yangwang825 commited on
Commit
c7656a3
·
verified ·
1 Parent(s): 6921f7f

Update wmms.py

Browse files
Files changed (1) hide show
  1. wmms.py +30 -6
wmms.py CHANGED
@@ -4,11 +4,13 @@
4
 
5
 
6
  import os
 
7
  import textwrap
8
  import datasets
9
  import itertools
10
  import typing as tp
11
  from pathlib import Path
 
12
  from sklearn.model_selection import train_test_split
13
 
14
  SAMPLE_RATE = 16_000
@@ -57,18 +59,40 @@ class WMMS(datasets.GeneratorBasedBuilder):
57
  """Returns SplitGenerators."""
58
  archive_path = dl_manager.extract(_COMPRESSED_FILENAME)
59
  extensions = ['.wav']
60
- _, _walker = fast_scandir(archive_path, extensions, recursive=True)
 
61
 
62
- train_walker, test_walker = train_test_split(
63
- _walker, test_size=0.2, random_state=914, stratify=[default_find_classes(f) for f in _walker]
64
- )
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
65
 
66
  return [
67
  datasets.SplitGenerator(
68
- name=datasets.Split.TRAIN, gen_kwargs={"audio_paths": train_walker, "split": "train"}
69
  ),
70
  datasets.SplitGenerator(
71
- name=datasets.Split.TEST, gen_kwargs={"audio_paths": test_walker, "split": "test"}
72
  ),
73
  ]
74
 
 
4
 
5
 
6
  import os
7
+ import random
8
  import textwrap
9
  import datasets
10
  import itertools
11
  import typing as tp
12
  from pathlib import Path
13
+ from collections import defaultdict
14
  from sklearn.model_selection import train_test_split
15
 
16
  SAMPLE_RATE = 16_000
 
59
  """Returns SplitGenerators."""
60
  archive_path = dl_manager.extract(_COMPRESSED_FILENAME)
61
  extensions = ['.wav']
62
+ _, filepaths = fast_scandir(archive_path, extensions, recursive=True)
63
+ labels = [default_find_classes(f) for f in filepaths]
64
 
65
+ # train_walker, test_walker = train_test_split(
66
+ # _walker, test_size=0.2, random_state=914, stratify=[default_find_classes(f) for f in _walker]
67
+ # )
68
+
69
+ # Step 1: Organize samples by class
70
+ class_to_files = defaultdict(list)
71
+ for filepath, label in zip(filepaths, labels):
72
+ class_to_files[label].append(filepath)
73
+
74
+ # Step 2: Select exactly 2 samples per class for the test set
75
+ test_files, test_labels = [], []
76
+ train_files, train_labels = [], []
77
+
78
+ for label, files in class_to_files.items():
79
+ if len(files) < 2:
80
+ raise ValueError(f"Not enough samples for class {label}") # Ensure each class has at least 2 samples
81
+
82
+ random.Random(914).shuffle(files) # Shuffle to ensure randomness
83
+
84
+ test_files.extend(files[:2]) # Pick first 2 for test
85
+ test_labels.extend([label] * 2)
86
+
87
+ train_files.extend(files[2:]) # Remaining go to train
88
+ train_labels.extend([label] * (len(files) - 2))
89
 
90
  return [
91
  datasets.SplitGenerator(
92
+ name=datasets.Split.TRAIN, gen_kwargs={"audio_paths": train_files, "split": "train"}
93
  ),
94
  datasets.SplitGenerator(
95
+ name=datasets.Split.TEST, gen_kwargs={"audio_paths": test_files, "split": "test"}
96
  ),
97
  ]
98