Version 2.0 - LTX 2.3/I2V/T2V
Version 2.0, LTX 2.3 ready, I2V and T2V
I run about 417 frames when I generate videos, the training set assumes video lengths of about 10-15 seconds to fit the motions trained.
If you make a good video with the lora all I ask is you post it to the gallery, share it as that is what helps me improve the lora in the next training session. I hope everyone enjoys this lora, any input you give me or suggestions ill take for the next version
I feel LTX 2.3 does better with a strong starting image (I2V) but both work. This version 2.0 was trained to do both T2V and I2V modes at the same time.
Version 2.0 was trained on 105 high res nude female videos in various positions and lighting conditions focused on female anatomy (breast jiggle, vulva, nipples, natural ass movement , walking, showering, masturbating etc). it should be a step up over version 1. further training and refinement on more diverse datasets will come next as this lora is improved. trigger word is 'LTXNUDES'.
Bouncing Tits:
"young nude woman jogging on the spot outdoors in public park, high-energy motion causing large natural breasts to bounce and jiggle wildly up and down with each step, heavy sag and rebound, erect pink nipples swaying prominently, wide pale areolas textured and bumpy"
Trigger word: LTXNUDES
Choose a workflow:
Work Flow 1: https://civarchive.com/models/2354193/ltx-23-all-in-one-workflow-for-rtx-3060-with-12-gb-vram-32-gb-ram
Workflow 2: https://civarchive.com/models/2445871/ltx-23-i2v-fp8bf16
Training captions:
close-up low-angle view of young nude woman on bed
smoothed shaved pussy
fingers parting slick outer labia and circling swollen clitoris
baby oil all over her toned body
actively thrusts her hups up and down in steady rhythm
duldo fully inserted into her shaved pussy
she energetically bounces up and down
breasts jiggling
striking large breasts
massages her large natural breasts with both hands
fingers circling nipples
she teases the camera by slowly running her hands all over her body
close-up of a beautiful woman's shaved pussy as she masturbates
small areolas and perky breasts
realistic dildo mounted on the floor
she firmly squeezes her large breasts together
stands in steamy shower under warm running water
intense pleasure expression
intimate erotic self-pleasure scene
both hands gently spreading vulva wide open, parting her labia fully to reveal pink inner folds
intimate risky public flash
using one hand to gently spread her firm butt cheeks wide apart
she slowly turns her body to the side in a sensual sway, hips twisting gracefully
one finger inserted deeply into her tight anus
fully exposing tight pink anus and wet shaved vulva from behind
large pale natural breasts with puffy pink areolas and small erect nipple
one hand sliding up along her side and cupping her breast teasingly
upward and over her large augmented breasts to fully expose them
nude woman on all fours outside
soft belly and wide hips
labia spreading motion, clitoris detail, arousal wetness, and wide-spread legs pose
thick outer labia majora glistening with arousal fluids
reaching back with both hands to spread large plump buttocks wide apart
close-up low-angle view of young woman nude from waist down
Description
version 1.5 was trained on 85 high res videos of nude females, with larger bucket size and rank 64. Better overall motion and anatomy than version 1.0. examples of training captions: masturbating with dildo, natural breasts, laying down, shaved pussy, indoor lighting, detailed vulva, on all fours, bending over, ass spreading, anus visible, nude woman, shaved pussy, indoor lighting, detailed vulva, detailed labia, baby oil on body, large natural breasts, nude woman, shaved pussy, indoor lighting, detailed vulva, detailed labia, rubbing fingers on breasts
FAQ
Comments (36)
will upload some tests soon <3
appreciate it! not easy making a lora, still needs lots of refinement
1.5 looks reallllly good!
thanks! Still needs more generalization, will train it on larger dataset with more angles and scenes but its getting the broad nudity part right now
@sexgod1979 will try to check if it can work on distilled wf later today in i2v mode. got small abliterated gemma around 7 gb that works well with gemma loader so will test that as well.
