U
9%e; @ s d Z ddlZddlZddlmZ ddlmZmZmZm Z m
Z
ddl Z ddl
mZmZ ddlmZmZ eeZdd iZdd
d iiZG d d
d
eZdS )zTokenization classes. N)copyfile)AnyDictListOptionalTuple )
AddedTokenPreTrainedTokenizer)SPIECE_UNDERLINElogging
vocab_filez spiece.modelzTsinghuaAI/CPM-GeneratezHhttps://huggingface.co/TsinghuaAI/CPM-Generate/resolve/main/spiece.modelc sL e Zd ZdZeZeZddddddddd d
d d gd
f ee e
e f d
d fdd
Z e
dd Zdd Zdd Zdd Zdd Ze
ee
dd d Zd d Zd d! Zd"d# Zd1ee eee ee d$d%d&Zd2ee eee eee d' fd(d)
Zd3ee eee ee d$d*d+Z d4e
ee
e e
d,d-d.Z fd/d0Z Z S )5 CpmTokenizerzMRuns pre-tokenization with Jieba segmentation tool. It is used in CPM models.FTzzzzzzzzzN)sp_model_kwargsreturnc s t | trt| dddn| } |
dkr(i n|
| _|| _|| _|| _|| _tj f | j| _
| j
| z ddl }W n, t
k
r } z|dW 5 d}~X Y nX || _ tdd| _t jf |||||||| |
| | | jd | d
| _dS ) a
Construct a CPM tokenizer. Based on [Jieba](https://pypi.org/project/jieba/) and
[SentencePiece](https://github.com/google/sentencepiece).
This tokenizer inherits from [`PreTrainedTokenizer`] which contains most of the main methods. Users should
refer to this superclass for more information regarding those methods.
Args:
vocab_file (`str`):
[SentencePiece](https://github.com/google/sentencepiece) file (generally has a .spm extension) that
contains the vocabulary necessary to instantiate a tokenizer.
do_lower_case (`bool`, *optional*, defaults to `True`):
Whether to lowercase the input when tokenizing.
remove_space (`bool`, *optional*, defaults to `True`):
Whether to strip the text when tokenizing (removing excess spaces before and after the string).
keep_accents (`bool`, *optional*, defaults to `False`):
Whether to keep accents when tokenizing.
bos_token (`str`, *optional*, defaults to `""`):
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 `""`):
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`.
unk_token (`str`, *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.
sep_token (`str`, *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.
pad_token (`str`, *optional*, defaults to `""`):
The token used for padding, for example when batching sequences of different lengths.
cls_token (`str`, *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.
mask_token (`str`, *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.
additional_special_tokens (`List[str]`, *optional*, defaults to `["", ""]`):
Additional special tokens used by the tokenizer.
Attributes:
sp_model (`SentencePieceProcessor`):
The *SentencePiece* processor that is used for every conversion (string, tokens and IDs).
TF)lstriprstripNr zxYou need to install jieba to use CpmTokenizer or CpmTokenizerFast. See https://pypi.org/project/jieba/ for installation.z
u ▂▃)
do_lower_case remove_space keep_accents bos_token eos_token unk_token sep_token pad_token cls_token
mask_tokenadditional_special_tokensr r )
isinstancestrr r r r r r
spmSentencePieceProcessorsp_modelLoadjiebaModuleNotFoundError __class__ maketransZ
translatorsuper__init__Z_pad_token_type_id)selfr
r r r r r r r r r r r r kwargsr$ errorr& g/var/www/html/Darija-Ai-API/env/lib/python3.8/site-packages/transformers/models/cpm/tokenization_cpm.pyr) , sB O
zCpmTokenizer.__init__c C s
t | jS )N)lenr" r* r. r. r/
vocab_size s zCpmTokenizer.vocab_sizec s( fddt jD }| j |S )Nc s i | ]} ||qS r. )Zconvert_ids_to_tokens).0ir1 r. r/
s z*CpmTokenizer.get_vocab..)ranger2 updateZadded_tokens_encoder)r* Zvocabr. r1 r/ get_vocab s zCpmTokenizer.get_vocabc C s | j }d |d< |S )Nr" )__dict__copy)r* stater. r. r/ __getstate__ s
zCpmTokenizer.__getstate__c C s8 || _ t| dsi | _tjf | j| _| j| j d S )Nr )r9 hasattrr r r! r" r# r
)r* dr. r. r/ __setstate__ s
zCpmTokenizer.__setstate__c C sj | j rd| }n|}|dddd}| jsXtd|}ddd |D }| jrf| }|S ) N z``"z''ZNFKD c S s g | ]}t |s|qS r. ) unicodedata combining)r3 cr. r. r/
s
z0CpmTokenizer.preprocess_text..)
