Add inference script

[JCBrouwer]

Just a simple script to sample from a model.

I'd like to add video init (vid2vid) and watermark removal later on as well.

[ExponentialML]

Thanks! I have one locally, but I'll merge this one. Just let me know when it's ready 👍 .

[JCBrouwer]

@ExponentialML alright, I've added watermark inpainting and video init functionality to the inference script, take a look and let me know what you think!

Here's some examples, the first generated with "a beautiful waterfall" and inpainted. The rest use that video as initialization with varying init_weight and prompt "cosmic nebula":

a.beautiful.waterfall.22c39ce5.mp4

init

0.6.a.beautiful.waterfall.22c39ce5.cosmic.nebula.885d144e.mp4

0.6

0.5.a.beautiful.waterfall.22c39ce5.cosmic.nebula.e6a9d554.mp4

0.5

0.4.a.beautiful.waterfall.22c39ce5.cosmic.nebula.222d1cbd.mp4

0.4

0.3.a.beautiful.waterfall.22c39ce5.cosmic.nebula.9e574cbb.mp4

0.3

0.2.a.beautiful.waterfall.22c39ce5.cosmic.nebula.de2d5cc4.mp4

0.2

0.1.a.beautiful.waterfall.22c39ce5.cosmic.nebula.f70a6e3e.mp4

0.1

[ExponentialML]

Excellent! I really appreciate your contribution 👍 .

[ExponentialML]

@JCBrouwer Forgot to add this before merging, but it's fine 🙂 .

Could you add a section to the end of the README of the preferred ways using the inference script?