Update README.md

README.md

@@ -18,11 +18,17 @@ MAmmoTH-VL2, the model trained with VisualWebInstruct.
 [Paper](https://arxiv.org/abs/2503.10582)|
 [Website](https://tiger-ai-lab.github.io/VisualWebInstruct/)
 
+
+
+
 # Example Usage
+## Requirements
+```python
+llava==1.7.0.dev0 # pip install git+https://github.com/LLaVA-VL/LLaVA-NeXT.git
+torch==2.5.1
+```
 To perform inference using MAmmoTH-VL2, you can use the following code snippet:
 ```python
-# pip install git+https://github.com/LLaVA-VL/LLaVA-NeXT.git
-
 from llava.model.builder import load_pretrained_model
 from llava.mm_utils import process_images
 from llava.constants import DEFAULT_IMAGE_TOKEN