update

main.py

@@ -122,14 +122,14 @@ def process(
     start = time.time()
 
     text = decode.decode_by_recognizer(recognizer=recognizer,
-                                       filename=out_filename,
+                                       filename=out_filename.as_posix(),
                                        )
 
     date_time = now.strftime("%Y-%m-%d %H:%M:%S.%f")
     end = time.time()
 
     # statistics
-    metadata = torchaudio.info(out_filename)
+    metadata = torchaudio.info(out_filename.as_posix())
     duration = metadata.num_frames / 16000
     rtf = (end - start) / duration