Update app.py

app.py

@@ -26,7 +26,6 @@ def transcribe(video_file, transcribe_to_text, transcribe_to_srt, language):
         return_timestamps=True,
         torch_dtype=torch_dtype,
         device=device,
-        generate_kwargs={"language": language}
     )
     
     # Handle the video file input
@@ -46,7 +45,7 @@ def transcribe(video_file, transcribe_to_text, transcribe_to_srt, language):
         temp_file_path = f"temp_audio_{i}.wav"
         audio_chunk.write_audiofile(temp_file_path, codec='pcm_s16le')
         with open(temp_file_path, "rb") as temp_file:
-            result = pipe(temp_file_path)
+            result = pipe(temp_file_path, generate_kwargs={"language": language})
             transcription_txt += result["text"]
             if transcribe_to_srt:
                 for chunk in result["chunks"]:
@@ -57,9 +56,10 @@ def transcribe(video_file, transcribe_to_text, transcribe_to_srt, language):
                         "text": chunk["text"]
                     })
         os.remove(temp_file_path)
-        yield f"Progress: {int(((i + 1) / n_chunks) * 100)}%"
-    
+        yield f"Progress: {int(((i + 1) / n_chunks) * 100)}%", None
+
     output = ""
+    srt_file_path = None
     if transcribe_to_text:
         output += "Text Transcription:\n" + transcription_txt + "\n\n"
     if transcribe_to_srt:
@@ -77,7 +77,7 @@ def transcribe(video_file, transcribe_to_text, transcribe_to_srt, language):
         
         output += f"\nSRT file saved as: {srt_file_path}"
     
-    yield output, srt_file_path if transcribe_to_srt else None
+    yield output, srt_file_path
 
 def format_time(seconds):
     m, s = divmod(seconds, 60)