<h2 id="heading-117">ModelDescription --------------------------</h2>[1] Lora, a minimalist flat-style Chinese style illustration, will add gold edges and noise effects to the picture, and will change the shape of mountains and plants;[2] Pure scenes are more likely to produce effects, and the character card draw rate is higher than that of scenes.<h2 id="heading-118">About Version &amp; Training Set --------------------------</h2>[1] Training base model: helloflatart[2] Trigger word: guochao[3] Training set: 42 images | Number of steps: 3780<h2 id="heading-119">Checkpoint matching suggestions ----------------------</h2>[1] More suitable for flat illustration checkpoint: hello series (helloFlatCute2D, helloFlatArt, helloyoung2D) Flat Anime-mix Aflatcolor Aflatcolor L-Flat-Fourteenthversion Flat national trend *The above base models are only some recommendations, you can also try them according to your own preferences~<h2 id="heading-120">Lora weight suggestions --------------------------</h2>[1] Architectural scene: It is recommended to be between 0.4-0.8; (use 0.8 or above with caution)[2] Persona: Recommended between 0.3-0.7. * Different base model adaptation weights will be slightly different, you can try starting with 0.6; it is recommended to use ADetailer to draw characters<h2 id="heading-121">XYZ Test Chart --------------------------</h2><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/8c4fdf73-bd0a-4ea6-ae64-5f00f2eaf869/width=525/8c4fdf73-bd0a-4ea6-ae64-5f00f2eaf869.jpeg" /><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/e4b4c7e6-df2e-4fb3-9cb2-8ccd28f53693/width=525/e4b4c7e6-df2e-4fb3-9cb2-8ccd28f53693.jpeg" /><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/17109810-bd10-4a9e-937f-27a45004c479/width=525/17109810-bd10-4a9e-937f-27a45004c479.jpeg" />

Files with SHA256: 37bebf38a05616df63f7845cb6c73e2a73b8b62b830e28a996c3a4c473691699