livekit-plugins-fal 1.0.0.dev4__tar.gz → 1.0.0rc1__tar.gz

{livekit_plugins_fal-1.0.0.dev4 → livekit_plugins_fal-1.0.0rc1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: livekit-plugins-fal
-Version: 1.0.0.dev4
+Version: 1.0.0rc1
 Summary: fal plugin template for LiveKit Agents
 Project-URL: Documentation, https://docs.livekit.io
 Project-URL: Website, https://livekit.io/
@@ -19,7 +19,7 @@ Classifier: Topic :: Multimedia :: Video
 Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
 Requires-Python: >=3.9.0
 Requires-Dist: fal-client
-Requires-Dist: livekit-agents>=1.0.0.dev4
+Requires-Dist: livekit-agents>=1.0.0.rc1
 Description-Content-Type: text/markdown
 
 # LiveKit Plugins fal

livekit_plugins_fal-1.0.0rc1/livekit/plugins/fal/stt.py

@@ -0,0 +1,80 @@
+from __future__ import annotations
+
+import os
+from dataclasses import dataclass
+
+import fal_client
+
+from livekit import rtc
+from livekit.agents import APIConnectionError, APIConnectOptions, stt
+from livekit.agents.stt import SpeechEventType, STTCapabilities
+from livekit.agents.types import (
+    NOT_GIVEN,
+    NotGivenOr,
+)
+from livekit.agents.utils import AudioBuffer, is_given
+
+
+@dataclass
+class _STTOptions:
+    language: str = "en"
+    task: str = "transcribe"
+    chunk_level: str = "segment"
+    version: str = "3"
+
+
+class WizperSTT(stt.STT):
+    def __init__(
+        self,
+        *,
+        language: NotGivenOr[str] = NOT_GIVEN,
+        api_key: NotGivenOr[str] = NOT_GIVEN,
+    ):
+        super().__init__(capabilities=STTCapabilities(streaming=False, interim_results=True))
+        self._api_key = api_key if is_given(api_key) else os.getenv("FAL_KEY")
+        if not self._api_key:
+            raise ValueError("fal AI API key is required. It should be set with env FAL_KEY")
+        self._opts = _STTOptions(language=language)
+        self._fal_client = fal_client.AsyncClient(key=self._api_key)
+
+    def update_options(self, *, language: NotGivenOr[str] = NOT_GIVEN) -> None:
+        if is_given(language):
+            self._opts.language = language
+
+    async def _recognize_impl(
+        self,
+        buffer: AudioBuffer,
+        *,
+        language: NotGivenOr[str] = NOT_GIVEN,
+        conn_options: APIConnectOptions,
+    ) -> stt.SpeechEvent:
+        try:
+            if is_given(language):
+                self._opts.language = language
+            data_uri = fal_client.encode(
+                rtc.combine_audio_frames(buffer).to_wav_bytes(), "audio/x-wav"
+            )
+            response = await self._fal_client.run(
+                "fal-ai/wizper",
+                arguments={
+                    "audio_url": data_uri,
+                    "task": self._opts.task,
+                    "language": self._opts.language,
+                    "chunk_level": self._opts.chunk_level,
+                    "version": self._opts.version,
+                },
+                timeout=conn_options.timeout,
+            )
+            text = response.get("text", "")
+            return self._transcription_to_speech_event(text=text)
+        except fal_client.client.FalClientError as e:
+            raise APIConnectionError() from e
+
+    def _transcription_to_speech_event(self, text: str) -> stt.SpeechEvent:
+        return stt.SpeechEvent(
+            type=SpeechEventType.FINAL_TRANSCRIPT,
+            alternatives=[stt.SpeechData(text=text, language=self._opts.language)],
+        )
+
+    async def aclose(self) -> None:
+        await self._fal_client._client.aclose()

{livekit_plugins_fal-1.0.0.dev4 → livekit_plugins_fal-1.0.0rc1}/livekit/plugins/fal/version.py

