AI Tutorials for spanking content

[xaodj]

Hi!

I noticed that there have been more and more AI content threads, so I thought I'd create my own to share what I have learned in the past few years.

Ideally you should have a GPU with at least 8-12 GB VRAM, but you can also use your CPU and system RAM (althought it's much slower).

The point of this thread is to give you guys the tools to create your own content, so feel free to ask any questions.

I'm hoping to give something back, given all the content that I downloaded from you guys on this board and the old one.

So first here is a brief overview of what I personally create for myself and the tools I use (they are all free and mostly uncensored):

1. Images
There are 2 interfaces I use: Invoke and Comfyui.
Invoke is the easiest to learn, as it kind of looks like photoshop and allows you to download the AI models directly from the interface.
Comfyui is a little more complex but allows you to create complex workflows to automate the image creation (for example: Generate an idea using a language model -> generate the image -> upscale the image -> improve details).
There are many different base models you can use depending on your hardware (SDXL, Flux, ...) and you can add what is called a lora model to get a custom output.
Example: Use Flux as a base model, add a Lora of your favorite singer/actress/crush, add a lora trained on spanking images aaand voilà... there she/he is getting spanked.

A good place to start to find AI models for image and video is:

You must be registered to view links and images

I think you gotta register to see the nsfw stuff.

2. Stories
If you like stories, you gotta try this!
I mainly use LM Studio to run the LLm and Silly Tavern to create characters/storylines.
LM Studio is basically like Chat-GPT, but it runs locally on your PC (totally private, nothing is shared).
Silly Tavern is a roleplay interface where you can create characters, download premade ones, and roleplay with them. I use it to write stories and it works great.

3. Audio
This is my new favorite thing, I use a Text-to-speech model called Chatterbox to generate audio of a person telling me a story.
You can clone a voice (please be responsable), using an audio clip of the person talking.
Ever wanted to hear Amber Pixie Wells tell you about that one spanking that you just made up? Or listen Kyle Johnson explain why the hell he talks so much during his videos?
This is super easy to use and very satisfying when it works properly.

4. Video
EDIT: Finally we have a decent model for generating videos as well!
Here is a Comfyui workflow, the model is Minimax H3:

You must be registered to view links and images

My idea is to make two other threads and add the links to this post.

• 1 for sharing content:
  
  You must be registered to view links and images
  
  
  
• 1 for sharing AI models, particularly custom made ones trained on spanking:
  
  You must be registered to view links and images

This thread will remain focused on tutorials, answering questions and if possible hearing from other members and their best practices.

Just a few rules:
I will not share anything illegal
I will not share any content involving real people

Please, let me know what you think.

 

[jackleson42069]

Regarding Video , the best way to start is to use the "Self forcing" model for Wan 2.1 .

The hardest part of getting started with video generation is that with most models/workflows, on excellent hardware it takes a couple minutes to generate a short clip, and on anything short of perfect hardware it takes significantly longer. Waiting 30-60 minutes to find out that your generation parameters are broken and the output is just noise is a miserable experience.

Using the Wan 2.1 , 1.3B model (self forcing) and a fairly vanilla workflow, on good hardware a video clip in moderate resolution (400 x 540 or something) is doable in 10 seconds. Being able to iterate at that speed will get you over the hump of "video is impossible, also I'm OOM all the time".

The other trick with video is to deliberately make low-resolution videos (think 200 x 300, 3 steps of stable diffusion instead of 20). Then you look at the low-resolution video and see if it's going to eventually be viable, then you hit the gas on a higher resolution generation, more steps, and run your video upscaler.

Check out this - no install needed.

You must be registered to view links and images

 

[xaodj]

Regarding Video , the best way to start is to use the "Self forcing" model for Wan 2.1 .

The hardest part of getting started with video generation is that with most models/workflows, on excellent hardware it takes a couple minutes to generate a short clip, and on anything short of perfect hardware it takes significantly longer. Waiting 30-60 minutes to find out that your generation parameters are broken and the output is just noise is a miserable experience.

Using the Wan 2.1 , 1.3B model (self forcing) and a fairly vanilla workflow, on good hardware a video clip in moderate resolution (400 x 540 or something) is doable in 10 seconds. Being able to iterate at that speed will get you over the hump of "video is impossible, also I'm OOM all the time".

