in the cross-attention if the model is configured as a decoder. - **encoder_attention_mask** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Mask to avoid performing attention on the padding token indices of the encoder input. This mask is used in the cross-attention if the model is configured as a decoder. Mask values selected in `[0, 1]`: - 1 for tokens that are **not masked**, - 0 for tokens that are **masked**. - **labels** (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*) -- Labels for computing the left-to-right language modeling loss (next word prediction). Indices should be in `[-100, 0, ..., config.vocab_size]` (see `input_ids` docstring) Tokens with indices set to `-100` are ignored (masked), the loss is only computed for the tokens with labels n `[0, ..., config.vocab_size]` - **past_key_values** (`~cache_utils.Cache`, *optional*) -- Pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention blocks) that can be used to speed up sequential decoding. This typically consists in the `past_key_values` returned by the model at a previous stage of decoding, when `use_cache=True` or `config.use_cache=True`. Only [Cache](/docs/transformers/v5.6.0/ko/internal/generation_utils#transformers.Cache) instance is allowed as input, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache). If no `past_key_values` are passed, [DynamicCache](/docs/transformers/v5.6.0/ko/internal/generation_utils#transformers.DynamicCache) will be initialized by default. The model will output the same cache format that is fed as input. If `past_key_values` are used, the user is expected to input only unprocessed `input_ids` (those that don't have their past key value states given to this model) of shape `(batch_size, unprocessed_length)` instead of all `input_ids` of shape `(batch_size, sequence_length)`. - **use_cache** (`bool`, *optional*) -- If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see `past_key_values`). - **logits_to_keep** (`Union[int, torch.Tensor]`, *optional*, defaults to `0`) -- If an `int`, compute logits for the last `logits_to_keep` tokens. If `0`, calculate logits for all `input_ids` (special case). Only last token logits are needed for generation, and calculating them only for that token can save memory, which becomes pretty significant for long sequences or large vocabulary size. If a `torch.Tensor`, must be 1D corresponding to the indices to keep in the sequence length dimension. This is useful when using packed tensor format (single dimension for batch and sequence length).0[CausalLMOutputWithCrossAttentions](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.CausalLMOutputWithCrossAttentions) or `tuple(torch.FloatTensor)`A [CausalLMOutputWithCrossAttentions](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.CausalLMOutputWithCrossAttentions) or a tuple of `torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various elements depending on the configuration ([BertConfig](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertConfig)) and inputs. The [BertLMHeadModel](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertLMHeadModel) forward method, overrides the `__call__` special method. Although the recipe for forward pass needs to be defined within this function, one should call the `Module` instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them. - **loss** (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided) -- Language modeling loss (for next-token prediction). - **logits** (`torch.FloatTensor` of shape `(batch_size, sequence_length, config.vocab_size)`) -- Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax). - **hidden_states** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`) -- Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`. Hidden-states of the model at the output of each layer plus the optional initial embedding outputs. - **attentions** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`) -- Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, sequence_length)`. Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads. - **cross_attentions** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`) -- Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, sequence_length)`. Cross attentions weights after the attention softmax, used to compute the weighted average in the cross-attention heads. - **past_key_values** (`Cache`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`) -- It is a [Cache](/docs/transformers/v5.6.0/ko/internal/generation_utils#transformers.Cache) instance. For more details, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache). Contains pre-computed hidden-states (key and values in the attention blocks) that can be used (see `past_key_values` input) to speed up sequential decoding. Example: ```python >>> import torch >>> from transformers import AutoTokenizer, BertLMHeadModel >>> tokenizer = AutoTokenizer.from_pretrained("google-bert/bert-base-uncased") >>> model = BertLMHeadModel.from_pretrained("google-bert/bert-base-uncased") >>> inputs = tokenizer("Hello, my dog is cute", return_tensors="pt") >>> outputs = model(**inputs, labels=inputs["input_ids"]) >>> loss = outputs.loss >>> logits = outputs.logits ``` **Parameters:** config ([BertLMHeadModel](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertLMHeadModel)) : Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the [from_pretrained()](/docs/transformers/v5.6.0/ko/main_classes/model#transformers.PreTrainedModel.from_pretrained) method to load the model weights. **Returns:** `[CausalLMOutputWithCrossAttentions](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.CausalLMOutputWithCrossAttentions) or `tuple(torch.FloatTensor)`` A [CausalLMOutputWithCrossAttentions](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.CausalLMOutputWithCrossAttentions) or a tuple of `torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various elements depending on the configuration ([BertConfig](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertConfig)) and inputs. ## BertForMaskedLM[[transformers.BertForMaskedLM]] #### transformers.BertForMaskedLM[[transformers.BertForMaskedLM]] [Source](https://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/modeling_bert.py#L913) The Bert Model with a `language modeling` head on top." This model inherits from [PreTrainedModel](/docs/transformers/v5.6.0/ko/main_classes/model#transformers.PreTrainedModel). Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.) This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior. forwardtransformers.BertForMaskedLM.forwardhttps://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/modeling_bert.py#L941[{"name": "input_ids", "val": ": torch.Tensor | None = None"}, {"name": "attention_mask", "val": ": torch.Tensor | None = None"}, {"name": "token_type_ids", "val": ": torch.Tensor | None = None"}, {"name": "position_ids", "val": ": torch.Tensor | None = None"}, {"name": "inputs_embeds", "val": ": torch.Tensor | None = None"}, {"name": "encoder_hidden_states", "val": ": torch.Tensor | None = None"}, {"name": "encoder_attention_mask", "val": ": torch.Tensor | None = None"}, {"name": "labels", "val": ": torch.Tensor | None = None"}, {"name": "**kwargs", "val": ": typing_extensions.Unpack[transformers.utils.generic.TransformersKwargs]"}]- **input_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Indices of input sequence tokens in the vocabulary. Padding will be ignored by default. Indices can be obtained using [AutoTokenizer](/docs/transformers/v5.6.0/ko/model_doc/auto#transformers.AutoTokenizer). See [PreTrainedTokenizer.encode()](/docs/transformers/v5.6.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.encode) and [PreTrainedTokenizer.__call__()](/docs/transformers/v5.6.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.__call__) for details. [What are input IDs?](../glossary#input-ids) - **attention_mask** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`: - 1 for tokens that are **not masked**, - 0 for tokens that are **masked**. [What are attention masks?](../glossary#attention-mask) - **token_type_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Segment token indices to indicate first and second portions of the inputs. Indices are selected in `[0, 1]`: - 0 corresponds to a *sentence A* token, - 1 corresponds to a *sentence B* token. [What are token type IDs?](../glossary#token-type-ids) - **position_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0, config.n_positions - 1]`. [What are position IDs?](../glossary#position-ids) - **inputs_embeds** (`torch.Tensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*) -- Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert `input_ids` indices into associated vectors than the model's internal embedding lookup matrix. - **encoder_hidden_states** (`torch.Tensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*) -- Sequence of hidden-states at the output of the last layer of the encoder. Used in the cross-attention if the model is configured as a decoder. - **encoder_attention_mask** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Mask to avoid performing attention on the padding token indices of the encoder input. This mask is used in the cross-attention if the model is configured as a decoder. Mask values selected in `[0, 1]`: - 1 for tokens that are **not masked**, - 0 for tokens that are **masked**. - **labels** (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*) -- Labels for computing the masked language modeling loss. Indices should be in `[-100, 0, ..., config.vocab_size]` (see `input_ids` docstring) Tokens with indices set to `-100` are ignored (masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`0[MaskedLMOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.MaskedLMOutput) or `tuple(torch.FloatTensor)`A [MaskedLMOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.MaskedLMOutput) or a tuple of `torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various elements depending on the configuration ([BertConfig](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertConfig)) and inputs. The [BertForMaskedLM](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForMaskedLM) forward method, overrides the `__call__` special method. Although the recipe for forward pass needs to be defined within this function, one should call the `Module` instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them. - **loss** (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided) -- Masked language modeling (MLM) loss. - **logits** (`torch.FloatTensor` of shape `(batch_size, sequence_length, config.vocab_size)`) -- Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax). - **hidden_states** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`) -- Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`. Hidden-states of the model at the output of each layer plus the optional initial embedding outputs. - **attentions** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`) -- Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, sequence_length)`. Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads. Example: ```python >>> from transformers import AutoTokenizer, BertForMaskedLM >>> import torch >>> tokenizer = AutoTokenizer.from_pretrained("google-bert/bert-base-uncased") >>> model = BertForMaskedLM.from_pretrained("google-bert/bert-base-uncased") >>> inputs = tokenizer("The capital of France is .", return_tensors="pt") >>> with torch.no_grad(): ... logits = model(**inputs).logits >>> # retrieve index of >>> mask_token_index = (inputs.input_ids == tokenizer.mask_token_id)[0].nonzero(as_tuple=True)[0] >>> predicted_token_id = logits[0, mask_token_index].argmax(axis=-1) >>> tokenizer.decode(predicted_token_id) ... >>> labels = tokenizer("The capital of France is Paris.", return_tensors="pt")["input_ids"] >>> # mask labels of non- tokens >>> labels = torch.where(inputs.input_ids == tokenizer.mask_token_id, labels, -100) >>> outputs = model(**inputs, labels=labels) >>> round(outputs.loss.item(), 2) ... ``` **Parameters:** config ([BertForMaskedLM](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForMaskedLM)) : Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the [from_pretrained()](/docs/transformers/v5.6.0/ko/main_classes/model#transformers.PreTrainedModel.from_pretrained) method to load the model weights. **Returns:** `[MaskedLMOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.MaskedLMOutput) or `tuple(torch.FloatTensor)`` A [MaskedLMOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.MaskedLMOutput) or a tuple of `torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various elements depending on the configuration ([BertConfig](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertConfig)) and inputs. ## BertForNextSentencePrediction[[transformers.BertForNextSentencePrediction]] #### transformers.BertForNextSentencePrediction[[transformers.BertForNextSentencePrediction]] [Source](https://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/modeling_bert.py#L994) Bert Model with a `next sentence prediction (classification)` head on top. This model inherits from [PreTrainedModel](/docs/transformers/v5.6.0/ko/main_classes/model#transformers.PreTrainedModel). Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.) This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior. forwardtransformers.BertForNextSentencePrediction.forwardhttps://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/modeling_bert.py#L1004[{"name": "input_ids", "val": ": torch.Tensor | None = None"}, {"name": "attention_mask", "val": ": torch.Tensor | None = None"}, {"name": "token_type_ids", "val": ": torch.Tensor | None = None"}, {"name": "position_ids", "val": ": torch.Tensor | None = None"}, {"name": "inputs_embeds", "val": ": torch.Tensor | None = None"}, {"name": "labels", "val": ": torch.Tensor | None = None"}, {"name": "**kwargs", "val": ": typing_extensions.Unpack[transformers.utils.generic.TransformersKwargs]"}]- **input_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Indices of input sequence tokens in the vocabulary. Padding will be ignored by default. Indices can be obtained using [AutoTokenizer](/docs/transformers/v5.6.0/ko/model_doc/auto#transformers.AutoTokenizer). See [PreTrainedTokenizer.encode()](/docs/transformers/v5.6.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.encode) and [PreTrainedTokenizer.__call__()](/docs/transformers/v5.6.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.__call__) for details. [What are input IDs?](../glossary#input-ids) - **attention_mask** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`: - 1 for tokens that are **not masked**, - 0 for tokens that are **masked**. [What are attention masks?](../glossary#attention-mask) - **token_type_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Segment token indices to indicate first and second portions of the inputs. Indices are selected in `[0, 1]`: - 0 corresponds to a *sentence A* token, - 1 corresponds to a *sentence B* token. [What are token type IDs?](../glossary#token-type-ids) - **position_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0, config.n_positions - 1]`. [What are position IDs?](../glossary#position-ids) - **inputs_embeds** (`torch.Tensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*) -- Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert `input_ids` indices into associated vectors than the model's internal embedding lookup matrix. - **labels** (`torch.LongTensor` of shape `(batch_size,)`, *optional*) -- Labels for computing the next sequence prediction (classification) loss. Input should be a sequence pair (see `input_ids` docstring). Indices should be in `[0, 1]`: - 0 indicates sequence B is a continuation of sequence A, - 1 indicates sequence B is a random sequence.0[NextSentencePredictorOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.NextSentencePredictorOutput) or `tuple(torch.FloatTensor)`A [NextSentencePredictorOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.NextSentencePredictorOutput) or a tuple of `torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various elements depending on the configuration ([BertConfig](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertConfig)) and inputs. The [BertForNextSentencePrediction](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForNextSentencePrediction) forward method, overrides the `__call__` special method. Although the recipe for forward pass needs to be defined within this function, one should call the `Module` instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them. - **loss** (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `next_sentence_label` is provided) -- Next sequence prediction (classification) loss. - **logits** (`torch.FloatTensor` of shape `(batch_size, 2)`) -- Prediction scores of the next sequence prediction (classification) head (scores of True/False continuation before SoftMax). - **hidden_states** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`) -- Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`. Hidden-states of the model at the output of each layer plus the optional initial embedding outputs. - **attentions** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`) -- Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, sequence_length)`. Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads. Example: ```python >>> from transformers import AutoTokenizer, BertForNextSentencePrediction >>> import torch >>> tokenizer = AutoTokenizer.from_pretrained("google-bert/bert-base-uncased") >>> model = BertForNextSentencePrediction.from_pretrained("google-bert/bert-base-uncased") >>> prompt = "In Italy, pizza served in formal settings, such as at a restaurant, is presented unsliced." >>> next_sentence = "The sky is blue due to the shorter wavelength of blue light." >>> encoding = tokenizer(prompt, next_sentence, return_tensors="pt") >>> outputs = model(**encoding, labels=torch.LongTensor([1])) >>> logits = outputs.logits >>> assert logits[0, 0] **Parameters:** config ([BertForNextSentencePrediction](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForNextSentencePrediction)) : Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the [from_pretrained()](/docs/transformers/v5.6.0/ko/main_classes/model#transformers.PreTrainedModel.from_pretrained) method to load the model weights. **Returns:** `[NextSentencePredictorOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.NextSentencePredictorOutput) or `tuple(torch.FloatTensor)`` A [NextSentencePredictorOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.NextSentencePredictorOutput) or a tuple of `torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various elements depending on the configuration ([BertConfig](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertConfig)) and inputs. ## BertForSequenceClassification[[transformers.BertForSequenceClassification]] #### transformers.BertForSequenceClassification[[transformers.BertForSequenceClassification]] [Source](https://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/modeling_bert.py#L1076) Bert Model transformer with a sequence classification/regression head on top (a linear layer on top of the pooled output) e.g. for GLUE tasks. This model inherits from [PreTrainedModel](/docs/transformers/v5.6.0/ko/main_classes/model#transformers.PreTrainedModel). Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.) This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior. forwardtransformers.BertForSequenceClassification.forwardhttps://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/modeling_bert.py#L1092[{"name": "input_ids", "val": ": torch.Tensor | None = None"}, {"name": "attention_mask", "val": ": torch.Tensor | None = None"}, {"name": "token_type_ids", "val": ": torch.Tensor | None = None"}, {"name": "position_ids", "val": ": torch.Tensor | None = None"}, {"name": "inputs_embeds", "val": ": torch.Tensor | None = None"}, {"name": "labels", "val": ": torch.Tensor | None = None"}, {"name": "**kwargs", "val": ": typing_extensions.Unpack[transformers.utils.generic.TransformersKwargs]"}]- **input_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Indices of input sequence tokens in the vocabulary. Padding will be ignored by default. Indices can be obtained using [AutoTokenizer](/docs/transformers/v5.6.0/ko/model_doc/auto#transformers.AutoTokenizer). See [PreTrainedTokenizer.encode()](/docs/transformers/v5.6.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.encode) and [PreTrainedTokenizer.__call__()](/docs/transformers/v5.6.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.__call__) for details. [What are input IDs?](../glossary#input-ids) - **attention_mask** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`: - 1 for tokens that are **not masked**, - 0 for tokens that are **masked**. [What are attention masks?](../glossary#attention-mask) - **token_type_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Segment token indices to indicate first and second portions of the inputs. Indices are selected in `[0, 1]`: - 0 corresponds to a *sentence A* token, - 1 corresponds to a *sentence B* token. [What are token type IDs?](../glossary#token-type-ids) - **position_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0, config.n_positions - 1]`. [What are position IDs?](../glossary#position-ids) - **inputs_embeds** (`torch.Tensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*) -- Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert `input_ids` indices into associated vectors than the model's internal embedding lookup matrix. - **labels** (`torch.LongTensor` of shape `(batch_size,)`, *optional*) -- Labels for computing the sequence classification/regression loss. Indices should be in `[0, ..., config.num_labels - 1]`. If `config.num_labels == 1` a regression loss is computed (Mean-Square loss), If `config.num_labels > 1` a classification loss is computed (Cross-Entropy).0[SequenceClassifierOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.SequenceClassifierOutput) or `tuple(torch.FloatTensor)`A [SequenceClassifierOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.SequenceClassifierOutput) or a tuple of `torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various elements depending on the configuration ([BertConfig](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertConfig)) and inputs. The [BertForSequenceClassification](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForSequenceClassification) forward method, overrides the `__call__` special method. Although the recipe for forward pass needs to be defined within this function, one should call the `Module` instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them. - **loss** (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided) -- Classification (or regression if config.num_labels==1) loss. - **logits** (`torch.FloatTensor` of shape `(batch_size, config.num_labels)`) -- Classification (or regression if config.num_labels==1) scores (before SoftMax). - **hidden_states** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`) -- Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`. Hidden-states of the model at the output of each layer plus the optional initial embedding outputs. - **attentions** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`) -- Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, sequence_length)`. Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads. Example of single-label classification: ```python >>> import torch >>> from transformers import AutoTokenizer, BertForSequenceClassification >>> tokenizer = AutoTokenizer.from_pretrained("google-bert/bert-base-uncased") >>> model = BertForSequenceClassification.from_pretrained("google-bert/bert-base-uncased") >>> inputs = tokenizer("Hello, my dog is cute", return_tensors="pt") >>> with torch.no_grad(): ... logits = model(**inputs).logits >>> predicted_class_id = logits.argmax().item() >>> model.config.id2label[predicted_class_id] ... >>> # To train a model on `num_labels` classes, you can pass `num_labels=num_labels` to `.from_pretrained(...)` >>> num_labels = len(model.config.id2label) >>> model = BertForSequenceClassification.from_pretrained("google-bert/bert-base-uncased", num_labels=num_labels) >>> labels = torch.tensor([1]) >>> loss = model(**inputs, labels=labels).loss >>> round(loss.item(), 2) ... ``` Example of multi-label classification: ```python >>> import torch >>> from transformers import AutoTokenizer, BertForSequenceClassification >>> tokenizer = AutoTokenizer.from_pretrained("google-bert/bert-base-uncased") >>> model = BertForSequenceClassification.from_pretrained("google-bert/bert-base-uncased", problem_type="multi_label_classification") >>> inputs = tokenizer("Hello, my dog is cute", return_tensors="pt") >>> with torch.no_grad(): ... logits = model(**inputs).logits >>> predicted_class_ids = torch.arange(0, logits.shape[-1])[torch.sigmoid(logits).squeeze(dim=0) > 0.5] >>> # To train a model on `num_labels` classes, you can pass `num_labels=num_labels` to `.from_pretrained(...)` >>> num_labels = len(model.config.id2label) >>> model = BertForSequenceClassification.from_pretrained( ... "google-bert/bert-base-uncased", num_labels=num_labels, problem_type="multi_label_classification" ... ) >>> labels = torch.sum( ... torch.nn.functional.one_hot(predicted_class_ids[None, :].clone(), num_classes=num_labels), dim=1 ... ).to(torch.float) >>> loss = model(**inputs, labels=labels).loss ``` **Parameters:** config ([BertForSequenceClassification](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForSequenceClassification)) : Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the [from_pretrained()](/docs/transformers/v5.6.0/ko/main_classes/model#transformers.PreTrainedModel.from_pretrained) method to load the model weights. **Returns:** `[SequenceClassifierOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.SequenceClassifierOutput) or `tuple(torch.FloatTensor)`` A [SequenceClassifierOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.SequenceClassifierOutput) or a tuple of `torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various elements depending on the configuration ([BertConfig](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertConfig)) and inputs. ## BertForMultipleChoice[[transformers.BertForMultipleChoice]] #### transformers.BertForMultipleChoice[[transformers.BertForMultipleChoice]] [Source](https://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/modeling_bert.py#L1157) The Bert Model with a multiple choice classification head on top (a linear layer on top of the pooled output and a softmax) e.g. for RocStories/SWAG tasks. This model inherits from [PreTrainedModel](/docs/transformers/v5.6.0/ko/main_classes/model#transformers.PreTrainedModel). Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.) This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior. forwardtransformers.BertForMultipleChoice.forwardhttps://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/modeling_bert.py#L1171[{"name": "input_ids", "val": ": torch.Tensor | None = None"}, {"name": "attention_mask", "val": ": torch.Tensor | None = None"}, {"name": "token_type_ids", "val": ": torch.Tensor | None = None"}, {"name": "position_ids", "val": ": torch.Tensor | None = None"}, {"name": "inputs_embeds", "val": ": torch.Tensor | None = None"}, {"name": "labels", "val": ": torch.Tensor | None = None"}, {"name": "**kwargs", "val": ": typing_extensions.Unpack[transformers.utils.generic.TransformersKwargs]"}]- **input_ids** (`torch.LongTensor` of shape `(batch_size, num_choices, sequence_length)`) -- Indices of input sequence tokens in the vocabulary. Indices can be obtained using [AutoTokenizer](/docs/transformers/v5.6.0/ko/model_doc/auto#transformers.AutoTokenizer). See [PreTrainedTokenizer.encode()](/docs/transformers/v5.6.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.encode) and [PreTrainedTokenizer.__call__()](/docs/transformers/v5.6.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.