qi5dZddlZddlZddlmZddlmZmZmZddl m Z ddl m Z m Z mZmZmZmZmZmZmZmZddlmZddlmZdd lmZmZdd lmZdd lm Z eGd d eZ!GddeZ"y)a/Deepgram text-to-speech service for AWS SageMaker. This module provides a Pipecat TTS service that connects to Deepgram models deployed on AWS SageMaker endpoints. Uses HTTP/2 bidirectional streaming for low-latency real-time speech synthesis with support for interruptions and streaming audio output. N) dataclass)AnyAsyncGeneratorOptional)logger) BotStoppedSpeakingFrame CancelFrameEndFrame ErrorFrameFrameInterruptionFrameLLMFullResponseEndFrame StartFrameTTSAudioRawFrameTTSStartedFrame)FrameDirection)SageMakerBidiClient) TTSSettings_warn_deprecated_param) TTSService) traced_ttsceZdZdZy)DeepgramSageMakerTTSSettingsz)Settings for DeepgramSageMakerTTSService.N)__name__ __module__ __qualname____doc__Y/opt/pipecat/venv/lib/python3.12/site-packages/pipecat/services/deepgram/sagemaker/tts.pyrr)s3rrc:eZdZUdZeZeed<ddddddededeed ee d ed eef fd Z d e fdZ de ffd Zdeffd Zdeffd Zdedeffd ZdZdZded eeefffd ZdZdedeffd ZddeefdZe deded e!edffdZ"xZ#S) DeepgramSageMakerTTSServiceaDeepgram text-to-speech service for AWS SageMaker. Provides real-time speech synthesis using Deepgram models deployed on AWS SageMaker endpoints. Uses HTTP/2 bidirectional streaming for low-latency audio generation with support for interruptions via the Clear message. Requirements: - AWS credentials configured (via environment variables, AWS CLI, or instance metadata) - A deployed SageMaker endpoint with Deepgram TTS model: https://developers.deepgram.com/docs/deploy-amazon-sagemaker - ``pipecat-ai[sagemaker]`` installed Example:: tts = DeepgramSageMakerTTSService( endpoint_name="my-deepgram-tts-endpoint", region="us-east-2", voice="aura-2-helena-en", ) _settingsNlinear16)voice sample_rateencodingsettings endpoint_nameregionr%r&r'r(c |tdtd|xsd}td|d}||j|t |d|ddd|d|||_||_||_d|_d|_ d|_ d|_ y) aInitialize the Deepgram SageMaker TTS service. Args: endpoint_name: Name of the SageMaker endpoint with Deepgram TTS model deployed (e.g., "my-deepgram-tts-endpoint"). region: AWS region where the endpoint is deployed (e.g., "us-east-2"). voice: Voice model to use for synthesis. Defaults to "aura-2-helena-en". .. deprecated:: 0.0.105 Use ``settings=DeepgramSageMakerTTSSettings(voice=...)`` instead. sample_rate: Audio sample rate in Hz. If None, uses the value from StartFrame. encoding: Audio encoding format. Defaults to "linear16". settings: Runtime-updatable settings. When provided alongside deprecated parameters, ``settings`` values take precedence. **kwargs: Additional arguments passed to the parent TTSService. Nr%zaura-2-helena-en)modelr%languageT)r&push_stop_framespause_frame_processingappend_trailing_spacer(Fr) rr apply_updatesuper__init___endpoint_name_region _encoding_client_response_task _context_id _ttfb_started) selfr)r*r%r&r'r(kwargsdefault_settings __class__s r r3z$DeepgramSageMakerTTSService.__init__Is8   "7,H' R++7     ) )( 3  #!#'"&%    , !6: 6:*.