dv-pipecat-ai 0.0.82.dev857__py3-none-any.whl → 0.0.82.dev867__py3-none-any.whl

{dv_pipecat_ai-0.0.82.dev857.dist-info → dv_pipecat_ai-0.0.82.dev867.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dv-pipecat-ai
-Version: 0.0.82.dev857
+Version: 0.0.82.dev867
 Summary: An open source framework for voice (and multimodal) assistants
 License-Expression: BSD-2-Clause
 Project-URL: Source, https://github.com/pipecat-ai/pipecat

{dv_pipecat_ai-0.0.82.dev857.dist-info → dv_pipecat_ai-0.0.82.dev867.dist-info}/RECORD

@@ -1,4 +1,4 @@
-dv_pipecat_ai-0.0.82.dev857.dist-info/licenses/LICENSE,sha256=DWY2QGf2eMCFhuu2ChairtT6CB7BEFffNVhXWc4Od08,1301
+dv_pipecat_ai-0.0.82.dev867.dist-info/licenses/LICENSE,sha256=DWY2QGf2eMCFhuu2ChairtT6CB7BEFffNVhXWc4Od08,1301
 pipecat/__init__.py,sha256=j0Xm6adxHhd7D06dIyyPV_GlBYLlBnTAERVvD_jAARQ,861
 pipecat/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 pipecat/adapters/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -17,6 +17,18 @@ pipecat/adapters/services/open_ai_realtime_adapter.py,sha256=VTGlynwVwtEl4OxMMSF
 pipecat/audio/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 pipecat/audio/utils.py,sha256=RVm0NrjHYY5xhUIOEdKAqM2VsTDkMi1Sv0hIi49QYyI,10729
 pipecat/audio/dtmf/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+pipecat/audio/dtmf/dtmf-0.wav,sha256=NwdoIMEPk-w9MH807_zZc6WXhcnArSAX5uKIC0h0Po0,8078
+pipecat/audio/dtmf/dtmf-1.wav,sha256=TXajUIUpM0U0Z4WHbeLc7lYUzX901Q6E6ilkSM9mPuY,8078
+pipecat/audio/dtmf/dtmf-2.wav,sha256=S8QO__zcaSUnQMFctEXJQl0qM4q_IgwnlnK0spjMWXY,8078
+pipecat/audio/dtmf/dtmf-3.wav,sha256=aEKGgbWGKOOvai8JLdnT8mAOzga52z-izXCyhmZd9yA,8078
+pipecat/audio/dtmf/dtmf-4.wav,sha256=VssGOi6Dp4-N1Hw8HccDUZjenmVhpO9T3W3CxnCYa0w,8078
+pipecat/audio/dtmf/dtmf-5.wav,sha256=lBpuSylX23jzIwPhJ_BSiarbN7cpWNG0wY4jFp1CMeY,8078
+pipecat/audio/dtmf/dtmf-6.wav,sha256=9AiGiJIvaNJUjj6rqvzbjWuCLNj-jpkiyoHJZo6lbEM,8078
+pipecat/audio/dtmf/dtmf-7.wav,sha256=pnMsACUTPoRF192wKS6waylhsp-oyHbXUEy10nU6a-w,8078
+pipecat/audio/dtmf/dtmf-8.wav,sha256=_EbMLXS2oyjmps4sWCluruTA-APvo_LaiO4EgLoBZPo,8078
+pipecat/audio/dtmf/dtmf-9.wav,sha256=BUxr58k8vLK4fVlvl8M3DWdg3IClXjNChxRVm4k9Dp4,8078
+pipecat/audio/dtmf/dtmf-pound.wav,sha256=xkpEUGQFuAhUWTnRI718d5EINMtFjg3RKS6pUnjjFiw,8078
+pipecat/audio/dtmf/dtmf-star.wav,sha256=mmQlcuOb5JP5MFXupSp4mWnGhuC0AMPNPW11eyrvtXs,8078
 pipecat/audio/dtmf/types.py,sha256=5Djj7p89eEYBLp6wnVzAxHLWHYTphKcXlJgo5exqf0M,1071
 pipecat/audio/dtmf/utils.py,sha256=bZNHnjU2zdQBDm02xrUoZXaMtzyxJInqEhDEj0soPdE,2238
 pipecat/audio/filters/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -187,7 +199,8 @@ pipecat/services/deepgram/stt.py,sha256=IvdKvo23PxhKoWTJDxuK4Uoo0wCtkFGAE_QrMUoG
