SD3 - Art Poster Composition Design LoRA-CADS-AiARTiST

AiARTiST Art Composition SD V3 Version LoRA Art Poster Layout Design Model

What is SD v* *oster layout design LoRA?

Following the positive feedback on the previously released 1.5 version "AiARTiST Art Poster Layout Design Model", after half a year, we finally have the "fragmented" SD3 m2b open-source version. To keep the community celebration going and bring back broken things, we started participating in the SD3 fine-tuning competition. Currently, this version is my first SD3 LoRA, used to test the training effect of SD3 Diffusers, hoping to inspire the community to progress together.

Since June 12, when the community released v3 weights, we have preemptively released the workflow and model files.

SD3-CADS Workflow  https://www.liblib.art/modelinfo/09ca04ae8dfe41b887fdaf1580bd93a4

It now supports text-to-image, image-to-image, text-to-video, and v3 LoRA loading. Although we are not perfect, we will try our best to be down-to-earth, minimize the impact of device, VRAM, and computing power limitations, and allow everyone to experience the latest v3 large models running smoothly on their machines as soon as possible.

Subsequently, alongside the release of many SD3 example images, we also released LoRA, which simulates SD3 image quality and style for XL models.

AiARTiST-SD3 Simulator-XL https://www.liblib.art/modelinfo/7449642bdf39443aa66341837025c690

Used to compare and evaluate the image quality difference between XL and SDv3 models, this serves to accelerate understanding of the new features of the SD3 version and hopefully sustain the vitality of the XL version.

----------------------------------------------------------------------

Currently, this LoRA is just a test version with limited image quality and generalization performance!

Rank 128 / lr 0.0001 / trained for 10,000 steps (6000-step version released)

Recommended base model: sd3_medium_incl_clips 5.6GB version (without T5)

Recommended weights: 0.6~0.8  CFG: 3~4.5

Recommended sampler: DPM++2m Scheduler: sgm_uniform

Recommended steps: 28~30

Recommended workflow:SD3-CADS-LoRA Workflow Link provided above

Recommended sizes: 768 x 1536 / 640 x 1280 / 576 x 1024

----------------------------------------------------------------------

Lib members can enjoy online acceleration for easy use with the CADS LoRA series.

Model list: https://www.liblib.art/search?keyword=AiARTiST

55% Free Models, 95% Open Source Models

40% Member Models, 5% Closed Source Models

Commercial models:

AiARTiST-CADS2.0 XL Commercial Advertising Assistance System (Enterprise Customized)

https://www.liblib.art/modelinfo/a56ebacdba7d4e30b97bb124bc3fc28f <<Click Here

----------------------------------------------------------------------

Model usage disclaimer:

1. You are not allowed to host this model and its derivative versions (such as fusion model versions) on websites/apps designed to earn revenue or donations.

2. You cannot directly sell this model and its derivative versions (such as fusion model versions) unless you have sufficiently modified the model to be considered your personal work in a legal sense. If you violate this rule, you will be solely responsible for any legal consequences, and I shall not be liable.

3. You cannot deliberately use the model to create or share illegal or harmful content. Please use this model for positive and constructive purposes, respecting ethical norms and moral standards.

----------------------------------------------------------------------

For testing questions, leave a message. For business cooperation, visit personal homepage +V Zyuan980

AiARTiST | Metaverse China · Shandong Digital Human/AiGC/Metaspace Direction

Serving artists as a tool, contact:https://afdian.net/a/AiARTiST-lib