Datasets:
QCRI
/

Modalities:
Image
Text
Formats:
parquet
Languages:
Arabic
ArXiv:
Libraries:
Datasets
pandas
License:
Firoj commited on
Commit
772bf5d
1 Parent(s): 8ac2686

Upload armeme_loader.py with huggingface_hub

Browse files
Files changed (1) hide show
  1. armeme_loader.py +69 -0
armeme_loader.py ADDED
@@ -0,0 +1,69 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import os
2
+ import json
3
+ import datasets
4
+ from datasets import Dataset, DatasetDict, load_dataset, Features, Value, Image, ClassLabel
5
+
6
+
7
+ # Define the paths to your dataset
8
+ image_root_dir = "./"
9
+ train_jsonl_file_path = "arabic_memes_categorization_train.jsonl"
10
+ dev_jsonl_file_path = "arabic_memes_categorization_dev.jsonl"
11
+ test_jsonl_file_path = "arabic_memes_categorization_test.jsonl"
12
+
13
+
14
+
15
+ # Function to load each dataset split
16
+ def load_armeme_split(jsonl_file_path, image_root_dir):
17
+ texts = []
18
+ images = []
19
+ ids=[]
20
+ class_labels=[]
21
+ image_file_paths = []
22
+
23
+ # Load JSONL file
24
+ with open(jsonl_file_path, 'r') as f:
25
+ for line in f:
26
+ item = json.loads(line)
27
+ ids.append(item['id'])
28
+ texts.append(item['text'])
29
+ image_file_path = os.path.join(image_root_dir, item['img_path'])
30
+ images.append(image_file_path)
31
+ image_file_paths.append(image_file_path)
32
+ class_labels.append(item['class_label'])
33
+
34
+ # Create a dictionary to match the dataset structure
35
+ data_dict = {
36
+ 'id':ids,
37
+ 'text': texts,
38
+ 'image': images,
39
+ 'img_path': image_file_paths,
40
+ 'class_label': class_labels
41
+ }
42
+
43
+ # Define the features
44
+ features = Features({
45
+ 'id': Value('string'),
46
+ 'text': Value('string'),
47
+ 'image': Image(),
48
+ 'img_path': Value('string'),
49
+ 'class_label': ClassLabel(names=['not_propaganda','propaganda','not-meme','other'])
50
+ })
51
+
52
+ # Create a Hugging Face dataset from the dictionary
53
+ dataset = Dataset.from_dict(data_dict, features=features)
54
+ return dataset
55
+
56
+ # Load each split
57
+ train_dataset = load_armeme_split(train_jsonl_file_path, image_root_dir)
58
+ dev_dataset = load_armeme_split(dev_jsonl_file_path, image_root_dir)
59
+ test_dataset = load_armeme_split(test_jsonl_file_path, image_root_dir)
60
+
61
+ # Create a DatasetDict
62
+ dataset_dict = DatasetDict({
63
+ 'train': train_dataset,
64
+ 'dev': dev_dataset,
65
+ 'test': test_dataset
66
+ })
67
+
68
+ # Push the dataset to Hugging Face Hub
69
+ dataset_dict.push_to_hub("QCRI/ArMeme")