 pipecat/services/deepgram/tts.py,sha256=H_2WCJEx3_L4ytrHHRNkA-6GKTd1coou_vvTfiEodpQ,3745
 pipecat/services/deepseek/__init__.py,sha256=bU5z_oNGzgrF_YpsD9pYIMtEibeZFaUobbRjJ9WcYyE,259
 pipecat/services/deepseek/llm.py,sha256=5KjpU2blmhUTM3LcRE1ymdsk6OmoFkIzeQgyNOGwQh8,3112
-pipecat/services/elevenlabs/__init__.py,sha256=FgA--iiHyoart9xZZGWTTrBaEjmFxJuugESjPXihI7A,263
+pipecat/services/elevenlabs/__init__.py,sha256=N9TITy2chKr6zoAI3JWabkqNkiND9EbXHIMZbaH5InY,306
+pipecat/services/elevenlabs/stt.py,sha256=yekKTu-ymsvP79LXYaXIFBF2uCBNfqqYmFoimlcSUIg,11291
 pipecat/services/elevenlabs/tts.py,sha256=RZ9thg_kFWk1xhfmW047Pk0FjtWdkgnRMiGYFv3_cWk,42777
 pipecat/services/fal/__init__.py,sha256=z_kfZETvUcKy68Lyvni4B-RtdkOvz3J3eh6sFDVKq6M,278
 pipecat/services/fal/image.py,sha256=vArKLKrIGoZfw_xeZY_E7zbUzfzVsScj-R7mOmVqjRQ,4585
@@ -283,7 +296,7 @@ pipecat/services/simli/video.py,sha256=fVMYsCE5epH9rTdhN_tyPPJw7W6TCMHCOe2akKHWd
 pipecat/services/soniox/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 pipecat/services/soniox/stt.py,sha256=AhJF2YOzmqgB80x22jocgzr3neYCBMyxzP_WjkYR9Gc,15441
 pipecat/services/speechmatics/__init__.py,sha256=Jgq1fqrZVkpWC21D79L1cn5Ub8PnYgnnCaqC5pOlbIc,89
-pipecat/services/speechmatics/stt.py,sha256=tw2Fo-Kjcp45EplaTrHtKwjvLL35GKVt0K-pzU9MUtM,44251
+pipecat/services/speechmatics/stt.py,sha256=Ljc7cqDMLxdg99NTP4Fcs9rJ5Xfl_GzxH4gMiVhmR88,44261
 pipecat/services/tavus/__init__.py,sha256=SNyyi2Xq6tXIihDG2Bwvmg6Srbd-uWd1RwG-NKWcPuI,257
 pipecat/services/tavus/video.py,sha256=67WX49Th7v6xRIPoTpvbe15CB84OD9ax2ta9jL0fWiQ,10714
 pipecat/services/together/__init__.py,sha256=hNMycJDDf3CLiL9WA9fwvMdYphyDWLv0OabYIMgvC1c,259
@@ -364,7 +377,7 @@ pipecat/utils/tracing/service_decorators.py,sha256=HwDCqLGijhYD3F8nxDuQmEw-YkRw0
 pipecat/utils/tracing/setup.py,sha256=7TEgPNpq6M8lww8OQvf0P9FzYc5A30xICGklVA-fua0,2892
 pipecat/utils/tracing/turn_context_provider.py,sha256=ikon3plFOx0XbMrH6DdeHttNpb-U0gzMZIm3bWLc9eI,2485
 pipecat/utils/tracing/turn_trace_observer.py,sha256=dma16SBJpYSOE58YDWy89QzHyQFc_9gQZszKeWixuwc,9725
-dv_pipecat_ai-0.0.82.dev857.dist-info/METADATA,sha256=jIoMduEyMAIzGMNkLs_duD4O14Kp9IRfktgOd2SqquI,32639
-dv_pipecat_ai-0.0.82.dev857.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-dv_pipecat_ai-0.0.82.dev857.dist-info/top_level.txt,sha256=kQzG20CxGf-nSsHmtXHx3hY2-8zHA3jYg8jk0TajqXc,8
-dv_pipecat_ai-0.0.82.dev857.dist-info/RECORD,,
+dv_pipecat_ai-0.0.82.dev867.dist-info/METADATA,sha256=LKUKnU9RwBl7LC6IC8nxe6KpKhf7GO9C5MAtdkA-esM,32639
+dv_pipecat_ai-0.0.82.dev867.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+dv_pipecat_ai-0.0.82.dev867.dist-info/top_level.txt,sha256=kQzG20CxGf-nSsHmtXHx3hY2-8zHA3jYg8jk0TajqXc,8
+dv_pipecat_ai-0.0.82.dev867.dist-info/RECORD,,

