Emanon14 commited on
Commit
81a3efa
·
verified ·
1 Parent(s): f426330

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +18 -19
README.md CHANGED
@@ -12,8 +12,7 @@ tags:
12
  - stable-diffusion-xl
13
  - merge
14
  base_model:
15
- - Raelina/Raehoshi-illust-XL-3
16
- - Laxhar/noobai-XL-Vpred-0.6
17
  ---
18
 
19
  <style>
@@ -38,8 +37,8 @@ base_model:
38
 
39
  ## NOTICE
40
  - This is a v-prediction model.
41
- - Plz read "How to Use the Model" section on [NoobAI-XL model page](https://civitai.com/models/833294).
42
- - Problem that not detected as a v_pred model is fixed.
43
 
44
  <div class="gallery">
45
  <img src="./sample01.webp" alt="sample_v11_1">
@@ -53,31 +52,30 @@ base_model:
53
  - **License**: [Fair AI Public License 1.0-SD](https://freedevproject.org/faipl-1.0-sd/)
54
 
55
  ## Merge Recipe
56
- - Raehoshi illust XL 3.0 (A)
57
- - (https://civitai.com/models/846917/raehoshi-illust-xl)
58
- - RouWei 0.6.1 vpred (B)
59
  - (https://civitai.com/models/950531/rouwei)
60
- - paruparu illustrious v3.2 (C)
61
- - NoobAI-XL Epsilon-Pred 0.75 (D)
62
- - NoobAI-XL Epsilon-Pred 1.1 (E)
63
- - (https://civitai.com/models/833294?modelVersionId=998979)
64
- - SPO-SDXL_4k-p_10ep_LoRA_webui (F)
65
  - (https://civitai.com/models/510261/spo-sdxl4k-p10eplorawebui)
 
 
66
  ```
67
- A + (C - D) * (0.2,0.2,0.3,0.2,0.3,0.3,0.7,0.3,0.2,0.2,0.2,0.2,0.3,0.3,0.2,0.2,0.5,0.3,0.2) = G
68
- B + (G - E) * (0,0.2,0.7,0.3,0.5,0.5,0.7,0.2,0.3,1,0.2,0.2,0.7,0.2,0.2,0,0.7,0.2,0.3) = H
69
- H + F * weight 1.0 = This Model v1.1
70
  ```
71
 
72
  ## Recommended settings
73
 
74
  - **Positive prompts:**
75
  ```
76
- masterpiece, best quality, very aesthetic,
77
  ```
78
  - **Negative prompts:**
79
  ```
80
- lowres, bad quality, worst quality, very displeasing, bad anatomy, sketch, jpeg artifacts, signature, watermark,
81
  ```
82
  - **CFG:** 3-5
83
  - **Sampling steps:** 28
@@ -92,6 +90,7 @@ lowres, bad quality, worst quality, very displeasing, bad anatomy, sketch, jpeg
92
  - **Hires step:** 10-15
93
  - **Denoising:** 0.3-0.4
94
 
95
- ## License
96
 
97
- [Fair AI Public License 1.0-SD](https://freedevproject.org/faipl-1.0-sd/)
 
 
 
12
  - stable-diffusion-xl
13
  - merge
14
  base_model:
15
+ - Minthy/RouWei-0.7
 
16
  ---
17
 
18
  <style>
 
37
 
38
  ## NOTICE
39
  - This is a v-prediction model.
40
+ - No longer includes noob AI.
41
+ - 99% of this checkpoint is made with RouWei.
42
 
43
  <div class="gallery">
44
  <img src="./sample01.webp" alt="sample_v11_1">
 
52
  - **License**: [Fair AI Public License 1.0-SD](https://freedevproject.org/faipl-1.0-sd/)
53
 
54
  ## Merge Recipe
55
+ - RouWei 0.7 vpred (A)
 
 
56
  - (https://civitai.com/models/950531/rouwei)
57
+ - CyberRealistic XL (B)
58
+ - (https://civitai.com/models/312530/cyberrealistic-xl)
59
+ - stable-diffusion-xl-base-1.0 (C)
60
+ - (https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0)
61
+ - SPO-SDXL_4k-p_10ep_LoRA_webui (D)
62
  - (https://civitai.com/models/510261/spo-sdxl4k-p10eplorawebui)
63
+ - My LoRA (E)
64
+ - Trained by 300 images AI-generated
65
  ```
66
+ A + (B - C) * 0.5 = F
67
+ F + D + E * 0.5 = This Model v2.0
 
68
  ```
69
 
70
  ## Recommended settings
71
 
72
  - **Positive prompts:**
73
  ```
74
+ masterpiece, best quality
75
  ```
76
  - **Negative prompts:**
77
  ```
78
+ worst quality, low quality, watermark
79
  ```
80
  - **CFG:** 3-5
81
  - **Sampling steps:** 28
 
90
  - **Hires step:** 10-15
91
  - **Denoising:** 0.3-0.4
92
 
 
93
 
94
+ # Thanks
95
+ - First of all, to Minthy for RouWei. That has the best prompt following I know of!
96
+ - And, of course, thanks for the Animagine. Animagine was the first signpost that led me to use SDXL, and it continues to do so!