Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -15,10 +15,8 @@ dataset_info:
|
|
| 15 |
features:
|
| 16 |
- name: id
|
| 17 |
dtype: string
|
| 18 |
-
- name:
|
| 19 |
-
|
| 20 |
-
- name: image_2
|
| 21 |
-
dtype: image
|
| 22 |
- name: question
|
| 23 |
dtype: string
|
| 24 |
- name: answer
|
|
@@ -26,6 +24,11 @@ dataset_info:
|
|
| 26 |
splits:
|
| 27 |
- name: train
|
| 28 |
num_examples: 1000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 29 |
---
|
| 30 |
|
| 31 |
# SenseBench β Representative Sample
|
|
@@ -55,9 +58,3 @@ A representative 1,000-instance sample of [SenseBench](https://huggingface.co/da
|
|
| 55 |
β βββ paired/ 110 images
|
| 56 |
βββ questions.jsonl
|
| 57 |
```
|
| 58 |
-
|
| 59 |
-
## Reproduce
|
| 60 |
-
|
| 61 |
-
```bash
|
| 62 |
-
python sample.py
|
| 63 |
-
```
|
|
|
|
| 15 |
features:
|
| 16 |
- name: id
|
| 17 |
dtype: string
|
| 18 |
+
- name: images
|
| 19 |
+
sequence: image
|
|
|
|
|
|
|
| 20 |
- name: question
|
| 21 |
dtype: string
|
| 22 |
- name: answer
|
|
|
|
| 24 |
splits:
|
| 25 |
- name: train
|
| 26 |
num_examples: 1000
|
| 27 |
+
configs:
|
| 28 |
+
- config_name: default
|
| 29 |
+
data_files:
|
| 30 |
+
- split: train
|
| 31 |
+
path: questions.jsonl
|
| 32 |
---
|
| 33 |
|
| 34 |
# SenseBench β Representative Sample
|
|
|
|
| 58 |
β βββ paired/ 110 images
|
| 59 |
βββ questions.jsonl
|
| 60 |
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|