Datasets:
Update README.md
Browse files
README.md
CHANGED
@@ -32,6 +32,27 @@ This release includes **200 open-source samples** from the full dataset, provide
|
|
32 |
|
33 |
---
|
34 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
35 |
## 📚 Citation
|
36 |
|
37 |
If you use this dataset in your research, please cite it as follows:
|
|
|
32 |
|
33 |
---
|
34 |
|
35 |
+
## 💡 Usage
|
36 |
+
|
37 |
+
This dataset is fully supported by [VLMEvalKit](https://github.com/THUDM/VLMEvalKit).
|
38 |
+
|
39 |
+
### 🔧 Evaluate with VLMEvalKit
|
40 |
+
|
41 |
+
Registered dataset names:
|
42 |
+
|
43 |
+
- `StaticEmbodiedBench` — for standard evaluation
|
44 |
+
- `StaticEmbodiedBench_circular` — for circular evaluation (multi-round)
|
45 |
+
|
46 |
+
To run evaluation in VLMEvalKit:
|
47 |
+
|
48 |
+
```bash
|
49 |
+
python run.py --data StaticEmbodiedBench --model <your_model_name> --verbose
|
50 |
+
```
|
51 |
+
For circular evaluation, simply use:
|
52 |
+
```bash
|
53 |
+
python run.py --data StaticEmbodiedBench_circular --model <your_model_name> --verbose
|
54 |
+
```
|
55 |
+
|
56 |
## 📚 Citation
|
57 |
|
58 |
If you use this dataset in your research, please cite it as follows:
|