o dZh| @sddlmZddlmZmZmZddlmZddlm Z er4ddl Z ddl m Z ddlmZddlmZGd d d eZGd d d ZGd ddeddZGdddeddZ   d$d%d d!Zed"krddlZddlZeZed#eZej ej!Z"ede"Z#e$e#dSdS)&) annotations) TYPE_CHECKINGLiteral TypedDict) AudioData)WhisperCompatibleRecognizerN WhisperModel)Segment)Unpackc@s&eZdZUded<ded<ded<dS)TranscribeOutputstrtextz list[Segment]segmentslanguageN__name__ __module__ __qualname____annotations__rrj/var/www/auris/lib/python3.10/site-packages/speech_recognition/recognizers/whisper_local/faster_whisper.pyr s  r c@s eZdZd ddZdd d Zd S)TranscribableAdaptermodelr returnNonecCs ||_dSN)r)selfrrrr__init__s zTranscribableAdapter.__init__ audio_array np.ndarrayr cKs>|jj|fi|\}}t|}ddd|D||jdS)N css|]}|jVqdSr)r).0segmentrrr !sz2TranscribableAdapter.transcribe..)rrr)r transcribelistjoinr)rrkwargsZsegments_generatorinforrrrr%s zTranscribableAdapter.transcribeN)rr rr)rr rr )rrrrr%rrrrrs rc@s&eZdZUded<ded<ded<dS)InitOptionalParameterszLiteral['cpu', 'gpu', 'auto']Zdevicer Z compute_typeZ download_rootNrrrrrr*'  r*F)totalc@s&eZdZUded<ded<ded<dS)TranscribeOptionalParametersr rz"Literal['transcribe', 'translate']taskintZ beam_sizeNrrrrrr-/r+r-base audio_datarrr show_dictbool init_optionsInitOptionalParameters | Nonetranscribe_options$Unpack[TranscribeOptionalParameters]rstr | TranscribeOutputcKsBddlm}||fi|p i}tt|}|j|fd|i|S)aPerforms speech recognition on ``audio_data`` (an ``AudioData`` instance), using Whisper. Pick ``model`` size (Same as Whisper). If ``show_dict`` is true, returns the detailed response from Whisper, including the detected language. Otherwise returns only the transcription. You can specify: * ``language``: recognition language, an uncapitalized 2 letters language name like "en" or "fr". * If not set, Faster Whisper will automatically detect the language. * ``task`` * If you want transcribe + **translate** to english, set ``task="translate"``. Other values are passed directly to whisper. See https://github.com/SYSTRAN/faster-whisper/blob/master/faster_whisper/transcribe.py for all options. rrr2)faster_whisperr rr recognize)Z recognizerr1rr2r4r6r Zwhisper_recognizerrrrr:7s r:__main__ audio_file)r0FN) r1rrr r2r3r4r5r6r7rr8)% __future__rtypingrrrZspeech_recognition.audiorZ1speech_recognition.recognizers.whisper_local.basernumpynpr9r Zfaster_whisper.transcriber Ztyping_extensionsr r rr*r-r:rargparseZspeech_recognitionsrArgumentParserparser add_argument parse_argsargs from_filer<r1Z transcriptionprintrrrrs6        %