This is a finetuned Illustrious-XL 2.0 checkpoint. It’s designed to be a reliable starting point for those interested in model merging or finetuning. The base style is a neutral anime look with no strong artistic bias.The finetune improves the anatomy and overall stability of the model compared to the original model.This model itself can already make quite good looking pictures, but it is primerly intended to be used as a base for merges or lora testing.<h3 id="update">Update</h3>Version 3.0 I did not originally plan to release this experimental version, but since this model works quite well, I decided to upload it anyway. This model uses similar optimizations then V2, but made some tweaks and added some new to the merge. Not sure if it is better then the previous build, but it is defiantly worth a try.<span style="background-color:rgb(26, 27, 30);color:rgb(193, 194, 197);font-family:-apple-system, BlinkMacSystemFont, &quot;Segoe UI&quot;, Roboto, Helvetica, Arial, sans-serif, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;;font-size:16px">Version 2.0 <span style="background-color:rgb(26, 27, 30);color:rgb(193, 194, 197);font-family:-apple-system, BlinkMacSystemFont, &quot;Segoe UI&quot;, Roboto, Helvetica, Arial, sans-serif, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;;font-size:16px">For this release, I applied a different finetuning to the model and works in my opinion better then V1. But that does not mean it is better in every way. Please check out for yourself which model version suits better for your tastes.Settings:You can use the usual settings like DPM2M++ SDE with Karras, CFG 4-5 or Euler A with SGM Uniform.

Illustrious-XL 2.0 finetuned

Version 2.0 For this release, I applied a different finetuning to the model and works in my opinion better then V1. But that does not mean it is better in every way. Please check out for yourself which model version suits better for your tastes.

ZSA_ILXL2_finetuned V2.fp16.safetensors

v2.0

v3.0

v1.0

creator: ZessicaThis is a finetuned Illustrious-XL 2.0 checkpoint. It’s designed to be a reliable starting point for those interested in model merging or finetuning. The base style is a neutral anime look with no strong artistic bias.The finetune improves the anatomy and overall stability of the model compared to the original model.This model itself can already make quite good looking pictures, but it is primerly intended to be used as a base for merges or lora testing.UpdateVersion 2.0 For this release, I applied a different finetuning to the model and works in my opinion better then V1. But that does not mean it is better in every way. Please check out for yourself which model version suits better for your tastes.Settings:You can use the usual settings like DPM2M++ SDE with Karras, CFG 4-5 or Euler A with SGM Uniform.

Files with SHA256: a47324c877bca2c38457d612d4b8265d8174881c1715c11c9dbe85c21b057961