Datasets:
Upload folder using huggingface_hub
Browse files- README.md +22 -3
- test/metadata.jsonl +0 -0
README.md
CHANGED
|
@@ -1,3 +1,22 @@
|
|
| 1 |
-
---
|
| 2 |
-
|
| 3 |
-
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
configs:
|
| 3 |
+
- config_name: default
|
| 4 |
+
data_files:
|
| 5 |
+
- split: test
|
| 6 |
+
path: test/metadata.jsonl
|
| 7 |
+
---
|
| 8 |
+
|
| 9 |
+
# SYNCR
|
| 10 |
+
|
| 11 |
+
A synthetic cross-video reasoning dataset spanning 4 categories and 8 tasks.
|
| 12 |
+
|
| 13 |
+
This version contains annotations only. Video files are not included.
|
| 14 |
+
|
| 15 |
+
Each row contains:
|
| 16 |
+
- category
|
| 17 |
+
- task
|
| 18 |
+
- question
|
| 19 |
+
- options
|
| 20 |
+
- answer
|
| 21 |
+
- videos
|
| 22 |
+
- video_ranges
|
test/metadata.jsonl
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|