The other trick with video is to deliberately make low-resolution videos (think 200 x 300, 3 steps of stable diffusion instead of 20). Then you look at the low-resolution video and see if it's going to eventually be viable, then you hit the gas on a higher resolution generation, more steps, and run your video upscaler.

Check out this - no install needed.

You must be registered to view links and images

Thanks a lot for this, I'll give it a try. I'm running a 3090, so it should work great.

edit: do you know if this will work with Wan 2.1 loras? or does it require it's own?

 

[Cro]

Thank you both for your effort to help us create! 🙏

 

[Xenon]

To add to this, here's another cool way you can get started (esp. helpful if you don't have a separate GPU in your machine) using Google Colab.

Setup is pretty easy (and free) – you should be able to generate your first AI image within 5 minutes!

I added some context about the tools being used in the spoiler below if you are interested; otherwise, feel free to dive in –

Spoiler: Background on Google Colab and Fooocus

Google Colab
Google offers a product called "Google Colab," which is a free service that let's you write and run python scripts using Google's computers instead of your own. Everything is done through the web browser, so you don't need to install anything on your computer. Colab has recently expanded its capabilities so that users can now access Google's GPUs, which has opened the door to some cool AI applications.

Fooocus
Fooocus is an AI image generator with a user interface based on stable diffusion. A more detailed description of the implementation of Fooocus can be found via their

You must be registered to view links and images

.

Of note, I chose Fooocus for this tutorial because it's super easy to get up and running, though it's worth noting that the models are a bit out of date at this point (last repository update was 10 months ago). I would encourage anyone who is interested to do some more digging – there are a bunch AI image generators that leverage Google Colab out there! Feel free to share any better / more recent notebooks you come across 🙂

Setting Up AI Image Generation via Google Colab

1. Open the
   You must be registered to view links and images
   (note: login to your Google account if prompted)
   
   
2. [Optional] At the top of the page, click "Copy to Drive" – this just creates a copy of the notebook to your account so you can easily access it later on
   
   
3. To start the notebook, click the "Run All" button (next to "Copy to Drive")
   
   You must be registered to view images
   
   
   
4. As the code runs, you should start to see text output on the page – you can ignore all of this (it's just showing you the steps it is taking to install the AI image generator on Google's computers)
   
   
5. After a minute or two, you should see two lines in the log similar to this:
   
   Running on local URL: http://127.0.0.1:7865
Running on public URL: https://fa7e224dd3030e0412.gradio.live
   
   This means that the image generator has been installed and is ready to use! Click on the second link ("public URL") to open your personal AI image generator in a new tab.
   
   
6. Once you have the image generator open, you can quickly get started by entering a prompt and pressing "Generate." However, I would encourage you to select the "Advanced" checkbox at the bottom, which will give you a lot more control over what the image will look like (including image size, number of image generated at once, etc.). Notably, you can also play around with different styles to make the images look more/less realistic

A few other notes worth mentioning:

• There shouldn't be any limits to the app, so feel free to play around with it as much as you want!
• Based on my initial testing, images take around 30 seconds to generate
• It's worth noting that Google Colab notebooks can run for a maximum of 12 hours, so you will need to restart the app once that time has elapsed. Since this is using the free version of Google Colab, there's also a chance that you may not be able to run the notebook if Google is experiencing a heavy load from other users (I've never actually had this happen to me, but it is outlined in Google's terms and conditions).

App UI Screenshot (w/ a few examples of image output)

You must be registered to view images

Enjoy!

 

[agentx69]

I have trouble finding working prompts for OTK spanking, thanks for your help

 

[xaodj]

I have trouble finding working prompts for OTK spanking, thanks for your help

I don't think there is a prompt that would really help, the best thing you can do is find a decent lora model trained on OTK.
Te best ones I have seen are for a model called Pony and another called Illustrious, but it's mostly anime.

You can see them here (but you have to be logged in with nsfw turned on):

You must be registered to view links and images

After opening one of the model pages, you can scroll down to the bottom of the page, to see images created with that model. That should give you an idea of what it can generate.

I downloaded a few images so you can see something before registering.

You must be registered to view images

You must be registered to view images

You must be registered to view images

You must be registered to view images

 

[jackleson42069]

Couple comments:

1. Illustrious and pony have "mixes" for realism. Basically people merge the model with other models to make a model that can do both anime and also photographic content.

Try e.g. "pony realism" or "ilustmix" and prepare to be amazed.

