Update app.py

app.py

@@ -21,33 +21,53 @@ languages = {
 }
 
 def generate_subtitles(video_path, language_name):
-    # Extract the target language code from the selected language name
-    target_language = languages[language_name]
-    
-    # Extract audio from video
-    video = mp.VideoFileClip(video_path)
-    audio = video.audio
-    audio.write_audiofile("temp_audio.wav", codec='pcm_s16le')
+    try:
+        # Extract the target language code from the selected language name
+        target_language = languages[language_name]
+        
+        # Log the file path for debugging
+        print(f"Received video file: {video_path}")
+        
+        # Extract audio from video
+        video = mp.VideoFileClip(video_path)
+        audio = video.audio
+        audio.write_audiofile("temp_audio.wav", codec='pcm_s16le')
+
+        # Log transcription start
+        print("Starting speech-to-text transcription")
+        
+        # Convert speech to text (ASR using Whisper)
+        with open("temp_audio.wav", "rb") as audio_file:
+            transcription = asr(audio_file)["text"]
 
-    # Convert speech to text (ASR using Whisper)
-    with open("temp_audio.wav", "rb") as audio_file:
-        transcription = asr(audio_file)["text"]
+        # Log translation start
+        print("Starting translation")
 
-    # Translate transcription to the target language using M2M100
-    translation_pipeline = pipeline('translation', model='facebook/m2m100_418M')
-    translated_subtitles = translation_pipeline(
-        transcription, 
-        forced_bos_token_id=translation_pipeline.tokenizer.get_lang_id(target_language)
-    )[0]["translation_text"]
+        # Translate transcription to the target language using M2M100
+        translation_pipeline = pipeline('translation', model='facebook/m2m100_418M')
+        translated_subtitles = translation_pipeline(
+            transcription, 
+            forced_bos_token_id=translation_pipeline.tokenizer.get_lang_id(target_language)
+        )[0]["translation_text"]
 
-    # Return subtitles (text for now)
-    subtitles = f"Original: {transcription}\nTranslated: {translated_subtitles}"
-    return subtitles
+        # Return subtitles
+        subtitles = f"Original: {transcription}\nTranslated: {translated_subtitles}"
+        return subtitles
+    
+    except Exception as e:
+        # Catch and log the error
+        print(f"Error occurred: {e}")
+        return f"Error occurred: {e}"
 
 # Define Gradio interface
 def subtitle_video(video_file, language_name):
-    video_path = video_file.name
-    return generate_subtitles(video_path, language_name)
+    try:
+        video_path = video_file.name
+        return generate_subtitles(video_path, language_name)
+    except Exception as e:
+        # Catch and display any high-level errors
+        print(f"Error in processing video: {e}")
+        return f"Error in processing video: {e}"
 
 # Gradio app layout
 interface = gr.Interface(