@sexgod1979 does it work on distilled,? As it takes forever at 720p 10sec on a 5090. Took me almost 15 minutes
@ylvlylvuyv374212 It SHOULD NOT be taking 15 minutes to generate a video on LTX-2 of a 5090, Im geting 2-5s/it on stage 1, and 10-30s/it on stage 2. videos at 768x1280 take less than 3-5 minutes to render even at 40 steps in phase 1. Something is off with your workflow. start with the official I2V workflow, hit me up on chat and ill help
@sexgod1979 oh I am using T2V. Is this only for I2V?
I keep defying deformation and artifacts
Hello. Are distilled models supported?
yes it works with the fp8-dev and the distilled version, I haven't tested GGUF yet but I dont see why it shouldnt as its a standard LTX-2 lora format
@sexgod1979 distilled. Its not following prompts at all. Gemma abliterated. It switches genders entirely and give boobs to men.
Although I am using T2V. Is this only for I2V?
@ylvlylvuyv374212 what happens if you use an image with say a nude woman on it and have her do something?
@sexgod1979 I am only doing T2V I have no tried I2V
@ylvlylvuyv374212 the base model has bias , by doing T2V I suspect it will be more challenging, try I2V first feed it a nude image or any image with a women clothed first, and see if you get different results, make sure the lora is in both model1 and 2 workflow paths to the sampler
@ylvlylvuyv374212 one of the videos in the gallery for v1.5 has a t2v prompt example and they got the woman to take her shirt off riding a dirt bike, so maybe the workflow is off?
@ylvlylvuyv374212 also FYI this is not a sex act lora, it only knows nudity. It doesnt know how to give a blowjob or do sex at all, it wasnt trained on it at all
Guide from the LTX team on how to train your own lora, they specialized their own free training system: https://www.youtube.com/watch?v=sL-T6dsO0v4
faces change alot, fake eyelashes, always caucasian, typical pornstar look. deleted it after a few tries
It does seem to impact the face, throwing off character loras a bit.
This isn't doing pubic hair for me. What do I need to do to get it?
Disgusting hairy cunts. That'll work.
@axicec What is your problem?
Great LORA! I'm really looking forward to the LTX 2.3 version of LORA. Thank you!
yes! me too, work starts soon
LTX 2.3 VERSION, please!! Thanks!
yes me too, work starts soon!
I do not understand how to make this work. I have downloaded usual template from ComfyUI for LTX2.3 replaced the Lora with your and I got a distorted pixalated non NSWF video. Is there a workflow I can use somewhere ???
it was trained for image to video I2V, give it a strong input image ideally with a nude or partially nude woman, dont rely on text prompt alone. A newer version of this lora is going to be released soon, so if this one is giving problems wait for the next verison and try again
I've definitely used this lora with LTX 2.3. I use the fp8/fp16 workflow for LTX 2.3 here: https://civitai.com/models/2445871/ltx-23-i2v-fp8bf16 and it works great for me with this lora, but I give it a starting image always. LTX 2.3 did change things and this lora was trained on LTX-2 not 2.3. I will release a LTX 2.3 lora as soon as im finished it
@sexgod1979 Why not train T2V as well? Otherwise I think it will have issues with new things entering the scene even in I2V
@kangaru861 totally hearing what you are saying! version 1.7 is trained on both I2V and T2V but is still better overall with I2V. I am going to train a version that is purely T2V native next. Stay tuned, we got this!
@sexgod1979 No, no, Laura needs i2v for undressing and other things, t2v is also interesting, but not as much as i2v, so we're waiting for i2v for itx 2.3
@sexgod1979 By the way, for some reason, my Laura is moving at an accelerated pace, as if in fast-forward mode. Is this a bug? I'm using the distilled version with 8 steps.
@soyv4 I havent seen that myself, im also using distilled lora but I am doing 377 frames. If you do low frames say like 81 then yes it would like super fast because it has to compress all the motions in this lora into short period. The lora was trained with the intention of 300-400 frame counts of motion. If you use the workflow from one of my videos do you get the same thing with that frame count?
You must not "replace" the distilled lora that's in the template. You need to chain a second lora loader behind it to apply both, the distilled AND this lora here. Without the distilled lora the video will be distorted.