V7.0所有示例图片都是ComfyUI默认工作流直出没有放大和修正
参数为：
All the example images in V7.0 are ComfyUI's default workflow, and they are not enlarged or corrected
The parameters are:
分辨率：1040 x 1520
采样器：Euler a
Steps: 30, CFG scale: 8,
V6.9不需要质量提示词
V6.9 does not require quality prompts
由于这个模型和我另一个模型的一些版本差别不太大所以决定把这个模型往动画风格靠。
融合了几个模型和自己的Lora。
Since this model is not very different from some versions of my other model, I decided to move this model towards Animation style.
Fuses several models and its own Lora.
DucHaiten-Pony-XL_3.0
PonyKaine_Paint_4.0
Copycat_2.0
SatPony-Niji2.0
出图参数推荐：DPM++ 2M Karras 和 Euler a
Plot parameter recommendation:DPM++ 2M Karras 和 Euler a
步数：30，CFG: 8, 分辨率：832x1216,
高清倍率: 1.5, 高清步数: 15, 高清模型: 4x_NMKD-Superscale-SP_178000_G,
Steps: 30, CFG scale: 8, Size: 832x1216,
Hires upscale: 1.5, Hires steps: 15, Hires upscaler: 4x_NMKD-Superscale-SP_178000_G,
推荐正反提示词：
Recommended positive and negative prompts:
cinematic film still,score_9,score_8_up,score_7_up,masterpiece,best quality,8k,score_6,score_5,score_4,lowres,bad_quality,signature,watermark,monochrome,logo,text,Description
FAQ
Comments (4)
3.0的3D质感最好，但面容固定而且有点糊，和LORA的配合不好。6.0的手绘风格过于明显，6.5有一些平衡。4.0基本上全面被5.0超越。目前来说，个人觉得最好的是5.0
确实，我个人也喜欢5.0，有时候想尽量往动画风格靠，但有时候弄着弄着就跑偏了哈哈哈。
I'd prefer the 6.0 over 6.5. While 6.5 has better interpretation to the prompts, the texture, depth, and lighting are much better in 6.0 in my opinion. I understand this is supposed to be more anime-like, but 6.5 feels a bit flat and the lighting are consistently very bright for some reason. I really like your checkpoints so I hope you'll take it as a positive criticism.
Thank you for liking my model, 6.5 is really too bland, I prefer 6.0 myself,5.0 is also good, there are a lot of ideas when merging models, the model quality is also uneven. -- From translation software