Datasets:

bruAristimunha commited on
Commit
fa469fe
·
verified ·
1 Parent(s): bbb5950

Restore direct load_dataset examples using native Hub caching

Browse files
Files changed (1) hide show
  1. README.md +2 -4
README.md CHANGED
@@ -31,11 +31,9 @@ pip install "datasets[bids] @ git+https://github.com/bruAristimunha/datasets@fea
31
 
32
  ```python
33
  from datasets import load_dataset
34
- from huggingface_hub import snapshot_download
35
 
36
- path = snapshot_download("EEGDash/nm000290", repo_type="dataset")
37
 
38
- dataset = load_dataset("bids", data_dir=path, split="train")
39
  row = dataset[0]
40
  recording = row["biosignal"]
41
  print(recording)
@@ -46,7 +44,7 @@ print(row["metadata"])
46
  There are 10 subjects and
47
  112 recordings in the complete mirror.
48
  The example may select one subject; all subjects are stored here.
49
- The snapshot download preserves companion files and sidecars. Local recordings are opened lazily.
50
 
51
 
52
  ## Source and attribution
 
31
 
32
  ```python
33
  from datasets import load_dataset
 
34
 
 
35
 
36
+ dataset = load_dataset("EEGDash/nm000290", split="train")
37
  row = dataset[0]
38
  recording = row["biosignal"]
39
  print(recording)
 
44
  There are 10 subjects and
45
  112 recordings in the complete mirror.
46
  The example may select one subject; all subjects are stored here.
47
+ Accessing a recording caches its subject/session/datatype folder and the repository’s JSON/TSV metadata. Other runs in that folder may also be downloaded. MNE then opens the cached recording lazily.
48
 
49
 
50
  ## Source and attribution