r joinstripsplitreplacer rC normalizer lower)r* inputsoutputsr. r. r/ preprocess_text s z CpmTokenizer.preprocess_text)textr c C s | |}| jj|td}g }|D ]}t|dkr|d tdkr|d r| j|dd td}|d tkr|d d tkrt|d dkr|dd }n|d dd |d< | |d |
| q"| | q"|S ) zTokenize a string.)Zout_type ,NrB r ) rO r" encoder r0 isdigitZEncodeAsPiecesrJ r appendextend)r* rP piecesZ
new_piecesZpieceZ
cur_piecesr. r. r/ _tokenize s
( zCpmTokenizer._tokenizec C s | j |S )z0Converts a token (str) in an id using the vocab.)r" Z PieceToId)r* tokenr. r. r/ _convert_token_to_id s z!CpmTokenizer._convert_token_to_idc C s | j |S )z=Converts an index (integer) in a token (str) using the vocab.)r" Z IdToPiece)r* indexr. r. r/ _convert_id_to_token s z!CpmTokenizer._convert_id_to_tokenc C s d |td }|S )zIConverts a sequence of tokens (strings for sub-words) in a single string.rB r@ )rG rJ r rH )r* tokensZ
out_stringr. r. r/ convert_tokens_to_string s z%CpmTokenizer.convert_tokens_to_string) token_ids_0 token_ids_1r c C s8 | j g}| jg}|dkr$|| | S || | | | S )a
Build model inputs from a sequence or a pair of sequence for sequence classification tasks by concatenating and
adding special tokens. An XLNet sequence has the following format:
- single sequence: `X `
- pair of sequences: `A B `
Args:
token_ids_0 (`List[int]`):
List of IDs to which the special tokens will be added.
token_ids_1 (`List[int]`, *optional*):
Optional second list of IDs for sequence pairs.
Returns:
`List[int]`: List of [input IDs](../glossary#input-ids) with the appropriate special tokens.
N) sep_token_idZ cls_token_id)r* ra rb sepclsr. r. r/ build_inputs_with_special_tokens s
z-CpmTokenizer.build_inputs_with_special_tokens)ra rb already_has_special_tokensr c s^ |rt j||ddS |dk rHdgt| dg dgt| ddg S dgt| ddg S )a
Retrieve sequence ids from a token list that has no special tokens added. This method is called when adding
special tokens using the tokenizer `prepare_for_model` method.
Args:
token_ids_0 (`List[int]`):
List of IDs.
token_ids_1 (`List[int]`, *optional*):
Optional second list of IDs for sequence pairs.
already_has_special_tokens (`bool`, *optional*, defaults to `False`):
Whether or not the token list is already formatted with special tokens for the model.
Returns:
`List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
T)ra rb rg Nr rQ )r( get_special_tokens_maskr0 )r* ra rb rg r- r. r/ rh s *z$CpmTokenizer.get_special_tokens_maskc C sT | j g}dg}|dkr,t|| dg | S t|| dg t|| dg | S )a
Create a mask from the two sequences passed to be used in a sequence-pair classification task. An XLNet
sequence pair mask has the following format:
```
0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1
| first sequence | second sequence |
```
If `token_ids_1` is `None`, this method only returns the first portion of the mask (0s).
Args:
token_ids_0 (`List[int]`):
List of IDs.
token_ids_1 (`List[int]`, *optional*):
Optional second list of IDs for sequence pairs.
Returns:
`List[int]`: List of [token type IDs](../glossary#token-type-ids) according to the given sequence(s).
Nr rQ )rc r0 )r* ra rb rd Zcls_segment_idr. r. r/ $create_token_type_ids_from_sequences, s
z1CpmTokenizer.create_token_type_ids_from_sequences)save_directoryfilename_prefixr c C s t j|s"td| d d S t j||r6|d ndtd }t j| jt j|krzt j | jrzt
| j| n8t j | jst |d}| j
}|| W 5 Q R X |fS )NzVocabulary path (z) should be a directory-rB r
wb)ospathisdirloggerr, rG VOCAB_FILES_NAMESabspathr
isfiler openr" Zserialized_model_protowrite)r* rk rl Zout_vocab_filefiZcontent_spiece_modelr. r. r/ save_vocabularyK s (
z CpmTokenizer.save_vocabularyc s. t j||}|dddddd}|S )Nr@ rB u ▂u ▃
)r( _decoderJ )r* argsr+ rP r- r. r/ r{ \ s zCpmTokenizer._decode)N)NF)N)N)!__name__
__module__ __qualname____doc__rs Zvocab_files_namesPRETRAINED_VOCAB_FILES_MAPZpretrained_vocab_files_mapr r r r r) propertyr2 r8 r< r? rO r rZ r\ r^ r` intrf boolrh rj r ry r{
__classcell__r. r. r- r/ r & sf w
r )r ro rC shutilr typingr r r r r Z
sentencepiecer Ztokenization_utilsr r
utilsr r Z
get_loggerr} rr rs r r r. r. r. r/ s