__call__) for details. [What are input IDs?](../glossary#input-ids) - **attention_mask** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`: - 1 for tokens that are **not masked**, - 0 for tokens that are **masked**. [What are attention masks?](../glossary#attention-mask) - **token_type_ids** (`torch.LongTensor` of shape `(batch_size, num_choices, sequence_length)`, *optional*) -- Segment token indices to indicate first and second portions of the inputs. Indices are selected in `[0, 1]`: - 0 corresponds to a *sentence A* token, - 1 corresponds to a *sentence B* token. [What are token type IDs?](../glossary#token-type-ids) - **position_ids** (`torch.LongTensor` of shape `(batch_size, num_choices, sequence_length)`, *optional*) -- Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0, config.max_position_embeddings - 1]`. [What are position IDs?](../glossary#position-ids) - **inputs_embeds** (`torch.FloatTensor` of shape `(batch_size, num_choices, sequence_length, hidden_size)`, *optional*) -- Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert `input_ids` indices into associated vectors than the model's internal embedding lookup matrix. - **labels** (`torch.LongTensor` of shape `(batch_size,)`, *optional*) -- Labels for computing the multiple choice classification loss. Indices should be in `[0, ..., num_choices-1]` where `num_choices` is the size of the second dimension of the input tensors. (See `input_ids` above)0[MultipleChoiceModelOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.MultipleChoiceModelOutput) or `tuple(torch.FloatTensor)`A [MultipleChoiceModelOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.MultipleChoiceModelOutput) or a tuple of `torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various elements depending on the configuration ([BertConfig](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertConfig)) and inputs. The [BertForMultipleChoice](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForMultipleChoice) forward method, overrides the `__call__` special method. Although the recipe for forward pass needs to be defined within this function, one should call the `Module` instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them. - **loss** (`torch.FloatTensor` of shape *(1,)*, *optional*, returned when `labels` is provided) -- Classification loss. - **logits** (`torch.FloatTensor` of shape `(batch_size, num_choices)`) -- *num_choices* is the second dimension of the input tensors. (see *input_ids* above). Classification scores (before SoftMax). - **hidden_states** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`) -- Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`. Hidden-states of the model at the output of each layer plus the optional initial embedding outputs. - **attentions** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`) -- Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, sequence_length)`. Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads. Example: ```python >>> from transformers import AutoTokenizer, BertForMultipleChoice >>> import torch >>> tokenizer = AutoTokenizer.from_pretrained("google-bert/bert-base-uncased") >>> model = BertForMultipleChoice.from_pretrained("google-bert/bert-base-uncased") >>> prompt = "In Italy, pizza served in formal settings, such as at a restaurant, is presented unsliced." >>> choice0 = "It is eaten with a fork and a knife." >>> choice1 = "It is eaten while held in the hand." >>> labels = torch.tensor(0).unsqueeze(0) # choice0 is correct (according to Wikipedia ;)), batch size 1 >>> encoding = tokenizer([prompt, prompt], [choice0, choice1], return_tensors="pt", padding=True) >>> outputs = model(**{k: v.unsqueeze(0) for k, v in encoding.items()}, labels=labels) # batch size is 1 >>> # the linear classifier still needs to be trained >>> loss = outputs.loss >>> logits = outputs.logits ``` **Parameters:** config ([BertForMultipleChoice](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForMultipleChoice)) : Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the [from_pretrained()](/docs/transformers/v5.6.0/ko/main_classes/model#transformers.PreTrainedModel.from_pretrained) method to load the model weights. **Returns:** `[MultipleChoiceModelOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.MultipleChoiceModelOutput) or `tuple(torch.FloatTensor)`` A [MultipleChoiceModelOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.MultipleChoiceModelOutput) or a tuple of `torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various elements depending on the configuration ([BertConfig](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertConfig)) and inputs. ## BertForTokenClassification[[transformers.BertForTokenClassification]] #### transformers.BertForTokenClassification[[transformers.BertForTokenClassification]] [Source](https://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/modeling_bert.py#L1255) The Bert transformer with a token classification head on top (a linear layer on top of the hidden-states output) e.g. for Named-Entity-Recognition (NER) tasks. This model inherits from [PreTrainedModel](/docs/transformers/v5.6.0/ko/main_classes/model#transformers.PreTrainedModel). Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.) This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior. forwardtransformers.BertForTokenClassification.forwardhttps://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/modeling_bert.py#L1270[{"name": "input_ids", "val": ": torch.Tensor | None = None"}, {"name": "attention_mask", "val": ": torch.Tensor | None = None"}, {"name": "token_type_ids", "val": ": torch.Tensor | None = None"}, {"name": "position_ids", "val": ": torch.Tensor | None = None"}, {"name": "inputs_embeds", "val": ": torch.Tensor | None = None"}, {"name": "labels", "val": ": torch.Tensor | None = None"}, {"name": "**kwargs", "val": ": typing_extensions.Unpack[transformers.utils.generic.TransformersKwargs]"}]- **input_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Indices of input sequence tokens in the vocabulary. Padding will be ignored by default. Indices can be obtained using [AutoTokenizer](/docs/transformers/v5.6.0/ko/model_doc/auto#transformers.AutoTokenizer). See [PreTrainedTokenizer.encode()](/docs/transformers/v5.6.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.encode) and [PreTrainedTokenizer.__call__()](/docs/transformers/v5.6.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.