RedRayz commited on
Commit
efa0eb2
·
verified ·
1 Parent(s): 77751aa

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +57 -5
README.md CHANGED
@@ -1,5 +1,57 @@
1
- ---
2
- license: other
3
- license_name: faipl-1.0-sd
4
- license_link: https://freedevproject.org/faipl-1.0-sd/
5
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ license_name: faipl-1.0-sd
4
+ license_link: https://freedevproject.org/faipl-1.0-sd/
5
+ pipeline_tag: text-to-image
6
+ base_model:
7
+ - Laxhar/noobai-XL-Vpred-1.0
8
+ tags:
9
+ - stable-diffusion
10
+ - stable-diffusion-xl
11
+ ---
12
+
13
+ # Abydos_Noob_v-pred-1.1.0
14
+
15
+ Modified NoobAI-XL(v-prediction) with Blue Archive style
16
+
17
+ [Civitai model page](https://civitai.com/models/923120)
18
+
19
+ ## About 1.1.0
20
+ Improved stability
21
+ Better shadow rendering?
22
+
23
+ ## Recommended Prompt
24
+ `official art, blue archive, game cg, non-web source, newest`
25
+
26
+ ## Recommended Negative Prompt
27
+ `lowres, oldest`
28
+
29
+
30
+ ## Recommended Settings
31
+ Steps: 14-24
32
+
33
+ Sampler: DPM++ 2M SDE, DPM++ 3M SDE or Euler a
34
+
35
+ Scheduler: Simple or SGM Uniform
36
+
37
+ Guidance Scale: 2-5
38
+
39
+ ### Hires.fix
40
+ Upscaler: 4x-UltraSharp or Latent(nearest-exact)
41
+
42
+ Denoising strength: 0.4(0.6-0.7 for latent)
43
+
44
+ ## Training steps
45
+ 1. Make 2 models from NoobAI-XL v-pred 1.0(=A,B), A with ZTSNR, B w/o ZTSNR
46
+ 2. Merge A and B MBW(0,0,0,0,0,0.3,0.3,0,0.5,0.5,0.5,0.5,0.5,0.5,0.3,0.3,0,0,0,0) Adjust(0,0,0,0,-0.1,0,0,0)=tmp1
47
+ 3. tmp1 + spo_sdxl_10ep_4k-data_lora_webui x 1 + sdxl-boldline x -0.5 = Result
48
+
49
+ ## Training scripts:
50
+ [sd-scripts](https://github.com/kohya-ss/sd-scripts)
51
+
52
+ ## Notice
53
+ This model is licensed under [Fair AI Public License 1.0-SD](https://freedevproject.org/faipl-1.0-sd/)
54
+
55
+ If you make modify this model, you must share both your changes and the original license.
56
+
57
+ You are prohibited from monetizing any close-sourced fine-tuned / merged model, which disallows the public from accessing the model's source code / weights and its usages.