yangwang825 commited on
Commit
042e24c
·
verified ·
1 Parent(s): dcd8bae

Update magnatagatune.py

Browse files
Files changed (1) hide show
  1. magnatagatune.py +57 -34
magnatagatune.py CHANGED
@@ -1,6 +1,6 @@
1
  # coding=utf-8
2
 
3
- """Medley-Solos-DB dataset."""
4
 
5
 
6
  import os
@@ -18,13 +18,14 @@ import urllib.request
18
  from pathlib import Path
19
  from copy import deepcopy
20
  from tqdm.auto import tqdm
 
21
  from rich.logging import RichHandler
22
 
23
  logger = logging.getLogger(__name__)
24
  logger.addHandler(RichHandler())
25
  logger.setLevel(logging.INFO)
26
 
27
- SAMPLE_RATE = 44_100
28
 
29
  # Cache location
30
  VERSION = "0.0.1"
@@ -36,38 +37,41 @@ DEFAULT_HF_DATASETS_CACHE = os.path.join(HF_CACHE_HOME, "datasets")
36
  HF_DATASETS_CACHE = Path(os.getenv("HF_DATASETS_CACHE", DEFAULT_HF_DATASETS_CACHE))
37
 
38
  CLASSES = [
39
- 'clarinet', 'distorted electric guitar', 'female singer', 'flute', 'piano', 'tenor saxophone', 'trumpet', 'violin'
 
 
 
 
40
  ]
41
- CLASS2INDEX = {cls:idx for idx, cls in enumerate(CLASSES)}
42
- INDEX2CLASS = {idx:cls for idx, cls in enumerate(CLASSES)}
43
 
44
 
45
- class MedleySolosDBConfig(datasets.BuilderConfig):
46
- """BuilderConfig for Medley-Solos-DB."""
47
 
48
  def __init__(self, features, **kwargs):
49
- super(MedleySolosDBConfig, self).__init__(version=datasets.Version(VERSION, ""), **kwargs)
50
  self.features = features
51
 
52
 
53
- class MedleySolosDB(datasets.GeneratorBasedBuilder):
54
 
55
  BUILDER_CONFIGS = [
56
- MedleySolosDBConfig(
57
  features=datasets.Features(
58
  {
59
  "file": datasets.Value("string"),
60
  "audio": datasets.Audio(sampling_rate=SAMPLE_RATE),
61
- "instrument": datasets.Value("string"),
62
- "label": datasets.features.ClassLabel(names=CLASSES),
63
  }
64
  ),
65
- name="v1.2",
66
  description="",
67
  ),
68
  ]
69
 
70
- DEFAULT_CONFIG_NAME = "v1.2"
71
 
72
  def _info(self):
73
  return datasets.DatasetInfo(
@@ -81,10 +85,10 @@ class MedleySolosDB(datasets.GeneratorBasedBuilder):
81
 
82
  def _split_generators(self, dl_manager):
83
  """Returns SplitGenerators."""
84
- zip_file_url = "https://zenodo.org/records/3464194/files/Medley-solos-DB.tar.gz"
85
  _filename = zip_file_url.split('/')[-1]
86
  _save_path = os.path.join(
87
- HF_DATASETS_CACHE, 'confit___medley-solos-db/v1.2', VERSION, _filename
88
  )
89
  download_file(zip_file_url, _save_path)
90
  logger.info(f"`{_filename}` is downloaded to {_save_path}")
@@ -104,46 +108,65 @@ class MedleySolosDB(datasets.GeneratorBasedBuilder):
104
  ]
105
 
106
  def _generate_examples(self, archive_path, split=None):
107
- metadata_df = pd.read_csv("https://zenodo.org/records/3464194/files/Medley-solos-DB_metadata.csv")
108
- train_df = metadata_df[metadata_df["subset"] == "training"].reset_index(drop=True)
109
- validation_df = metadata_df[metadata_df["subset"] == "validation"].reset_index(drop=True)
110
- test_df = metadata_df[metadata_df["subset"] == "test"].reset_index(drop=True)
 
