dsmueller commited on
Commit
3c75e7b
1 Parent(s): bfcd978

Added directory for dataset

Browse files
Files changed (2) hide show
  1. Dockerfile +1 -0
  2. run.py +1 -1
Dockerfile CHANGED
@@ -14,6 +14,7 @@ RUN pip install datasets
14
 
15
  COPY . .
16
  RUN mkdir -p /code/.cache
 
17
  RUN chmod -R 777 /code
18
  RUN python prepare.py
19
  CMD ["python", "run.py"]
 
14
 
15
  COPY . .
16
  RUN mkdir -p /code/.cache
17
+ RUN mkdir -p /code/hf_data
18
  RUN chmod -R 777 /code
19
  RUN python prepare.py
20
  CMD ["python", "run.py"]
run.py CHANGED
@@ -7,7 +7,7 @@ import os
7
  if __name__ == "__main__":
8
  dataset_name="ai-aerospace/ac-text-embedding-ada-002-ams-test"
9
  # Load dataset using datasets.load_dataset()
10
- dataset = datasets.load_dataset(dataset_name, split="train",data_dir='/code')
11
  print("Dataset loaded using datasets.load_dataset().")
12
 
13
  df = dataset.to_pandas()
 
7
  if __name__ == "__main__":
8
  dataset_name="ai-aerospace/ac-text-embedding-ada-002-ams-test"
9
  # Load dataset using datasets.load_dataset()
10
+ dataset = datasets.load_dataset(dataset_name, split="train",data_dir=os.environ.get("HF_HOME"))
11
  print("Dataset loaded using datasets.load_dataset().")
12
 
13
  df = dataset.to_pandas()