~Haven't we all fantasized about wake up sex~

Very excited to present a brand new concept, Sleepy Male POV! Very little examples of this concept out there in the NSFW art space, so I made my own dataset for the LoRA! I go over some very important information in the description, please take the time to read everything below if you want to get the most out of this model :)

Key Features v1.1:

• Wake up/Sleeping POV sex

• POV with distinct eye-vignette

• Cowgirl sex

• Reverse cowgirl sex

• Fellatio

• Penis Hugging

• Distension

• Voyeur/Spying POV

• Almost Caught

• Footjob

-Sleepy Male POV v1.1 *NEW*-

- Parameters

• Weights: 1.0

• Triggers: 5L33PYM4L3P0V, vignette, head out of frame,

• Checkpoints: Base Pony for best functionality. See Suggested Resources below for other recommendations that I've tested.

• Resolutions: Stick to widescreen, all examples are 1152 width 896 height. Although, V1.1 did bring some pretty decent portrait mode support.

• Other LoRA's: Compatible with some, not with others. You'll have to do your own testing, but be sure to read over my section below about style changing!

• Limitations: Right now cowgirl sex is the most reliable position, the dataset includes various ~fun~ activities and positions, but none of them produce the eye-vignette that I'm going for as consistently as cowgirl sex(they just tend to produce dark images without a vignette). Working on making the model more flexible, but am happy with the results so far and will be sharing my progress going forward. Version 1.1 brings incredible reliability not only more so with what it could already do, but with more positions and scenarios. Actually very pleased with the results.

• Style Changing: I HIGHLY suggest checking out this resource regarding Pony. Has so much good info, but I want to draw your attention to the hashed tokens section. Download the highlighted .txt file. You can use it as a wildcard and find what you like.

• Hashed Tokens I love for this LoRA: zsh, jtv, jwt, amui, zgq, rsd, uvm, lnp, zro

  Take a look at this!

• Image with jwt hashed token:

• Image WITHOUT jwt hashed token:

Literally the only difference between these two images in terms of generation parameters is one, three-letter hashed token placed early in the prompt(eg. score_9, score_8, jwt, {prompt}). This is an insane difference in style while retaining ALL integrity of my LoRA, and that's just one of MANY hashed tokens. Note: these hashed tokens will have a greater impact using Base Pony, as any other merges will have their own biases that can be very strong.

Version 1.1 also brings more compatibility with other Pony Checkpoints. Here are some examples:

• SweetMix(incompatible in previous versions)

• Stylized Ponyocalypse(my own merge), before, this LoRA was VERY incompatible with this checkpoint. However, with 1.1 we get the desired effect. I still wouldn't use this checkpoint/lora combo due to its heavy bias towards stylized, kinda is going to make the image nonsensical. Still good tho.

• GRLE mix

I think it's safe to say you can use this LoRA with your favorite Pony checkpoints :)

XYZ Plot Grid - All Base Pony

Check down below for more images!

-

-

-

-

-

-

-

-

Authors Notes

-I love v1.1 and I think if I were to do a v1.2, I would customize vignette sizes and tag accordingly allowing to prompt for the size of the vignette. Let me know in the comments if you would like to see this, as well as what the model currently lacks and suggestions on what you'd like to see from this model. TIA! <3

Appreciate what I do? Here's a few ways to support me <3

• Contribute to a new GPU on my Patreon

• Give me some gas money on PayPal