Generative AI for the Blender VSE: Text, video or image to video, image and audio in Blender Video Sequence Editor.
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

1.6 KiB

Text to Video

Adding text to video to the Blender Video Sequence Editor using ModelScope. The full version needs 20 GB VRAM to work. An alternative version with a pruned data set(smaller) will download as default, for now - see code to change to 20 GB VRAM version.

How to install

Run Blender as Administrator and open the system console(Blender system console processsing can be stopped with Ctrl + C) before clicking the Generate Movie button, so you can see how it is progressing the module installation. It'll take a lot of time, since it is a huge and slow download.

Alternative download for the weighted set which needs 20 GB of VRAM to run: https://modelscope.cn/models/damo/text-to-video-synthesis/files

Or a pruned weighted set which should be able to run on 6+ GB VRAM: https://huggingface.co/kabachuha/modelscope-damo-text2video-pruned-weights/tree/main

The downloaded models should be placen in this folder on Windows: C:\Users\Your_user_name\AppData\Roaming\Blender Foundation\Blender\3.4\scripts\addons\text_to_video-main\model

Location

Video Sequence Editor > Sidebar > Text to Video

image

Text to Audio

Currently, not working. Produces a file without any content.

Modules

Diffusers: ModelScope: https://modelscope.cn/models/damo/text-to-video-synthesis/summary AudioLDM: https://github.com/haoheliu/AudioLDM If some additional python modules are missing, write about it here(so I can add them), and use the Blender PIP add-on to manually install the missing modules: https://github.com/amb/blender_pip