o ei7@sdZddlZddlZddlZddlZddlmZddlmZm Z ddl Z ddl m Z ddlmZer8ddl mZdd lmZmZdd lmZeeZd d iZd ZeddGdddeZdgZdS)z$Tokenization class for SigLIP model.N)copyfile) TYPE_CHECKINGAny) AddedToken)SentencePieceBackend) TextInput)loggingrequires_backends)requires vocab_filez spiece.modelu▁) sentencepiece)backendsc seZdZdZeZddgZ       d9d eee fdBd dffd d Z e d dZ ddZ d:deedeedBded eeffdd Zdeed eefddZ d;deedeedBd eefddZ d;deedeedBd eefddZddZd d!Zd"ed efd#d$Zdd%d&d'Zd"`): The end of sequence 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. pad_token (`str`, *optional*, defaults to `""`): The token used for padding, for example when batching sequences of different lengths. additional_special_tokens (`list[str]`, *optional*): Additional special tokens used by the tokenizer. sp_model_kwargs (`dict`, *optional*): Will be passed to the `SentencePieceProcessor.__init__()` method. The [Python wrapper for SentencePiece](https://github.com/google/sentencepiece/tree/master/python) can be used, among other things, to set: - `enable_sampling`: Enable subword regularization. - `nbest_size`: Sampling parameters for unigram. Invalid for BPE-Dropout. - `nbest_size = {0,1}`: No sampling is performed. - `nbest_size > 1`: samples from the nbest_size results. - `nbest_size < 0`: assuming that nbest_size is infinite and samples from the all hypothesis (lattice) using forward-filtering-and-backward-sampling algorithm. - `alpha`: Smoothing parameter for unigram sampling, and dropout probability of merge operations for BPE-dropout. model_max_length (`int`, *optional*, defaults to 64): The maximum length (in number of tokens) for model inputs. do_lower_case (`bool`, *optional*, defaults to `True`): Whether or not to lowercase the input when tokenizing. input_idsattention_maskN@Tsp_model_kwargsreturnc st|dt|trt|dddddn|}t|tr#t|dddddn|}t|tr3t|dddddn|}|dur;in||_||_tjd||||||j||d| dS)NprotobufTF)rstriplstrip normalizedspecial)r eos_token unk_token pad_tokenadditional_special_tokensrmodel_max_length do_lower_case)r isinstancestrrrr!super__init__) selfr rrrrrr r!kwargs __class__r"l/home/ubuntu/transcripts/venv/lib/python3.10/site-packages/transformers/models/siglip/tokenization_siglip.pyr&Xs6  zSiglipTokenizer.__init__cCs |jSN)sp_modelget_piece_sizer'r"r"r+ vocab_sizes zSiglipTokenizer.vocab_sizecs(fddtjD}|j|S)Ncsi|]}||qSr")convert_ids_to_tokens).0ir/r"r+ sz-SiglipTokenizer.get_vocab..)ranger0updateadded_tokens_encoder)r'vocabr"r/r+ get_vocabs zSiglipTokenizer.get_vocabF token_ids_0 token_ids_1already_has_special_tokenscsZ|r tj||ddS|durdgt|dgSdgt|dgdgt|dgS)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)r:r;r<Nr)r%get_special_tokens_masklen)r'r:r;r<r)r"r+r>s(z'SiglipTokenizer.get_special_tokens_mask token_idscCs>t|dkr|d|jkrtd|jd|S||jgS)z.Do not add eos again if user already added it.rzThis sequence already has zQ. In future versions this behavior may lead to duplicated eos tokens being added.)r? eos_token_idwarningswarnr)r'r@r"r"r+_add_eos_if_not_presents   z'SiglipTokenizer._add_eos_if_not_presentcCs<|jg}|durt||dgSt||||dgS)a Create a mask from the two sequences passed to be used in a sequence-pair classification task. T5 does not make use of token type ids, therefore a list of zeros is returned. 