version 3

app.py

@@ -1,48 +1,59 @@
 import gradio as gr
 import librosa
-import os
-import tempfile
-import shutil
+import io
 from asr import transcribe, ASR_LANGUAGES, model
 from lid import identify, LID_EXAMPLES
+import logging
+
+# Set up logging
+logging.basicConfig(level=logging.DEBUG)
+logger = logging.getLogger(__name__)
 
 def safe_process_file(file_obj):
     try:
-        # Create a temporary directory
-        with tempfile.TemporaryDirectory() as temp_dir:
-            # Generate a safe file name
-            safe_name = f"audio_{hash(file_obj.name)}.wav"
-            temp_path = os.path.join(temp_dir, safe_name)
-            
-            # Copy the file to the temporary directory
-            shutil.copy(file_obj.name, temp_path)
-            
-            # Load the audio from the temporary file
-            audio, sr = librosa.load(temp_path)
-            
-            return audio, sr, safe_name
+        logger.debug(f"Processing file: {file_obj.name}")
+        file_content = file_obj.read()
+        
+        with io.BytesIO(file_content) as temp_file:
+            logger.debug("Loading audio with librosa")
+            audio, sr = librosa.load(temp_file)
+        
+        safe_name = f"audio_{hash(file_obj.name)}.wav"
+        logger.debug(f"File processed successfully: {safe_name}")
+        return audio, sr, safe_name
     except Exception as e:
-        raise Exception(f"Error processing file: {str(e)}")
+        logger.error(f"Error processing file {file_obj.name}: {str(e)}")
+        raise
 
 def transcribe_multiple_files(audio_files, lang, transcription):
     transcriptions = []
     for audio_file in audio_files:
         try:
             audio, sr, safe_name = safe_process_file(audio_file)
+            logger.debug(f"Transcribing file: {safe_name}")
+            logger.debug(f"Language selected: {lang}")
+            logger.debug(f"User-provided transcription: {transcription}")
+            
             result = transcribe(model, audio, lang, transcription)
+            logger.debug(f"Transcription result: {result}")
+            
             transcriptions.append(f"File: {safe_name}\nTranscription: {result}\n")
         except Exception as e:
+            logger.error(f"Error in transcription process: {str(e)}")
             transcriptions.append(f"Error processing file: {str(e)}\n")
     return "\n".join(transcriptions)
 
+# Prepare language options for Dropdown
+language_options = [f"{k} ({v})" for k, v in ASR_LANGUAGES.items()]
+
 mms_transcribe = gr.Interface(
     fn=transcribe_multiple_files,
     inputs=[
         gr.File(label="Audio Files", file_count="multiple"),
         gr.Dropdown(
-            [f"{k} ({v})" for k, v in ASR_LANGUAGES.items()],
+            choices=language_options,
             label="Language",
-            value="eng English",
+            value=language_options[0] if language_options else None,
         ),
         gr.Textbox(label="Optional: Provide your own transcription"),
     ],