livekit-plugins-elevenlabs 0.7.1__py3-none-any.whl → 0.7.3__py3-none-any.whl

livekit/plugins/elevenlabs/__init__.py

@@ -28,13 +28,12 @@ __all__ = [
 
 from livekit.agents import Plugin
 
+from .log import logger
+
 
 class ElevenLabsPlugin(Plugin):
     def __init__(self):
-        super().__init__(__name__, __version__, __package__)
-
-    def download_files(self):
-        pass
+        super().__init__(__name__, __version__, __package__, logger)
 
 
 Plugin.register_plugin(ElevenLabsPlugin())

livekit/plugins/elevenlabs/models.py

@@ -5,6 +5,7 @@ TTSModels = Literal[
     "eleven_multilingual_v1",
     "eleven_multilingual_v2",
     "eleven_turbo_v2",
+    "eleven_turbo_v2_5",
 ]
 
 TTSEncoding = Literal[

livekit/plugins/elevenlabs/tts.py

@@ -93,7 +93,7 @@ class TTS(tts.TTS):
         self,
         *,
         voice: Voice = DEFAULT_VOICE,
-        model_id: TTSModels = "eleven_turbo_v2",
+        model_id: TTSModels = "eleven_turbo_v2_5",
         api_key: str | None = None,
         base_url: str | None = None,
         encoding: TTSEncoding = "mp3_22050_32",
@@ -156,6 +156,8 @@ class ChunkedStream(tts.ChunkedStream):
     ) -> None:
         super().__init__()
         self._text, self._opts, self._session = text, opts, session
+        if _encoding_from_format(self._opts.encoding) == "mp3":
+            self._mp3_decoder = utils.codecs.Mp3StreamDecoder()
 
     @utils.log_exceptions(logger=logger)
     async def _main_task(self) -> None:
@@ -166,7 +168,7 @@ class ChunkedStream(tts.ChunkedStream):
         segment_id = utils.shortuuid()
 
         voice_settings = (
-            dataclasses.asdict(self._opts.voice.settings)
+            _strip_nones(dataclasses.asdict(self._opts.voice.settings))
             if self._opts.voice.settings
             else None
         )
@@ -181,21 +183,39 @@ class ChunkedStream(tts.ChunkedStream):
             headers={AUTHORIZATION_HEADER: self._opts.api_key},
             json=data,
         ) as resp:
-            async for bytes_data, _ in resp.content.iter_chunks():
-                for frame in bstream.write(bytes_data):
+            if not resp.content_type.startswith("audio/"):
+                content = await resp.text()
+                logger.error("11labs returned non-audio data: %s", content)
+                return
+            encoding = _encoding_from_format(self._opts.encoding)
+            if encoding == "mp3":
+                async for bytes_data, _ in resp.content.iter_chunks():
+                    for frame in self._mp3_decoder.decode_chunk(bytes_data):
+                        self._event_ch.send_nowait(
+                            tts.SynthesizedAudio(
+                                request_id=request_id,
+                                segment_id=segment_id,
+                                frame=frame,
+                            )
+                        )
+            else:
+                async for bytes_data, _ in resp.content.iter_chunks():
+                    for frame in bstream.write(bytes_data):
+                        self._event_ch.send_nowait(
+                            tts.SynthesizedAudio(
+                                request_id=request_id,
+                                segment_id=segment_id,
+                                frame=frame,
+                            )
+                        )
+
+                for frame in bstream.flush():
                     self._event_ch.send_nowait(
                         tts.SynthesizedAudio(
                             request_id=request_id, segment_id=segment_id, frame=frame
                         )
                     )
 
-            for frame in bstream.flush():
-                self._event_ch.send_nowait(
-                    tts.SynthesizedAudio(
-                        request_id=request_id, segment_id=segment_id, frame=frame
-                    )
-                )
-
 
 class SynthesizeStream(tts.SynthesizeStream):
     """Streamed API using websockets"""
@@ -280,7 +300,7 @@ class SynthesizeStream(tts.SynthesizeStream):
         init_pkt = dict(
             text=" ",
             try_trigger_generation=True,
-            voice_settings=dataclasses.asdict(self._opts.voice.settings)
+            voice_settings=_strip_nones(dataclasses.asdict(self._opts.voice.settings))
             if self._opts.voice.settings
             else None,
             generation_config=dict(
@@ -333,7 +353,15 @@ class SynthesizeStream(tts.SynthesizeStream):
                     segment_id=segment_id,
                 )
 
-        await asyncio.gather(send_task(), recv_task())
+        tasks = [
+            asyncio.create_task(send_task()),
+            asyncio.create_task(recv_task()),
+        ]
+
+        try:
+            await asyncio.gather(*tasks)
+        finally:
+            await utils.aio.gracefully_cancel(*tasks)
 
     def _process_stream_event(
         self, *, data: dict, request_id: str, segment_id: str
@@ -384,15 +412,19 @@ def _dict_to_voices_list(data: dict[str, Any]):
     return voices
 
