[Sora行业应用] Open-Sora-Plan v1.0.0 released: Peking University AI team reproduces Sora, available to try online

https://v.douyin.com/i2Pnyanb/
Open-Sora-Plan is an open-source project initiated by the Yuan Lab at Peking University and the Tuzhan AI Joint Lab, with the goal of reproducing the text-to-video model Sora. The project hopes to improve video generation quality, text control, and AI chip support through the joint efforts of the open-source community. The project is still in its early stages, and the community is welcome to contribute via Pull Requests. The project uses the MIT License, meaning anyone can freely use and contribute to it.
[baike]Hardware Support Team[/baike]
At present, the PKU team has received support from Huawei's Ascend NPU team. Thanks to this, inference is now supported on domestic AI chips (Huawei Ascend 910; we look forward to more domestic computing power chips).
[baike]Some Results[/baike]
The content shown below consists of compressed GIFs provided by the PKU AI team (they were not compressed by us).

A tranquil underwater scene, with a sea turtle swimming......



A tranquil waterfall cascading down over moss-covered rocks



The dynamic interaction between the ocean and large rocks



[baike]Project URL[/baike]
https://github.com/PKU-YuanGroup/Open-Sora-Plan
[baike]Open-Sora1.0 Online Address[/baike]
In addition to open-sourcing the system, the PKU team also provided a ready-to-use online address for people to try.
https://huggingface.co/spaces/LanguageBind/Open-Sora-Plan-v1.0.0