Datasets:
Update README.md
Browse files
README.md
CHANGED
@@ -30,29 +30,6 @@ dataset_info:
|
|
30 |
num_examples: 932
|
31 |
download_size: 2198653805
|
32 |
dataset_size: 2201624804.0
|
33 |
-
- config_name: char_coherence
|
34 |
-
features:
|
35 |
-
- name: context
|
36 |
-
sequence: image
|
37 |
-
- name: options
|
38 |
-
sequence: image
|
39 |
-
- name: index
|
40 |
-
dtype: int32
|
41 |
-
- name: solution_index
|
42 |
-
dtype: int32
|
43 |
-
- name: split
|
44 |
-
dtype: string
|
45 |
-
- name: task_type
|
46 |
-
dtype: string
|
47 |
-
splits:
|
48 |
-
- name: val
|
49 |
-
num_bytes: 379247043
|
50 |
-
num_examples: 143
|
51 |
-
- name: test
|
52 |
-
num_bytes: 1139804961.0
|
53 |
-
num_examples: 489
|
54 |
-
download_size: 1518604969
|
55 |
-
dataset_size: 1519052004.0
|
56 |
- config_name: sequence_filling
|
57 |
features:
|
58 |
- name: sample_id
|
@@ -135,8 +112,8 @@ Given a caption from the previous panel, select the panel that best continues th
|
|
135 |
```python
|
136 |
from datasets import load_dataset
|
137 |
|
138 |
-
skill = "
|
139 |
-
split = "val" # "test"
|
140 |
dataset = load_dataset("VLR-CVC/ComPAP", skill, split=split)
|
141 |
```
|
142 |
|
|
|
30 |
num_examples: 932
|
31 |
download_size: 2198653805
|
32 |
dataset_size: 2201624804.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
33 |
- config_name: sequence_filling
|
34 |
features:
|
35 |
- name: sample_id
|
|
|
112 |
```python
|
113 |
from datasets import load_dataset
|
114 |
|
115 |
+
skill = "sequence_filling" # "sequence_filling", "char_coherence", "visual_closure", "text_closure", "caption_relevance"
|
116 |
+
split = "val" # "val", "test"
|
117 |
dataset = load_dataset("VLR-CVC/ComPAP", skill, split=split)
|
118 |
```
|
119 |
|