Sentinel7 commited on
Commit
db5df3b
·
verified ·
1 Parent(s): 160d388

Upload 2058077/2328988/README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. 2058077/2328988/README.md +71 -4
2058077/2328988/README.md CHANGED
@@ -1,11 +1,78 @@
1
-
2
  ---
3
  license: other
4
  tags:
5
- - civitai
 
6
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7
  Author: [fok3827](https://civitai.com/user/fok3827)
8
 
9
- Model: [https://civitai.com/models/2058077?modelVersionId=2328988](https://civitai.com/models/2058077?modelVersionId=2328988)
 
 
10
 
11
- Mirror: [https://civarchive.com/models/2058077?modelVersionId=2328988](https://civarchive.com/models/2058077?modelVersionId=2328988)
 
 
1
  ---
2
  license: other
3
  tags:
4
+ - base model
5
+ - nsfw
6
  ---
7
+
8
+ # Qwen-image_NSFW_Adv.1 - v1.0 - 2328988
9
+
10
+ **Model Type**: LORA
11
+
12
+ **Base Model**: Qwen
13
+
14
+ **Trigger Words**: None
15
+
16
+ **Tags**: base model, nsfw
17
+
18
+ ## Gallery
19
+
20
+ <table>
21
+ <tr>
22
+ <td><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/15f13c31-81a9-4837-95e3-e62f636458d8/original=true/106745065.jpeg" width="200"></td>
23
+ <td><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/e586c6c2-4da6-4dc3-a8d9-8a3ca2b45121/original=true/106745086.jpeg" width="200"></td>
24
+ <td><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/fef03b75-0d2c-48ef-b016-39a6a1ff0dca/original=true/106745057.jpeg" width="200"></td>
25
+ </tr>
26
+ <tr>
27
+ <td><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/9f38dcf5-13ac-4bbe-a6c5-10614d4c3d14/original=true/106745063.jpeg" width="200"></td>
28
+ <td><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/6a9dd102-944d-42a4-b8d0-dd82896843f6/original=true/106746819.jpeg" width="200"></td>
29
+ <td><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/bc1b260c-fda0-4360-82e0-82e11fadf1ed/original=true/106745079.jpeg" width="200"></td>
30
+ </tr>
31
+ <tr>
32
+ <td><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/34c2fa20-d445-4b55-a483-bd49a592acbb/original=true/106745048.jpeg" width="200"></td>
33
+ <td><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/312af04b-f0a6-4a56-a436-f6d14cd56b22/original=true/106745052.jpeg" width="200"></td>
34
+ <td><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/ea8468cb-c8af-4253-b3ea-2902f5cbbdfc/original=true/106745055.jpeg" width="200"></td>
35
+ </tr>
36
+ <tr>
37
+ <td><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/a08acc0c-9735-4d80-9f75-e27d8ecbe82f/original=true/106745059.jpeg" width="200"></td>
38
+ </tr>
39
+ </table>
40
+
41
+ ## Description
42
+
43
+ NSFW Adv. is a LoKr trained on Qwen-image FP8, focused on refining and adding details to the human body. It was trained on a curated dataset of over 3000 images. I had to scrap it and start over from scratch more than 10 times, with this final version being trained for over 100,000 steps.
44
+
45
+ Even with all that effort, it's still got plenty of flaws. But after grinding on it for over two months, I had to push it out. The whole process was honestly super frustrating—tweaking, testing, culling, and repeating the process with a 10,000+ image dataset... it was just so damn hard. I seriously thought I might never finish it. But anyway, it's finally out. Hope you guys have fun with it.
46
+
47
+ **IMPORTANT USAGE TIPS:**
48
+ First, you'll need to install this plugin for your ComfyUI: [**https://github.com/ClownsharkBatwing/RES4LYF**](https://github.com/ClownsharkBatwing/RES4LYF)
49
+ Once it's installed, you'll get a bunch of new, high-quality samplers and schedulers in your ComfyUI.
50
+ I recommend using **Sampler:** **res\_2s** and **Scheduler:** **bong\_tangent**.
51
+
52
+ I also recommend using it with this distilled LoRA for faster generation:
53
+ [**https://huggingface.co/lightx2v/Qwen-Image-Lightning/resolve/main/Qwen-Image-Lightning-8steps-V2.0-bf16.safetensors?download=true**](https://huggingface.co/lightx2v/Qwen-Image-Lightning/resolve/main/Qwen-Image-Lightning-8steps-V2.0-bf16.safetensors?download=true)
54
+
55
+ All demo images were generated using:
56
+ **Steps: 10
57
+ CFG: 1
58
+ Sampler:** **res\_2s** **Scheduler:** **bong\_tangent**
59
+
60
+ **NSFW Adv lora strength 1.0**
61
+
62
+ **Qwen-Image-Lightning-8steps lora strength 0.4~0.5**
63
+
64
+ PS: If you plan on merging this, please give credit or link back to this page. Thanks
65
+
66
+ PS2: This LoRA training session was done with [**https://github.com/ostris/ai-toolkit**](https://github.com/ostris/ai-toolkit), but I incorporated the new sampling methods for timesteps from **sdbds** ([**https://github.com/kohya-ss/musubi-tuner/pull/407**](https://github.com/kohya-ss/musubi-tuner/pull/407)).
67
+
68
+ After several experiments, I found this method to be significantly effective for this training task, as it notably accelerated convergence. A big shoutout to them for their excellent contribution.
69
+
70
+ ---
71
+
72
  Author: [fok3827](https://civitai.com/user/fok3827)
73
 
74
+ Model: [CivitAI Model Page](https://civitai.com/models/2058077?modelVersionId=2328988)
75
+
76
+ Archive: [CivArchive Page](https://civarchive.com/models/2058077?modelVersionId=2328988)
77
 
78
+ <!-- Version: 20260425_upload -->