 
 
 
 
 
 
 
 
 
 
 
 
111
 
112
- extensions = ['.wav']
 
 
 
 
 
 
113
  _, _walker = fast_scandir(archive_path, extensions, recursive=True)
114
 
 
 
 
115
  if split == 'train':
116
  fileid2class = {}
117
  for idx, row in train_df.iterrows():
118
- fileid = row['uuid4']
119
- class_ = row['instrument']
120
  fileid2class[fileid] = class_
121
  elif split == 'validation':
122
  fileid2class = {}
123
  for idx, row in validation_df.iterrows():
124
- fileid = row['uuid4']
125
- class_ = row['instrument']
126
  fileid2class[fileid] = class_
127
  elif split == 'test':
128
  fileid2class = {}
129
  for idx, row in test_df.iterrows():
130
- fileid = row['uuid4']
131
- class_ = row['instrument']
132
  fileid2class[fileid] = class_
133
 
134
- _walker = [fileid for fileid in _walker if not Path(fileid).name.startswith('._Medley')]
135
  for guid, audio_path in enumerate(_walker):
136
- fileid = Path(audio_path).stem
137
- fileid = fileid.split('_')[-1]
138
- if fileid not in fileid2class:
139
  continue
140
- instrument = fileid2class.get(fileid)
141
  yield guid, {
142
  "id": str(guid),
143
  "file": audio_path,
144
  "audio": audio_path,
145
- "instrument": instrument,
146
- "label": instrument,
147
  }
148
 
149
 
 
1
  # coding=utf-8
2
 
3
+ """MagnaTagATune dataset."""
4
 
5
 
6
  import os
 
18
  from pathlib import Path
19
  from copy import deepcopy
20
  from tqdm.auto import tqdm
21
+ from rich import print
22
  from rich.logging import RichHandler
23
 
24
  logger = logging.getLogger(__name__)
25
  logger.addHandler(RichHandler())
26
  logger.setLevel(logging.INFO)
27
 
28
+ SAMPLE_RATE = 16_000
29
 
30
  # Cache location
31
  VERSION = "0.0.1"
 
37
  HF_DATASETS_CACHE = Path(os.getenv("HF_DATASETS_CACHE", DEFAULT_HF_DATASETS_CACHE))
38
 
39
  CLASSES = [
40
+ "guitar", "classical", "slow", "techno", "strings", "drums", "electronic", "rock", "fast", "piano",
41
+ "ambient", "beat", "violin", "vocal", "synth", "female", "indian", "opera", "male", "singing", "vocals",
42
+ "no vocals", "harpsichord", "loud", "quiet", "flute", "woman", "male vocal", "no vocal", "pop", "soft",
43
+ "sitar", "solo", "man", "classic", "choir", "voice", "new age", "dance", "male voice", "female vocal",
44
+ "beats", "harp", "cello", "no voice", "weird", "country", "metal", "female voice", "choral"
45
  ]
46
+ CLASSES = sorted(CLASSES)
 
47
 
48
 
49
+ class MagnaTagATuneConfig(datasets.BuilderConfig):
50
+ """BuilderConfig for MagnaTagATune."""
51
 
52
  def __init__(self, features, **kwargs):
53
+ super(MagnaTagATuneConfig, self).__init__(version=datasets.Version(VERSION, ""), **kwargs)
54
  self.features = features
55
 
56
 
57
+ class MagnaTagATune(datasets.GeneratorBasedBuilder):
58
 
59
  BUILDER_CONFIGS = [
60
+ MagnaTagATuneConfig(
61
  features=datasets.Features(
62
  {
63
  "file": datasets.Value("string"),
64
  "audio": datasets.Audio(sampling_rate=SAMPLE_RATE),
65
+ "sound": datasets.Sequence(datasets.Value("string")),
66
+ "label": datasets.Sequence(datasets.features.ClassLabel(names=CLASSES)),
67
  }
68
  ),
69
+ name="top50",
70
  description="",
71
  ),
72
  ]
73
 
