How to Use
This is a style LoRA, designed to be the base layer of your LoRA stack. It creates the foundational aesthetic of realism, upon which you can add character or concept LoRAs.
Note: The included ZIP archive contains both the high-noise and low-noise LoRA variants, along with our recommended ComfyUI workflows.
Trigger Word:
InstacamRecommended Strength:
1.0. Start here and adjust in small increments.
I would also like to thank Danrisi, who originally taught us how to train LoRAs and helped make our work possible.
Description
Instagirl V2 is a complete overhaul, trained from the ground up to push the boundaries of realism on Wan 2.2.
Unprecedented Realism: We've massively upgraded the dataset and training method with a focus on photorealistic skin textures, natural lighting, and flawless environmental details.
Greater Diversity: V2 is trained on a much wider and more inclusive dataset, enabling a greater variety of faces, ethnicities, and styles right out of the box.
Better Composition: The model now has a deeper understanding of world composition, resulting in more coherent and believable scenes with fewer artifacts.
This model is trained on WAN 2.2, meaning there are two versions; a high and low noise variant.
FAQ
Comments (65)
possible make some young Asian girl next time?
The workflow in the file: Credits to model creator:
{"id":"11e19437-56ee-487e-bb1f-3bd053f2f3c2","revision":0,"last_node_id":74,"last_link_id":136,"nodes":[{"id":68,"type":"LoraLoader","pos":[3345.027099609375,-345.39691162109375],"size":[259.98895263671875,126],"flags":{},"order":6,"mode":0,"inputs":[{"localized_name":"model","name":"model","type":"MODEL","link":135},{"localized_name":"clip","name":"clip","type":"CLIP","link":121},{"localized_name":"lora_name","name":"lora_name","type":"COMBO","widget":{"name":"lora_name"},"link":null},{"localized_name":"strength_model","name":"strength_model","type":"FLOAT","widget":{"name":"strength_model"},"link":null},{"localized_name":"strength_clip","name":"strength_clip","type":"FLOAT","widget":{"name":"strength_clip"},"link":null}],"outputs":[{"localized_name":"MODEL","name":"MODEL","type":"MODEL","links":[134]},{"localized_name":"CLIP","name":"CLIP","type":"CLIP","links":[124,125]}],"properties":{"cnr_id":"comfy-core","ver":"0.3.46","Node name for S&R":"LoraLoader","widget_ue_connectable":{}},"widgets_values":["Instagirlv5_high_noise.safetensors",1,1]},{"id":70,"type":"LoraLoaderModelOnly","pos":[3342.34765625,-164.2708282470703],"size":[270,82],"flags":{},"order":7,"mode":0,"inputs":[{"localized_name":"model","name":"model","type":"MODEL","link":136},{"localized_name":"lora_name","name":"lora_name","type":"COMBO","widget":{"name":"lora_name"},"link":null},{"localized_name":"strength_model","name":"strength_model","type":"FLOAT","widget":{"name":"strength_model"},"link":null}],"outputs":[{"localized_name":"MODEL","name":"MODEL","type":"MODEL","links":[128]}],"properties":{"cnr_id":"comfy-core","ver":"0.3.46","Node name for S&R":"LoraLoaderModelOnly","widget_ue_connectable":{}},"widgets_values":["wanWan21_T2V_14B_lightx2v_cfg_step_distill_lora_rank32.safetensors",0.6000000000000001]},{"id":72,"type":"LoraLoaderModelOnly","pos":[3348.833251953125,103.12315368652344],"size":[270,82],"flags":{},"order":12,"mode":0,"inputs":[{"localized_name":"model","name":"model","type":"MODEL","link":129},{"localized_name":"lora_name","name":"lora_name","type":"COMBO","widget":{"name":"lora_name"},"link":null},{"localized_name":"strength_model","name":"strength_model","type":"FLOAT","widget":{"name":"strength_model"},"link":null}],"outputs":[{"localized_name":"MODEL","name":"MODEL","type":"MODEL","links":[132]}],"properties":{"cnr_id":"comfy-core","ver":"0.3.46","Node