T^(hETdZddlmZmZddlZddlmZddlm Z GddeZ dgZ y) z) Text/audio processor class for MusicGen )ListOptionalN)ProcessorMixin)to_numpycpeZdZdZdZdZfdZd dZdZdZ dZ d d e d e e jfd ZxZS)MusicgenProcessoral Constructs a MusicGen processor which wraps an EnCodec feature extractor and a T5 tokenizer into a single processor class. [`MusicgenProcessor`] offers all the functionalities of [`EncodecFeatureExtractor`] and [`TTokenizer`]. See [`~MusicgenProcessor.__call__`] and [`~MusicgenProcessor.decode`] for more information. Args: feature_extractor (`EncodecFeatureExtractor`): An instance of [`EncodecFeatureExtractor`]. The feature extractor is a required input. tokenizer (`T5Tokenizer`): An instance of [`T5Tokenizer`]. The tokenizer is a required input. EncodecFeatureExtractor) T5TokenizerT5TokenizerFastcVt||||j|_d|_y)NF)super__init__feature_extractorcurrent_processor_in_target_context_manager)selfr tokenizer __class__s n/var/www/skyplay_api_hub/venv/lib/python3.12/site-packages/transformers/models/musicgen/processing_musicgen.pyrzMusicgenProcessor.__init__-s) *I6!%!7!7*/'c>|jj|||S)N)tasklanguage no_timestamps)rget_decoder_prompt_ids)rrrrs rrz(MusicgenProcessor.get_decoder_prompt_ids2s~~44$an4oorc|jr|j|i|S|jdd}|jdd}|jdd}t|dkDr |d}|dd}| | t d||j |fi|}||j |g|d|i|}|S|Sdd<d |vr|d |d <|S) a Forwards the `audio` argument to EncodecFeatureExtractor's [`~EncodecFeatureExtractor.__call__`] and the `text` argument to [`~T5Tokenizer.__call__`]. Please refer to the docstring of the above two methods for more information. audioN sampling_ratetextrzAYou need to specify either an `audio` or `text` input to process. input_values padding_mask)rrpoplen ValueErrorrr)rargskwargsrrr inputs audio_inputss r__call__zMusicgenProcessor.__call__5s  * *)4))4:6: : 7D) ?D9 zz&$' t9q=GE8D =T\`a a  #T^^D3F3F  1411%f$fmf_efL =M \ &2.%AF> "-)5n)E~&Mrc|jdd}|jdd}t|dkDr |d}|dd}||j||S|jj|i|S)ai This method is used to decode either batches of audio outputs from the MusicGen model, or batches of token ids from the tokenizer. In the case of decoding token ids, this method forwards all its arguments to T5Tokenizer's [`~PreTrainedTokenizer.batch_decode`]. Please refer to the docstring of this method for more information. rNr#rr!)r#)r$r% _decode_audior batch_decode)rr'r( audio_valuesr#s rr.zMusicgenProcessor.batch_decode[s{ zz'40 zz.$7 t9q=7L8D  #%%l%N N.4>>..?? ?rc:|jj|i|S)z This method forwards all its arguments to T5Tokenizer's [`~PreTrainedTokenizer.decode`]. Please refer to the docstring of this method for more information. )rdecode)rr'r(s rr1zMusicgenProcessor.decodems %t~~$$d5f55rr#returnct|}|j\}}}| t|St|}||jdz }d|jjz }t j |dd|ffd|}|j}t|D]S}t j||||dddf|jjk7} | j|d||<U|S)zn This method strips any padding from the audio values to return a list of numpy audio arrays. Nr!)rrrconstant)constant_values) rshapelistr padding_valuenppadtolistrangeasarrayreshape) rr/r#bszchannelsseq_len differencer9i sliced_audios rr-zMusicgenProcessor._decode_audiots - !-!3!3Xw   % % - |11"55 D22@@@ vvlVa_,Ezcpq #**, s AA::l1o6Qa(D,B,B,P,PPL+228R@LO  A r)NNT)N)__name__ __module__ __qualname____doc__feature_extractor_classtokenizer_classrrr+r.r1rrr:ndarrayr- __classcell__)rs@rr r sP 88O0 p$L@$6DQSQ[Q[L\rr ) rItypingrrnumpyr:processing_utilsrutilsrr __all__rrrTs1".rrj  r