2. Colab is excellent. I haven't read the terms of service but I do remember a couple years ago people being banned for doing stable diffusion inference on Colab. Explicit content might be specifically disallowed.

 

[jackleson42069]

I don't think there is a prompt that would really help, the best thing you can do is find a decent lora model trained on OTK.

ControlNet (depth) is also really useful

 

[xaodj]

Yes, with a good reference image.

 

[jackleson42069]

I came across this today

You must be registered to view links and images

Not that new a merge, but honestly a bit of a sleeper. I find it pretty creative and needs very short prompts to get the job done. Works well with PONY Loras. Realism is very solid IMO. Anatomy is the usual stable diffusion jank that you solve with rerolls until you get a good seed.

 

[BrushBoy]

Hi!

I noticed that there have been more and more AI content threads, so I thought I'd create my own to share what I have learned in the past few years.

Ideally you should have a GPU with at least 8-12 GB VRAM, but you can also use your CPU and system RAM (althought it's much slower).

The point of this thread is to give you guys the tools to create your own content, so feel free to ask any questions.

I'm hoping to give something back, given all the content that I downloaded from you guys on this board and the old one.

So first here is a brief overview of what I personally create for myself and the tools I use (they are all free and mostly uncensored):

1. Images
There are 2 interfaces I use: Invoke and Comfyui.
Invoke is the easiest to learn, as it kind of looks like photoshop and allows you to download the AI models directly from the interface.
Comfyui is a little more complex but allows you to create complex workflows to automate the image creation (for example: Generate an idea using a language model -> generate the image -> upscale the image -> improve details).
There are many different base models you can use depending on your hardware (SDXL, Flux, ...) and you can add what is called a lora model to get a custom output.
Example: Use Flux as a base model, add a Lora of your favorite singer/actress/crush, add a lora trained on spanking images aaand voilà... there she/he is getting spanked.

A good place to start to find AI models for image and video is:

You must be registered to view links and images

I think you gotta register to see the nsfw stuff.

2. Stories
If you like stories, you gotta try this!
I mainly use LM Studio to run the LLm and Silly Tavern to create characters/storylines.
LM Studio is basically like Chat-GPT, but it runs locally on your PC (totally private, nothing is shared).
Comfyui is a roleplay interface where you can create characters, download premade ones, and roleplay with them. I use it to write stories and it works great.

3. Audio
This is my new favorite thing, I use a Text-to-speech model called Chatterbox to generate audio of a person telling me a story.
You can clone a voice (please be responsable), using an audio clip of the person talking.
Ever wanted to hear Amber Pixie Wells tell you about that one spanking that you just made up? Or listen Kyle Johnson explain why the hell he talks so much during his videos?
This is super easy to use and very satisfying when it works properly.

4. Video
Videos are really hard to make, as the available models are not properly trained on spanking, the hardware requirements are high and the generation times can be very long for very short clips.
I experimented with comfyui and models like Hunyuan and Wan 2.1, but got unsatisfying results.


My idea is to make two other threads and add the links to this post.

• 1 for sharing content:
  
  You must be registered to view links and images
  
  
  
• 1 for sharing AI models, particularly custom made ones trained on spanking

This thread will remain focused on tutorials, answering questions and if possible hearing from other members and their best practices.

Just a few rules:
I will not share anything illegal
I will not share any content involving real people

Please, let me know what you think.

I would love to hear more about your experiences with text to speech (Chatterbox and other possible options).

Elevenlabs is the best I know of, but they record and listen to all content so nothing is private, and they forbid any spanking or fetish content on pain of a permanent ban!

How private is Chatterbox? Are there any decent offline/local options (ideally not requiring an Nvidia GPU which is always my problem with local AI stuff, since I have an AMD GPU).

I'd love to be able to start making audios again; it's far more important to me personally than making videos or even images.

 

[beritjunk]

I would love to hear more about your experiences with text to speech (Chatterbox and other possible options).

Elevenlabs is the best I know of, but they record and listen to all content so nothing is private, and they forbid any spanking or fetish content on pain of a permanent ban!

How private is Chatterbox? Are there any decent offline/local options (ideally not requiring an Nvidia GPU which is always my problem with local AI stuff, since I have an AMD GPU).

I'd love to be able to start making audios again; it's far more important to me personally than making videos or even images.