@@ -12,4 +12,4 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-__version__ = "1.0.0.dev4"
+__version__ = '1.0.0.rc1'

{livekit_plugins_fal-1.0.0.dev4 → livekit_plugins_fal-1.0.0rc1}/pyproject.toml

@@ -22,7 +22,7 @@ classifiers = [
     "Programming Language :: Python :: 3.10",
     "Programming Language :: Python :: 3 :: Only",
 ]
-dependencies = ["livekit-agents>=1.0.0.dev4", "fal_client"]
+dependencies = ["livekit-agents>=1.0.0.rc1", "fal_client"]
 
 [project.urls]
 Documentation = "https://docs.livekit.io"

livekit_plugins_fal-1.0.0.dev4/livekit/plugins/fal/stt.py

@@ -1,100 +0,0 @@
-from __future__ import annotations
-
-import dataclasses
-import os
-from dataclasses import dataclass
-
-import fal_client
-
-from livekit import rtc
-from livekit.agents import APIConnectionError, APIConnectOptions, stt
-from livekit.agents.stt import SpeechEventType, STTCapabilities
-from livekit.agents.utils import AudioBuffer
-
-
-@dataclass
-class _STTOptions:
-    language: str
-    task: str
-    chunk_level: str
-    version: str
-
-
-class WizperSTT(stt.STT):
-    def __init__(
-        self,
-        *,
-        language: str | None = "en",
-        task: str | None = "transcribe",
-        chunk_level: str | None = "segment",
-        version: str | None = "3",
-    ):
-        super().__init__(capabilities=STTCapabilities(streaming=False, interim_results=True))
-        self._api_key = os.getenv("FAL_KEY")
-        self._opts = _STTOptions(
-            language=language or "en",
-            task=task or "transcribe",
-            chunk_level=chunk_level or "segment",
-            version=version or "3",
-        )
-        self._fal_client = fal_client.AsyncClient()
-
-        if not self._api_key:
-            raise ValueError("fal AI API key is required. It should be set with env FAL_KEY")
-
-    def update_options(self, *, language: str | None = None) -> None:
-        self._opts.language = language or self._opts.language
-
-    def _sanitize_options(
-        self,
-        *,
-        language: str | None = None,
-        task: str | None = None,
-        chunk_level: str | None = None,
-        version: str | None = None,
-    ) -> _STTOptions:
-        config = dataclasses.replace(self._opts)
-        config.language = language or config.language
-        config.task = task or config.task
-        config.chunk_level = chunk_level or config.chunk_level
-        config.version = version or config.version
-        return config
-
-    async def _recognize_impl(
-        self,
-        buffer: AudioBuffer,
-        *,
-        language: str | None,
-        conn_options: APIConnectOptions,
-    ) -> stt.SpeechEvent:
-        try:
-            config = self._sanitize_options(language=language)
-            data_uri = fal_client.encode(
-                rtc.combine_audio_frames(buffer).to_wav_bytes(), "audio/x-wav"
-            )
-            response = await self._fal_client.run(
-                "fal-ai/wizper",
-                arguments={
-                    "audio_url": data_uri,
-                    "task": config.task,
-                    "language": config.language,
-                    "chunk_level": config.chunk_level,
-                    "version": config.version,
-                },
-                timeout=conn_options.timeout,
-            )
-            text = response.get("text", "")
-            return self._transcription_to_speech_event(text=text)
-        except fal_client.client.FalClientError as e:
-            raise APIConnectionError() from e
-
-    def _transcription_to_speech_event(
-        self, event_type=SpeechEventType.FINAL_TRANSCRIPT, text=None
-    ) -> stt.SpeechEvent:
-        return stt.SpeechEvent(
-            type=event_type,
-            alternatives=[stt.SpeechData(text=text, language=self._opts.language)],
-        )
-
-    async def aclose(self) -> None:
-        await self._fal_client._client.aclose()