diff --git a/README.md b/README.md index 13b5a9c..a29378e 100644 --- a/README.md +++ b/README.md @@ -40,7 +40,7 @@ With Open-Sora, our goal is to foster innovation, creativity, and inclusivity wi ## ๐Ÿ“ฐ News -- **[2025.03.12]** ๐Ÿ”ฅ We released **Open-Sora 2.0** (11B). ๐ŸŽฌ 11B model achieves [on-par performance](#evaluation) with 14B HunyuanVideo & 30B Step-Video on ๐Ÿ“VBench & ๐Ÿ“ŠHuman Preference. ๐Ÿ› ๏ธ Fully open-source: checkpoints and training codes for training with only **$200K**. [[report]](https://github.com/hpcaitech/Open-Sora-Demo/blob/main/paper/Open_Sora_2_tech_report.pdf) +- **[2025.03.12]** ๐Ÿ”ฅ We released **Open-Sora 2.0** (11B). ๐ŸŽฌ 11B model achieves [on-par performance](#evaluation) with 11B HunyuanVideo & 30B Step-Video on ๐Ÿ“VBench & ๐Ÿ“ŠHuman Preference. ๐Ÿ› ๏ธ Fully open-source: checkpoints and training codes for training with only **$200K**. [[report]](https://github.com/hpcaitech/Open-Sora-Demo/blob/main/paper/Open_Sora_2_tech_report.pdf) - **[2025.02.20]** ๐Ÿ”ฅ We released **Open-Sora 1.3** (1B). With the upgraded VAE and Transformer architecture, the quality of our generated videos has been greatly improved ๐Ÿš€. [[checkpoints]](#open-sora-13-model-weights) [[report]](/docs/report_04.md) [[demo]](https://huggingface.co/spaces/hpcai-tech/open-sora) - **[2024.12.23]** The development cost of video generation models has saved by 50%! Open-source solutions are now available with H200 GPU vouchers. [[blog]](https://company.hpc-ai.com/blog/the-development-cost-of-video-generation-models-has-saved-by-50-open-source-solutions-are-now-available-with-h200-gpu-vouchers) [[code]](https://github.com/hpcaitech/Open-Sora/blob/main/scripts/train.py) [[vouchers]](https://colossalai.org/zh-Hans/docs/get_started/bonus/) - **[2024.06.17]** We released **Open-Sora 1.2**, which includes **3D-VAE**, **rectified flow**, and **score condition**. The video quality is greatly improved. [[checkpoints]](#open-sora-12-model-weights) [[report]](/docs/report_03.md) [[arxiv]](https://arxiv.org/abs/2412.20404) @@ -285,7 +285,7 @@ On [VBench](https://huggingface.co/spaces/Vchitect/VBench_Leaderboard), Open-Sor ![VBench](https://github.com/hpcaitech/Open-Sora-Demo/blob/main/readme/v2_vbench.png) -Human preference results show our model is on par with HunyuanVideo 14B and Step-Video 30B. +Human preference results show our model is on par with HunyuanVideo 11B and Step-Video 30B. ![Win Rate](https://github.com/hpcaitech/Open-Sora-Demo/blob/main/readme/v2_winrate.png) @@ -327,12 +327,11 @@ Here we only list a few of the projects. For other works and datasets, please re ## Citation ```bibtex -@software{opensora, - author = {Zangwei Zheng and Xiangyu Peng and Tianji Yang and Chenhui Shen and Shenggui Li and Hongxin Liu and Yukun Zhou and Tianyi Li and Yang You}, - title = {Open-Sora: Democratizing Efficient Video Production for All}, - month = {March}, - year = {2024}, - url = {https://github.com/hpcaitech/Open-Sora} +@article{opensora, + title={Open-sora: Democratizing efficient video production for all}, + author={Zheng, Zangwei and Peng, Xiangyu and Yang, Tianji and Shen, Chenhui and Li, Shenggui and Liu, Hongxin and Zhou, Yukun and Li, Tianyi and You, Yang}, + journal={arXiv preprint arXiv:2412.20404}, + year={2024} } ```