Datasets:
QCRI
/

Modalities:
Image
Text
Formats:
parquet
Languages:
Arabic
ArXiv:
Libraries:
Datasets
pandas
License:
Firoj commited on
Commit
5198cc6
1 Parent(s): 257976b

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -4
README.md CHANGED
@@ -55,16 +55,13 @@ import json
55
  output_dir="./ArMeme/"
56
  os.makedirs(output_dir, exist_ok=True)
57
 
58
- # Assuming the images are in a column named "image"
59
- image_column = "image"
60
-
61
  # Iterate over the dataset and save each image
62
  for split in ['train','dev','test']:
63
  jsonl_path = os.path.join(output_dir, f"arabic_memes_categorization_{split}.jsonl")
64
  with open(jsonl_path, 'w', encoding='utf-8') as f:
65
  for idx, item in enumerate(dataset[split]):
66
  # Access the image directly as it's already a PIL.Image object
67
- image = item[image_column]
68
  image_path = os.path.join(output_dir, item['img_path'])
69
  # Ensure the directory exists
70
  os.makedirs(os.path.dirname(image_path), exist_ok=True)
 
55
  output_dir="./ArMeme/"
56
  os.makedirs(output_dir, exist_ok=True)
57
 
 
 
 
58
  # Iterate over the dataset and save each image
59
  for split in ['train','dev','test']:
60
  jsonl_path = os.path.join(output_dir, f"arabic_memes_categorization_{split}.jsonl")
61
  with open(jsonl_path, 'w', encoding='utf-8') as f:
62
  for idx, item in enumerate(dataset[split]):
63
  # Access the image directly as it's already a PIL.Image object
64
+ image = item['image']
65
  image_path = os.path.join(output_dir, item['img_path'])
66
  # Ensure the directory exists
67
  os.makedirs(os.path.dirname(image_path), exist_ok=True)