Update main.py

main.py

@@ -14,7 +14,7 @@ def hello():
 # Load the Whisper model
 print("Loading Whisper model...\n", flush=True)
 model = whisper.load_model("tiny")
-print("Whisper model loaded.\n", flush=True)
+print("\nWhisper model loaded.\n", flush=True)
 
 # Get time of request
 def get_time():
@@ -38,36 +38,36 @@ def convert_size(bytes):
         return f"{bytes / 1024**2:.2f} MB"
 
 def transcribe(audio_path):
-    print(f"Transcribing audio from: {audio_path}\n", flush=True)
+    print(f"Transcribing audio from: {audio_path}", flush=True)
 
     # Load audio and pad/trim it to fit 30 seconds
-    print("Loading and processing audio...\n", flush=True)
+    print("        Loading and processing audio...", flush=True)
     audio = whisper.load_audio(audio_path)
     audio = whisper.pad_or_trim(audio)
 
     # Make log-Mel spectrogram and move to the same device as the model
-    print("Creating log-Mel spectrogram...\n", flush=True)
+    print("        Creating log-Mel spectrogram...", flush=True)
     mel = whisper.log_mel_spectrogram(audio).to(model.device)
 
     # Detect the spoken language
-    print("Detecting language...", flush=True)
+    print("        Detecting language...", flush=True)
     _, probs = model.detect_language(mel)
     language = max(probs, key=probs.get)
-    print(f"Detected language: {language}\n", flush=True)
+    print(f"        Detected language: {language}", flush=True)
 
     # Decode the audio
-    print("Decoding audio...", flush=True)
+    print("        Decoding audio...", flush=True)
     options = whisper.DecodingOptions(fp16=False)
     result = whisper.decode(model, mel, options)
     
-    print("Transcription complete.\n", flush=True)
+    print("        Transcription complete.", flush=True)
     return result.text, language
 
 @app.route('/transcribe', methods=['POST'])
 def transcribe_audio():
     # Record the time when the request was received
     request_received_time, _ = get_time()
-    print(f"Received request at /transcribe at {request_received_time}\n", flush=True)
+    print(f"Received request at /transcribe at {request_received_time}", flush=True)
 
     if 'audio' not in request.files:
         print("Error: No audio file provided", flush=True)
@@ -82,7 +82,7 @@ def transcribe_audio():
     audio_path = os.path.join("temp_audio", audio_file.filename)
     os.makedirs("temp_audio", exist_ok=True)
     audio_file.save(audio_path)
-    print(f"Audio file saved to: {audio_path} (Size: {audio_file_size})\n", flush=True)
+    print(f"        Audio file saved to: {audio_path} (Size: {audio_file_size})", flush=True)
     
     # Record the time before starting transcription
     transcription_start_time = time.time()
@@ -91,7 +91,7 @@ def transcribe_audio():
     try:
         transcription, language = transcribe(audio_path)
     except Exception as e:
-        print(f"Error during transcription: {str(e)}", flush=True)
+        print(f"        Error during transcription: {str(e)}", flush=True)
         return jsonify({"error": f"An error occurred: {str(e)}"}), 500
     
     # Calculate the time taken for transcription
@@ -100,13 +100,13 @@ def transcribe_audio():
     
     # Clean up the saved file
     os.remove(audio_path)
-    print(f"Audio file removed from: {audio_path}\n", flush=True)
+    print(f"        Audio file removed from: {audio_path}\n", flush=True)
     
     # Record the time when the response is being sent
     response_sent_time, _ = get_time()
 
     # Return the transcription, detected language, and timing information
-    print(f"Transcription: {transcription}, Language: {language}\n", flush=True)
+    print(f"        Transcription: {transcription}, Language: {language}\n", flush=True)
     return jsonify({
         "transcription": transcription,
         "language": language,