U &c9@sdZddlZddlmZmZddlmZddlmZddl m Z m Z ddl m Z eeZd d d Zd d ddd d ddddddddd ZdddddddZGddde ZGddde ZdS)z!Tokenization classes for RoBERTa.N)ListOptional) AddedToken)RobertaProcessing) GPT2TokenizerGPT2TokenizerFast)PreTrainedTokenizerz vocab.jsonz merges.txt) vocab_file merges_filezKhttps://s3.amazonaws.com/models.huggingface.co/bert/roberta-base-vocab.jsonzLhttps://s3.amazonaws.com/models.huggingface.co/bert/roberta-large-vocab.jsonzQhttps://s3.amazonaws.com/models.huggingface.co/bert/roberta-large-mnli-vocab.jsonzQhttps://s3.amazonaws.com/models.huggingface.co/bert/distilroberta-base-vocab.json)z roberta-basez roberta-largezroberta-large-mnlizdistilroberta-basezroberta-base-openai-detectorzroberta-large-openai-detectorzKhttps://s3.amazonaws.com/models.huggingface.co/bert/roberta-base-merges.txtzLhttps://s3.amazonaws.com/models.huggingface.co/bert/roberta-large-merges.txtzQhttps://s3.amazonaws.com/models.huggingface.co/bert/roberta-large-mnli-merges.txtzQhttps://s3.amazonaws.com/models.huggingface.co/bert/distilroberta-base-merges.txticseZdZdZeZeZeZ dgZ dfd d Z de e ee e e e d d dZde e ee e ee e dddZde e ee e e e d ddZdddZZS)RobertaTokenizera+ Constructs a RoBERTa BPE tokenizer, derived from the GPT-2 tokenizer. Peculiarities: - Byte-level Byte-Pair-Encoding - Requires a space to start the input string => the encoding methods should be called with the ``add_prefix_space`` flag set to ``True``. Otherwise, this tokenizer ``encode`` and ``decode`` method will not conserve the absence of a space at the beginning of a string: :: tokenizer.decode(tokenizer.encode("Hello")) = " Hello" This tokenizer inherits from :class:`~transformers.PreTrainedTokenizer` which contains most of the methods. Users should refer to the superclass for more information regarding methods. Args: vocab_file (:obj:`str`): Path to the vocabulary file. merges_file (:obj:`str`): Path to the merges file. errors (:obj:`str`, `optional`, defaults to "replace"): Paradigm to follow when decoding bytes to UTF-8. See `bytes.decode `__ for more information. bos_token (:obj:`string`, `optional`, defaults to ""): The beginning of sequence token that was used during pre-training. Can be used a sequence classifier token. .. note:: 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 :obj:`cls_token`. eos_token (:obj:`string`, `optional`, defaults to ""): The end of sequence token. .. note:: 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 :obj:`sep_token`. sep_token (:obj:`string`, `optional`, defaults to ""): 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 (:obj:`string`, `optional`, defaults to ""): 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 (:obj:`string`, `optional`, defaults to ""): 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 (:obj:`string`, `optional`, defaults to ""): The token used for padding, for example when batching sequences of different lengths. mask_token (:obj:`string`, `optional`, defaults to ""): 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. attention_maskreplacec s,tjf||||||||| | d | dS)N) r r errors bos_token eos_token unk_token sep_token cls_token pad_token mask_token)super__init__) selfr r rrrrrrrrkwargs __class__E/tmp/pip-unpacked-wheel-ymerj3tt/transformers/tokenization_roberta.pyr~s zRobertaTokenizer.__init__N) token_ids_0 token_ids_1returncCsD|dkr|jg||jgS|jg}|jg}||||||S)a Build model inputs from a sequence or a pair of sequence for sequence classification tasks by concatenating and adding special tokens. A RoBERTa sequence has the following format: - single sequence: `` X `` - pair of sequences: `` A B `` Args: token_ids_0 (:obj:`List[int]`): List of IDs to which the special tokens will be added token_ids_1 (:obj:`List[int]`, `optional`, defaults to :obj:`None`): Optional second list of IDs for sequence pairs. Returns: :obj:`List[int]`: list of `input IDs <../glossary.html#input-ids>`__ with the appropriate special tokens. N) cls_token_id sep_token_id)rr$r%clssepr"r"r# build_inputs_with_special_tokenss z1RobertaTokenizer.build_inputs_with_special_tokensF)r$r%already_has_special_tokensr&cs||r*|dk rtdttfdd|S|dkrLdgdgt|dgSdgdgt|ddgdgt|dgS)a  Retrieves 