qanastek commited on
Commit
6d23e28
·
verified ·
1 Parent(s): 5ebc683

Delete convert_to_parquet.py

Browse files
Files changed (1) hide show
  1. convert_to_parquet.py +0 -64
convert_to_parquet.py DELETED
@@ -1,64 +0,0 @@
1
- import os
2
- import csv
3
- from datasets import Dataset, Audio, Features, Value, Version
4
-
5
- # Point to the current directory
6
- dataset_path = os.path.dirname(os.path.abspath(__file__))
7
-
8
- # Dataset metadata
9
- VERSION = Version("1.0.0")
10
- DESCRIPTION = "Background noise dataset."
11
- HOMEPAGE = "https://huggingface.co/datasets/sdialog/background"
12
- LICENSE = "CC BY-NC-SA 4.0"
13
-
14
- # Define features
15
- features = Features({
16
- "audio": Audio(sampling_rate=16000),
17
- "file_path": Value("string"),
18
- "label": Value("string"),
19
- "description": Value("string"),
20
- })
21
-
22
-
23
- def dataset_generator():
24
- csv_path = os.path.join(dataset_path, "metadata.csv")
25
- if not os.path.exists(csv_path):
26
- raise FileNotFoundError(
27
- f"{csv_path} not found. Run create_metadata.py first."
28
- )
29
-
30
- with open(csv_path, "r", encoding="utf-8") as f:
31
- reader = csv.DictReader(f)
32
- for row in reader:
33
- audio_path = row["audio"]
34
- # Ensure the path is absolute for reading
35
- # audio_path is relative to the dataset root
36
- full_path = os.path.join(dataset_path, audio_path)
37
-
38
- # Read audio bytes to ensure embedding
39
- with open(full_path, "rb") as audio_file:
40
- audio_bytes = audio_file.read()
41
-
42
- yield {
43
- "audio": {"path": None, "bytes": audio_bytes},
44
- "file_path": audio_path,
45
- "label": row["label"],
46
- "description": row["description"]
47
- }
48
-
49
-
50
- print("Creating dataset from generator (embedding bytes)...")
51
- ds = Dataset.from_generator(dataset_generator, features=features)
52
-
53
- # Set dataset info
54
- ds.info.version = VERSION
55
- ds.info.description = DESCRIPTION
56
- ds.info.homepage = HOMEPAGE
57
- ds.info.license = LICENSE
58
-
59
- print("Saving to Parquet...")
60
- output_path = os.path.join(dataset_path, "data/train-00000-of-00001.parquet")
61
- os.makedirs(os.path.dirname(output_path), exist_ok=True)
62
- ds.to_parquet(output_path)
63
-
64
- print(f"Saved to {output_path}")