Version History:
Beta 1: Initial Release
LazyEye-XTC Beta
The goal for this merge was to have a stable and clean model for lora training with good knowledge about personalities
and also being able to produce natural real-life like output for generation.
This model was tuned to be used with DMD2 Lora, CFG=1 at 12 Steps,
but it should work with common basic settings like Euler-A, CFG:7, 25 Steps.
If you use Character-Loras with DMD2 lower the Character-Lora weight to about 0.65.
If you want to use DMD2 with Euler-A or other unifrom schedulers, lower the DMD2 Lora weight (i.e. 0.75).
To improve further i need your feedback. Please test and report.
Thank you.
Additional notes:
For example images i simply used the positive prompt of some of the most popular SDXL gens on this site from this year.
Just the raw model without Loras, Embeddings or negative prompt. Yes i am a lazy fart.
Description
Beta 2: fp16 download
Beta 3: bf16 download
Comments (2)
Came here to give a thumbs up for Beta 1 even though only halfway through testing, saw that it had been made 'Beta 1: Obsolete'. Started testing Beta 2, initial impressions. Beta 1 for my purposes is superior, sharper, better faces, expressions with no enhancements in testing. Beta 2 is weighted into NSFW and has much more aberrations with hands,feet, and eyes. It needs more steps than Beta 1. Not saying at the end of testing that Beta 2 wont get a thumbs up, it will, it needs tuning , just the original version seems like a gem, as said only 50% testing done on it but it is definitely looking very promising, at least in my testing with forge.
Looking forward to the next iteration
Pics are from Beta 1
Removed Beta 2 because adding more NSFW broke everything the base had. Need to gain more skills to improve this further. Thanks for your feedback.
Looks like we don't have an active mirror for this file right now.
CivArchive is a community-maintained index — we catalog mirrors that volunteers upload to HuggingFace, torrents, and other public hosts. Looks like no one has uploaded a copy of this file yet.
Some files do get recovered over time through contributions. If you're looking for this one, feel free to ask in Discord, or help preserve it if you have a copy.