/usr/local/lib/python3.6/site-packages/transformers/data/processors/__pycache__
NameSizeModeActions
glue.cpython-36.pyc241670644editdlrm
squad.cpython-36.pyc204180644editdlrm
utils.cpython-36.pyc121110644editdlrm
xnli.cpython-36.pyc25520644editdlrm
__init__.cpython-36.pyc7170644editdlrm
Edit: /usr/local/lib/python3.6/site-packages/transformers/data/processors/__pycache__/glue.cpython-36.pyc (24167B)
3 EgZ @sdZddlZddlZddlmZddlmZddlmZm Z m Z ddl m Z ddl mZmZd d l mZmZmZer~ddlZejeZd Zd0e eed fe e ed ddZeredfejje e eejjdddZd1eee e ed ddZGdddeZ GdddeZ!GdddeZ"Gddde"Z#GdddeZ$Gdd d eZ%Gd!d"d"eZ&Gd#d$d$eZ'Gd%d&d&eZ(Gd'd(d(eZ)Gd)d*d*eZ*d+dd+d+d d+d+d+d+d, Z+e$e"e#e!e%e&e'e(e)e*d- Z,d.d.d.d.d.d/d.d.d.d.d- Z-dS)2z GLUE processors and helpersN)asdict)Enum)ListOptionalUnion)PreTrainedTokenizer)is_tf_availablelogging) DataProcessor InputExample InputFeaturesuThis {0} will be removed from the library soon, preprocessing should be handled with the 🤗 Datasets library. You can have a look at this example script for pointers: https://github.com/huggingface/transformers/blob/main/examples/pytorch/text-classification/run_glue.pyztf.data.Dataset)examples tokenizer max_lengthcCsZtjtjdttrFt|tjj rF|dkr6t dt ||||dSt ||||||dS)a= Loads a data file into a list of `InputFeatures` Args: examples: List of `InputExamples` or `tf.data.Dataset` containing the examples. tokenizer: Instance of a tokenizer that will tokenize the examples max_length: Maximum example length. Defaults to the tokenizer's max_len task: GLUE task label_list: List of labels. Can be obtained from the processor using the `processor.get_labels()` method output_mode: String indicating the output mode. Either `regression` or `classification` Returns: If the `examples` input is a `tf.data.Dataset`, will return a `tf.data.Dataset` containing the task-specific features. If the input is a list of `InputExamples`, will return a list of task-specific `InputFeatures` which can be fed to the model. functionNzWWhen calling glue_convert_examples_to_features from TF, the task parameter is required.)rtask)rr label_list output_mode) warningswarnDEPRECATION_WARNINGformat FutureWarningr isinstancetfdataDataset ValueError%_tf_glue_convert_examples_to_features"_glue_convert_examples_to_features)rrrrrrr"K/usr/local/lib/python3.6/site-packages/transformers/data/processors/glue.py!glue_convert_examples_to_features)sr$)rrrreturncst|fdd|D}t||||d|dkr:tjntj}fdd}|j}tjjj|dd|D|fd d|Dtj gfS) zb Returns: A `tf.data.Dataset` containing the task-specific features. csg|]}jj|qSr")Ztfds_mapget_example_from_tensor_dict).0example) processorr"r# Zsz9_tf_glue_convert_examples_to_features..)rrzsts-bc3s<x6D].