v2.4.0 - LCM
Now that the AI Horde natively supports LCM and multiple lora versions, Lucid Creations has likewise been updated to handle it!
This means you can select the one of the LCM loras and generate images with just 3-8 steps! Make sure you select the correct lora version for the sampler you’re using!
My suggestion is to always use the LCM&Turbomix lora and the AlbedoXL model which has been recently added to the AI Horde. Adjust your steps based on your sampler.
- lcm: 3-5 steps
- euler_a/dpm++ sde: 6-12 steps
Don’t forget to turn your cfg way down to 1-2 as well!
Enjoy and let me know how it works our for you!
I haven’t posted a devlog for a while here so in case you didn’t realize, the last two versions also added support for SDXL loras and worker selection (in case you want to target specific workers with your generations)
Also reminder to follow us on Mastodon or Lemmy if you haven’t already.
Files
Get Lucid Creations - Stable Diffusion GUI without GPU
Lucid Creations - Stable Diffusion GUI without GPU
A free, open source, crowdsourced GUI.
Status | In development |
Category | Tool |
Author | db0 |
Tags | ai, Godot, gpl, Open Source, stable-diffusion |
Languages | English |
More posts
- 2.6.x - QR CodesMay 26, 2024
- v2.5.0 - Stable CascadeMar 11, 2024
- v2.1.0 - Textual Inversions!Aug 31, 2023
- v2.0.0 - New UI!Aug 31, 2023
- 1.23.0 - Kudos Display, Clip Skip, SDXL BackgroundsJul 15, 2023
- Stable Diffusion XL on Lucid Creations!Jul 08, 2023
- 1.21.0 - Load generation settings from previous generationJun 08, 2023
- Using LoRas has never been easierJun 01, 2023
Leave a comment
Log in with itch.io to leave a comment.