pipecat/services/elevenlabs/__init__.py

@@ -9,5 +9,8 @@ import sys
 from pipecat.services import DeprecatedModuleProxy
 
 from .tts import *
+from .stt import *
 
-sys.modules[__name__] = DeprecatedModuleProxy(globals(), "elevenlabs", "elevenlabs.tts")
+sys.modules[__name__] = DeprecatedModuleProxy(
+    globals(), "elevenlabs", "elevenlabs.tts", "elevenlabs.stt"
+)

pipecat/services/elevenlabs/stt.py

@@ -0,0 +1,351 @@
+#
+# Copyright (c) 2024–2025, Daily
+#
+# SPDX-License-Identifier: BSD 2-Clause License
+#
+
+"""ElevenLabs speech-to-text service implementation."""
+
+import asyncio
+from typing import AsyncGenerator, Optional
+
+from loguru import logger
+
+from pipecat.frames.frames import ErrorFrame, Frame, TranscriptionFrame
+from pipecat.services.stt_service import SegmentedSTTService
+from pipecat.transcriptions.language import Language
+from pipecat.utils.time import time_now_iso8601
+from pipecat.utils.tracing.service_decorators import traced_stt
+
+try:
+    from elevenlabs.client import ElevenLabs
+except ModuleNotFoundError as e:
+    logger.error(f"Exception: {e}")
+    logger.error("In order to use ElevenLabs, you need to `pip install pipecat-ai[elevenlabs]`.")
+    raise Exception(f"Missing module: {e}")
+
+
+def language_to_elevenlabs_language(language: Language) -> Optional[str]:
+    """Maps pipecat Language enum to ElevenLabs language codes.
+
+    Args:
+        language: A Language enum value representing the input language.
+
+    Returns:
+        str or None: The corresponding ElevenLabs language code, or None if not supported.
+    """
+    language_map = {
+        # English
+        Language.EN: "eng",
+        Language.EN_US: "eng",
+        Language.EN_GB: "eng",
+        Language.EN_AU: "eng",
+        Language.EN_CA: "eng",
+        Language.EN_IN: "eng",
+        Language.EN_IE: "eng",
+        Language.EN_NZ: "eng",
+        Language.EN_ZA: "eng",
+        Language.EN_SG: "eng",
+        Language.EN_HK: "eng",
+        Language.EN_PH: "eng",
+        Language.EN_KE: "eng",
+        Language.EN_NG: "eng",
+        Language.EN_TZ: "eng",
+        # Spanish
+        Language.ES: "spa",
+        Language.ES_ES: "spa",
+        Language.ES_MX: "spa",
+        Language.ES_AR: "spa",
+        Language.ES_CO: "spa",
+        Language.ES_CL: "spa",
+        Language.ES_VE: "spa",
+        Language.ES_PE: "spa",
+        Language.ES_EC: "spa",
+        Language.ES_GT: "spa",
+        Language.ES_CU: "spa",
+        Language.ES_BO: "spa",
+        Language.ES_DO: "spa",
+        Language.ES_HN: "spa",
+        Language.ES_PY: "spa",
+        Language.ES_SV: "spa",
+        Language.ES_NI: "spa",
+        Language.ES_CR: "spa",
+        Language.ES_PA: "spa",
+        Language.ES_UY: "spa",
+        Language.ES_PR: "spa",
+        Language.ES_US: "spa",
+        Language.ES_GQ: "spa",
+        # French
+        Language.FR: "fra",
+        Language.FR_FR: "fra",
+        Language.FR_CA: "fra",
+        Language.FR_BE: "fra",
+        Language.FR_CH: "fra",
+        # German
+        Language.DE: "deu",
+        Language.DE_DE: "deu",
+        Language.DE_AT: "deu",
+        Language.DE_CH: "deu",
+        # Italian
+        Language.IT: "ita",
+        Language.IT_IT: "ita",
+        # Portuguese
+        Language.PT: "por",
+        Language.PT_PT: "por",
+        Language.PT_BR: "por",
+        # Hindi
+        Language.HI: "hin",
+        Language.HI_IN: "hin",
+        # Arabic
+        Language.AR: "ara",
+        Language.AR_SA: "ara",
+        Language.AR_EG: "ara",
+        Language.AR_AE: "ara",
+        Language.AR_BH: "ara",
+        Language.AR_DZ: "ara",
+        Language.AR_IQ: "ara",
+        Language.AR_JO: "ara",
+        Language.AR_KW: "ara",
+        Language.AR_LB: "ara",
+        Language.AR_LY: "ara",
+        Language.AR_MA: "ara",
+        Language.AR_OM: "ara",
+        Language.AR_QA: "ara",
+        Language.AR_SY: "ara",
+        Language.AR_TN: "ara",
+        Language.AR_YE: "ara",
+        # Japanese
+        Language.JA: "jpn",
+        Language.JA_JP: "jpn",
+        # Korean
+        Language.KO: "kor",
+        Language.KO_KR: "kor",
+        # Chinese
+        Language.ZH: "cmn",
+        Language.ZH_CN: "cmn",
+        Language.ZH_TW: "cmn",
+        Language.ZH_HK: "cmn",
