This is a NSFW-hentai-themed concept (you can also use it for SFW content...),

based on the concept of NTR. Main purpose is to help you add some age-difference content into your masterpiece.

Trained on 100+ realistic images. Generated AI pictures DO NOT depict any real individual.

Instruction for use

LoRA weight to use: 0.6-1.1 (0.7 is optimal)

Main prompt (activation): rodo (from 労働者 rodo-sha)

rodo, old worker, retired,

enhancing prompts: 2boys, surrounded by men, multiple boys, hetero,

actions: ?????? grab, grab ???????????, (naked),

e.g. 1girl, rodo, (old worker), retired, 2boys, surrounded by men, multiple boys, hetero, ?????? grab, grab ???????????, <lora:rodo_retired_old_workers_JP:0.7>

Interactions with other LoRA

Generating japanese ojiisan that is also fat and ugly

I have also tested the interaction of this LoRA with the existing ugly_bastard (fat man LoRA; with a tendency to generate caucassian 'uncles')

to see how these LoRAs can be implemented simultaneously.
(and I had a lot of fun seeing sweet girls turning into chubby 'body-positive' eurasians...)

Advice: tweak both in a contra-positive way
(i.e. high weight for rodo, low weight for ugly_bastard; vice versa)

Recommendation: Keep the LoRA weights between 0.6 and 1.3.

Ideally:0.7 for both, and experiment in (+/-)0.1 from there.

Similarly, both this LoRA and Ugly_bastard go crazy when prompt words like (multiple boys, surrounded by boys) is added, so be mindful of how many chubby dudes can fit into your image.

Have not tried this LoRA extensively on Anime models (apart from animilf/semi-realistic models). Let me know how it goes.