Recommending NTR, netorare, cuckold - resources, discussion, development

5.00 star(s) 6 Votes

Bofadeezus

Member
Jul 24, 2020
232
555
I've been lightly looking into the LLMs and playing around with Ollama and using Silly Tavern as my WebUI.

Im not too clued up but what application do you use to run these GGUF models? (LM Studio, Kobold AI. etc)

If possible id still like to use Silly Tavern as its UI.

yeah, the R1 Distill-Qwen-14B is kinda poop for RP. for me personally, Magnum V4 is still the best 12B model for RP. I almost dropped all of my VN's thanks to it.
 

blablablagivemethemoney

Active Member
Jul 20, 2019
554
871
View attachment 4483515

Perhaps you tried too hard to insert yourself into the main character, overlooking the broader narrative. Haven't you missed the subplots involving the protagonist's sister and the princess? Just like the main heroine herself had learned to ignore them. It's not just about the relationships falling apart; it's about how people break and how everyone, in a twisted way, seems to deserve the pain they eventually get.

You don't have permission to view the spoiler content. Log in or register now.
, highly recommended.
I already said it's not a bad story, but the feeling of NTR is weak.
 

Sieglinnde

Tifa's Dark Heaven - Developer
Game Developer
Dec 2, 2019
2,881
6,338
View attachment 4483515

Perhaps you tried too hard to insert yourself into the main character, overlooking the broader narrative. Haven't you missed the subplots involving the protagonist's sister and the princess? Just like the main heroine herself had learned to ignore them. It's not just about the relationships falling apart; it's about how people break and how everyone, in a twisted way, seems to deserve the pain they eventually get.

You don't have permission to view the spoiler content. Log in or register now.
, highly recommended.
That's another thing I didn't like, apart from what blablablagivemethemoney mentioned... It was the way the game handles the characters

I didn't like the game either, speaking from the perspective that I expected it to be an NTR game... That's why I always disagree when some people say it's the darkest NTR game and all that... It's weird to me because it's as if other people have seen something in the game that for me doesn't exist

To begin with, the whole conflict starts right away, and I didn't have time to get interested in the characters enough for me to care if something bad happens to them or not... And from then on everything went from bad to worse for me, because immediately afterwards we go to a scene where the protagonist's sister is abused and even beaten while they do it... That scene made me feel horrified and disgusted... And in general that's how I felt the whole game played out... They present you with unpleasant and exaggerated situations so that you feel that the MC is completely useless and it doesn't matter what he does because something bad is always going to happen or everything is going to go wrong... And from there on I know what I'm going to find, so I don't feel any emotion... No, from the beginning I had no emotions because I didn't care about the characters at all, because I didn't have time for them to be properly introduced to me so that I would care about them... The game is very sad, yes, but I can't say it's an NTR game, at least not for what it made me feel... I feel that it relies more on trying hard to be sad and traumatic and that disappointed me, because many said it was the saddest NTR game so I was expecting something like that
 

Yambidexter

Member
Jul 16, 2020
310
589
I've been lightly looking into the LLMs and playing around with Ollama and using Silly Tavern as my WebUI.

Im not too clued up but what application do you use to run these GGUF models? (LM Studio, Kobold AI. etc)

If possible id still like to use Silly Tavern as its UI.
I started thanks to melantha's small guide or a couple you can find here. Started with LM Studio + Silly Tavern as sort of back end and front end, so yes you can use both, but some time ago turned to Kobold + Silly Tavern, as with LM Studio connection was somehow breaking periodically and generation stopped after that. I.e. Kobold, LM Studio and Ollama are used for running a model, but usually people chat and have their char cards in Silly Tavern.

The models for me are usually (based on ) or sometimes . I tried some more like Gemmasutra, Stheno, Nemomix unleashed but wasn't particularly impressed - mostly due to some blatant logical lapses like cheating GF/wife nonchalantly telling you about how she's had a nice time with her affair partner in the first messages, quickly starting to mess names/sexes, or just the language style. BTW I started actually online on Yodayo, but then moved on to Janitor and LLM, and Yodayo is kind of monetizing more aggressively now as far as I understand, with every message costing you some internal currency of theirs a bit of which you can get daily or just buy, so you can also find their own model which is one of the options at Youdayo, and use it locally. It's a simple one, but still.

I import char cards from Janitor AI via web links, or Chub AI via downloading jsons and then importing them into Silly Tavern. Set temp to 0.7 to curb down creative nonsense of LLM, that's the value that is used by Janitor AI if you use it online by default. There is also a thing called Lorebooks on Chub AI that you can import into Silly Tavern. A lorebook is kind of additional layer of events that can kick in depending on appearance of key words in your chat. Might fine-tune your chat or liven them up some. Search e.g. for Cuckworld, Slow corruption, Humiliation methods, Manipulation methods lorebooks and some more, if interested.

