feat(parser, audio): fixed formatting for remote faster whisper server

langchain-ai · Nov 10, 2024 · d9ca838 · d9ca838
1 parent 4ae8542
commit d9ca838
Showing 1 changed file with 15 additions and 5 deletions.
diff --git a/libs/community/langchain_community/document_loaders/parsers/audio.py b/libs/community/langchain_community/document_loaders/parsers/audio.py
@@ -708,7 +708,9 @@ class RemoteFasterWhisperParser(BaseBlobParser):
     Example: Load a local audio file and remotely transcribe it into a document.
         .. code-block:: python
         from langchain.document_loaders.generic import GenericLoader
-        from langchain_community.document_loaders.parsers.audio import RemoteFasterWhisperParser
+        from langchain_community.document_loaders.parsers.audio import (
+            RemoteFasterWhisperParser
+        )
 
         server_url = 'http://localhost:8000'  # add you server url here
 
@@ -747,7 +749,8 @@ def lazy_parse(self, blob: Blob) -> Iterator[Document]:
             blob (Blob): The audio blob to be transcribed.
 
         Yields:
-            Document: A Document object containing the transcribed text and associated metadata.
+            Document: A Document object containing the transcribed text and
+            associated metadata.
         """
         audio_bytes = self._load_audio_from_blob(blob=blob)
         transcription = self._transcribe_audio(file_bytes=audio_bytes)
@@ -772,9 +775,15 @@ def _transcribe_audio(self, file_bytes: bytes) -> dict[str, str]:
         Raises:
             RuntimeError: If the transcription process fails.
         """
-        url = self._get_transcription_url()
         process = subprocess.Popen(
-            ['curl', url, '-F', 'file=@-;type=audio/mp3', '-F', f'model={self.model_size}'],
+            [
+                'curl',
+                self.transcription_url,
+                '-F',
+                'file=@-;type=audio/mp3',
+                '-F',
+                f'model={self.model_size}'
+            ],
             stdin=subprocess.PIPE,
             stdout=subprocess.PIPE,
             stderr=subprocess.PIPE
@@ -787,7 +796,8 @@ def _transcribe_audio(self, file_bytes: bytes) -> dict[str, str]:
 
         return json.loads(stdout.decode())
 
-    def _get_transcription_url(self) -> str:
+    @property
+    def transcription_url(self) -> str:
         return f"{self.base_url}/{self.TRANSCRIPTION_ENDPOINT}"
 
     def _load_audio_from_blob(self, blob: Blob) -> bytes: