hJJjYddlZddlZddlmZmZmZmZddlmZ ddl Z ddl Z ddl mZmZmZmZmZmZmZddlmZmZddlmZddlmZddlmZmZd efd Z d ee!d eefd Z"GddZ#dddddddddddddd dee$e jJe jLfde$dee'deeeedffd eed!eed"eed#e'd$ee$d%e'd&e$d'e$d(ee$eefd)eefd*Z(y)+N)ListOptionalTupleUnion)FRAMES_PER_SECOND HOP_LENGTHN_FRAMES N_SAMPLES SAMPLE_RATElog_mel_spectrogram pad_or_trim)DecodingOptionsDecodingResult) load_model)add_word_timestamps) LANGUAGES get_tokenizersecondsc|dk\sJdt|dz}|dz}||dzz}|dz}||dzz}|dz}||dzz}|dkDr|ddnd }||dd|dd |d S) Nrznon-negative timestamp expectedg@@i6i`i02d:.03d)round)r millisecondshoursminutes hours_markers `/Users/ahmed/devFolder/claude-voice/.venv/lib/python3.12/site-packages/mlx_whisper/transcribe.py_format_timestampr"s a<:::<6)*L I %EEI%%Lf$GGf$$Le#GGeO#L&+aieC[?RL^GC='#a S7I JJsegmentsreturncVtdt|D|r |ddSdS)Nc3LK|]}t|dD] }|d yw)wordsendN)reversed).0sws r! z_get_end..-s)L-ax' 7K!57K-s"$r))nextr*r$s r!_get_endr2+s7 L8H-L' U -1 r#cBeZdZdZdZededejfdZ y) ModelHolderN model_pathdtypec|j||jk7rt|||_||_|jS)N)r6)modelr5r)clsr5r6s r! get_modelzModelHolder.get_model6s6 99  cnn <":U;CI'CNyyr#) __name__ __module__ __qualname__r8r5 classmethodstrmxDtyper:r#r!r4r42s0 EJ3rxxr#r4zmlx-community/whisper-tiny)g?g?333333?g??g333333@grDTFu "'“¿([{-u"'.。,，!！?？:：”)]}、0) path_or_hf_repoverbose temperaturecompression_ratio_thresholdlogprob_thresholdno_speech_thresholdcondition_on_previous_textinitial_promptword_timestampsprepend_punctuationsappend_punctuationsclip_timestampshallucination_silence_thresholdaudiorGrHrI.rJrKrLrMrNrOrPrQrRrSc >RSTUVWjddrtjntj}tj ||Rt |Rjjt}|jdtz }t|tztz }|r"tj UUdk7rUfd}nd}jdd Rj"sd d<n|r t%d t'|td j)|}Rj+|\}}t-||j d<|&t%dt.dj1d}jdd}t3Rj"Rj4||Vt7| t8r+| r| j;dngDcgc] }t|} }| Dcgc]}t=|t>z}}tA|dk(r|jCdtA|dzdk(r|jC|ntE||d|d<tGtI|d d d|dd d}dS| r|dk(rtKjLddtjNdtPfRfd }d}||dTtRjjRz}|tztz }g} g}!d}"|4VjUd|jWz}#| jY|#ng}#dtdtd tjNd!tPfTVfd" }$t[jZ|d#|d$u%5}%d&}&|D]\}'}(T|(ks tTtztz })tTtztztz }*tEt|Tz |(Tz }+|TT|+z}|+tztz },t'|td j)|}| |"d d'<||}-t]jN|-j^}.*|-j`kD}/|-jbkDrd$}/|/rT|+z TT}0g}1d(tddtfd)WdtftddthfSWfd* }2d+tjtddtftdfd,}3|.Vjlk\}4|4dd jod$dgk(}5t]jpt]jr|4d d|4dd d}6|6dz }6tA|6dkDr|6jo}7|5r|7jCtA|.d}8|7D]p}9|.|8|9}:|:djuVjlz };|:djuVjlz }<|1jC|$|)|;|zz|)|<|zz|:|--|9}8r|5rT|+z Tn|.|8dz juVjlz }=T|=|zz Tn|,}>|.|4jwd}?tA|?dkDrE|?djuVjlk7r%|?djuVjlz }=|=|z}>|1jC|$|)|)|>z|.|--T|+z T| rty|1RV||+| | |&.|5s$t{|1}@|@@|)kDrt=@t>zT| @| }A|5s4t{|1}@|@'@|)kDr"|*@z }B|BAkDrt=@t>zTn|0|+zT|3|1}C|C,|2Cr$Cd|)z }D|DAkDr|0t=Dt>zzT|&}Et}tA|1D]}F|1|F}G|Gd/s|2Gr|3|1Fdzd }H|H Hd/dd}In|)|,z}IGdEz AkDxsGdAkxs Gd|)z d0k}JIGdz AkDxs|2Hxs |*Gdz d0k}KJr8Kr6t=t-|)dzGdt>zT||Gdz Akr|Tg|1Fd nGd}Et{|1}@|@@}&|rD|1D]?}G|Gd|Gd|Gd1}N}M}Ld2t|Ld3t|Md4|N}Ot%|OAt|1D]6\}P}G|Gd|Gdk(sGd1jWd5k(s(d5Gd1<g|Gd <g|Gd/<8|!jYt|1tA|!6DPGcgc] \}P}Gd7|Pi|G c}G}P| jY|1DGQcgc]}G|Gd D]}Q|Qc}Q}G|r|-jd8kDr tA| }"|%jtE|T|0z T|(kr d d d teVj| tA|#d |!