Update README.md
Browse files
README.md
CHANGED
@@ -13,7 +13,7 @@ tags:
|
|
13 |
|
14 |
It has better performance in txt2img when using positive prompts describing what you want without any quality tags, and you can use `EasyNegative` as negative prompt.
|
15 |
|
16 |
-
`
|
17 |
|
18 |
## How to Use
|
19 |
|
@@ -37,7 +37,7 @@ From Clam:
|
|
37 |
|
38 |
<img src="https://huggingface.co/Johnhex/Clam/resolve/main/Images/Clam-1-1.png" alt="Result From Clam" style="zoom: 33%;" />
|
39 |
|
40 |
-
From
|
41 |
|
42 |
<img src="https://huggingface.co/Johnhex/Clam/resolve/main/Images/Clam-2-1.png" alt="Result From Clam" style="zoom: 33%;" />
|
43 |
|
@@ -55,7 +55,7 @@ From Clam:
|
|
55 |
|
56 |
<img src="https://huggingface.co/Johnhex/Clam/resolve/main/Images/Clam-1-2.png" style="zoom:33%;" />
|
57 |
|
58 |
-
From
|
59 |
|
60 |
<img src="https://huggingface.co/Johnhex/Clam/resolve/main/Images/Clam-2-2.png" style="zoom:33%;" />
|
61 |
|
@@ -73,7 +73,7 @@ From Clam:
|
|
73 |
|
74 |
<img src="https://huggingface.co/Johnhex/Clam/resolve/main/Images/Clam-1-3.png" style="zoom:33%;" />
|
75 |
|
76 |
-
From
|
77 |
|
78 |
<img src="https://huggingface.co/Johnhex/Clam/resolve/main/Images/Clam-2-3.png" style="zoom:33%;" />
|
79 |
|
@@ -87,37 +87,7 @@ Steps: 26, Sampler: DPM++ 2S a Karras, CFG scale: 7, Seed: 2773744106, Size: 512
|
|
87 |
|
88 |
|
89 |
|
90 |
-
## Clam
|
91 |
|
92 |
-
### Use Models
|
93 |
-
|
94 |
-
Sorry that I forgot to record using models' weight.
|
95 |
-
|
96 |
-
- Color Box Model [a80fee84]
|
97 |
-
- Ligne Claire Anime Diffusion V1 [ee5b91da]
|
98 |
-
- Anything-V3.0-pruned [2700c435]
|
99 |
-
- tmndMix-tmndMixPlus [65b8d0b2]
|
100 |
-
- momoko-e [a2a802b2]
|
101 |
-
|
102 |
-
|
103 |
-
|
104 |
-
## Clam2
|
105 |
-
|
106 |
-
I remerge models from above, and add some another models that I think they may generate more detailed results, but it really destroyed the stability of `Clam`.
|
107 |
-
|
108 |
-
### Use Models
|
109 |
-
|
110 |
-
| Models | Weights |
|
111 |
-
| ------------------------------------- | ---------- |
|
112 |
-
| NyanMix_230303Absurd2 [466a84a7] | 0.1152 |
|
113 |
-
| Anything-V3.0-pruned [2700c435] | 0.08257536 |
|
114 |
-
| EimisAnimeDiffusion_1-0v [4f828a15] | 0.0589824 |
|
115 |
-
| LigneClaireAnimeDiffusion [ee5b91da] | 0.2029824 |
|
116 |
-
| pastelmix-better-vae [260de5e8] | 0.18 |
|
117 |
-
| momoko-e [a2a802b2] | 0.1 |
|
118 |
-
| IripasteDiffusionSD1_lines [8aa756f3] | 0.073728 |
|
119 |
-
| colorBoxModel_colorBOX [a80fee84] | 0.0589824 |
|
120 |
-
| TmndMix_TmndMixPlus [65b8d0b2] | 0.12754944 |
|
121 |
|
122 |
|
123 |
|
|
|
13 |
|
14 |
It has better performance in txt2img when using positive prompts describing what you want without any quality tags, and you can use `EasyNegative` as negative prompt.
|
15 |
|
16 |
+
`Clam1.1` has more detailed background but it is less stable than `Clam`, so I recommend to try to use `Clam` first.
|
17 |
|
18 |
## How to Use
|
19 |
|
|
|
37 |
|
38 |
<img src="https://huggingface.co/Johnhex/Clam/resolve/main/Images/Clam-1-1.png" alt="Result From Clam" style="zoom: 33%;" />
|
39 |
|
40 |
+
From Clam11:
|
41 |
|
42 |
<img src="https://huggingface.co/Johnhex/Clam/resolve/main/Images/Clam-2-1.png" alt="Result From Clam" style="zoom: 33%;" />
|
43 |
|
|
|
55 |
|
56 |
<img src="https://huggingface.co/Johnhex/Clam/resolve/main/Images/Clam-1-2.png" style="zoom:33%;" />
|
57 |
|
58 |
+
From Clam1.1:
|
59 |
|
60 |
<img src="https://huggingface.co/Johnhex/Clam/resolve/main/Images/Clam-2-2.png" style="zoom:33%;" />
|
61 |
|
|
|
73 |
|
74 |
<img src="https://huggingface.co/Johnhex/Clam/resolve/main/Images/Clam-1-3.png" style="zoom:33%;" />
|
75 |
|
76 |
+
From Clam1.1:
|
77 |
|
78 |
<img src="https://huggingface.co/Johnhex/Clam/resolve/main/Images/Clam-2-3.png" style="zoom:33%;" />
|
79 |
|
|
|
87 |
|
88 |
|
89 |
|
|
|
90 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
91 |
|
92 |
|
93 |
|