Blowjob in the style of "MHBHJ." Trained on 14B T2V, I2V also works. You randomly get tip sucking, licking, edging. You will occasionally get a logo or some text (bad source data by me!). I'll refine this version later, I sped through the captions and training.
I like to prompt with:
POV, first person, front view, A woman lying on her stomach in between a man's legs. She is mhb_suck1ng his penis.
I also add to my negative to suppress any text (not sure if this actually helps):
text, logo
I like to combine it with the dick lora at a low weight: https://civarchive.com/models/1358184/wan-21-penis-cock-dick-lora-t2v-i2v?modelVersionId=1534254
These flux loras were my inspiration:
https://civarchive.com/models/802369/mhb-tip-sucking-flux
https://civarchive.com/models/810290/mhb-licking-flux
Trained using hearmeman's Runpod diffusion-pipe template and tested using his WAN template:
Description
FAQ
Comments (5)
Thank you for the great work.
is it good with just a pic of a woman (front view)?
It's only trained in the front view on the specific style of video "marks head bobbers and hand jobbers", but I can see from the gallery it works on side and 3/4 view.
@jalenbrunson yes, i know the video style and i like it! But I am not getting the same result (no slow sucking, no feet, etc.) can you share the prompt and the lora you used please? also the diffusion models if you have used any. It would help me a lot
@CumBlaster9522 Sure, I have it in the info for the gallery samples. I used Wan 14B T2V and combined it with the "dicks" lora, check the first pick in the gallery for links or the suggested resources. I trained an I2V version but haven't released it yet - need to find some time to test it.
Positive Prompt: POV, first person, front view, A brunette woman, naked, wearing large hoop earrings, lying on her stomach in between a man's legs. She is mhb_suck1ng, licking and stroking his penis.
Negative prompt: text, logos, 色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走