I don't do much of RP now though, as I got to think that:

1) Chars are essentially non-entities, they'll do whatever they want them to do, they're too reactive. But it's kind of natural as chats are designed to respond to your inputs and don't know what you want.

2) Once you hit you context limit in tokens of about 4-12k locally, first point will get even worse along with decreased performance, LLM forgetting events and people that lie outside of these last 4-12k tokens in chat, which - together - considerably deteriorate the experience further.

A way around this is to use some powerful online models with huge token limits like ChatGPT, Claude Sonnet or perhaps new Deepseek v3, I don't know how latter will perform with NSFW stuff though. But recently Pango_12 translated a couple of novels and shared them here, so I guess it has to be fine in terms of censorship. And use really well written char cards with setting and lore, and characters' descriptions worth of a few thousand tokens (not that there are many of them, if any, from what I've seen). Or wait for LLM & char cards to mature some more...

P.S. I'd appreciate a link to quantized Magnum V4 model on Hugginface that was mentioned here to try it out, though.
 
  • Heart
Reactions: Bofadeezus

Bofadeezus

Member
Jul 24, 2020
232
555
I also started out on Yoyodaro/Moescape but left cause of the increasing lack of free tokens.

Last used JanitorAI for online gen but decided that going local would be best.

I quite liked the chat memory feature as it worked decently for long terms NTR scenes but was a pain to always update.

Currently running through Ollama and Silly Tavern but wanted to give others a shot as the long-term memory was an issue for NTR events.

Downloaded today but I'm having issues with getting it to output both character text and scene description so probably gonna dump it soon.

Thanks, I'll give the Moistral model a try next and try it out.

I started thanks to melantha's small guide or a couple you can find here. Started with LM Studio + Silly Tavern as sort of back end and front end, so yes you can use both, but some time ago turned to Kobold + Silly Tavern, as with LM Studio connection was somehow breaking periodically and generation stopped after that. I.e. Kobold, LM Studio and Ollama are used for running a model, but usually people chat and have their char cards in Silly Tavern.

The models for me are usually (based on ) or sometimes . I tried some more like Gemmasutra, Stheno, Nemomix unleashed but wasn't particularly impressed - mostly due to some blatant logical lapses like cheating GF/wife nonchalantly telling you about how she's had a nice time with her affair partner in the first messages, quickly starting to mess names/sexes, or just the language style. BTW I started actually online on Yodayo, but then moved on to Janitor and LLM, and Yodayo is kind of monetizing more aggressively now as far as I understand, with every message costing you some internal currency of theirs a bit of which you can get daily or just buy, so you can also find their own model which is one of the options at Youdayo, and use it locally. It's a simple one, but still.

I import char cards from Janitor AI via web links, or Chub AI via downloading jsons and then importing them into Silly Tavern. Set temp to 0.7 to curb down creative nonsense of LLM, that's the value that is used by Janitor AI if you use it online by default. There is also a thing called Lorebooks on Chub AI that you can import into Silly Tavern. A lorebook is kind of additional layer of events that can kick in depending on appearance of key words in your chat. Might fine-tune your chat or liven them up some. Search e.g. for Cuckworld, Slow corruption, Humiliation methods, Manipulation methods lorebooks and some more, if interested.

I don't do much of RP now though, as I got to think that:

1) Chars are essentially non-entities, they'll do whatever they want them to do, they're too reactive. But it's kind of natural as chats are designed to respond to your inputs and don't know what you want.

2) Once you hit you context limit in tokens of about 4-12k locally, first point will get even worse along with decreased performance, LLM forgetting events and people that lie outside of these last 4-12k tokens in chat, which - together - considerably deteriorate the experience further.

A way around this is to use some powerful online models with huge token limits like ChatGPT, Claude Sonnet or perhaps new Deepseek v3, I don't know how latter will perform with NSFW stuff though. But recently Pango_12 translated a couple of novels and shared them here, so I guess it has to be fine in terms of censorship. And use really well written char cards with setting and lore, and characters' descriptions worth of a few thousand tokens (not that there are many of them, if any, from what I've seen). Or wait for LLM & char cards to mature some more...

P.S. I'd appreciate a link to quantized Magnum V4 model on Hugginface that was mentioned here to try it out, though.
 

Kulman

Engaged Member
Sep 28, 2017
2,839
7,297

Penumbral Evanescence

|☽◯☾| Somnium of the Night ✧ Bête Noire
Moderator
GFX Designer
Donor
Jul 16, 2021
3,261
11,455
That isnt a single 4 hour movie, right? Just 4 hour compliation.