+        # Russian
+        Language.RU: "rus",
+        Language.RU_RU: "rus",
+        # Dutch
+        Language.NL: "nld",
+        Language.NL_NL: "nld",
+        Language.NL_BE: "nld",
+        # Polish
+        Language.PL: "pol",
+        Language.PL_PL: "pol",
+        # Turkish
+        Language.TR: "tur",
+        Language.TR_TR: "tur",
+        # Swedish
+        Language.SV: "swe",
+        Language.SV_SE: "swe",
+        # Norwegian
+        Language.NO: "nor",
+        Language.NB: "nor",
+        Language.NN: "nor",
+        # Danish
+        Language.DA: "dan",
+        Language.DA_DK: "dan",
+        # Finnish
+        Language.FI: "fin",
+        Language.FI_FI: "fin",
+        # Czech
+        Language.CS: "ces",
+        Language.CS_CZ: "ces",
+        # Hungarian
+        Language.HU: "hun",
+        Language.HU_HU: "hun",
+        # Greek
+        Language.EL: "ell",
+        Language.EL_GR: "ell",
+        # Hebrew
+        Language.HE: "heb",
+        Language.HE_IL: "heb",
+        # Thai
+        Language.TH: "tha",
+        Language.TH_TH: "tha",
+        # Vietnamese
+        Language.VI: "vie",
+        Language.VI_VN: "vie",
+        # Indonesian
+        Language.ID: "ind",
+        Language.ID_ID: "ind",
+        # Malay
+        Language.MS: "msa",
+        Language.MS_MY: "msa",
+        # Ukrainian
+        Language.UK: "ukr",
+        Language.UK_UA: "ukr",
+        # Bulgarian
+        Language.BG: "bul",
+        Language.BG_BG: "bul",
+        # Croatian
+        Language.HR: "hrv",
+        Language.HR_HR: "hrv",
+        # Slovak
+        Language.SK: "slk",
+        Language.SK_SK: "slk",
+        # Slovenian
+        Language.SL: "slv",
+        Language.SL_SI: "slv",
+        # Estonian
+        Language.ET: "est",
+        Language.ET_EE: "est",
+        # Latvian
+        Language.LV: "lav",
+        Language.LV_LV: "lav",
+        # Lithuanian
+        Language.LT: "lit",
+        Language.LT_LT: "lit",
+    }
+    return language_map.get(language)
+
+
+class ElevenlabsSTTService(SegmentedSTTService):
+    """ElevenLabs speech-to-text service using Scribe v1 model.
+
+    This service uses ElevenLabs' batch STT API to transcribe audio segments.
+    It extends SegmentedSTTService to handle VAD-based audio segmentation.
+
+    Args:
+        api_key: ElevenLabs API key for authentication.
+        model_id: Model to use for transcription (default: "scribe_v1").
+        language: Default language for transcription.
+        tag_audio_events: Whether to tag audio events like laughter (default: False).
+        diarize: Whether to enable speaker diarization (default: False).
+        **kwargs: Additional arguments passed to SegmentedSTTService.
+    """
+
+    def __init__(
+        self,
+        *,
+        api_key: str,
+        model_id: str = "scribe_v1",
+        language: Language = Language.EN,
+        tag_audio_events: bool = False,
+        sample_rate: Optional[int] = None,
+        diarize: bool = False,
+        **kwargs,
+    ):
+        super().__init__(**kwargs)
+
+        self._client = ElevenLabs(api_key=api_key)
+        self._model_id = model_id
+        self._tag_audio_events = tag_audio_events
+        self._diarize = diarize
+
+        self._settings = {
+            "language": language,
+            "model_id": self._model_id,
+            "tag_audio_events": self._tag_audio_events,
+            "diarize": self._diarize,
+        }
+        self.set_model_name(model_id)
+
+    def can_generate_metrics(self) -> bool:
+        """Check if this service can generate processing metrics.
+
+        Returns:
+            True, as ElevenLabs service supports metrics generation.
+        """
+        return True
+
+    def language_to_service_language(self, language: Language) -> Optional[str]:
+        """Convert from pipecat Language to ElevenLabs language code.
+
+        Args:
+            language: The Language enum value to convert.
+
+        Returns:
+            str or None: The corresponding ElevenLabs language code, or None if not supported.
+        """
+        return language_to_elevenlabs_language(language)
+
+    async def set_language(self, language: Language):
+        """Set the language for transcription.
+
+        Args:
+            language: The Language enum value to use for transcription.
