The Collage Atlas Influence

LORA
Original


Updated:

I am admittedly kind of a noob at this, but if you enjoy my humble work, any donations are greatly appreciated: https://ko-fi.com/insult_to_ninjary

Your generosity helps me stay motivated to keep improving and to keep trying new things!

At a Glance:

  • Base model: Illustrated (but you can experiment with other kinds!)

  • Use a VAE?: Yes (baked or your choice)

  • Clip skip: 1 or 2 (experiment with both!)

  • LoRA weight: 1 - 0.7 (go lower if your images appear too noisy/busy/garbled)

  • Hires. fix: use non-default upscaler; denoising strength 0.5 - 0.4

Trigger Words:

  • monochrome - this is necessary to ensure that you get a black and white image

  • You may need to weight the trigger word with numerical values or parentheses if the effect is not strong enough. I generally put the trigger word inside 3 sets of parentheses, like so: (((monochrome))).

Description:

This LoRA is intended to create images in the style of former Apple Arcade game, and current Steam game, The Collage Atlas.

Honestly, I don't think it came out too great. I don't think SD1.5 is capable of truly capturing the intricate detail and nuance of the original art style, but back when I originally made this, SD1.5 was the only show in town. It might turn out better retraining it now with one of the newer base models, but I'm on a mission to post my backlog of 1.5 models no matter what, so for better or for worse, here this is.

Troubleshooting:

  • It tends to get a little messy with generating extraneous stuff, especially in the background. This seems to be exacerbated by the fact that this is a black and white model, so things can often look really busy and garbled. Lowering the weight of the LoRA can help with this, but can also sometimes cause images to lose too much of the style. Just stay patient and experiment with different settings.

  • The LoRA also struggles with anatomy at times, most notably hands and fingers (surprise, surprise!). Just try your best with all the usual fixes - different prompting, try a different base model, use a VAE, try different clip skip values, lower the LoRA weight, use adetailer, etc.

Outro:

If you make anything awesome, please consider posting them to the gallery! That goes for all of my LoRAs! I really like seeing what kind of stuff you all are able to make!

If you have any questions, feel free to ask! I will try my best to help! Also, if you have any ideas for future LoRAs, please don't hesitate to share them with me!

Version Detail

SD 1.5
The OG version

Project Permissions

    Use Permissions

  • Use in TENSOR Online

  • As a online training base model on TENSOR

  • Use without crediting me

  • Share merges of this model

  • Use different permissions on merges

    Commercial Use

  • Sell generated contents

  • Use on generation services

  • Sell this model or merges

Related Posts

Describe the image you want to generate, then press Enter to send.