__call__) for details. [What are input IDs?](../glossary#input-ids) - **attention_mask** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`: - 1 for tokens that are **not masked**, - 0 for tokens that are **masked**. [What are attention masks?](../glossary#attention-mask) - **token_type_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Segment token indices to indicate first and second portions of the inputs. Indices are selected in `[0, 1]`: - 0 corresponds to a *sentence A* token, - 1 corresponds to a *sentence B* token. [What are token type IDs?](../glossary#token-type-ids) - **position_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0, config.n_positions - 1]`. [What are position IDs?](../glossary#position-ids) - **inputs_embeds** (`torch.Tensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*) -- Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert `input_ids` indices into associated vectors than the model's internal embedding lookup matrix. - **labels** (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*) -- Labels for computing the token classification loss. Indices should be in `[0, ..., config.num_labels - 1]`.0[TokenClassifierOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.TokenClassifierOutput) or `tuple(torch.FloatTensor)`A [TokenClassifierOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.TokenClassifierOutput) or a tuple of `torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various elements depending on the configuration ([BertConfig](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertConfig)) and inputs. The [BertForTokenClassification](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForTokenClassification) forward method, overrides the `__call__` special method. Although the recipe for forward pass needs to be defined within this function, one should call the `Module` instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them. - **loss** (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided) -- Classification loss. - **logits** (`torch.FloatTensor` of shape `(batch_size, sequence_length, config.num_labels)`) -- Classification scores (before SoftMax). - **hidden_states** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`) -- Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`. Hidden-states of the model at the output of each layer plus the optional initial embedding outputs. - **attentions** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`) -- Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, sequence_length)`. Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads. Example: ```python >>> from transformers import AutoTokenizer, BertForTokenClassification >>> import torch >>> tokenizer = AutoTokenizer.from_pretrained("google-bert/bert-base-uncased") >>> model = BertForTokenClassification.from_pretrained("google-bert/bert-base-uncased") >>> inputs = tokenizer( ... "HuggingFace is a company based in Paris and New York", add_special_tokens=False, return_tensors="pt" ... ) >>> with torch.no_grad(): ... logits = model(**inputs).logits >>> predicted_token_class_ids = logits.argmax(-1) >>> # Note that tokens are classified rather then input words which means that >>> # there might be more predicted token classes than words. >>> # Multiple token classes might account for the same word >>> predicted_tokens_classes = [model.config.id2label[t.item()] for t in predicted_token_class_ids[0]] >>> predicted_tokens_classes ... >>> labels = predicted_token_class_ids >>> loss = model(**inputs, labels=labels).loss >>> round(loss.item(), 2) ... ``` **Parameters:** config ([BertForTokenClassification](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForTokenClassification)) : Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the [from_pretrained()](/docs/transformers/v5.6.0/ko/main_classes/model#transformers.PreTrainedModel.from_pretrained) method to load the model weights. **Returns:** `[TokenClassifierOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.TokenClassifierOutput) or `tuple(torch.FloatTensor)`` A [TokenClassifierOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.TokenClassifierOutput) or a tuple of `torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various elements depending on the configuration ([BertConfig](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertConfig)) and inputs. ## BertForQuestionAnswering[[transformers.BertForQuestionAnswering]] #### transformers.BertForQuestionAnswering[[transformers.BertForQuestionAnswering]] [Source](https://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/modeling_bert.py#L1315) The Bert transformer with a span classification head on top for extractive question-answering tasks like SQuAD (a linear layer on top of the hidden-states output to compute `span start logits` and `span end logits`). This model inherits from [PreTrainedModel](/docs/transformers/v5.6.0/ko/main_classes/model#transformers.PreTrainedModel). Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.) This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior. forwardtransformers.BertForQuestionAnswering.forwardhttps://github.com/huggingface/transformers/blob/v5.6.0/src/transformers/models/bert/modeling_bert.py#L1326[{"name": "input_ids", "val": ": torch.Tensor | None = None"}, {"name": "attention_mask", "val": ": torch.Tensor | None = None"}, {"name": "token_type_ids", "val": ": torch.Tensor | None = None"}, {"name": "position_ids", "val": ": torch.Tensor | None = None"}, {"name": "inputs_embeds", "val": ": torch.Tensor | None = None"}, {"name": "start_positions", "val": ": torch.Tensor | None = None"}, {"name": "end_positions", "val": ": torch.Tensor | None = None"}, {"name": "**kwargs", "val": ": typing_extensions.Unpack[transformers.utils.generic.TransformersKwargs]"}]- **input_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Indices of input sequence tokens in the vocabulary. Padding will be ignored by default. Indices can be obtained using [AutoTokenizer](/docs/transformers/v5.6.0/ko/model_doc/auto#transformers.AutoTokenizer). See [PreTrainedTokenizer.encode()](/docs/transformers/v5.6.