jacklangerman commited on
Commit
a3a1af7
·
1 Parent(s): 403b0be
Files changed (1) hide show
  1. script.py +2 -1
script.py CHANGED
@@ -56,13 +56,14 @@ if __name__ == "__main__":
56
  data_path = data_path_local
57
 
58
  print(data_path)
59
- print([str(p) for p in data_path.rglob('*validation*.(arrow|tar)')])
60
 
61
  # dataset = load_dataset(params['dataset'], trust_remote_code=True, use_auth_token=params['token'])
62
  data_files = {
63
  "validation": [str(p) for p in [*data_path.rglob('*validation*.arrow')]+[*data_path.rglob('*validation*.tar')]],
64
  "test": [str(p) for p in [*data_path.rglob('*test*.arrow')]+[*data_path.rglob('*test*.tar')]],
65
  }
 
 
66
  try:
67
  dataset = load_dataset(
68
  "arrow",
 
56
  data_path = data_path_local
57
 
58
  print(data_path)
 
59
 
60
  # dataset = load_dataset(params['dataset'], trust_remote_code=True, use_auth_token=params['token'])
61
  data_files = {
62
  "validation": [str(p) for p in [*data_path.rglob('*validation*.arrow')]+[*data_path.rglob('*validation*.tar')]],
63
  "test": [str(p) for p in [*data_path.rglob('*test*.arrow')]+[*data_path.rglob('*test*.tar')]],
64
  }
65
+ print(data_files)
66
+
67
  try:
68
  dataset = load_dataset(
69
  "arrow",