Datasets:

Modalities:
Image
Text
Formats:
parquet
ArXiv:
Libraries:
Datasets
Dask
HugoLaurencon commited on
Commit
4ba0b3c
1 Parent(s): c6010d5

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +22 -0
README.md CHANGED
@@ -36,6 +36,24 @@ dataset_info:
36
  num_examples: 1308
37
  download_size: 25388245
38
  dataset_size: 25528774.867153686
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
39
  - config_name: vqarad
40
  features:
41
  - name: images
@@ -63,6 +81,10 @@ configs:
63
  data_files:
64
  - split: train
65
  path: intergps/train-*
 
 
 
 
66
  - config_name: vqarad
67
  data_files:
68
  - split: train
 
36
  num_examples: 1308
37
  download_size: 25388245
38
  dataset_size: 25528774.867153686
39
+ - config_name: tat_qa
40
+ features:
41
+ - name: images
42
+ sequence: image
43
+ - name: texts
44
+ list:
45
+ - name: user
46
+ dtype: string
47
+ - name: assistant
48
+ dtype: string
49
+ - name: source
50
+ dtype: string
51
+ splits:
52
+ - name: train
53
+ num_bytes: 73213942.0
54
+ num_examples: 2199
55
+ download_size: 70862028
56
+ dataset_size: 73213942.0
57
  - config_name: vqarad
58
  features:
59
  - name: images
 
81
  data_files:
82
  - split: train
83
  path: intergps/train-*
84
+ - config_name: tat_qa
85
+ data_files:
86
+ - split: train
87
+ path: tat_qa/train-*
88
  - config_name: vqarad
89
  data_files:
90
  - split: train