Update README.md

README.md

@@ -28,9 +28,9 @@
 
 
 ## 1. News
-- 2025-5-16:✨✨We release our 4 stages prograssive training code (supporting Huawei's NPU and NVIDIA's GPU). You can refer to [LVM/script/train](LVM/script/train) and [LVM/train](LVM/train) for detailed training information.
-- 2025-5-16:✨✨We release the inference code in [LVM/script/inference](LVM/script/inference) and [LVM/inference](LVM/inference).
-- 2025-5-16:🔥🔥We release the first version of Video-GPT. Model Weight: [Video-GPT](https://huggingface.co/GrayShine/Video-GPT)
+- 2025-5-21:✨✨We release our 4 stages prograssive training code (supporting Huawei's NPU and NVIDIA's GPU). You can refer to [LVM/script/train](LVM/script/train) and [LVM/train](LVM/train) for detailed training information.
+- 2025-5-21:✨✨We release the inference code in [LVM/script/inference](LVM/script/inference) and [LVM/inference](LVM/inference).
+- 2025-5-21:🔥🔥We release the first version of Video-GPT. Model Weight: [Video-GPT](https://huggingface.co/GrayShine/Video-GPT)
 
 
 ## 2. Overview
@@ -41,7 +41,7 @@ Previous works on visual generation relies heavily on supervisory signals from t
 
 <!-- ![demo](https://github.com/zhuangshaobin/Video-GPT/tree/main/imgs/teaser.png) -->
 <p align="left">
-  <img src="https://github.com/zhuangshaobin/Video-GPT/tree/main/imgs/teaser.png" alt="demo" width="640"/>
+  <img src="./imgs/teaser.png" alt="demo" width="640"/>
 </p>
 
 In addition, compared with the previous model architecture with many special designs for diffusion model (e.g., UNet, DiT, MM-DiT), we adopted the simplest vanilla transformer architecture. On the one hand, it is more conducive to the exploration of scaling law in the future. On the other hand, it is also more convenient for the community to follow up.