Version 2.0
Load any short clip of a woman and atmosphere you'd like. The resolution I've been resizing to is 640 x 896, which is the recommended video resolution for outputs. Use my workflow
Link: SingularUnity Super Simple Workflow.
Finetuned: for better motion and longform videos. Slight loss in visual acuity but it's hardly noticeable.
Included camera motions: It's hit-or-miss but more likely with denoise 0.9+: zoom in, zoom out, and pan around.
So please try it out and post your results; it'll help me fine-tune this and generate better Loras in the future. THANK YOU!!
Version 1.0 Original Post.
This is a LoRA model built for Hunyuan Video, designed to render A red-haired, fair-skinned elf woman, natural curves in high-quality motion. Character inspiration from @ElvishGlade.
Still very much a work in progress. Posting it here to see what people are getting for further fine-tuning. So please try it out and post your results; it'll help me fine-tune this and generate better Loras in the future.
Kling not required.
Load the LoRA in your preferred Hunyuan workflow (V2V or I2V). ( Refer to my Super Simple HV2V - Now with UPSCALER and INTERPOLATION with Super Smooth videos.)
Load any short clip of a woman and atmosphere you'd like. The resolution I've been resizing to is 640 x 896, which is the recommended video resolution for outputs.
Use the trigger prompt(s):
A red-haired, fair-skinned elf womanA red-haired, fair-skinned elf woman, nude, with exposed vulvaA red-haired, fair-skinned elf woman with large, curly hair, wearing glassesA red-haired, fair-skinned elf woman with large, curly hair, wearing glasses, stands nude against a plain background, showcasing her slender physique and small breasts.Add your scene description and adjust CFG settings, steps, and denoise as needed.
Example prompt style:
A red-haired, fair-skinned elf woman, face down bottom up, presents her ample buttocks and vulva, nude, on a white bed. Showcasing her ample buttocks and vulva as she jiggles and bounces her hips. The camera captures her from behind, focusing in her vulva and buttocks as they bounce.A red-haired, fair-skinned elf woman in a standing leg split pose, wearing a shiny blue latex outfit and black lace stockings. showcasing her flexibility and toned physique. The background is plain white, emphasizing her form and pose.A red-haired, fair-skinned elf woman, face down bottom up, presents her ample buttocks and vulva, wearing black lingerie and fishnet stockings, on a white bed. Showcasing her ample buttocks and vulva as she jiggles and bounces her hips. The camera captures her from behind, focusing in her vulva and buttocks as they bounce.A red-haired, fair-skinned elf woman walks along the beach, her back to the camera, with the ocean waves crashing behind her.A red-haired, fair-skinned elf woman stands confidently, her long, wavy hair cascading down her back. She wears black fishnet stockings and a black bow around her waist, adding a touch of elegance to her appearance. Her expression is serious, and she gazes directly at the camera, exuding an air of mystery and allure.CFG Scale: 1
CFG values in this range significantly improve animation quality and coherence. Higher values require significantly more VRam and tend to over-constrain the model, leading to stiff or unnatural motion. If generating smaller resolutions like 544 x 720 try CFG 2 and 49 frames.
Negative Prompt:
low quality, bad hands, bad teeth, bad eyes, bad limbs, distortion, talking, speaking
Description
Finetune for better motion and long form videos. Slight loss in visual acuity, hardly noticeable.
FAQ
Comments (10)
Hell yeah!!!! V2 BABY- LETS GOOOOOOOO!!!!
day 35 of still no camel lora around.
This is the only that may comes remotely close.
unfortunately it has a strong bias toward disproportionate anatomical sizes that make half body looks like fat ass cows🤦♂️
.. 615mb? wtf? 🤣
Does it? perhaps change the input video and prompting to emphasise skinny thighs.
I just noticed the 615mb, haha it's a higher rank than previous Lora's I've made, which means larger number of trained parameters and larger file size.
Hi! Thanks fo your Workflow (Super Simple Workflows - v2.0)!
Is is possible to have prompt/workflow for this video: https://civitai.com/images/60394981
I try to do the same video, but in your workflow v2.0, its upscale of an existing video...
So I don't have the prompt or the workflow for the first video
PS: I start using Comfy and Hunyuan, sorry for noobs questions...
PS2: is it possible to leave the metadata in your videos so we can try generation too?
Thanks! :)
If you want to DM me your discord I'd be happy to help you out. I can look for that generation and send you meta with workflow, but I don't typically leave my meta in gens when posting them and I probably won't start.
On occasion folks have asked and I'm happy to share in that circumstance. In the future I will be providing detailed summaries and meta info to Patreon supporters. Until Civitai figures out its creator program this is the only mechanism aside from selling digital assets for creators to monetize. I enjoy releasing as much as possible for free but a completely free system unfortunately it's not sustainable unless Civitai is willing to pay creators by some other means.
@SingularUnity OK thanks! I will send you a PM
Hope this eventually gets a WAN version. This creates the cutest elf girls and WAN really struggles with both cute and elves.
Is there a PONY or XL lora for the elf in your videos? How was she generated?
Can you make a Wan 2.2 version?