I've been experimenting a bit with Chatterbox and it's actually quite good. I'm very impressed by the voice cloning feature. You can run it fully locally with no internet access, and it can run on CPU if you don't have Nvidia GPU. Only downside is that clips at limited to 300 characters, so if you want to voice a full story for example there'll be a lot of small clips you need to merge.

 

[BrushBoy]

I don't think there is a prompt that would really help, the best thing you can do is find a decent lora model trained on OTK.
Te best ones I have seen are for a model called Pony and another called Illustrious, but it's mostly anime.

You can see them here (but you have to be logged in with nsfw turned on):

You must be registered to view links and images

After opening one of the model pages, you can scroll down to the bottom of the page, to see images created with that model. That should give you an idea of what it can generate.

I downloaded a few images so you can see something before registering.

You must be registered to view images

You must be registered to view images

You must be registered to view images

You must be registered to view images

I just wanted to say as a big Evangelion fan that I love that Rei/Asuka one. Rei's expression is priceless, and Asuka's is spot on.

I've been experimenting a bit with Chatterbox and it's actually quite good. I'm very impressed by the voice cloning feature. You can run it fully locally with no internet access, and it can run on CPU if you don't have Nvidia GPU. Only downside is that clips at limited to 300 characters, so if you want to voice a full story for example there'll be a lot of small clips you need to merge.

That's awesome news! I can deal with the character limit as that was initially an issue with Elevenlabs too until they moved to allowing full audio books... which is ironically how I got banned (getting them to read existing published spanking novels).

I've really struggled to find any sort of guide or tutorial for setting it up (especially without saying an NVIDIA GPU is required; I am fine with CPU it taking longer if it at least works on CPU only). Can you point me towards something please, if possible? I am on Windows 11 if that makes a difference.

Privacy is my big issue with all this AI stuff, honestly; I would gladly pay for Chatterbox if it was private but after Elevenlabs, I don't trust any of these companies not to be recording and viewing/listening to everything.

 

[xaodj]

I use this for Chatterbox:

You must be registered to view links and images

The cool thing about this webui, is that you can also use it to set up a voice chat with an LLm on Sillytavern and have the AI anwer you with the voice you choose.

If you click on donwload installer, it downloads a zip with an easy to use method to install the app.

Here are the steps to follow:

Current base installation size is around 10.7 GB. Each model will require 2-8 GB of space in addition.

• Download the
  You must be registered to view links and images
  and extract it.
• Run start_tts_webui.bat or start_tts_webui.sh to start the server. It will ask you to select the GPU/Chip you are using. Once everything has installed, it will start the Gradio server at
  You must be registered to view links and images
  and the React UI at
  You must be registered to view links and images
  .
• Output log will be available in the installer_scripts/output.log file.
• Note: The start script sets up a conda environment and a python virtual environment. Thus you don't need to make a venv before that, and in fact, launching from another venv might break this script.

 

[Cro]

Hey, can you guys show more things you have created? 🙂

 

[xaodj]

Hey, can you guys show more things you have created? 🙂

If you mean images, I am posting mine in this thread:

You must be registered to view links and images

Feel free to make requests, if you want me to try and generate something.

 

[BrushBoy]

I use this for Chatterbox:

You must be registered to view links and images

The cool thing about this webui, is that you can use it to set up a voice chat with an LLm on Sillytavern and have the AI anwer you with the voice you choose.

If you click on donwload installer, it downloads a zip with an easy to use method to install the app.

Here are the steps to follow:

Current base installation size is around 10.7 GB. Each model will require 2-8 GB of space in addition.

• Download the
  You must be registered to view links and images
  and extract it.
• Run start_tts_webui.bat or start_tts_webui.sh to start the server. It will ask you to select the GPU/Chip you are using. Once everything has installed, it will start the Gradio server at
  You must be registered to view links and images
  and the React UI at
  You must be registered to view links and images
  .
• Output log will be available in the installer_scripts/output.log file.
• Note: The start script sets up a conda environment and a python virtual environment. Thus you don't need to make a venv before that, and in fact, launching from another venv might break this script.

Thanks for this, but it says Linux only for AMD GPUs when installing. 🙁

 

[xaodj]

Thanks for this, but it says Linux only for AMD GPUs when installing. 🙁

You should be able to use it with CPU, it's slower but still works.

 

[BrushBoy]

You should be able to use it with CPU, it's slower but still works.

Thanks, I am installing now. I can deal with a bit slower but it just bugs me having a GPU but not being able to use it. Appreciate the guide! I miss making audios so much.