Card: drop seed bullet
Browse files
README.md
CHANGED
|
@@ -51,7 +51,6 @@ VLM-as-judge baselines on the same items.
|
|
| 51 |
- 3 tasks (classification, detection, segmentation) x 8 domains (everyday, street, aerial, medical, industrial, agriculture, documents/UI, wildlife/underwater)
|
| 52 |
- 6,584 items: 5,988 test, 596 dev (separate photos, for prompt tuning only)
|
| 53 |
- 50% correct / 50% corrupted in every task x domain cell; 16 error types, easy / medium / hard
|
| 54 |
-
- Seed 42, rebuild gives an identical dataset; all sources redistributable
|
| 55 |
|
| 56 |
## Quick start
|
| 57 |
|
|
|
|
| 51 |
- 3 tasks (classification, detection, segmentation) x 8 domains (everyday, street, aerial, medical, industrial, agriculture, documents/UI, wildlife/underwater)
|
| 52 |
- 6,584 items: 5,988 test, 596 dev (separate photos, for prompt tuning only)
|
| 53 |
- 50% correct / 50% corrupted in every task x domain cell; 16 error types, easy / medium / hard
|
|
|
|
| 54 |
|
| 55 |
## Quick start
|
| 56 |
|