#(rreturncy)zCheck if this service can generate processing metrics. Returns: True, as Deepgram SageMaker TTS service supports metrics generation. Tr)r;s r can_generate_metricsz0DeepgramSageMakerTTSService.can_generate_metricss rframectKt||d{|jd{y77w)zStart the Deepgram SageMaker TTS service. Args: frame: The start frame containing initialization parameters. N)r2start_connectr;rBr>s r rDz!DeepgramSageMakerTTSService.starts3 gmE"""mmo #848688ctKt||d{|jd{y77w)zbStop the Deepgram SageMaker TTS service. Args: frame: The end frame. N)r2stop _disconnectrFs r rIz DeepgramSageMakerTTSService.stops6 gl5!!!    " rGctKt||d{|jd{y77w)zgCancel the Deepgram SageMaker TTS service. Args: frame: The cancel frame. N)r2cancelrJrFs r rLz"DeepgramSageMakerTTSService.cancels6 gnU###    $ rG directioncKt|||d{t|ttfr|j d{yt|t rd|_yy7L7 w)zProcess frames with special handling for LLM response end. Args: frame: The frame to process. direction: The direction of frame processing. NF)r2 process_frame isinstancerr flush_audiorr:)r;rBrMr>s r rOz)DeepgramSageMakerTTSService.process_framescg#E9555 e5x@ A""$ $ $ 6 7!&D 8 6 %s!A(A$-A(A&A(&A(cKtjdd|jjd|jd|j }t |j|jd||_ |jjd{|j|j|_ tjd|jd d{y7W7#t$rM}|j!d || d{7|jd t#|d{7Yd}~yd}~wwxYww) zConnect to the SageMaker endpoint and start the BiDi session. Builds the Deepgram TTS query string, creates the BiDi client, starts the streaming session, and launches a background task for processing responses. z*Connecting to Deepgram TTS on SageMaker...zmodel=z &encoding=z &sample_rate=zv1/speak)r)r*model_invocation_pathmodel_query_stringNz&Connected to Deepgram TTS on SageMaker on_connectedUnknown error occurred:  error_msg exceptionon_connection_error)rdebugr#r%r6r&rr4r5r7 start_session create_task_process_responsesr8_call_event_handler Exception push_errorstr)r; query_stringes r rEz$DeepgramSageMakerTTSService._connects  ABT^^))**T^^4DMRVRbRbQc d +--<<",+    J,,,,. . ."&"2"243J3J3L"MD  LLA B**>: : : / ; J//.Fqc,JVW/X X X**+@#a&I I I JsmA1E4C-C)AC-#C+$C-(E)C-+C-- E6D>D#D>3D64D>9E>EEchK|jr|jjrtjd |jj ddid{|jr=|jjs#|j|jd{|jjd{tjd|jdd{yyy7#t $r"}tj d|Yd}~d}~wwxYw77g7;w)zDisconnect from the SageMaker endpoint. Sends a Close message to Deepgram, cancels the response processing task, and closes the BiDi session. Safe to call multiple times. z/Disconnecting from Deepgram TTS on SageMaker...typeCloseNzFailed to send Close message: z+Disconnected from Deepgram TTS on SageMakeron_disconnected) r7 is_activerr[ send_jsonr`warningr8done cancel_task close_sessionr_)r;rds r rJz'DeepgramSageMakerTTSService._disconnects <???""4+>+>+C+C+E&&t':':;;;,,,,. . . LLF G**+<= = =3<@ E!?sCDD E< . >sq8D2 C>C<C> AD2$D,%!D2D.-D24D05D2<C>> D)D$D2$D))D2.D20D2deltacKt||d{}|s|Sd|vr5|jj|j_|j |j ||S7Tw)zApply a settings delta and reconnect if necessary. Since all settings are part of the SageMaker session query string, any setting change requires reconnecting to apply the new values. Nr%)r2_update_settingsr#r%r,_sync_model_name_to_metrics _warn_unhandled_updated_settings)r;rochangedr>s r rqz,DeepgramSageMakerTTSService._update_settingssi 077N g #'>>#7#7DNN  , , . --g6#8sA-A+AA-cHK |jr|jjrt|jjd{}|nNt|drt|jdr|jj r|jj } |j d}tj|}|jd}|dk(rtjd|n|dk(rtjd |nf|d k(rtjd |nH|d k(r+tj|d |jddntjd||jr|jjrttjdy7m#ttjf$rY|j!d{7t#||j$d|j&}|j)|d{7YwxYw#t*j,$rtjdYt.