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.encode) and [PreTrainedTokenizer.__call__()](/docs/transformers/v5.6.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.__call__) for details. [What are input IDs?](../glossary#input-ids) - **attention_mask** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`: - 1 for tokens that are **not masked**, - 0 for tokens that are **masked**. [What are attention masks?](../glossary#attention-mask) - **token_type_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Segment token indices to indicate first and second portions of the inputs. Indices are selected in `[0, 1]`: - 0 corresponds to a *sentence A* token, - 1 corresponds to a *sentence B* token. [What are token type IDs?](../glossary#token-type-ids) - **position_ids** (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) -- Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0, config.n_positions - 1]`. [What are position IDs?](../glossary#position-ids) - **inputs_embeds** (`torch.Tensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*) -- Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert `input_ids` indices into associated vectors than the model's internal embedding lookup matrix. - **start_positions** (`torch.Tensor` of shape `(batch_size,)`, *optional*) -- Labels for position (index) of the start of the labelled span for computing the token classification loss. Positions are clamped to the length of the sequence (`sequence_length`). Position outside of the sequence are not taken into account for computing the loss. - **end_positions** (`torch.Tensor` of shape `(batch_size,)`, *optional*) -- Labels for position (index) of the end of the labelled span for computing the token classification loss. Positions are clamped to the length of the sequence (`sequence_length`). Position outside of the sequence are not taken into account for computing the loss.0[QuestionAnsweringModelOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.QuestionAnsweringModelOutput) or `tuple(torch.FloatTensor)`A [QuestionAnsweringModelOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.QuestionAnsweringModelOutput) or a tuple of `torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various elements depending on the configuration ([BertConfig](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertConfig)) and inputs. The [BertForQuestionAnswering](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForQuestionAnswering) forward method, overrides the `__call__` special method. Although the recipe for forward pass needs to be defined within this function, one should call the `Module` instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them. - **loss** (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided) -- Total span extraction loss is the sum of a Cross-Entropy for the start and end positions. - **start_logits** (`torch.FloatTensor` of shape `(batch_size, sequence_length)`) -- Span-start scores (before SoftMax). - **end_logits** (`torch.FloatTensor` of shape `(batch_size, sequence_length)`) -- Span-end scores (before SoftMax). - **hidden_states** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`) -- Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`. Hidden-states of the model at the output of each layer plus the optional initial embedding outputs. - **attentions** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`) -- Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, sequence_length)`. Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads. Example: ```python >>> from transformers import AutoTokenizer, BertForQuestionAnswering >>> import torch >>> tokenizer = AutoTokenizer.from_pretrained("google-bert/bert-base-uncased") >>> model = BertForQuestionAnswering.from_pretrained("google-bert/bert-base-uncased") >>> question, text = "Who was Jim Henson?", "Jim Henson was a nice puppet" >>> inputs = tokenizer(question, text, return_tensors="pt") >>> with torch.no_grad(): ... outputs = model(**inputs) >>> answer_start_index = outputs.start_logits.argmax() >>> answer_end_index = outputs.end_logits.argmax() >>> predict_answer_tokens = inputs.input_ids[0, answer_start_index : answer_end_index + 1] >>> tokenizer.decode(predict_answer_tokens, skip_special_tokens=True) ... >>> # target is "nice puppet" >>> target_start_index = torch.tensor([14]) >>> target_end_index = torch.tensor([15]) >>> outputs = model(**inputs, start_positions=target_start_index, end_positions=target_end_index) >>> loss = outputs.loss >>> round(loss.item(), 2) ... ``` **Parameters:** config ([BertForQuestionAnswering](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertForQuestionAnswering)) : Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the [from_pretrained()](/docs/transformers/v5.6.0/ko/main_classes/model#transformers.PreTrainedModel.from_pretrained) method to load the model weights. **Returns:** `[QuestionAnsweringModelOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.QuestionAnsweringModelOutput) or `tuple(torch.FloatTensor)`` A [QuestionAnsweringModelOutput](/docs/transformers/v5.6.0/ko/main_classes/output#transformers.modeling_outputs.QuestionAnsweringModelOutput) or a tuple of `torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various elements depending on the configuration ([BertConfig](/docs/transformers/v5.6.0/ko/model_doc/bert#transformers.BertConfig)) and inputs.