Update README.md

README.md

@@ -4,6 +4,35 @@ datasets:
 - VITA-MLLM/Long-VITA-Training-Data
 ---
 
+
+# Long-VITA-128K
+
+Github: https://github.com/VITA-MLLM/Long-VITA
+
+## 👀 Overview
+Long-VITA is a strong long-context visual language model and supports more than 1 million tokens.
+
+- This weight is trained on Ascend NPU with MindSpeed.
+
+- To infer and evaluate on Nvidia GPU, we also implement Long-VITA on Megatron with Transformer Engine.
+
+- The converted weight is in https://huggingface.co/VITA-MLLM/Long-VITA-128K_MG.
+
+
+## 📈 Experimental Results
+- **Comparison of image understanding**.
+
+![image](https://github.com/user-attachments/assets/30f62f51-675e-4dac-9f18-f743c311f9be)
+
+
+
+- **Comparison of video understanding**.
+
+![image](https://github.com/user-attachments/assets/fee848d5-da20-4a30-9172-2ec9746ada25)
+
+
+
+
 ## ACCEPTABLE USE POLICY
 
 Any license on the model is subject to your compliance with the Acceptable Use Policy, and You must not violate (or encourage or permit anyone else to violate) any term of the Acceptable Use Policy. Tencent reserves the right to update this Acceptable Use Policy from time to time.