Update app.py

app.py

@@ -27,7 +27,7 @@ def youtube_url_to_text(url, model_id, language_choice):
     return transcript, video_path
 
 
-def speaker_diarization(url, model_id, device, num_speakers, min_speaker, max_speaker):
+def speaker_diarization(url, model_id, num_speakers, min_speaker, max_speaker):
     """
     Main function that downloads and converts a video to MP3 format, performs speech-to-text conversion using
     a specified model, and returns the transcript along with the video path.
@@ -47,7 +47,7 @@ def speaker_diarization(url, model_id, device, num_speakers, min_speaker, max_sp
         diarizer_model="pyannote/speaker-diarization",
         use_auth_token="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",
         chunk_length_s=30,
-        device=device,
+        device="cuda",
     )
 
     audio_path = download_and_convert_to_mp3(url)
@@ -140,11 +140,6 @@ def speaker_diarization_app():
                     value="openai/whisper-large-v3",
                     label="Whisper Model",
                 )
-                device = gr.Dropdown(
-                    choices=["cpu", "cuda", "mps"],
-                    value="cuda",
-                    label="Device",
-                )
                 num_speakers = gr.Number(value=2, label="Number of Speakers")
                 min_speaker = gr.Number(value=1, label="Minimum Number of Speakers")
                 max_speaker = gr.Number(value=2, label="Maximum Number of Speakers")
@@ -159,7 +154,6 @@ def speaker_diarization_app():
             inputs=[
                 youtube_url_path,
                 whisper_model_id,
-                device,
                 num_speakers,
                 min_speaker,
                 max_speaker,
@@ -181,7 +175,6 @@ def speaker_diarization_app():
             inputs=[
                 youtube_url_path,
                 whisper_model_id,
-                device,
                 num_speakers,
                 min_speaker,
                 max_speaker,