Convert dataset to Parquet

GTZAN_genre_classification.py

@@ -1,116 +0,0 @@
-import os
-import csv
-import datasets
-
-logger = datasets.logging.get_logger(__name__)
-
-_CITATION = """\
-@misc{gtzan2023,
-  title={GTZAN Music Genre Classification Dataset},
-  author={Your Name},
-  year={2023},
-  url={https://example.com},
-}
-"""
-
-_DESCRIPTION = """\
-The GTZAN dataset is a music genre classification dataset.
-It consists of 10 genres, each represented by 100 tracks.
-"""
-
-_HOMEPAGE_URL = "https://example.com"
-_DATA_URL = "data/gtzan.zip"
-
-class GTZANConfig(datasets.BuilderConfig):
-    """BuilderConfig for GTZAN"""
-
-    def __init__(self, **kwargs):
-        super(GTZANConfig, self).__init__(**kwargs)
-
-
-class GTZAN(datasets.GeneratorBasedBuilder):
-    BUILDER_CONFIGS = [
-        GTZANConfig(name="gtzan", version=datasets.Version("1.0.0"), description="GTZAN Music Genre Classification Dataset")
-    ]
-
-    def _info(self):
-        return datasets.DatasetInfo(
-            description=_DESCRIPTION,
-            features=datasets.Features(
-                {
-                    "id": datasets.Value("int32"),
-                    "audio": datasets.Audio(sampling_rate=44100),
-                    "genre": datasets.Value("string"),
-                    "title": datasets.Value("string"),
-                    "artist": datasets.Value("string"),
-                    "tempo": datasets.Value("float"),
-                    "keys": datasets.Value("string"),
-                    "loudness": datasets.Value("float"),
-                    "sorted_pred_genres": datasets.Value("string"),
-                    "x_tsne": datasets.Value("float"),
-                    "y_tsne": datasets.Value("float"),
-                    "z_tsne": datasets.Value("float"),
-                    "x_umap": datasets.Value("float"),
-                    "y_umap": datasets.Value("float"),
-                    "z_umap": datasets.Value("float"),
-                    "album_cover_path": datasets.Value("string"),
-                    "key": datasets.Value("string"),
-                    "filepath": datasets.Value("string"),
-                }
-            ),
-            supervised_keys=("genre", "title"),
-            homepage=_HOMEPAGE_URL,
-            citation=_CITATION,
-        )
-
-    def _split_generators(self, dl_manager):
-        archive_path = dl_manager.download_and_extract(_DATA_URL)
-        data_dir = os.path.join(archive_path, "gtzan")
-
-        # Debugging the paths
-        logger.info(f"Archive path: {archive_path}")
-        logger.info(f"Data directory: {data_dir}")
-
-        return [
-            datasets.SplitGenerator(
-                name=datasets.Split.TRAIN,
-                gen_kwargs={
-                    "metadata_file": os.path.join(data_dir, "metadata.csv"),
-                    "data_dir": data_dir,
-                },
-            ),
-        ]
-
-    def _generate_examples(self, metadata_file, data_dir):
-        with open(metadata_file, "r", encoding="utf-8") as f:
-            reader = csv.DictReader(f)
-            for id_, row in enumerate(reader):
-                # Correct the file paths by removing the 'data/gtzan/' prefix
-                file_path = os.path.join(data_dir, row["filepath"].replace("data/gtzan/", ""))
-                album_cover_path = os.path.join(data_dir, row["album_cover_path"].replace("data/gtzan/", ""))
-
-                # Debugging the file paths
-                logger.info(f"Processing file: {file_path}")
-                if not os.path.exists(file_path):
-                    logger.error(f"File not found: {file_path}")
-
-                yield id_, {
-                    "id": int(row["id"]),
-                    "audio": file_path,  # Include the audio path next to the ID
-                    "genre": row["genre"],
-                    "title": row["title"],
-                    "artist": row["artist"],
-                    "tempo": float(row["tempo"]),
-                    "keys": row["keys"],
-                    "loudness": float(row["loudness"]),
-                    "sorted_pred_genres": row["sorted_pred_genres"],
-                    "x_tsne": float(row["x_tsne"]),
-                    "y_tsne": float(row["y_tsne"]),
-                    "z_tsne": float(row["z_tsne"]),
-                    "x_umap": float(row["x_umap"]),
-                    "y_umap": float(row["y_umap"]),
-                    "z_umap": float(row["z_umap"]),
-                    "album_cover_path": album_cover_path,
-                    "key": row["key"],
-                    "filepath": file_path,
-                }

README.md

@@ -5,6 +5,10 @@ dataset_info:
   features:
   - name: id
     dtype: int32
+  - name: audio
+    dtype:
+      audio:
+        sampling_rate: 44100
   - name: genre
     dtype: string
   - name: title
@@ -39,10 +43,10 @@ dataset_info:
     dtype: string
   splits:
   - name: train
-    num_bytes: 680045
+    num_bytes: 1323465590.0
     num_examples: 999
-  download_size: 209606
-  dataset_size: 680045
+  download_size: 1305799028
+  dataset_size: 1323465590.0
 configs:
 - config_name: gtzan
   data_files:

gtzan/{train-00000-of-00001.parquet → train-00000-of-00003.parquet}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aeaec9a21c64eec0fffc3823641178a13f2c71001010e36423c2fc22ce8b12c9
-size 209606
+oid sha256:3ceaf06c3f89cb21e68a6bf1d9a2175460cb4d8cd8634bb460fa54becf66d364
+size 440637794

data/gtzan.zip → gtzan/train-00001-of-00003.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0e62a3b63c083298de9d7238edd74c4f2c4908d81c6514ef9d1f421f1c87f9ba
-size 1244166188
+oid sha256:4ddd536bf8d06e082aaa4c5fc76e42e71b46ac28d1f55806732a04f56fd0cae3
+size 429099034

gtzan/train-00002-of-00003.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd2bd2736e359f24b1ff6d80232b94f09b59acd3f0b3de9ad6c6c40f3e049dd5
+size 436062200