|9Scc}wcc}wcc}G}Pwcc}Q}Gw#1swYHxYw):a Transcribe an audio file using Whisper Parameters ---------- audio: Union[str, np.ndarray, mx.array] The path to the audio file to open, or the audio waveform path_or_hf_repo: str The localpath to the Whisper model or HF Hub repo with the MLX converted weights. verbose: bool Whether to display the text being decoded to the console. If True, displays all the details, If False, displays minimal details. If None, does not display anything temperature: Union[float, Tuple[float, ...]] Temperature for sampling. It can be a tuple of temperatures, which will be successively used upon failures according to either `compression_ratio_threshold` or `logprob_threshold`. compression_ratio_threshold: float If the gzip compression ratio is above this value, treat as failed logprob_threshold: float If the average log probability over sampled tokens is below this value, treat as failed no_speech_threshold: float If the no_speech probability is higher than this value AND the average log probability over sampled tokens is below `logprob_threshold`, consider the segment as silent condition_on_previous_text: bool if True, the previous output of the model is provided as a prompt for the next window; disabling may make the text inconsistent across windows, but the model becomes less prone to getting stuck in a failure loop, such as repetition looping or timestamps going out of sync. word_timestamps: bool Extract word-level timestamps using the cross-attention pattern and dynamic time warping, and include the timestamps for each word in each segment. prepend_punctuations: str If word_timestamps is True, merge these punctuation symbols with the next word append_punctuations: str If word_timestamps is True, merge these punctuation symbols with the previous word initial_prompt: Optional[str] Optional text to provide as a prompt for the first window. This can be used to provide, or "prompt-engineer" a context for transcription, e.g. custom vocabularies or proper nouns to make it more likely to predict those word correctly. decode_options: dict Keyword arguments to construct `DecodingOptions` instances clip_timestamps: Union[str, List[float]] Comma-separated list start,end,start,end,... timestamps (in seconds) of clips to process. The last end timestamp defaults to the end of the file. hallucination_silence_threshold: Optional[float] When word_timestamps is True, skip silent periods longer than this threshold (in seconds) when a possible hallucination is detected Returns ------- A dictionary containing the resulting text ("text") and segment-level details ("segments"), and the spoken language ("language"), which is detected when `decode_options["language"]` is None. fp16T)n_melspaddingzutf-8cH|jdjS)Nreplace)errors)encodedecode)xsystem_encodings r!ztranscribe..s !((?9("M"T"T#r#c|SNrB)r_s r!raztranscribe..s!r#languageNenzoDetecting language using up to the first 30 seconds. Use the `language` decoding option to specify the language)axis)keyzDetected language: task transcribe) num_languagesrdrh,rrr/u*"'“¿([{-"'.。,，!！?？:：”)]}、 translatez:Word-level timestamps on translations may not be reliable.segmentr%ct ttfr gn }d}|D]}i}|dkDr%|jdd|jddn|jddt di|d|i} j ||}d}|j kDrd} |j krd} |j kDrd}|r|S|S) Nr beam_sizepatiencebest_ofrIFTrB) isinstanceintfloatpoprr^compression_ratio avg_logprobno_speech_prob) rn temperatures decode_resulttkwargsoptionsneeds_fallbackrJdecode_optionsrKr8rLrIs r!decode_with_fallbackz(transcribe..decode_with_fallbacks' c5\B[M   A''F1u ;- :t, 9d+%>>A>G!LL':M"N+7!336QQ!