$r)}|j1d||d{7Yd}~d}~wwxYw#tjdwxYww)ahProcess streaming responses from Deepgram TTS on SageMaker. Continuously receives responses from the BiDi stream. Attempts to decode each payload as UTF-8 JSON for control messages (Flushed, Cleared, Metadata, Warning). If decoding fails, treats the payload as raw audio bytes and pushes a TTSAudioRawFrame downstream. Nvaluebytes_zutf-8rfMetadatazReceived metadata: FlushedzReceived Flushed: ClearedzReceived Cleared: Warningz warning: descriptionzUnknown warningzReceived unknown message type:  context_idz TTS response processor cancelledrVrWzTTS response processor stopped)r7rireceive_responsehasattrrvrwdecodejsonloadsgetrtracerkr[UnicodeDecodeErrorJSONDecodeErrorstop_ttfb_metricsrr&r9 push_frameasyncioCancelledErrorr`ra)r;resultpayload response_dataparsedmsg_typerBrds r r^z.DeepgramSageMakerTTSService._process_responsess%/ ;,,4<<#9#9#||<<>>>67+ h0O||**"(,,"5"59,3NN7,CM%)ZZ %>F'-zz&'9H':5 & /B6(-K L!)Y!6 & /A&-J K!)Y!6 & /A&-J K!)Y!6 &'+fJ'-zz-AR'S&T%V!" !' /Nvh-W X9,,4<<#9#9\ LL9 :[?:!3D4H4HI 9"&"8"8":::$4 ' $ 0 0 !+/+;+; %E #'//%"888 9%% = LL; < Y//.Fqc,JVW/X X X Y LL9 :sJ"AH*F2AH*CF58"H*J"2H*5,H'!G$"|jjr' |jj ddid{yyy7V7 #t $r$}tj|d|Yd}~yd}~wwxYww)zHandle interruption by sending Clear message to Deepgram. The Clear message will clear Deepgram's internal text buffer and stop sending audio, allowing for a new response to be generated. NFrfClearz error sending Clear message: ) r2_handle_interruptionr:r7rirjr`rerror)r;rBrMrdr>s r rz0DeepgramSageMakerTTSService._handle_interruption=s g*5)<<<" <???3< = @ I v%CA3GHH IsKB"A.-B" A2&A0'A2+B"0A22 B;BB"BB"rcK|jr>|jjr' |jjddid{yyy7#t$r$}t j |d|Yd}~yd}~wwxYww)zFlush any pending audio synthesis by sending Flush command. This should be called when the LLM finishes a complete response to force generation of audio from Deepgram's internal text buffer. rfFlushNz error sending Flush message: )r7rirjr`rr)r;rrds r rQz'DeepgramSageMakerTTSService.flush_audioLst <???3<? I v%CA3GHH Is@#B AAA BA A=A83B8A==BtextcKtj|d|d |j|sS|j|d{|js|j d{d|_t |||_|jjd|dd{dy7q7O7#t$r}td| Yd}~yd}~wwxYww) aNGenerate speech from text using Deepgram TTS on SageMaker. Args: text: The text to synthesize into speech. context_id: The context ID for tracking audio frames. Yields: Frame: TTSStartedFrame, then None (audio comes asynchronously via the response processor). z: Generating TTS []NTr~Speak)rfrrV)r) rr[audio_context_availablecreate_audio_contextr:start_ttfb_metricsrr9r7rjr`r )r;rrrds r run_ttsz#DeepgramSageMakerTTSService.run_ttsXs  v/vQ78 C// ;// ;;;))11333)-D&%<<)D ,,(('4)HI I IJ<3 J C%=aS#AB B B CseC$%B<B6#B<(B8)AB<+B:, B<5C$6B<8B<:B<< C!CC$C!!C$)N)$rrrrrSettings__annotations__rbrintr3boolrArrDr rIr rLr rrOrErJrdictrrqr^r rrQrrr __classcell__)r>s@r r"r"0sD*,H++ $%)";?9)9) 9) } 9) c] 9)9)789)vd!!!+! ' '> 'J@>,KDcN27;r I0A In I IHSM IC#C3C>%QU+;VCCrr")#rrr dataclassesrtypingrrrlogururpipecat.frames.framesrr r r r r rrrr"pipecat.processors.frame_processorr*pipecat.services.aws.sagemaker.bidi_clientrpipecat.services.settingsrrpipecat.services.tts_servicer(pipecat.utils.tracing.service_decoratorsrrr"rrr rsi !00   >JI3?  ;   DC*DCr