script used to remaster the original dataset in the audiofolder format

darija10.py

@@ -16,7 +16,7 @@ for file in Path("dvoice-v1.0/transcriptions/csv").iterdir():
         columns = reader.fieldnames
         for row in reader:
             audio_path = Path(row["path"])
-            row["path"] = audio_path.name
+            row["path"] = Path("data") / split / audio_path.name
             shutil.copy(Path("dvoice-v1.0/audios_fixed") / audio_path, split_audios_path)
             all_rows.append(row)
 
@@ -28,5 +28,4 @@ with (Path("darija10") / "metadata.csv").open("w", newline='', encoding="utf-8")
         row["file_name"] = row["path"]
         del row["path"]
         writer.writerow(row)
-        
-
+