From ba28f6e5ee2ccffdb27e54e80950d86699f8a8d5 Mon Sep 17 00:00:00 2001 From: Juha Jeronen Date: Fri, 22 Dec 2023 11:47:37 +0200 Subject: [PATCH] update docstring --- talkinghead/tha3/app/manual_poser.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/talkinghead/tha3/app/manual_poser.py b/talkinghead/tha3/app/manual_poser.py index f43e72f..7c8c2f6 100644 --- a/talkinghead/tha3/app/manual_poser.py +++ b/talkinghead/tha3/app/manual_poser.py @@ -13,7 +13,7 @@ There are two motivations: - Much faster than inpainting all 28 expressions manually in Stable Diffusion. Enables agile experimentation on the look of your character, since you only need to produce one new image to change the look. - - No CPU or GPU load while running SillyTavern, unlike the live plugin mode, which is cool, but slow. + - No CPU or GPU load while running SillyTavern, unlike the live plugin mode. For best results for generating the static input image in Stable Diffusion, consider the various vtuber checkpoints available on the internet. These should reduce the amount of work it takes to get SD to render your character in