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`` or ``encode_plus`` methods. Args: token_ids_0 (:obj:`List[int]`): List of ids. token_ids_1 (:obj:`List[int]`, `optional`, defaults to :obj:`None`): Optional second list of IDs for sequence pairs. already_has_special_tokens (:obj:`bool`, `optional`, defaults to :obj:`False`): Set to True if the token list is already formatted with special tokens for the model Returns: :obj:`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token. Nz~You should not supply a second sequence if the provided sequence of ids is already formated with special tokens for the model.cs|jjfkrdSdS)Nrr)r(r')xrr"r#z:RobertaTokenizer.get_special_tokens_mask..rr) ValueErrorlistmaplen)rr$r%r,r"r.r#get_special_tokens_masksz(RobertaTokenizer.get_special_tokens_maskcCsP|jg}|jg}|dkr.t|||dgSt||||||dgS)a Creates a mask from the two sequences passed to be used in a sequence-pair classification task. RoBERTa does not make use of token type ids, therefore a list of zeros is returned. Args: token_ids_0 (:obj:`List[int]`): List of ids. token_ids_1 (:obj:`List[int]`, `optional`, defaults to :obj:`None`): Optional second list of IDs for sequence pairs. Returns: :obj:`List[int]`: List of zeros. Nr)r(r'r4)rr$r%r*r)r"r"r#$create_token_type_ids_from_sequencess z5RobertaTokenizer.create_token_type_ids_from_sequencescKs2d|kr|d}n|}|r.|ds.d|}|S)Nadd_prefix_spacer )isspace)rtextadd_special_tokensrr7r"r"r#prepare_for_tokenizations  z)RobertaTokenizer.prepare_for_tokenization)rrrrrrrr)N)NF)N)F)__name__ __module__ __qualname____doc__VOCAB_FILES_NAMESvocab_files_namesPRETRAINED_VOCAB_FILES_MAPpretrained_vocab_files_map&PRETRAINED_POSITIONAL_EMBEDDINGS_SIZESmax_model_input_sizesmodel_input_namesrrintrr+boolr5r6r< __classcell__r"r"r r#r @sF8      r c sNeZdZdZeZeZeZ dgZ dfd d Z e j jd d Z dddZZS)RobertaTokenizerFasta Constructs a "Fast" RoBERTa BPE tokenizer (backed by HuggingFace's `tokenizers` library). Peculiarities: - Byte-level Byte-Pair-Encoding - Requires a space to start the input string => the encoding methods should be called with the ``add_prefix_space`` flag set to ``True``. Otherwise, this tokenizer ``encode`` and ``decode`` method will not conserve the absence of a space at the beginning of a string: :: tokenizer.decode(tokenizer.encode("Hello")) = " Hello" This tokenizer inherits from :class:`~transformers.PreTrainedTokenizerFast` which contains most of the methods. Users should refer to the superclass for more information regarding methods. Args: vocab_file (:obj:`str`): Path to the vocabulary file. merges_file (:obj:`str`): Path to the merges file. errors (:obj:`str`, `optional`, defaults to "replace"): Paradigm to follow when decoding bytes to UTF-8. See `bytes.decode `__ for more information. unk_token (:obj:`string`, `optional`, defaults to `<|endoftext|>`): 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. bos_token (:obj:`string`, `optional`, defaults to `<|endoftext|>`): The beginning of sequence token. eos_token (:obj:`string`, `optional`, defaults to `<|endoftext|>`): The end of sequence token. add_prefix_space (:obj:`bool`, `optional`, defaults to `False`): Whether to add a leading space to the first word. This allows to treat the leading word just as any other word. (GPT2 tokenizer detect beginning of words by the preceeding space) trim_offsets (:obj:`bool`, `optional`, defaults to `True`): Whether the post processing step should trim offsets to avoid including whitespaces. r rrrrrrTc  s| d| | d|| d|| d| tjf|||||| | d| t||jf||jf| | d|jj_|j | dgdS)Nrrrr)r r rrrr7 trim_offsets)r*r)r7rL) setdefaultrrrr(r'Zbackend_tokenizer _tokenizerZpost_processorr;)rr r rrrrrrrrr7rLrr r"r#r#s,      zRobertaTokenizerFast.__init__cCs0t|tst|dd}t||_||gdS)NT)lstrip) isinstancerstrZ _mask_tokenZ_maybe_update_backend)rvaluer"r"r#rLs   zRobertaTokenizerFast.mask_tokenNcCs8|jg||jg}|dkr |S||jg||jgS)N)Z bos_token_idZ eos_token_id)rr$r%outputr"r"r#r+Tsz5RobertaTokenizerFast.build_inputs_with_special_tokens) rrrrrrrrTT)N)r=r>r?r@rArBrCrDrErFrGrr rsetterr+rJr"r"r r#rKs&)) rK)r@loggingtypingrrZ tokenizersrZtokenizers.processorsrZtokenization_gpt2rrZtokenization_utilsr getLoggerr=loggerrArCrEr rKr"r"r"r#sF      5