Update README.md
Browse files
README.md
CHANGED
@@ -12,8 +12,7 @@ Buruma(not bloomers) is a world heritage.
|
|
12 |
|
13 |
## Updates
|
14 |
|
15 |
-
- **
|
16 |
-
- **buruma20renew_test-230323**: Test for renewal 2.0 version.Activeted taraining tags.
|
17 |
|
18 |
|
19 |
## Features
|
@@ -26,52 +25,41 @@ It will give us highly-detailed buruma images.
|
|
26 |
This model is an experimental. It will be improved for your support. see Donations
|
27 |
## How to Use
|
28 |
|
29 |
-
Recommended LoRA weight is 0.3.
|
30 |
```prompt example
|
31 |
-
<lora:
|
32 |
```
|
33 |
```prompt example
|
34 |
Negative Prompt: monochrome
|
35 |
```
|
36 |
**Notes:** This model is **Double Trigger**.Both of "buruma" and "gym_uniform" are nessesary.
|
37 |
|
38 |
-
|
39 |
-
```prompt example
|
40 |
-
red_buruma, gym_uniform
|
41 |
-
```
|
42 |
-
But it's hard to draw.Please wait a verious clolored version.
|
43 |
-
|
44 |
-
Clothes Lifting looks good...
|
45 |
-
```prompt example
|
46 |
-
buruma, gym_uniform, clothes_lift, nipples
|
47 |
-
```
|
48 |
-
### LoRA Block Weight Plugin Tequinic
|
49 |
|
50 |
-
|
51 |
-
So these are can be(but a little bit hard.I can't
|
52 |
- Wearing another upper wear
|
53 |
-
- Getting more huge breasts
|
54 |
-
- Coexsisting with another LoRAs
|
55 |
|
56 |
```prompt example
|
57 |
-
<lora:
|
58 |
```
|
59 |
-
|
60 |
|
61 |
### Recommended SD models
|
62 |
- Anything V3(I had confirmed good outputs)
|
63 |
- BloodOrangeMix(I love it) https://huggingface.co/WarriorMama777/OrangeMixs
|
64 |
- **Realistic models also available**
|
65 |
|
66 |
-
### LoRA Training Parameters
|
67 |
<!-- Provide a longer summary of what this model is. -->
|
68 |
- **Max Resolution:** 512,512
|
69 |
-
- **
|
70 |
-
- **
|
71 |
-
- **
|
72 |
-
- **
|
|
|
73 |
- **Training Based Model:** ACertainty https://huggingface.co/JosephusCheung/ACertainty
|
74 |
-
- **Source Pictures:**
|
75 |
- **Calculating Graphic Card:** RTX2060(6GB) It's so hard!
|
76 |
|
77 |
## Term of Use
|
|
|
12 |
|
13 |
## Updates
|
14 |
|
15 |
+
- **TrueBurumaV20**: Version 2.0 Latest
|
|
|
16 |
|
17 |
|
18 |
## Features
|
|
|
25 |
This model is an experimental. It will be improved for your support. see Donations
|
26 |
## How to Use
|
27 |
|
28 |
+
Recommended LoRA weight is 0.2 or 0.3.
|
29 |
```prompt example
|
30 |
+
<lora:TrueBurumaV20:0.3>, buruma, gym_uniform
|
31 |
```
|
32 |
```prompt example
|
33 |
Negative Prompt: monochrome
|
34 |
```
|
35 |
**Notes:** This model is **Double Trigger**.Both of "buruma" and "gym_uniform" are nessesary.
|
36 |
|
37 |
+
### LoRA Block Weight Plugin Tequinic(V2.0)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
38 |
|
39 |
+
Decrease **OUT04, OUT05**, you can reduce the upper_body's effect.
|
40 |
+
So these are can be(but a little bit hard.I can't help it)
|
41 |
- Wearing another upper wear
|
|
|
|
|
42 |
|
43 |
```prompt example
|
44 |
+
<lora:TrueBurumaV20:0.3:1,1,1,1,1,1,1,1,1,0,0.5,1,1,1,1,1,1>
|
45 |
```
|
46 |
+
|
47 |
|
48 |
### Recommended SD models
|
49 |
- Anything V3(I had confirmed good outputs)
|
50 |
- BloodOrangeMix(I love it) https://huggingface.co/WarriorMama777/OrangeMixs
|
51 |
- **Realistic models also available**
|
52 |
|
53 |
+
### LoRA Training Parameters(V2.0)
|
54 |
<!-- Provide a longer summary of what this model is. -->
|
55 |
- **Max Resolution:** 512,512
|
56 |
+
- **Optimizer:** AdamW
|
57 |
+
- **Dimeisions:** 9
|
58 |
+
- **Network Alpha:** 4
|
59 |
+
- **Learning Rate:** 5e-6
|
60 |
+
- **Epoch:** 16
|
61 |
- **Training Based Model:** ACertainty https://huggingface.co/JosephusCheung/ACertainty
|
62 |
+
- **Source Pictures:** 126 max
|
63 |
- **Calculating Graphic Card:** RTX2060(6GB) It's so hard!
|
64 |
|
65 |
## Term of Use
|