}ddt|jD}|jd}||fVqWdS)NcSsi|]\}}|dk r||qS)Nr")r'kvr"r"r# `szF_tf_glue_convert_examples_to_features..gen..label)ritemspop)exdr.)featuresr"r#gen^s  z2_tf_glue_convert_examples_to_features..gencSsi|] }tj|qSr")rZint32)r'r+r"r"r#r-hsz9_tf_glue_convert_examples_to_features..cSsi|]}tjdg|qS)N)r TensorShape)r'r+r"r"r#r-is) glue_processorsr$rZfloat32Zint64Zmodel_input_namesrrZfrom_generatorr5)rrrrZ label_typer4Z input_namesr")r3r)r#r Ns  r c sl|dkr|j}|dk rlt|}|dkrF|j}tjd|d|dkrlt|tjdd|ddt|Dttt t dfdfdd fd d |D}|d d |D|d d dg}xHt t |D]8fddD} t f| d|i} |j| qWxNt|ddD]:\} tjdtjd| jtjd|q*W|S)NzUsing label list z for task zUsing output mode cSsi|]\}}||qSr"r")r'ir.r"r"r#r-sz6_glue_convert_examples_to_features..)r(r%cs>|jdkrdSdkr |jSdkr2t|jStdS)Nclassification regression)r.floatKeyError)r() label_maprr"r#label_from_examples   z>_glue_convert_examples_to_features..label_from_examplecsg|] }|qSr"r")r'r()r=r"r#r*sz6_glue_convert_examples_to_features..cSsg|]}|j|jfqSr")text_atext_b)r'r(r"r"r#r*srT)rpaddingZ truncationcsi|]}||qSr"r")r'r+)batch_encodingr7r"r#r-sr.z*** Example ***zguid: z features: )Zmodel_max_lengthr6 get_labelsloggerinfoglue_output_modes enumerater rintr:rangelenrappendguid) rrrrrrr)labelsr3inputsfeaturer(r")rAr7r=r<rr#r!ms8    r!c@seZdZdZdZdS) OutputModer8r9N)__name__ __module__ __qualname__r8r9r"r"r"r#rPsrPcsPeZdZdZfddZddZddZdd Zd d Zd d Z ddZ Z S) MrpcProcessorz/Processor for the MRPC data set (GLUE version).cs$tj||tjtjdtdS)Nr))super__init__rrrrr)selfargskwargs) __class__r"r#rVszMrpcProcessor.__init__cCs>t|dj|djjd|djjdt|djS)zSee base class.idx sentence1zutf-8 sentence2r.)r numpydecodestr)rW tensor_dictr"r"r#r&s  z*MrpcProcessor.get_example_from_tensor_dictcCs6tjdtjj|d|j|jtjj|ddS)zSee base class.z LOOKING AT z train.tsvtrain)rDrEospathjoin_create_examples _read_tsv)rWdata_dirr"r"r#get_train_examplessz MrpcProcessor.get_train_examplescCs|j|jtjj|ddS)zSee base class.zdev.tsvdev)rfrgrcrdre)rWrhr"r"r#get_dev_examplesszMrpcProcessor.get_dev_examplescCs|j|jtjj|ddS)zSee base class.ztest.tsvtest)rfrgrcrdre)rWrhr"r"r#get_test_examplesszMrpcProcessor.get_test_examplescCsddgS)zSee base class.01r")rWr"r"r#rCszMrpcProcessor.get_labelsc Cspg}xft|D]Z\}}|dkr q|d|}|d}|d}|dkrJdn|d} |jt|||| dqW|S)z5Creates examples for the training, dev and test sets.r-rrlN)rLr>r?r.)rGrKr ) rWlinesset_typerr7linerLr>r?r.r"r"r#rfszMrpcProcessor._create_examples) rQrRrS__doc__rVr&rirkrmrCrf __classcell__r"r")rZr#rTs  rTcsPeZdZdZfddZddZddZdd Zd d Zd d Z ddZ Z S) MnliProcessorz3Processor for the MultiNLI data set (GLUE version).cs$tj||tjtjdtdS)Nr))rUrVrrrrr)rWrXrY)rZr"r#rVszMnliProcessor.