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 zeros. Nr)rBr?)r'r:r;eosr"r"r+$create_token_type_ids_from_sequencessz4SiglipTokenizer.create_token_type_ids_from_sequencescCs(||}|dur |S||}||S)a Build model inputs from a sequence or a pair of sequence for sequence classification tasks by concatenating and adding special tokens. A sequence has the following format: - single sequence: `X ` - pair of sequences: `A B ` Args: token_ids_0 (`list[int]`): List of IDs to which the special tokens will be added. token_ids_1 (`list[int]`, *optional*): Optional second list of IDs for sequence pairs. Returns: `list[int]`: List of [input IDs](../glossary#input-ids) with the appropriate special tokens. N)rE)r'r:r;r"r"r+ build_inputs_with_special_tokenss  z0SiglipTokenizer.build_inputs_with_special_tokenscCs|j}d|d<|S)Nr-)__dict__copy)r'stater"r"r+ __getstate__s zSiglipTokenizer.__getstate__cCs<||_t|ds i|_tjdi|j|_|j|jdS)Nrr")rIhasattrrspmSentencePieceProcessorr-Loadr )r'dr"r"r+ __setstate__s  zSiglipTokenizer.__setstate__textcCs|tddtjS)N) translater$ maketransstring punctuation)r'rSr"r"r+remove_punctuationsz"SiglipTokenizer.remove_punctuationkeep_punctuation_exact_stringcsVjr|}|r|fdd||D}n|}tdd|}|}|S)aReturns canonicalized `text` (puncuation removed). Args: text (`str`): String to be canonicalized. keep_punctuation_exact_string (`str`, *optional*): If provided, then this exact string is kept. For example providing '{}' will keep any occurrences of '{}' (but will still remove '{' and '}' that appear separately). c3s|]}|VqdSr,)rY)r2partr/r"r+ s  z4SiglipTokenizer.canonicalize_text..z\s+ )r!lowerjoinsplitrYresubstrip)r'rSr[r"r/r+canonicalize_texts  z!SiglipTokenizer.canonicalize_textrc sVtjt|tdfi|}t|dkr)|dtkr)|d|jvr)|dd}|S)z8 Converts a string to a list of tokens. r^r=rN)r%tokenizeSPIECE_UNDERLINEreplacer?all_special_tokens)r'rSadd_special_tokensr(tokensr)r"r+rfs & zSiglipTokenizer.tokenizecCst|jt|jSr,)r?r-encoder$rr/r"r"r+unk_token_lengthsz SiglipTokenizer.unk_token_lengthcKsT|j|dd}|jj|td}|jj|j|td}t||jkr(||jdS|S)u* Returns a tokenized string. We de-activated the `add_dummy_prefix` option, thus the sentencepiece internals will always strip any SPIECE_UNDERLINE. For example: `self.sp_model.encode(f"{SPIECE_UNDERLINE}Hey", out_type = str)` will give `['H', 'e', 'y']` instead of `['▁He', 'y']`. Thus we always encode `f"{unk_token}text"` and strip the `unk_token`. Here is an example with `unk_token = ""` and `unk_token_length = 4`. `self.tokenizer.sp_model.encode(" Hey", out_type = str)[4:]`. NrZ)out_type)rer-rlr$rr?rm)r'rSr(rkr"r"r+ _tokenizes  zSiglipTokenizer._tokenizecCs |j|S)z0Converts a token (str) in an id using the vocab.)r- piece_to_id)r'tokenr"r"r+_convert_token_to_id2s z$SiglipTokenizer._convert_token_to_idcCs|j|}|S)z=Converts an index (integer) in a token (str) using the vocab.)r- IdToPiece)r'indexrqr"r"r+_convert_id_to_token6s z$SiglipTokenizer._convert_id_to_tokencCspg}d}d}|D]#}||jvr$|s|d7}||j||7}d}g}q||d}q||j|7}|S)z:Converts a sequence of tokens (string) in a single string.rTFr^T)rir-decodeappendrd)r'rkcurrent_sub_tokens out_stringprev_is_specialrqr"r"r+convert_tokens_to_string;s  z(SiglipTokenizer.convert_tokens_to_stringsave_directoryfilename_prefixcCstj|std|ddStj||r|dndtd}tj|jtj|kr?tj |jr?t |j||fStj |jsgt |d}|j }||Wd|fS1sbwY|fS)NzVocabulary path (z) should be a directory-rTr wb)ospathisdirloggererrorr`VOCAB_FILES_NAMESabspathr isfileropenr-serialized_model_protowrite)r'r|r}out_vocab_fileficontent_spiece_modelr"r"r+save_vocabularyNs" (     zSiglipTokenizer.save_vocabulary)rrrNNrT)NFr,)F)"__name__ __module__ __qualname____doc__rvocab_files_namesmodel_input_namesdictr$rr&propertyr0r9listintboolr>rErGrHrLrRrYrerfrmrorrrur{tupler __classcell__r"r"r)r+r+st( -         (r)rrrbrWrCshutilrtypingrrr rNtokenization_utils_baser tokenization_utils_sentencepiecerrutilsr r utils.import_utilsr get_loggerrrrrgr__all__r"r"r"r+s,        6