cogstudio added and #323

README.md

@@ -277,7 +277,10 @@ pipe.vae.enable_tiling()
 We highly welcome contributions from the community and actively contribute to the open-source community. The following
 works have already been adapted for CogVideoX, and we invite everyone to use them:
 
-+ [CogVideoX-Fun](https://github.com/aigc-apps/CogVideoX-Fun): CogVideoX-Fun is a modified pipeline based on the CogVideoX architecture, supporting flexible resolutions and multiple launch methods.
++ [CogVideoX-Fun](https://github.com/aigc-apps/CogVideoX-Fun): CogVideoX-Fun is a modified pipeline based on the
+  CogVideoX architecture, supporting flexible resolutions and multiple launch methods.
++ [CogStudio](https://github.com/pinokiofactory/cogstudio): A separate repository for CogVideo's Gradio Web UI, which
+  supports more functional Web UIs.
 + [Xorbits Inference](https://github.com/xorbitsai/inference): A powerful and comprehensive distributed inference
   framework, allowing you to easily deploy your own models or the latest cutting-edge open-source models with just one
   click.
@@ -288,7 +291,8 @@ works have already been adapted for CogVideoX, and we invite everyone to use the
   techniques.
 + [AutoDL Space](https://www.codewithgpu.com/i/THUDM/CogVideo/CogVideoX-5b-demo): A one-click deployment Huggingface
   Space image provided by community members.
-+ [Interior Design Fine-Tuning Model](https://huggingface.co/collections/bertjiazheng/koolcogvideox-66e4762f53287b7f39f8f3ba): is a fine-tuned model based on CogVideoX, specifically designed for interior design.
++ [Interior Design Fine-Tuning Model](https://huggingface.co/collections/bertjiazheng/koolcogvideox-66e4762f53287b7f39f8f3ba):
+  is a fine-tuned model based on CogVideoX, specifically designed for interior design.
 
 ## Project Structure
 

README_ja.md

@@ -262,6 +262,7 @@ pipe.vae.enable_tiling()
 コミュニティからの貢献を大歓迎し、私たちもオープンソースコミュニティに積極的に貢献しています。以下の作品はすでにCogVideoXに対応しており、ぜひご利用ください：
 
 + [CogVideoX-Fun](https://github.com/aigc-apps/CogVideoX-Fun): CogVideoX-Funは、CogVideoXアーキテクチャを基にした改良パイプラインで、自由な解像度と複数の起動方法をサポートしています。
++ [CogStudio](https://github.com/pinokiofactory/cogstudio): CogVideo の Gradio Web UI の別のリポジトリ。より高機能な Web UI をサポートします。
 + [Xorbits Inference](https://github.com/xorbitsai/inference):
   強力で包括的な分散推論フレームワークであり、ワンクリックで独自のモデルや最新のオープンソースモデルを簡単にデプロイできます。
 + [ComfyUI-CogVideoXWrapper](https://github.com/kijai/ComfyUI-CogVideoXWrapper)

README_zh.md

@@ -249,6 +249,7 @@ pipe.vae.enable_tiling()
 
 我们非常欢迎来自社区的贡献，并积极的贡献开源社区。以下作品已经对CogVideoX进行了适配，欢迎大家使用:
 + [CogVideoX-Fun](https://github.com/aigc-apps/CogVideoX-Fun): CogVideoX-Fun是一个基于CogVideoX结构修改后的的pipeline，支持自由的分辨率，多种启动方式。
++ [CogStudio](https://github.com/pinokiofactory/cogstudio): CogVideo 的 Gradio Web UI单独实现仓库，支持更多功能的 Web UI。
 + [Xorbits Inference](https://github.com/xorbitsai/inference): 性能强大且功能全面的分布式推理框架，轻松一键部署你自己的模型或内置的前沿开源模型。
 + [ComfyUI-CogVideoXWrapper](https://github.com/kijai/ComfyUI-CogVideoXWrapper) 使用ComfyUI框架，将CogVideoX加入到你的工作流中。
 + [VideoSys](https://github.com/NUS-HPC-AI-Lab/VideoSys): VideoSys 提供了易用且高性能的视频生成基础设施，支持完整的管道，并持续集成最新的模型和技术。

sat/dit_video_concat.py

@@ -308,6 +308,10 @@ class Rotary3DPositionEmbeddingMixin(BaseMixin):
         return t * freqs_cos + rotate_half(t) * freqs_sin
 
     def position_embedding_forward(self, position_ids, **kwargs):
+        if self.pos_embedding is not None:
+            print(self.pos_embedding)
+            return self.pos_embedding[:, :self.text_length + kwargs["seq_length"]]
+        else:
             return None
 
     def attention_fn(