+        """
+        self.logger.info(f"Switching STT language to: [{language}]")
+        self._settings["language"] = language
+
+    @traced_stt
+    async def _handle_transcription(
+        self, transcript: str, is_final: bool, language: Optional[Language] = None
+    ):
+        """Handle a transcription result with tracing."""
+        pass
+
+    async def run_stt(self, audio: bytes) -> AsyncGenerator[Frame, None]:
+        """Transcribe the provided audio using ElevenLabs STT.
+
+        Args:
+            audio: Audio data (WAV format) to transcribe.
+
+        Yields:
+            Frame: TranscriptionFrame containing the transcribed text or ErrorFrame on failure.
+        """
+        try:
+            await self.start_processing_metrics()
+            await self.start_ttfb_metrics()
+
+            # Get language code for ElevenLabs API
+            language = self._settings["language"]
+            elevenlabs_lang = self.language_to_service_language(language)
+
+            # Prepare API parameters
+            params = {
+                "file": audio,
+                "model_id": self._model_id,
+                "tag_audio_events": self._tag_audio_events,
+                "diarize": self._diarize,
+            }
+
+            # Add language if specified
+            if elevenlabs_lang:
+                params["language_code"] = elevenlabs_lang
+
+            # Call ElevenLabs STT API in thread pool to avoid blocking
+            transcription = await asyncio.to_thread(self._client.speech_to_text.convert, **params)
+
+            await self.stop_ttfb_metrics()
+
+            # Process transcription result
+            if transcription and hasattr(transcription, "text") and transcription.text:
+                transcript_text = transcription.text.strip()
+
+                if transcript_text:
+                    # Determine language if available from response
+                    response_language = language
+                    if hasattr(transcription, "language_code") and transcription.language_code:
+                        # Try to map back from ElevenLabs language code to pipecat Language
+                        try:
+                            # This is a simplified mapping - you might want to create a reverse map
+                            response_language = language  # For now, keep the original
+                        except ValueError:
+                            self.logger.warning(
+                                f"Unknown language detected: {transcription.language_code}"
+                            )
+
+                    # Handle transcription with tracing
+                    await self._handle_transcription(transcript_text, True, response_language)
+
+                    self.logger.debug(f"ElevenLabs transcription: [{transcript_text}]")
+
+                    yield TranscriptionFrame(
+                        text=transcript_text,
+                        user_id="",
+                        timestamp=time_now_iso8601(),
+                        language=response_language,
+                        result=transcription,
+                    )
+
+            await self.stop_processing_metrics()
+
+        except Exception as e:
+            self.logger.error(f"ElevenLabs STT error: {e}")
+            await self.stop_all_metrics()
+            yield ErrorFrame(f"ElevenLabs STT error: {str(e)}")

pipecat/services/speechmatics/stt.py

@@ -551,7 +551,7 @@ class SpeechmaticsSTTService(STTService):
 
             @self._client.on(ServerMessageType.END_OF_UTTERANCE)
             def _evt_on_end_of_utterance(message: dict[str, Any]):
-                logger.debug("End of utterance received from STT")
+                self.logger.debug("End of utterance received from STT")
                 asyncio.run_coroutine_threadsafe(
                     self._handle_end_of_utterance(), self.get_event_loop()
                 )
@@ -775,7 +775,7 @@ class SpeechmaticsSTTService(STTService):
             ]
 
             # Log transcript(s)
-            logger.debug(f"Finalized transcript: {[f.text for f in downstream_frames]}")
+            self.logger.debug(f"Finalized transcript: {[f.text for f in downstream_frames]}")
 
         # Return as interim results (unformatted)
         else: