Syoy commited on
Commit
c3c460b
1 Parent(s): 7827b9e

ignore embeddings while streaming

Browse files
Files changed (1) hide show
  1. dcase23-task2-enriched.py +2 -3
dcase23-task2-enriched.py CHANGED
@@ -63,7 +63,7 @@ DATA_URLS = {
63
 
64
  EMBEDDING_URLS = {
65
  "dev": {
66
- "ast-finetuned-audioset-10-10-0-4593-embeddings": {
67
  "train": "data/MIT_ast-finetuned-audioset-10-10-0-4593-embeddings_dev_train.npz",
68
  "test": "data/MIT_ast-finetuned-audioset-10-10-0-4593-embeddings_dev_test.npz",
69
  "size": (1, 768),
@@ -263,7 +263,7 @@ class DCASE2023Task2Dataset(datasets.GeneratorBasedBuilder):
263
  audio_path[split] = dl_manager.download(self.config.data_urls[split])
264
  local_extracted_archive[split] = dl_manager.extract(
265
  audio_path[split]) if not dl_manager.is_streaming else None
266
- if self.config.embeddings_urls is not None:
267
  for emb_name, emb_data in self.config.embeddings_urls.items():
268
  downloaded_embeddings = dl_manager.download(emb_data[split])
269
  embeddings[split][emb_name] = np.load(downloaded_embeddings, allow_pickle=True)["arr_0"].item()
@@ -310,4 +310,3 @@ class DCASE2023Task2Dataset(datasets.GeneratorBasedBuilder):
310
 
311
  if __name__ == "__main__":
312
  ds = load_dataset("dcase23-task2-enriched.py", "dev", split="train", streaming=True)
313
-
 
63
 
64
  EMBEDDING_URLS = {
65
  "dev": {
66
+ "ast-finetuned-audioset-10-10-0.4593-embeddings": {
67
  "train": "data/MIT_ast-finetuned-audioset-10-10-0-4593-embeddings_dev_train.npz",
68
  "test": "data/MIT_ast-finetuned-audioset-10-10-0-4593-embeddings_dev_test.npz",
69
  "size": (1, 768),
 
263
  audio_path[split] = dl_manager.download(self.config.data_urls[split])
264
  local_extracted_archive[split] = dl_manager.extract(
265
  audio_path[split]) if not dl_manager.is_streaming else None
266
+ if self.config.embeddings_urls is not None and not dl_manager.is_streaming:
267
  for emb_name, emb_data in self.config.embeddings_urls.items():
268
  downloaded_embeddings = dl_manager.download(emb_data[split])
269
  embeddings[split][emb_name] = np.load(downloaded_embeddings, allow_pickle=True)["arr_0"].item()
 
310
 
311
  if __name__ == "__main__":
312
  ds = load_dataset("dcase23-task2-enriched.py", "dev", split="train", streaming=True)