74
+ DEFAULT_CONFIG_NAME = "top50"
75
 
76
  def _info(self):
77
  return datasets.DatasetInfo(
 
85
 
86
  def _split_generators(self, dl_manager):
87
  """Returns SplitGenerators."""
88
+ zip_file_url = "https://huggingface.co/datasets/confit/magnatagatune/resolve/main/mp3.zip"
89
  _filename = zip_file_url.split('/')[-1]
90
  _save_path = os.path.join(
91
+ HF_DATASETS_CACHE, 'confit___magnatagatune/top50', VERSION, _filename
92
  )
93
  download_file(zip_file_url, _save_path)
94
  logger.info(f"`{_filename}` is downloaded to {_save_path}")
 
108
  ]
109
 
110
  def _generate_examples(self, archive_path, split=None):
111
+ df = pd.read_csv(
112
+ 'https://huggingface.co/datasets/confit/magnatagatune/resolve/main/annotations_final.csv', sep="\t"
113
+ )
114
+ # Filter only the songs that have at least one of the top 50 tags
115
+ df = df[df[CLASSES].sum(axis=1) > 0].reset_index(drop=True)
116
+ df = df[CLASSES + ["mp3_path", "clip_id"]]
117
+ train_ids_df = pd.read_csv(
118
+ "https://huggingface.co/datasets/confit/magnatagatune/resolve/main/train_gt_mtt.tsv", sep="\t", header=None
119
+ )
120
+ train_ids = train_ids_df[0].tolist()
121
+ train_df = df[df["clip_id"].isin(train_ids)].reset_index(drop=True)
122
+
123
+ validation_ids_df = pd.read_csv(
124
+ "https://huggingface.co/datasets/confit/magnatagatune/resolve/main/val_gt_mtt.tsv", sep="\t", header=None
125
+ )
126
+ validation_ids = validation_ids_df[0].tolist()
127
+ validation_df = df[df["clip_id"].isin(validation_ids)].reset_index(drop=True)
128
 
129
+ test_ids_df = pd.read_csv(
130
+ "https://huggingface.co/datasets/confit/magnatagatune/resolve/main/test_gt_mtt.tsv", sep="\t", header=None
131
+ )
132
+ test_ids = test_ids_df[0].tolist()
133
+ test_df = df[df["clip_id"].isin(test_ids)].reset_index(drop=True)
134
+
135
+ extensions = ['.mp3']
136
  _, _walker = fast_scandir(archive_path, extensions, recursive=True)
137
 
138
+ # Extract the list of column names where the value is 1 for each row
139
+ result = df.apply(lambda row: [col for col in df.columns if row[col] == 1], axis=1).tolist()
140
+
141
  if split == 'train':
142
  fileid2class = {}
143
  for idx, row in train_df.iterrows():
144
+ fileid = os.path.join(archive_path, str(row['mp3_path']))
145
+ class_ = result[idx]
146
  fileid2class[fileid] = class_
147
  elif split == 'validation':
148
  fileid2class = {}
149
  for idx, row in validation_df.iterrows():
150
+ fileid = os.path.join(archive_path, str(row['mp3_path']))
151
+ class_ = result[idx]
152
  fileid2class[fileid] = class_
153
  elif split == 'test':
154
  fileid2class = {}
155
  for idx, row in test_df.iterrows():
156
+ fileid = os.path.join(archive_path, str(row['mp3_path']))
157
+ class_ = result[idx]
158
  fileid2class[fileid] = class_
159
 
 
160
  for guid, audio_path in enumerate(_walker):
161
+ if audio_path not in fileid2class:
 
 
162
  continue
163
+ tags = fileid2class.get(audio_path)
164
  yield guid, {
165
  "id": str(guid),
166
  "file": audio_path,
167
  "audio": audio_path,
168
+ "sound": tags,
169
+ "label": tags,
170
  }
171
 
172