%!-!--0AA!%#/!003FF!&!A@r# startr)tokensresultc |j}|Dcgc]}|jks|}}||j|||j|j|j |j d Scc}w)N) seekrr)textrrIrxrwry)tolisteotr^rIrxrwry)rr)rrtoken text_tokensr tokenizers r! new_segmentztranscribe..new_segments}*0J&EIMM4Iu& J$$[1!--!--!'!9!9$33  Ks A4A4framesF)totalunitdisablerCpromptwordc|jdd}|d|dz }d}|dkr|dz }|dkr |d|z dzz }|d kDr||d z z }|S) N probabilityrCr)rg333333?rEg/$?@)get)rrdurationscores r!word_anomaly_scorez&transcribe..word_anomaly_scoreAsp"&((=#">K#E{T']:HE"T) %'%("2b!88#~C/ Lr#c||dsy|dDcgc] }|dvs |}}|dd}tfd|D}|dk\xs|dzt|k\Scc}w)Nr(Frc3.K|] }|ywrcrB)r+r-rs r!r.z9transcribe..is_segment_anomaly..TsEu! 21 5usg{Gz?)sumlen)rnr-r(r punctuationrs r!is_segment_anomalyz&transcribe..is_segment_anomalyMs}gg.>$#*7#3#3aqy 7S#3""1IEEuEEE A:CU)CC s AAr$c(td|DdS)Nc3,K|] }|ds |yw)r(NrB)r+r,s r!r.z9transcribe..next_words_segment..Xs CHq' Hs )r0r1s r!next_words_segmentz&transcribe..next_words_segmentWs CH CTJJr#)rr)rr)r$r8rmel num_framesrPrQlast_speech_timestampr(rr[z --> z] r)ridg?)rr$rd)Drr@float16float32r4r:r dimsrWr shaper rur r sysgetdefaultencodingis_multilingualprintrastypedetect_languagemaxrtitlerrjrsr?splitrrrappendminlistzipwarningswarnarrayr n_audio_ctxr]stripextendtqdmnprryrxdictrboolrtimestamp_beginrwhere logical_anditemnonzerorr2ranger" enumeraterIupdater^)XrTrGrHrIrJrKrLrMrNrOrPrQrRrSrr6rcontent_framescontent_duration make_safe mel_segment_probsrdrhts seek_points seek_clipsrclip_idx input_stridetime_precision all_tokens all_segmentsprompt_reset_sinceinitial_prompt_tokensrpbarrseek_clip_start seek_clip_end time_offsetwindow_end_time segment_sizesegment_durationrr should_skip previous_seekcurrent_segmentsrrtimestamp_tokenssingle_timestamp_ending consecutiveslices last_slice current_slice sliced_tokensstart_timestamp_posend_timestamp_poslast_timestamp_posr timestamps last_word_end thresholdremaining_duration first_segmentgap hal_last_endsirn next_segmenthal_next_startsilence_before silence_afterrr)rlineirr8rrr`rrsX ```` ` @@@@@@r!riri>sU h),,VT:BJJ E  ! !/5 9E eEJJ,=,=y QCYYr]X-N^j8;FG002 g %I$I*d+3$$)-N: &Q&c8"=DDUKK,,[9HAu),U )BN: &"))N:4N*O*U*U*W)XY#:.H""6<8D ))  I/3'?N!6!6s!;TV!V !V"E"I!V  GVVoeB):$:;oKV ;11 ;!q >*nk"o> B(,SSqS1A;qtRStCT-U(VJ@K4;. RS&bhh&>&&PH h  "Duzz555Lz!K/JL! ) 0 0~7K7K7M1M N/0 "  # -/XX ?M $ 8WE5I  #.8 *O]&#D:$5 $CD "'J(F(T"U"nt3]T5I  "$ )<= #/*#<{#J )+xbIPPQVW +56H6I+Jx()=k)J&--0&2"("7"7:M"MK)5"..1BB', "(! $ #%  !T !e !DD4DKdKK$*Y-F-F#F *:23*?*F*F*HM+' !hhNN#3CR#8:J12:NO q {#a'(//1F. c&k2!"J)/ (.z-(H )!,113i6O6OO,*"-224y7P7PP*)//'&1"5"F'G$/2Cn2T$T'4'- &3 #*0&/ ,#:>2779IB&6r&:G#*7#3 (1':/A$4R!VX$>0" $0#;5A'5J15Mg5VN5@CS5SN$+G$4|$Ci$O%L'.w'7)'C%L'.w'7+'E'K!/ %3WU^$Ci$O%N'9,'G%N'6'G#'M!. $2m+0(+K!OWW=M(N*;)<,&D(8'%.'H9'T/=<>$4RS$9$)+25>L?#?B%--=$>M$00=-#3#G,#EN#FO%)s "##4U#;"##+4,C 4E++JAwq,G,+!!(8'7G%,X%6E%6'72V5G5G#5M),Z& C5 EFw&/9  D    j-B)C)EF G {  WH i  sJe=.f5fTf5fff,f ?Aff ff))rrtypingrrrrmlx.corecorer@numpyrrrTrr r r r r rdecodingrr load_modelsrtimingrrrrrur"rr2r4r?ndarrayrrrirBr#r!r s // 6#'/KuK"tDzhuo  8"3Q36)-+.'+$(! .A/27;a bjj"((* +aad^ a ueE3J//0 a "*% a a"%a!%aSMaaaa3U +,a&.e_ar#