The model is for testing only, ha. The picture results are average. If you like it, please give it a lot of praise. Contact: jmai.art

prompt examples:

historical fantasy, ultra high res, 1girl, aesthetic, tearful face, refined and daunting, heavy duty, night view, rain,

reverse prompt examples are as follows:

worst quality, low quality, cgi, bad eye, worst eye, illustration, cartoon, deformed, distorted, disorganized, drawn, bad anatomy, long anatomy, bad hand,

other parameters: step:

30+

cfg: 6.5-8

sampling

method:

euler

a or dpm++ sde karras, etc.

The 9th version is ga, and the first eight are all test versions. Actually,

the t8 version is good, so I kept 1.

More sensitive to some parameters of photography; 2.

Optimized light; 3.

It is more sensitive to tags, so some conflicting tags will be less effective, but good tags will be more effective; 4.

If you want a different style, it is recommended to use Lora, which will be more stable; of course, some keywords can also work, but they may not achieve your expected results; 5.

Don't do bad things with the generated images; 6.

Don't do bad things with the generated images; 7.

Don't do bad things with the generated images; notes: It is recommended to write some tags for the camera that took the picture, which will have different effects. Also, the general effect

for style tags is as follows

✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️

✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️

🔮

Author

✳️ Thank You!

✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️

✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️✳️