home models images videos posts articles bounties challenges events updates shop

Noisy_Element_XL

Name: Noisy_Element_XL
Rating: 5 (40 reviews)
Author: Monet_Einsley

561

Updated: Jan 5, 2025

base model

anime victorian sdxl 1,0 lcm

Download

1 variant available

fp32 SafeTensor

Full precision, largest file • 6.62 GB

Verified: 2 years ago

Download (6.62 GB)

Details

Type

Checkpoint Trained

Stats

313

Reviews

Positive

(20)

Published

Mar 21, 2024

Base Model

SDXL 1.0 LCM

Hash

AutoV2

97AABC70C2

Trigger Words

all fours

seiza

wariza

squatting

kneeling

high kick

flying kick

Monet

real life

anime

+6 more

About this version

This checkpoint came about as a result of me messing with settings I don't fully understand lol. I liked the character style, and wanted to tweak it further to get a certain style, but since it got praise for how it is, I'm keeping it as is. Was intended to be a version 2 for VAXL, but since it doesn't quite meet that criteria, I'm calling it version 2.5 for Noisy Element XL.

In short this is a VAXL prototype fused with all my SDXL LoRAs, as well as the LCM LoRA. As such, it inherits all the trigger words from all my current LoRAs: Victorian Anime Art XL, STFD, and Realistic V (a prototype LoRA I tried to make to get realistic styles as well as good hands/feet).

Note:

Best to use the hand/foot focus triggers for Adetailer prompts... unless you want a closeup of hands, or a foot-focused image lol

If you don't adequately prompt in a background, it will default to a plain white background, or a drab grey wall. This is the issue I was trying to remedy. Perhaps using some sort of scenery-based LoRA could fix that, but I've yet to test it.

This, like other Noisy Element checkpoints, is an LCM model, so there's no guarantee you'll get good results using non-LCM samplers. No VAE required, no LCM LoRA required, as both are baked into this checkpoint.

The trigger words are completely optional. It's recommended that you don't use any of my current LoRAs in conjunction with this checkpoint, since all of them are already merged in.

default creator card background decoration

Monet_Einsley

License:

CreativeML Open RAIL++-M

Spent a literal week training this model locally, since I had great difficulty training it online.

This was trained on 500+ images in 4 specific categories: full body, close up, Victorian, and Monet (the painter, not me lol). Therefore, this checkpoint technically has 4 triggers, just like a LoRA.

This model has the SDXL LCM LoRA and SDXL VAE baked into it, therefore there's no need to use external VAEs or LCM LoRA to get a good result. I'd intended to prune this model, but I'm not sure if it actually did prune it. (To be fair, I'm not sure what a pruned SDXL checkpoint would look like; I simply tried it to see if I couldn't reduce the file size.)

If you're doing a close-up portrait of a face, or simply using ADetailer to refine faces, it's highly recommended to use the "close up" trigger prompt, though it can perform without the trigger word.

I'd had rough experiences with SDXL in the past whenever I tried making images with full body shots, thus I tried to rectify it here. The prompt "full body" is a trigger word which should facilitate the full body landscape shots I'm fond of.

Recommended Usage:

Set the sampler to LCM, reduce the CFG to 1-2, reduce the steps to 8-16, and then generate at will.

If you insist on using non-LCM samplers, then be sure to increase the steps to about 30. But it is of my opinion that this model has worse artifacts than the SD1.5 version when using non-LCM samplers, so be aware of that.