Does VideoSys support the "Open Sora Plan v.1.2.0" for image-to-video generation?

[trouble-maker007]

No description provided.

[oahzxl]

will soon

[trouble-maker007]

@oahzxl thanks response, btw, is there any plan support Open Sora Plan v.1.2.0 and cogvideox, make it more efficient training, by utilizing full 3D attention, the requirement for memory storage is very high

[oahzxl]

yeah but efficient inference for now. we don't have plan to optimize their training currently.

[devtony10]

Good day, please is there any news on the image to video generation?

[oahzxl]

will be in late oct or early nov, alone with some other video enhancements

[devtony10]

Okay, thanks for your response. Could I get a workaround I can implement?

[oahzxl]

it will be so good if you'd like to implement image-to-video for open-sora-plan v1.2! if you're interested, we can create a new issue and talk about the details!

[devtony10]

I am

[devtony10]

Thanks for the offer

[oahzxl]

Thanks for the offer

hi, i create a new issue here #226

look forward to your feedback