Refine task categories and tags
#2
by
nielsr
HF Staff
- opened
README.md
CHANGED
@@ -2,10 +2,11 @@
|
|
2 |
language:
|
3 |
- en
|
4 |
license: mit
|
5 |
-
|
|
|
6 |
task_categories:
|
7 |
-
- video-
|
8 |
-
|
9 |
tags:
|
10 |
- video
|
11 |
- multimodal
|
@@ -14,15 +15,11 @@ tags:
|
|
14 |
- multi-step-reasoning
|
15 |
- long-form-reasoning
|
16 |
- large-video-language-model
|
17 |
-
- large-multimodal-model
|
18 |
-
- multimodal-large-language-model
|
19 |
-
size_categories:
|
20 |
-
- n<1K
|
21 |
configs:
|
22 |
- config_name: default
|
23 |
data_files:
|
24 |
- split: test
|
25 |
-
path:
|
26 |
---
|
27 |
|
28 |
# VCRBench: Exploring Long-form Causal Reasoning Capabilities of Large Video Language Models
|
@@ -36,42 +33,37 @@ configs:
|
|
36 |
|
37 |
Authors: [Pritam Sarkar](https://pritamsarkar.com) and [Ali Etemad](https://www.aiimlab.com/ali-etemad)
|
38 |
|
39 |
-
This repository provides the official implementation of **[VCRBench](https://arxiv.org/abs/2505.08455)**.
|
40 |
|
41 |

|
42 |
|
43 |
## Usage
|
44 |
|
45 |
-
|
46 |
|
47 |
-
|
|
|
|
|
48 |
from dataset import VCRBench
|
49 |
-
dataset=VCRBench(question_file="data.json",
|
50 |
-
video_root="./",
|
51 |
-
mode='default',
|
52 |
-
)
|
53 |
-
|
54 |
for sample in dataset:
|
55 |
-
print(sample['question']
|
56 |
-
print(sample['answer']
|
57 |
print('*'*10)
|
58 |
-
|
59 |
break
|
60 |
-
|
61 |
```
|
62 |
|
63 |
|
64 |
### Licensing Information
|
65 |
|
66 |
-
This dataset incorporates samples from [CrossTask](https://github.com/DmZhukov/CrossTask/blob/master/LICENSE)
|
67 |
-
|
68 |
-
This repository is released under the **MIT**. See [LICENSE](LICENSE) for details.
|
69 |
|
70 |
### Citation Information
|
71 |
|
72 |
If you find this work useful, please use the given bibtex entry to cite our work:
|
73 |
|
74 |
-
```
|
75 |
@misc{sarkar2025vcrbench,
|
76 |
title={VCRBench: Exploring Long-form Causal Reasoning Capabilities of Large Video Language Models},
|
77 |
author={Pritam Sarkar and Ali Etemad},
|
|
|
2 |
language:
|
3 |
- en
|
4 |
license: mit
|
5 |
+
size_categories:
|
6 |
+
- n<1K
|
7 |
task_categories:
|
8 |
+
- video-classification
|
9 |
+
pretty_name: VCRBench
|
10 |
tags:
|
11 |
- video
|
12 |
- multimodal
|
|
|
15 |
- multi-step-reasoning
|
16 |
- long-form-reasoning
|
17 |
- large-video-language-model
|
|
|
|
|
|
|
|
|
18 |
configs:
|
19 |
- config_name: default
|
20 |
data_files:
|
21 |
- split: test
|
22 |
+
path: data.json
|
23 |
---
|
24 |
|
25 |
# VCRBench: Exploring Long-form Causal Reasoning Capabilities of Large Video Language Models
|
|
|
33 |
|
34 |
Authors: [Pritam Sarkar](https://pritamsarkar.com) and [Ali Etemad](https://www.aiimlab.com/ali-etemad)
|
35 |
|
36 |
+
This repository provides the official implementation of **[VCRBench](https://arxiv.org/abs/2505.08455)**. VCRBench is a benchmark dataset for evaluating the causal reasoning capabilities of Large Video Language Models (LVLMs) in visually grounded, goal-driven scenarios. It consists of procedural videos with shuffled steps, requiring LVLMs to identify, reason about, and correctly sequence events to achieve a specific goal.
|
37 |
|
38 |

|
39 |
|
40 |
## Usage
|
41 |
|
42 |
+
For detailed usage instructions, please refer to the GitHub repository: [VCRBench](https://github.com/pritamqu/VCRBench)
|
43 |
|
44 |
+
A basic example:
|
45 |
+
|
46 |
+
```python
|
47 |
from dataset import VCRBench
|
48 |
+
dataset = VCRBench(question_file="data.json", video_root="./", mode='default')
|
|
|
|
|
|
|
|
|
49 |
for sample in dataset:
|
50 |
+
print(sample['question'])
|
51 |
+
print(sample['answer'])
|
52 |
print('*'*10)
|
|
|
53 |
break
|
|
|
54 |
```
|
55 |
|
56 |
|
57 |
### Licensing Information
|
58 |
|
59 |
+
This dataset incorporates samples from [CrossTask](https://github.com/DmZhukov/CrossTask/blob/master/LICENSE) and is subject to their respective original licenses. This repository is released under the **MIT License**. See [LICENSE](LICENSE) for details. Users must adhere to the terms and conditions specified by these licenses.
|
60 |
+
|
|
|
61 |
|
62 |
### Citation Information
|
63 |
|
64 |
If you find this work useful, please use the given bibtex entry to cite our work:
|
65 |
|
66 |
+
```bibtex
|
67 |
@misc{sarkar2025vcrbench,
|
68 |
title={VCRBench: Exploring Long-form Causal Reasoning Capabilities of Large Video Language Models},
|
69 |
author={Pritam Sarkar and Ali Etemad},
|