o ix:@s&ddlZddlZddlZddlZddlZddlZddlmZmZddl m Z m Z m Z m Z mZmZmZddlZddlZddlZddlZddlZddlmZddlmZmZddlmZddlmZm Z ddl!m"Z"m#Z#dd l$m%Z%dd l&m'Z'dd l(m)Z)Gd d d ejjZ*GdddZ+d0ddZ,d1de-fddZ.ddZ/e0e,dgdde0e.dgdde0egdde0ej1e"ddgdde0ej1e"ddgd de0ej1e"d!dgd"de0ej1e"d#dgd$de0e#gd%de0e+gd&de0egd'de0e/d(gd)de0e%gd*dd+ Z2Gd,d-d-eeZ3Gd.d/d/e3Z4dS)2N)ABCabstractmethod)AnyCallable CollectionDictMappingTupleUnion)Dataset)check_argument_typescheck_return_type) NpyScpReader)FloatRandomGenerateDatasetIntRandomGenerateDataset)load_num_sequence_textread_2column_text) RttmReader)SoundScpReader) SizedDictc@sBeZdZdddZddZddZdd Zd ed ej fd d Z dS)AdapterForSoundScpReaderNcCs tsJ||_||_d|_dSN)r loaderdtyperate)selfrrrI/home/ubuntu/.local/lib/python3.10/site-packages/espnet2/train/dataset.py__init__s  z!AdapterForSoundScpReader.__init__cCs |jSr)rkeysrrrrr$ zAdapterForSoundScpReader.keyscC t|jSr)lenrr rrr__len__'r!z AdapterForSoundScpReader.__len__cCr"r)iterrr rrr__iter__*r!z!AdapterForSoundScpReader.__iter__keyreturncCs6|j|}t|trtt|dksJt|t|dtr*t|dtjr*|\}}n&t|dtr>t|dtjr>|\}}ntdt|ddt|d|j dure|j |kretd|j d|||_ |j durs| |j }nt|tjsJt||}|j dur| |j }t|tjsJt||S)NrzUnexpected type: z, zSampling rates are mismatched: z != ) r isinstancetupler#intnpndarray RuntimeErrortyperrastype)rr'retvalrarrayrrr __getitem__-s0        z$AdapterForSoundScpReader.__getitem__r) __name__ __module__ __qualname__rrr$r&strr.r/r5rrrrrs  rc@sNeZdZdefddZdefddZdefddZd d Zde j fd d Z d S) H5FileWrapperpathcCs||_t|d|_dS)Nr)r;h5pyFileh5_file)rr;rrrrSszH5FileWrapper.__init__r(cCr"r)r9r?r rrr__repr__Wr!zH5FileWrapper.__repr__cCr"r)r#r?r rrrr$Zr!zH5FileWrapper.__len__cCr"r)r%r?r rrrr&]r!zH5FileWrapper.__iter__cCs|j|}|dS)Nr)r?)rr'valuerrrr5`s zH5FileWrapper.__getitem__N) r6r7r8r9rr@r-r$r&r.r/r5rrrrr:Rs r:cCst|ddd}t||S)NTF) normalize always_2d)rr)r; float_dtyperrrr sound_loaderes rE max_cache_fdcCstj||d}t||S)N)rF)kaldiioload_scpr)r;rDrFrrrr kaldi_loaderts rIcCsNztt|tddd\}}Wnty td|wt|||S)N rand_int__ze.g rand_int_3_10: but got )mapr-r#split ValueErrorr0r)filepath loader_typelowhighrrrrand_int_loaderys $  rSrDzuAudio format types which supported by sndfile wav, flac, etc. utterance_id_a a.wav utterance_id_b b.wav ...)funckwargshelpzlKaldi-ark file type. utterance_id_A /some/where/a.ark:123 utterance_id_B /some/where/a.ark:456 ...z`Npy file format. utterance_id_A /some/where/a.npy utterance_id_B /some/where/b.npy ...text_int)rPzA text file in which is written a sequence of interger numbers separated by space. utterance_id_A 12 0 1 3 utterance_id_B 3 3 1 ...csv_intzA text file in which is written a sequence of interger numbers separated by comma. utterance_id_A 100,80 utterance_id_B 143,80 ... text_floatzA text file in which is written a sequence of float numbers separated by space. utterance_id_A 12. 3.1 3.4 4.4 utterance_id_B 3. 3.12 1.1 ... csv_floatzA text file in which is written a sequence of float numbers separated by comma. utterance_id_A 12.,3.1,3.4,4.4 utterance_id_B 3.,3.12,1.1 ...zReturn text as is. The text must be converted to ndarray by 'preprocess'. utterance_id_A hello world utterance_id_B foo bar ...zA HDF5 file which contains arrays at the first level or the second level. >>> f = h5py.File('file.h5') >>> array1 = f['utterance_id_A'] >>> array2 = f['utterance_id_B'] zzGenerate random float-ndarray which has the given shapes in the file. utterance_id_A 3,4 utterance_id_B 10,4 ...rPze.g. 