Update README.md

README.md

@@ -19,10 +19,10 @@ should probably proofread and complete it, then remove this comment. -->
 ## Model description
 
 
-This model is a fine-tuned version of [Qwen/Qwen2.5-VL-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-VL-7B-Instruct) on the cam_motion_balanced_vqa, the cam_motion_imbalanced_vqa and the cam_motion_caption datasets.
+This model is a fine-tuned version of [Qwen/Qwen2.5-VL-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-VL-7B-Instruct) on the current most, high-quality camera motion dataset that is publically available. This preview model is the current SOTA for classifying camera motion or being used for video-text retrieval with camera motion captions using [VQAScore](https://arxiv.org/pdf/2404.01291). Find more information about our work on our Github page for [CameraBench](https://github.com/sy77777en/CameraBench). *More updates to the benchmark and models will come in the future. Stay tuned!*
 ## Intended uses & limitations
 
- The usage is identical to a [Qwen2.5-VL](https://github.com/QwenLM/Qwen2.5-VL) model. Our model is primarily useful for camera motion classification in videos as well as video-text retrieval.
+ The usage is identical to a [Qwen2.5-VL](https://github.com/QwenLM/Qwen2.5-VL) model. Our model is primarily useful for camera motion classification in videos as well as video-text retrieval (current SOTA in both tasks).
  
  **A quick demo is shown below:**
 <details>