update

GTZAN_genre_classification.py

@@ -1,6 +1,7 @@
 import os
 import csv
 import datasets
+import librosa
 
 logger = datasets.logging.get_logger(__name__)
 
@@ -95,6 +96,13 @@ class GTZAN(datasets.GeneratorBasedBuilder):
                 if not os.path.exists(file_path):
                     logger.error(f"File not found: {file_path}")
 
+                # Use librosa to load the audio file
+                try:
+                    audio, sr = librosa.load(file_path, sr=44100)
+                except Exception as e:
+                    logger.error(f"Error loading audio file {file_path}: {e}")
+                    continue
+
                 yield id_, {
                     "id": int(row["id"]),
                     "genre": row["genre"],
@@ -114,5 +122,8 @@ class GTZAN(datasets.GeneratorBasedBuilder):
                     "album_cover_path": album_cover_path,
                     "key": row["key"],
                     "filepath": file_path,
-                    "audio": file_path,
+                    "audio": {
+                        "array": audio,
+                        "sampling_rate": sr,
+                    },
                 }