Magic Wonderland - Fantasy Art Design - Unmanned beauty is enhanced by humans - Artistic creation - Tencent Huan Yuan

Dandan staring blankly: Magic Space- Fantasy Art Design- Unmanned Beauty Pictures are Better with People- Artistic Conception Creation- Tencent Hunyuan DiT_v1.2

Model Introduction: More suitable for creating original materials for Chinese ancient style images, novel promotions, artistic illustrations, and graphic videos.

Creative Overview: Dandan has a strong interest in creating with positive prompts related to magic.

Model Usage Instructions: The checkpoint membrane is based on the Tencent Hunyuan iT_v1.2 large model training.

When testing images with this small Lora model, other second-dimensional large models, third-dimensional large models, and realistic human-like large models were also tested, and the ink effect was very good.

Prompt Word Usage Notes: Trigger words are dandan, magic

Other related parameter settings suggestions:

1. Sampling method: Recommended DPM++2M Karras, Eula, Eula A, SDE, Iteration steps: 20-30, are sufficient. Other sampling methods can also try LCM, Iteration steps can be ≤15, and so on.

2. Enable High-Resolution Repair: For realistic large models, recommend using 8x-NMKD-Superscale, 4x-UltraSharp enlargement algorithm; For anime, 3D large models, you can use R-ESRGAN_4x+ Anime6B, R-ESRGAN_4x+ and other enlargement algorithms.

3. Test image resolution sizes: 1:1 or 9:16 or 16:9 or 3:4 or 4:3, 512*512, 1024*1024; or 512*768; or 768*1024, 600*800, etc.

4. It is recommended to enable: If drawing human characters, turn on the After Detailer face retouch plug-in, face_yolov8m, face_yolov8n, face_yolov8n_v2, or multi-unit, open hand parts like hand_yolov8n, hand_yolov8s, etc.

5. Characters: Mainly trained 1 Chinese Boy, 1 Chinese Girl. However, mature or middle-aged women, men, elderly, children have not been fully trained, there may be unexpected results.

Anyway, before opening Pandora's Box, Schrödinger's cat, always half surprise and half horror natural probability. Whatever image it generates, I can accept it~~~~

Welcome all creators to use, provide feedback, and have friendly exchanges.