 
+def _strip_nones(data: dict[str, Any]):
+    return {k: v for k, v in data.items() if v is not None}
+
+
 def _synthesize_url(opts: _TTSOptions) -> str:
     base_url = opts.base_url
     voice_id = opts.voice.id
     model_id = opts.model_id
-    sample_rate = _sample_rate_from_format(opts.encoding)
+    output_format = opts.encoding
     latency = opts.streaming_latency
     return (
         f"{base_url}/text-to-speech/{voice_id}/stream?"
-        f"model_id={model_id}&output_format=pcm_{sample_rate}&optimize_streaming_latency={latency}"
+        f"model_id={model_id}&output_format={output_format}&optimize_streaming_latency={latency}"
     )
 
 

livekit/plugins/elevenlabs/version.py

@@ -12,4 +12,4 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-__version__ = "0.7.1"
+__version__ = "0.7.3"

{livekit_plugins_elevenlabs-0.7.1.dist-info → livekit_plugins_elevenlabs-0.7.3.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: livekit-plugins-elevenlabs
-Version: 0.7.1
+Version: 0.7.3
 Summary: Agent Framework plugin for voice synthesis with ElevenLabs' API.
 Home-page: https://github.com/livekit/agents
 License: Apache-2.0

livekit_plugins_elevenlabs-0.7.3.dist-info/RECORD

@@ -0,0 +1,10 @@
+livekit/plugins/elevenlabs/__init__.py,sha256=cYRVIPXkRvB3-jK9bKZ9rYiMBACytWlCSq6yoZXaSgA,1080
+livekit/plugins/elevenlabs/log.py,sha256=hIuXqDsEB5GBa7rQY3z4Uqi1oCqc_lRmCHZEmXz0LHw,73
+livekit/plugins/elevenlabs/models.py,sha256=ddBUlDT4707f64WDJASR0B60X0yQ-LRHK1ZpTuBJXK8,387
+livekit/plugins/elevenlabs/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+livekit/plugins/elevenlabs/tts.py,sha256=ZSR6WxSBhntZmdK4i9U8SKcxHwNk3_4qiZNRZc5jP28,14641
+livekit/plugins/elevenlabs/version.py,sha256=yJeG0VwiekDJAk7GHcIAe43ebagJgloe-ZsqEGZnqzE,600
+livekit_plugins_elevenlabs-0.7.3.dist-info/METADATA,sha256=hdSuPch445_jz_Z-Uzt6CgU0Eb1H0ZVZ9ZA50hHYsBM,1311
+livekit_plugins_elevenlabs-0.7.3.dist-info/WHEEL,sha256=R0nc6qTxuoLk7ShA2_Y-UWkN8ZdfDBG2B6Eqpz2WXbs,91
+livekit_plugins_elevenlabs-0.7.3.dist-info/top_level.txt,sha256=OoDok3xUmXbZRvOrfvvXB-Juu4DX79dlq188E19YHoo,8
+livekit_plugins_elevenlabs-0.7.3.dist-info/RECORD,,

{livekit_plugins_elevenlabs-0.7.1.dist-info → livekit_plugins_elevenlabs-0.7.3.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (71.1.0)
+Generator: setuptools (72.1.0)
 Root-Is-Purelib: true
 Tag: py3-none-any
 

livekit_plugins_elevenlabs-0.7.1.dist-info/RECORD

@@ -1,10 +0,0 @@
-livekit/plugins/elevenlabs/__init__.py,sha256=ez1ybDPt7GfKAKgPkxZFRB7Vyd-_i-0hfUMI79GQ5w4,1091
-livekit/plugins/elevenlabs/log.py,sha256=hIuXqDsEB5GBa7rQY3z4Uqi1oCqc_lRmCHZEmXz0LHw,73
-livekit/plugins/elevenlabs/models.py,sha256=8jTchztgpiTokHEaWUK8PPxWWfvm5SMrOGsJpzxbYAw,362
-livekit/plugins/elevenlabs/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-livekit/plugins/elevenlabs/tts.py,sha256=fEqtmbzvuJ0Pso0kzJ_37_2aCHES7W1kKUwTycLRGpM,13318
-livekit/plugins/elevenlabs/version.py,sha256=JOBYrlKcxbTTRXkUKH0921GsmV-i71_KHczg2cgQiLc,600
-livekit_plugins_elevenlabs-0.7.1.dist-info/METADATA,sha256=PuFr70N0Y4YzxtzkeMmxwnyLkrQbynCUN0YKFu6gQV0,1311
-livekit_plugins_elevenlabs-0.7.1.dist-info/WHEEL,sha256=Wyh-_nZ0DJYolHNn1_hMa4lM7uDedD_RGVwbmTjyItk,91
-livekit_plugins_elevenlabs-0.7.1.dist-info/top_level.txt,sha256=OoDok3xUmXbZRvOrfvvXB-Juu4DX79dlq188E19YHoo,8
-livekit_plugins_elevenlabs-0.7.1.dist-info/RECORD,,