__init__cCs>t|dj|djjd|djjdt|djS)zSee base class.r[Zpremisezutf-8Z hypothesisr.)r r^r_r`)rWrar"r"r#r&s  z*MnliProcessor.get_example_from_tensor_dictcCs|j|jtjj|ddS)zSee base class.z train.tsvrb)rfrgrcrdre)rWrhr"r"r#risz MnliProcessor.get_train_examplescCs|j|jtjj|ddS)zSee base class.zdev_matched.tsvZ dev_matched)rfrgrcrdre)rWrhr"r"r#rkszMnliProcessor.get_dev_examplescCs|j|jtjj|ddS)zSee base class.ztest_matched.tsvZ test_matched)rfrgrcrdre)rWrhr"r"r#rmszMnliProcessor.get_test_examplescCs dddgS)zSee base class.Z contradiction entailmentZneutralr")rWr"r"r#rCszMnliProcessor.get_labelsc Csvg}xlt|D]`\}}|dkr q|d|d}|d}|d}|jdrPdn|d } |jt|||| dqW|S) z5Creates examples for the training, dev and test sets.rrp rlNr )rLr>r?r.)rG startswithrKr ) rWrrrsrr7rtrLr>r?r.r"r"r#rfszMnliProcessor._create_examples) rQrRrSrurVr&rirkrmrCrfrvr"r")rZr#rws  rwcs0eZdZdZfddZddZddZZS)MnliMismatchedProcessorz>Processor for the MultiNLI Mismatched data set (GLUE version).cs$tj||tjtjdtdS)Nr))rUrVrrrrr)rWrXrY)rZr"r#rV sz MnliMismatchedProcessor.__init__cCs|j|jtjj|ddS)zSee base class.zdev_mismatched.tsvZdev_mismatched)rfrgrcrdre)rWrhr"r"r#rk sz(MnliMismatchedProcessor.get_dev_examplescCs|j|jtjj|ddS)zSee base class.ztest_mismatched.tsvZtest_mismatched)rfrgrcrdre)rWrhr"r"r#rmsz)MnliMismatchedProcessor.get_test_examples)rQrRrSrurVrkrmrvr"r")rZr#r}s r}csPeZdZdZfddZddZddZdd Zd d Zd d Z ddZ Z S) ColaProcessorz/Processor for the CoLA data set (GLUE version).cs$tj||tjtjdtdS)Nr))rUrVrrrrr)rWrXrY)rZr"r#rVszColaProcessor.__init__cCs0t|dj|djjddt|djS)zSee base class.r[sentencezutf-8Nr.)r r^r_r`)rWrar"r"r#r&s  z*ColaProcessor.get_example_from_tensor_dictcCs|j|jtjj|ddS)zSee base class.z train.tsvrb)rfrgrcrdre)rWrhr"r"r#ri&sz ColaProcessor.get_train_examplescCs|j|jtjj|ddS)zSee base class.zdev.tsvrj)rfrgrcrdre)rWrhr"r"r#rk*szColaProcessor.get_dev_examplescCs|j|jtjj|ddS)zSee base class.ztest.tsvrl)rfrgrcrdre)rWrhr"r"r#rm.szColaProcessor.get_test_examplescCsddgS)zSee base class.rnror")rWr"r"r#rC2szColaProcessor.get_labelsc Cs~|dk}|r|dd}|r dnd}g}xPt|D]D\}}|d|}||} |rXdn|d} |jt|| d| dq2W|S)z5Creates examples for the training, dev and test sets.rlr Nrrp)rLr>r?r.)rGrKr ) rWrrrs test_mode text_indexrr7rtrLr>r.r"r"r#rf6s  zColaProcessor._create_examples) rQrRrSrurVr&rirkrmrCrfrvr"r")rZr#r~s  r~csPeZdZdZfddZddZddZdd Zd d Zd d Z ddZ Z S) Sst2Processorz0Processor for the SST-2 data set (GLUE version).cs$tj||tjtjdtdS)Nr))rUrVrrrrr)rWrXrY)rZr"r#rVHszSst2Processor.__init__cCs0t|dj|djjddt|djS)zSee base class.r[rzutf-8Nr.)r r^r_r`)rWrar"r"r#r&Ls  z*Sst2Processor.get_example_from_tensor_dictcCs|j|jtjj|ddS)zSee base class.z train.tsvrb)rfrgrcrdre)rWrhr"r"r#riUsz Sst2Processor.get_train_examplescCs|j|jtjj|ddS)zSee base class.zdev.tsvrj)rfrgrcrdre)rWrhr"r"r#rkYszSst2Processor.get_dev_examplescCs|j|jtjj|ddS)zSee base class.ztest.tsvrl)rfrgrcrdre)rWrhr"r"r#rm]szSst2Processor.get_test_examplescCsddgS)zSee base class.rnror")rWr"r"r#rCaszSst2Processor.get_labelsc Csxg}|dkrdnd}x^t|D]R\}}|dkr0q|d|}||}|dkrRdn|d} |jt||d| dqW|S)z5Creates examples for the training, dev and test sets.rlr rrpN)rLr>r?r.)rGrKr ) rWrrrsrrr7rtrLr>r.r"r"r#rfeszSst2Processor._create_examples) rQrRrSrurVr&rirkrmrCrfrvr"r")rZr#rEs  rcsPeZdZdZfddZddZddZdd Zd d Zd d Z ddZ Z S) StsbProcessorz0Processor for the STS-B data set (GLUE version).cs$tj||tjtjdtdS)Nr))rUrVrrrrr)rWrXrY)rZr"r#rVvszStsbProcessor.__init__cCs>t|dj|djjd|djjdt|djS)zSee base class.r[r\zutf-8r]r.)r r^r_r`)rWrar"r"r#r&zs  z*StsbProcessor.get_example_from_tensor_dictcCs|j|jtjj|ddS)zSee base class.z train.tsvrb)rfrgrcrdre)rWrhr"r"r#risz StsbProcessor.get_train_examplescCs|j|jtjj|ddS)zSee base class.zdev.tsvrj)rfrgrcrdre)rWrhr"r"r#rkszStsbProcessor.get_dev_examplescCs|j|jtjj|ddS)zSee base class.ztest.tsvrl)rfrgrcrdre)rWrhr"r"r#rmszStsbProcessor.get_test_examplescCsdgS)zSee base class.Nr")rWr"r"r#rCszStsbProcessor.get_labelsc Cstg}xjt|D]^\}}|dkr q|d|d}|d}|d}|dkrNdn|d } |jt|||| dqW|S) z5Creates examples for the training, dev and test sets.rrpryrlNr )rLr>r?r.r{)rGrKr ) rWrrrsrr7rtrLr>r?r.r"r"r#rfszStsbProcessor._create_examples) rQrRrSrurVr&rirkrmrCrfrvr"r")rZr#rss  rcsPeZdZdZfddZddZddZdd Zd d Zd d Z ddZ Z S) QqpProcessorz.Processor for the QQP data set (GLUE version).cs$tj||tjtjdtdS)Nr))rUrVrrrrr)rWrXrY)rZr"r#rVszQqpProcessor.__init__cCs>t|dj|djjd|djjdt|djS)zSee base class.r[Z question1zutf-8Z question2r.)r r^r_r`)rWrar"r"r#r&s  z)QqpProcessor.get_example_from_tensor_dictcCs|j|jtjj|ddS)zSee base class.z train.tsvrb)rfrgrcrdre)rWrhr"r"r#riszQqpProcessor.get_train_examplescCs|j|jtjj|ddS)zSee base class.zdev.tsvrj)rfrgrcrdre)rWrhr"r"r#rkszQqpProcessor.get_dev_examplescCs|j|jtjj|ddS)zSee base class.ztest.tsvrl)rfrgrcrdre)rWrhr"r"r#rmszQqpProcessor.get_test_examplescCsddgS)zSee base class.rnror")rWr"r"r#rCszQqpProcessor.get_labelsc Cs|dk}|rdnd}|rdnd}g}xt|D]v\}}|dkr@q.|d|d} y$||} ||} |rldn|d } Wntk rw.YnX|jt| | | | d q.W|S) z5Creates examples for the training, dev and test sets.rlr rrqrrpNrB)rLr>r?r.)rG IndexErrorrKr ) rWrrrsrZq1_indexZq2_indexrr7rtrLr>r?r.r"r"r#rfs   zQqpProcessor._create_examples) rQrRrSrurVr&rirkrmrCrfrvr"r")rZr#rs  rcsPeZdZdZfddZddZddZdd Zd d Zd d Z ddZ Z S) QnliProcessorz/Processor for the QNLI data set (GLUE version).cs$tj||tjtjdtdS)Nr))rUrVrrrrr)rWrXrY)rZr"r#rVszQnliProcessor.__init__cCs>t|dj|djjd|djjdt|djS)zSee base class.r[Zquestionzutf-8rr.)r r^r_r`)rWrar"r"r#r&s  z*QnliProcessor.get_example_from_tensor_dictcCs|j|jtjj|ddS)zSee base class.z train.tsvrb)rfrgrcrdre)rWrhr"r"r#risz QnliProcessor.get_train_examplescCs|j|jtjj|ddS)zSee base class.zdev.tsvrj)rfrgrcrdre)rWrhr"r"r#rkszQnliProcessor.get_dev_examplescCs|j|jtjj|ddS)zSee base class.ztest.tsvrl)rfrgrcrdre)rWrhr"r"r#rmszQnliProcessor.get_test_examplescCsddgS)zSee base class.rxnot_entailmentr")rWr"r"r#rCszQnliProcessor.get_labelsc Cstg}xjt|D]^\}}|dkr q|d|d}|d}|d}|dkrNdn|d} |jt|||| dqW|S) z5Creates examples for the training, dev and test sets.rrpr rrlN)rLr>r?r.r{)rGrKr ) rWrrrsrr7rtrLr>r?r.r"r"r#rfszQnliProcessor._create_examples) rQrRrSrurVr&rirkrmrCrfrvr"r")rZr#rs  rcsPeZdZdZfddZddZddZdd Zd d Zd d Z ddZ Z S) RteProcessorz.Processor for the RTE data set (GLUE version).cs$tj||tjtjdtdS)Nr))rUrVrrrrr)rWrXrY)rZr"r#rVszRteProcessor.__init__cCs>t|dj|djjd|djjdt|djS)zSee base class.r[r\zutf-8r]r.)r r^r_r`)rWrar"r"r#r& s  z)RteProcessor.get_example_from_tensor_dictcCs|j|jtjj|ddS)zSee base class.z train.tsvrb)rfrgrcrdre)rWrhr"r"r#riszRteProcessor.get_train_examplescCs|j|jtjj|ddS)zSee base class.zdev.tsvrj)rfrgrcrdre)rWrhr"r"r#rkszRteProcessor.get_dev_examplescCs|j|jtjj|ddS)zSee base class.ztest.tsvrl)rfrgrcrdre)rWrhr"r"r#rmszRteProcessor.get_test_examplescCsddgS)zSee base class.rxrr")rWr"r"r#rCszRteProcessor.get_labelsc Cstg}xjt|D]^\}}|dkr q|d|d}|d}|d}|dkrNdn|d} |jt|||| dqW|S) z5Creates examples for the training, dev and test sets.rrpr rrlN)rLr>r?r.r{)rGrKr ) rWrrrsrr7rtrLr>r?r.r"r"r#rf#szRteProcessor._create_examples) rQrRrSrurVr&rirkrmrCrfrvr"r")rZr#rs  rcsPeZdZdZfddZddZddZdd Zd d Zd d Z ddZ Z S) WnliProcessorz/Processor for the WNLI data set (GLUE version).cs$tj||tjtjdtdS)Nr))rUrVrrrrr)rWrXrY)rZr"r#rV4szWnliProcessor.__init__cCs>t|dj|djjd|djjdt|djS)zSee base class.r[r\zutf-8r]r.)r r^r_r`)rWrar"r"r#r&8s  z*WnliProcessor.get_example_from_tensor_dictcCs|j|jtjj|ddS)zSee base class.z train.tsvrb)rfrgrcrdre)rWrhr"r"r#riAsz WnliProcessor.get_train_examplescCs|j|jtjj|ddS)zSee base class.zdev.tsvrj)rfrgrcrdre)rWrhr"r"r#rkEszWnliProcessor.get_dev_examplescCs|j|jtjj|ddS)zSee base class.ztest.tsvrl)rfrgrcrdre)rWrhr"r"r#rmIszWnliProcessor.get_test_examplescCsddgS)zSee base class.rnror")rWr"r"r#rCMszWnliProcessor.get_labelsc Cstg}xjt|D]^\}}|dkr q|d|d}|d}|d}|dkrNdn|d} |jt|||| dqW|S) z5Creates examples for the training, dev and test sets.rrpr rrlN)rLr>r?r.r{)rGrKr ) rWrrrsrr7rtrLr>r?r.r"r"r#rfQszWnliProcessor._create_examples) rQrRrSrurVr&rirkrmrCrfrvr"r")rZr#r1s  rr) colamnlimrpczsst-2zsts-bqqpqnlirtewnli) rrzmnli-mmrzsst-2zsts-brrrrr8r9)NNNN)NNNN).rurcrZ dataclassesrenumrtypingrrrZtokenization_utilsrutilsr r r r rZ tensorflowr get_loggerrQrDrrHr$r`rrr r!rPrTrwr}r~rrrrrrZglue_tasks_num_labelsr6rFr"r"r"r#s     $1/./..4../