Model Information




Description
I must abandon creating new for now, my pc is freasing.
i will not send models for update, beacuse my harddrive is broken

What is LoRA?
-A new (ish) way of training SD models that results in a tiny output file (6-100MB) that you apply to other models to transfer that knowledge. It's super fast to train (5-10x faster than dreambooth), and the results are nearly as good as dreambooth from what I've seen
How to use (how to mix)?
This A1111 LORA extension allows you to mix a LoRA trained model with any other model here, it adds a new button on the main page for
Additional Models:
https://github.com/kohya-ss/sd-webui-additional-networks
LoRA models so far:
https://civitai.com/user/throwawayjm
Tips:
-Since LoRA models are mixed in real time, you can adjust the strength on the fly if you think the results are too much or too little.
-You can mix multiple LoRA models at once with the extension
-To train your own model, I currently recommend https://rentry.org/lora_train if you have at least 8GB of VRAM
Neg & Prompt:
Neg:
((bad prompt version2, bad-artist-anime, bad-artist, bad anatomy, too wet, too muscular, text, error, extra digit, fewer digits, cropped, out of frame, worst quality, low quality, artist name, bad hands, text, logo, error, missing fingers, bob cut, jpeg artifacts, signature, watermark, username, blush:1.4), (monochrome, multiple views, normal quality, blurry, depth of field, lowres:1.1), painting (medium), sketch, see-through, no humans, chibi, greyscale), black thighhighs, bad art, graphite (medium), poop sketch, pointy ears, (bad multiple views, ugly, duplicate), morbid, mutilated, extra limbs, gross proportions, mutated hands, poorly drawn hands, poorly drawn face, (mutation, deformed, bad proportions, extra limbs, cloned face, disfigured),
Prompt:
Recomended:
photo background, masterpiece, best quality, super realistic, high resolution, absurdres, bisexual vibrant ligths, perfect smooth translucent skin, photoshop, airbrush,
Example:
{% for type in ['red', 'green ', 'blue'] %}
{% prompt %}medieval {{ type }} playcard with description and portrait
{% if type == 'red' %}
{{ type }} rose at savanna
{% elif type == 'blue' %}
ocean and {{type}} rose
{% else %}
huge breasts in {{type}} jungle
{% endif %}
(anime screencap, photo background, anime coloring, masterpiece, best quality, huge breasts, huge ass, ({{ choice('mature', 'furry', 'elf', 'demon', 'monster', 'dragon') }} female, {{ choice('long', 'short') }} open {{ choice('black', 'blue', 'yellow', 'orange', 'green') }} {{ choice('jacket', 'sleeves', 'dress', 'shirt', 'coat', 'kimono') }}, feet, tight fit, ahegao, large areolae, (barefoot, {{ choice('clothing cutout', 'lactation through clothes', 'torn clothes', 'striped', 'ill-fitting clothes', 'clothes lift', 'shirt lift', 'open clothes', 'striped panties', 'striped thighhighs') }}, outdoors, {{ choice('exposing top', 'topless', 'exposing pussy', 'bottomless', 'completely nude') }}, {{ choice('cat girl / nekomimi', 'maid', 'ribbed sweater') }}, masterpiece, sequential, horizontal comic,, (anime screencap, photo background, anime coloring, masterpiece, best quality, breast expansion, milk cannon, huge breasts, huge ass, ({{ choice('mature', 'furry', 'elf', 'demon', 'monster', 'dragon') }} female, {{ choice('long', 'short') }} open {{ choice('black', 'blue', 'yellow', 'orange', 'green') }} {{ choice('jacket', 'sleeves', 'dress', 'shirt', 'coat', 'kimono') }}, feet, tight fit, ahegao, large areolae, (barefoot, {{ choice('clothing cutout', 'lactation through clothes', 'torn clothes', 'striped', 'ill-fitting clothes', 'clothes lift', 'shirt lift', 'open clothes', 'striped panties', 'striped thighhighs') }}, outdoors, {{ choice('exposing top', 'topless', 'exposing pussy', 'bottomless', 'completely nude') }}, {{ choice('cat girl / nekomimi', 'maid', 'ribbed sweater') }}, masterpiece, sequential, horizontal comic,
{% endprompt %}
{% endfor %}
How to use my prompt:

Embandings:
Bad_prompt negative embedding:
https://huggingface.co/datasets/Nerfgun3/bad_prompt/tree/main
Bad-artist negative embedding:
https://huggingface.co/nick-x-hacker/bad-artist/tree/main
HyperHavenOrange
Model Details
- Type
- AI Model
- Task
- text-to-image
- Subtype
- Safetensors / Checkpoint AI Model
- Created
- Updated
- June 6, 2026