hao-ai-lab
[Feature] The NPU version adaptation work of Hunyunavideo
Motivation
Sorry ，Excuse me!Your work is meaningful, and we hope to see it on the NPU!So we have done some NPU adaptation and tuning of HunyuanVideo on the warehouse, can it be sent to your warehouse? If so, what form do you recommend? If not, please reply too! Thank you!
Related resources
No response
Could you explain more? Are you referring to contributing to the FastVideo repository?
Yeah! We have done some NPU adaptation and performance optimization work on the HunyuanVideo model of FastVideo.May I ask if this contribution can be accepted here?
Sure, we welcome contributions! You can find our contribution guide here. However, please note that we've migrated to V1, and there may be significant differences in the codebase compared to earlier versions.
If your PR is huge, you can join FastVideo slack: https://github.com/hao-ai-lab/FastVideo?tab=readme-ov-file or WeChat: https://ibb.co/qqPzbrw and we can open up a channel to discuss.
OKOK. We mainly want to discuss in what form this side is willing to undertake the NPU version code.We also need to take a look at the v1 version code repository!
If your PR is huge, you can join FastVideo slack: https://github.com/hao-ai-lab/FastVideo?tab=readme-ov-file or WeChat: https://ibb.co/qqPzbrw and we can open up a channel to discuss. OK. Due to environmental limitations, we'll join in when we go back at noon. Thank you!