Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -46,8 +46,10 @@ This is a newly curated dataset to evaluate multimodal language models' capabili
|
|
| 46 |
This evaluation dataset contains more than 200 human-annotated challenging multi-image reasoning problems.
|
| 47 |
|
| 48 |
### Leaderboard
|
|
|
|
|
|
|
| 49 |
| Models | Size | Mantis-Eval |
|
| 50 |
-
|
| 51 |
| GPT-4V | - | 62.67 |
|
| 52 |
| Mantis-SigLIP | 8B | 59.45 |
|
| 53 |
| Mantis-Idefics2 | 8B | 57.14 |
|
|
|
|
| 46 |
This evaluation dataset contains more than 200 human-annotated challenging multi-image reasoning problems.
|
| 47 |
|
| 48 |
### Leaderboard
|
| 49 |
+
We list the current results as follows:
|
| 50 |
+
|
| 51 |
| Models | Size | Mantis-Eval |
|
| 52 |
+
|:------------------|:-----|:------------|
|
| 53 |
| GPT-4V | - | 62.67 |
|
| 54 |
| Mantis-SigLIP | 8B | 59.45 |
|
| 55 |
| Mantis-Idefics2 | 8B | 57.14 |
|