transformers/main/ko/model_doc/bertweet.md

# BERTweet [[bertweet]]

## 개요 [[overview]]

BERTweet 모델은 Dat Quoc Nguyen, Thanh Vu, Anh Tuan Nguyen에 의해 [BERTweet: A pre-trained language model for English Tweets](https://www.aclweb.org/anthology/2020.emnlp-demos.2.pdf) 에서 제안되었습니다.

해당 논문의 초록 :

*영어 트윗을 위한 최초의 공개 대규모 사전 학습된 언어 모델인 BERTweet을 소개합니다. 
BERTweet은 BERT-base(Devlin et al., 2019)와 동일한 아키텍처를 가지고 있으며, RoBERTa 사전 학습 절차(Liu et al., 2019)를 사용하여 학습되었습니다. 
실험 결과, BERTweet은 강력한 기준 모델인 RoBERTa-base 및 XLM-R-base(Conneau et al., 2020)의 성능을 능가하여 세 가지 트윗 NLP 작업(품사 태깅, 개체명 인식, 텍스트 분류)에서 이전 최신 모델보다 더 나은 성능을 보여주었습니다.*

이 모델은 [dqnguyen](https://huggingface.co/dqnguyen) 께서 기여하셨습니다. 원본 코드는 [여기](https://github.com/VinAIResearch/BERTweet).에서 확인할 수 있습니다.

## 사용 예시 [[usage-example]]

```python
>>> import torch
>>> from transformers import AutoModel, AutoTokenizer

>>> bertweet = AutoModel.from_pretrained("vinai/bertweet-base")

>>> # 트랜스포머 버전 4.x 이상 :
>>> tokenizer = AutoTokenizer.from_pretrained("vinai/bertweet-base", use_fast=False)

>>> # 트랜스포머 버전 3.x 이상:
>>> # tokenizer = AutoTokenizer.from_pretrained("vinai/bertweet-base")

>>> # 입력된 트윗은 이미 정규화되었습니다!
>>> line = "SC has first two presumptive cases of coronavirus , DHEC confirms HTTPURL via @USER :cry:"

>>> input_ids = torch.tensor([tokenizer.encode(line)])

>>> with torch.no_grad():
...     features = bertweet(input_ids)  # Models outputs are now tuples

>>> # With TensorFlow 2.0+:
>>> # from transformers import TFAutoModel
>>> # bertweet = TFAutoModel.from_pretrained("vinai/bertweet-base")
```

 

이 구현은 토큰화 방법을 제외하고는 BERT와 동일합니다. API 참조 정보는 [BERT 문서](bert) 를 참조하세요.

## Bertweet 토큰화(BertweetTokenizer) [[transformers.BertweetTokenizer]][[transformers.BertweetTokenizer]]

#### transformers.BertweetTokenizer[[transformers.BertweetTokenizer]]

[Source](https://github.com/huggingface/transformers/blob/main/src/transformers/models/bertweet/tokenization_bertweet.py#L51)

Constructs a BERTweet tokenizer, using Byte-Pair-Encoding.

This tokenizer inherits from [PreTrainedTokenizer](/docs/transformers/main/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.

add_from_filetransformers.BertweetTokenizer.add_from_filehttps://github.com/huggingface/transformers/blob/main/src/transformers/models/bertweet/tokenization_bertweet.py#L332[{"name": "f", "val": ""}]

Loads a pre-existing dictionary from a text file and adds its symbols to this instance.

**Parameters:**

vocab_file (`str`) : Path to the vocabulary file.

merges_file (`str`) : Path to the merges file.

normalization (`bool`, *optional*, defaults to `False`) : Whether or not to apply a normalization preprocess.

bos_token (`str`, *optional*, defaults to `"<s>"`) : The beginning of sequence token that was used during pretraining. Can be used a sequence classifier token.    When building a sequence using special tokens, this is not the token that is used for the beginning of sequence. The token used is the `cls_token`.   

eos_token (`str`, *optional*, defaults to `"</s>"`) : The end of sequence token.    When building a sequence using special tokens, this is not the token that is used for the end of sequence. The token used is the `sep_token`.   

sep_token (`str`, *optional*, defaults to `"</s>"`) : 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.

cls_token (`str`, *optional*, defaults to `"<s>"`) : 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.

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.

pad_token (`str`, *optional*, defaults to `"<pad>"`) : The token used for padding, for example when batching sequences of different lengths.

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.
#### convert_tokens_to_string[[transformers.BertweetTokenizer.convert_tokens_to_string]]

[Source](https://github.com/huggingface/transformers/blob/main/src/transformers/models/bertweet/tokenization_bertweet.py#L291)

Converts a sequence of tokens (string) in a single string.
#### normalizeToken[[transformers.BertweetTokenizer.normalizeToken]]

[Source](https://github.com/huggingface/transformers/blob/main/src/transformers/models/bertweet/tokenization_bertweet.py#L264)

Normalize tokens in a Tweet
#### normalizeTweet[[transformers.BertweetTokenizer.normalizeTweet]]

[Source](https://github.com/huggingface/transformers/blob/main/src/transformers/models/bertweet/tokenization_bertweet.py#L230)

Normalize a raw Tweet
#### save_vocabulary[[transformers.BertweetTokenizer.save_vocabulary]]

[Source](https://github.com/huggingface/transformers/blob/main/src/transformers/models/bertweet/tokenization_bertweet.py#L302)

Save the vocabulary and merges files to a directory.