'rand_int_0_10'. Generate random int-ndarray which has the given shapes in the path. Give the lower and upper value by the file type. e.g. rand_int_0_10 -> Generate integers from 0 to 10. utterance_id_A 3,4 utterance_id_B 10,4 ...arttm file loader, currently support for speaker diarization SPEAKER file1 1 0 1023 spk1 SPEAKER file1 2 4000 3023 spk2 SPEAKER file1 3 500 4023 spk1 END file1 4023 ...) sound kaldi_arknpyrWrXrYrZtexthdf5 rand_floatzrand_int_\d+_\d+rttmc@s\eZdZedefddZedeedffddZedee e ee j fffddZ d S) AbsDatasetr(cCtrNotImplementedErrorrnamerrrhas_namezAbsDataset.has_name.cCrcrrdr rrrnamesrizAbsDataset.namescCrcrrd)ruidrrrr5rizAbsDataset.__getitem__N)r6r7r8rboolrhr r9rjrrr.r/r5rrrrrbs&rbc@seZdZdZ     d deeeeefdeeeee j fgeee j ffd ed ed e e e efd e f d dZdededeee e j ejeejfffddZdefddZdeedffddZddZddZde ee fdeeeee j fffddZdS)! ESPnetDatasetaOPytorch Dataset class for ESPNet. Examples: >>> dataset = ESPnetDataset([('wav.scp', 'input', 'sound'), ... ('token_int', 'output', 'text_int')], ... ) ... uttid, data = dataset['uttid'] {'input': per_utt_array, 'output': per_utt_array} Nfloat32longrpath_name_type_list preprocessrD int_dtypemax_cache_sizerFc CstsJt|dkrtdt|}||_||_||_||_i|_ i|_ |D]4\}}} ||j vr:t d|d| || } | |j |<|| f|j |<t|j |dkr\t |dq(t |trgt|}||_|dkrvtdd|_dSd|_dS)Nrz91 or more elements are required for "path_name_type_list""z" is duplicated for data-keyz has no samplesT)shared)r r#rNcopydeepcopyrrrDrsrF loader_dict debug_infor0 _build_loaderr+r9 humanfriendly parse_sizertrcache) rrqrrrDrsrtrFr;rg_typerrrrrs6        zESPnetDataset.__init__r;rPr(c CstD]q\}}t||rui}|dD]/}|dkr||d<q|dkr)|j|d<q|dkr3|j|d<q|dkr=|j|d<qtd||d}z ||fi|WStytt |drc|j }nt |}t d |d |d wqtd |) zHelper function to instantiate Loader. Args: path: The file path loader_type: loader_type. sound, npy, text_int, text_float, etc rUrPrDrsrFz"Not implemented keyword argument: rTr6zAn error happened with ()zNot supported: loader_type=) DATA_TYPESitemsrematchrDrsrFr0 Exceptionhasattrr6r9loggingerror) rr;rPr'dicrUkey2rTrgrrrr{Cs2        zESPnetDataset._build_loadercCs ||jvSr)ryrfrrrrhjr!zESPnetDataset.has_name.cCr"r)r,ryr rrrrjmr!zESPnetDataset.namescCsttt|jSr)r%nextryvaluesr rrrr&pszESPnetDataset.__iter__c CsZ|jj}|d7}|jD]\}\}}|d|d|d|d7}q |d|jd7}|S)Nrz z : {"path": "z ", "type": "z"}z preprocess: r) __class__r6rzrrr)r_mesrgr;rrrrr@ss zESPnetDataset.__repr__rkc CstsJt|trtt|j}t||}|jdur,||jvr,|j|}||fSi}|j D]d\}}z'||}t|tt frHt |}t|t j tjttjfs]tdt|Wnty{|j|\}}td|d|d|wt|tjr|}n t|tjrt |g}|||<q3|jdur|||}|D]?}||}t|t j std|dt|d|jjdkr||j}n|jjd kr||j }nt!d |j|||<q|jdur|jj"|j#kr||j|<||f} t$| sJ| S) Nz.Must be ndarray, torch.Tensor, str or Number: zError happened with path=z, type=z, id=zIAll values must be converted to np.ndarray object by preprocessing, but "z " is still .fizNot supported dtype: )%r r+r-rr%ryrlistr~rr,r.r4r/torchTensorr9numbersNumber TypeErrorr1rrzrrnumpyrrr0rkindr2rDrsresizertr ) rrkddatargrrAr;rr3rrrr5{sn                   zESPnetDataset.__getitem__)Nrnrorpr)r6r7r8__doc__rr r9rrr.r/r floatr-rrrrrrr{rlrhrjr&r@r5rrrrrm sD  . '0rmr)Nr)5 collectionsrw functoolsrrrabcrrtypingrrrrrr r r=r|rGrr.rtorch.utils.data.datasetr typeguardr r espnet2.fileio.npy_scprespnet2.fileio.rand_gen_datasetrrespnet2.fileio.read_textrrespnet2.fileio.rttmrespnet2.fileio.sound_scprespnet2.utils.sized_dictrrr:rEr-rIrSdictpartialrrbrmrrrrs$     5            z