When I saw 4 hours, I realized I would really love to watch a 4 hour Jav lol. Feel like that could really spread out the corruption.
lmao no, its just a 4 hour compilation

i think at most, a lotta JAV porn I've seen go up to the 2-2 and a half hour mark, but since this is a compilation it makes sense why its 4 hours

lmao so many JAV porn vids are longer than some Hollywood movies in terms of run time, because they wanna build a narrative and make u feel something for the characters (especially for the NTR films) :KEK: :HideThePain:
 

Kulman

Engaged Member
Sep 28, 2017
2,839
7,297
lmao no, its just a 4 hour compilation

i think at most, a lotta JAV porn I've seen go up to the 2-2 and a half hour mark, but since this is a compilation it makes sense why its 4 hours

lmao so many JAV porn vids are longer than some Hollywood movies in terms of run time, because they wanna build a narrative and make u feel something for the characters (especially for the NTR films) :KEK: :HideThePain:
Yeah thats why I kinda pogged when I saw 4 hours. Imagine.
 

archedio

Well-Known Member
Modder
Apr 18, 2019
1,202
3,825
I also started out on Yoyodaro/Moescape but left cause of the increasing lack of free tokens.

Last used JanitorAI for online gen but decided that going local would be best.
Try
You can customize or add story cards for AI to remember things long term.
It's my fav so far.
 

archedio

Well-Known Member
Modder
Apr 18, 2019
1,202
3,825
Is there a guide?
The site already offers a guide for you to learn ->
Or, if you just want to go right into scenarios, here are some recommendations:




But you kinda need to be a bit creative.
For ex: The popular kids condition is to let the boys fuck your GF OR sister (if you are into incest).
The possibilities are endless.
 
Last edited:

melantha

Member
Jan 21, 2019
277
649
what application do you use to run these GGUF models?
If possible id still like to use Silly Tavern as its UI.
I started thanks to melantha's small guide or a couple you can find here.

KoboldCpp guide included and other chatbot alternatives

1) Chars are essentially non-entities, they'll do whatever they want them to do, they're too reactive. But it's kind of natural as chats are designed to respond to your inputs and don't know what you want.
2) Once you hit you context limit in tokens of about 4-12k locally, first point will get even worse along with decreased performance, LLM forgetting events and people that lie outside of these last 4-12k tokens in chat, which - together - considerably deteriorate the experience further.
i observe it's more on the system prompt quality, and additional memory enchancement, like RAG.

system prompts are very important, more than i thought. it's not just the 1 paragraph, "Write {{char}} next response. Any act of role play scenarios will be described in details.". There should be alot more behind the scenes, tell it how to write, enclose actions and and narrations in symbols, never act or depict actions of {{user}}, etc. There's alot of pre-made system prompts made by users shared on reddit and the like.

also, i dont think SillyTavern has RAG (kind of like a memory between the chat interface and the LLM).
RAG helps the LLM remember stuff, kind of like lorebooks / knowledgebase, but automatic and learns as you chat.
that's one of the advantages of online services.

perhaps new Deepseek v3, I don't know how latter will perform with NSFW stuff though. But recently Pango_12 translated a couple of novels and shared them here, so I guess it has to be fine in terms of censorship. And use really well written char cards with setting and lore, and characters' descriptions worth of a few thousand tokens (not that there are many of them, if any, from what I've seen).
yodayo/moescape added DeepSeek V3, R1, and distilled Qwen 32B just now.
requires paid beans so i aint trying them out.
but we can check their feedback channels if people are enjoying it for RP.
there are also talks in SillyTavern discord about it but too technical for me to understand lol

BTW I started actually online on Yodayo, but then moved on to Janitor and LLM, and Yodayo is kind of monetizing more aggressively now as far as I understand
I also started out on Yoyodaro/Moescape but left cause of the increasing lack of free tokens.
glad to see people continuing a path outside of yodayo/moescape after their huge blunder.

The models for me are usually (based on ) or sometimes . I tried some more like Gemmasutra, Stheno, Nemomix unleashed but wasn't particularly impressed
Currently running through Ollama and Silly Tavern but wanted to give others a shot as the long-term memory was an issue for NTR events.
Downloaded today but I'm having issues with getting it to output both character text and scene description so probably gonna dump it soon.
Magnum V4 seems to be preferred by SillyTavern users.
also try or .


...im actually planning to make a combined LM Studio + SillyTavern in one desktop app. you can run your local LM, and chat in the same interface. it should also be simplified, and interface like discord, unlike the SillyTavern mess of a UI.
 
5.00 star(s) 6 Votes