name for S&R":"LoraLoaderModelOnly","widget_ue_connectable":{}},"widgets_values":["l3n0v0.safetensors",0.6000000000000001]},{"id":4,"type":"CLIPTextEncode","pos":[3642.8974609375,58.09259796142578],"size":[450,150],"flags":{"collapsed":false},"order":9,"mode":0,"inputs":[{"localized_name":"clip","name":"clip","type":"CLIP","link":125},{"localized_name":"text","name":"text","type":"STRING","widget":{"name":"text"},"link":null}],"outputs":[{"localized_name":"CONDITIONING","name":"CONDITIONING","type":"CONDITIONING","links":[59,64]}],"title":"Negative Prompt","properties":{"cnr_id":"comfy-core","ver":"0.3.43","Node name for S&R":"CLIPTextEncode","widget_ue_connectable":{"text":true}},"widgets_values":["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, censored, sunburnt skin, rashy skin, red cheeks\n"],"color":"#2a363b","bgcolor":"#3f5159"},{"id":5,"type":"EmptyHunyuanLatentVideo","pos":[3647.158935546875,260.2233581542969],"size":[450,150],"flags":{},"order":0,"mode":0,"inputs":[{"localized_name":"width","name":"width","type":"INT","widget":{"name":"width"},"link":null},{"localized_name":"height","name":"height","type":"INT","widget":{"name":"height"},"link":null},{"localized_name":"length","name":"length","type":"INT","widget":{"name":"length"},"link":null},{"localized_name":"batch_size","name":"batch_size","type":"INT","widget":{"name":"batch_size"},"link":null}],"outputs":[{"localized_name":"LATENT","name":"LATENT","type":"LATENT","links":[60]}],"properties":{"cnr_id":"comfy-core","ver":"0.3.43","Node name for S&R":"EmptyHunyuanLatentVideo","widget_ue_connectable":{"width":true,"height":true,"length":true,"batch_size":true}},"widgets_values":[1088,1440,1,1],"color":"#323","bgcolor":"#535"},{"id":67,"type":"Seed Generator","pos":[3659.67333984375,464.1074523925781],"size":[270,82],"flags":{},"order":1,"mode":0,"inputs":[{"localized_name":"seed","name":"seed","type":"INT","widget":{"name":"seed"},"link":null}],"outputs":[{"localized_name":"INT","name":"INT","type":"INT","links":[115,116]}],"properties":{"cnr_id":"comfy-image-saver","ver":"65e6903eff274a50f8b5cd768f0f96baf37baea1","Node name for S&R":"Seed Generator","widget_ue_connectable":{}},"widgets_values":[669817269655594,"increment"]},{"id":36,"type":"KSamplerAdvanced","pos":[4392.58154296875,-335.0544738769531],"size":[250,652.4117431640625],"flags":{},"order":13,"mode":0,"inputs":[{"localized_name":"model","name":"model","type":"MODEL","link":132},{"localized_name":"positive","name":"positive","type":"CONDITIONING","link":63},{"localized_name":"negative","name":"negative","type":"CONDITIONING","link":64},{"localized_name":"latent_image","name":"latent_image","type":"LATENT","link":61},{"localized_name":"add_noise","name":"add_noise","type":"COMBO","widget":{"name":"add_noise"},"link":null},{"localized_name":"noise_seed","name":"noise_seed","type":"INT","widget":{"name":"noise_seed"},"link":116},{"localized_name":"steps","name":"steps","type":"INT","widget":{"name":"steps"},"link":null},{"localized_name":"cfg","name":"cfg","type":"FLOAT","widget":{"name":"cfg"},"link":null},{"localized_name":"sampler_name","name":"sampler_name","type":"COMBO","widget":{"name":"sampler_name"},"link":null},{"localized_name":"scheduler","name":"scheduler","type":"COMBO","widget":{"name":"scheduler"},"link":null},{"localized_name":"start_at_step","name":"start_at_step","type":"INT","widget":{"name":"start_at_step"},"link":null},{"localized_name":"end_at_step","name":"end_at_step","type":"INT","widget":{"name":"end_at_step"},"link":null},{"localized_name":"return_with_leftover_noise","name":"return_with_leftover_noise","type":"COMBO","widget":{"name":"return_with_leftover_noise"},"link":null}],"outputs":[{"localized_name":"LATENT","name":"LATENT","type":"LATENT","links":[65]}],"properties":{"cnr_id":"comfy-core","ver":"0.3.46","Node name for S&R":"KSamplerAdvanced","widget_ue_connectable":{"add_noise":true,"noise_seed":true,"steps":true,"cfg":true,"sampler_name":true,"scheduler":true,"start_at_step":true,"end_at_step":true,"return_with_leftover_noise":true}},"widgets_values":["enable",100,"fixed",10,1,"res_2s","beta57",4,999,"disable",""],"color":"#223","bgcolor":"#335"},{"id":22,"type":"CLIPLoader","pos":[3014.672119140625,-88.59371185302734],"size":[302.9598083496094,125.00395202636719],"flags":{},"order":2,"mode":0,"inputs":[{"localized_name":"clip_name","name":"clip_name","type":"COMBO","widget":{"name":"clip_name"},"link":null},{"localized_name":"type","name":"type","type":"COMBO","widget":{"name":"type"},"link":null},{"localized_name":"device","name":"device","shape":7,"type":"COMBO","widget":{"name":"device"},"link":null}],"outputs":[{"localized_name":"CLIP","name":"CLIP","type":"CLIP","links":[121]}],"properties":{"cnr_id":"comfy-core","ver":"0.3.43","Node name for S&R":"CLIPLoader","widget_ue_connectable":{"clip_name":true,"type":true,"device":true}},"widgets_values":["umt5_xxl_fp8_e4m3fn_scaled.safetensors","wan","default"],"color":"#2a363b","bgcolor":"#3f5159"},{"id":35,"type":"KSamplerAdvanced","pos":[4114.67236328125,-337.7133483886719],"size":[250,652.4117431640625],"flags":{},"order":11,"mode":0,"inputs":[{"localized_name":"model","name":"model","type":"MODEL","link":134},{"localized_name":"positive","name":"positive","type":"CONDITIONING","link":58},{"localized_name":"negative","name":"negative","type":"CONDITIONING","link":59},{"localized_name":"latent_image","name":"latent_image","type":"LATENT","link":60},{"localized_name":"add_noise","name":"add_noise","type":"COMBO","widget":{"name":"add_noise"},"link":null},{"localized_name":"noise_seed","name":"noise_seed","type":"INT","widget":{"name":"noise_seed"},"link":115},{"localized_name":"steps","name":"steps","type":"INT","widget":{"name":"steps"},"link":null},{"localized_name":"cfg","name":"cfg","type":"FLOAT","widget":{"name":"cfg"},"link":null},{"localized_name":"sampler_name","name":"sampler_name","type":"COMBO","widget":{"name":"sampler_name"},"link":null},{"localized_name":"scheduler","name":"scheduler","type":"COMBO","widget":{"name":"scheduler"},"link":null},{"localized_name":"start_at_step","name":"start_at_step","type":"INT","widget":{"name":"start_at_step"},"link":null},{"localized_name":"end_at_step","name":"end_at_step","type":"INT","widget":{"name":"end_at_step"},"link":null},{"localized_name":"return_with_leftover_noise","name":"return_with_leftover_noise","type":"COMBO","widget":{"name":"return_with_leftover_noise"},"link":null}],"outputs":[{"localized_name":"LATENT","name":"LATENT","type":"LATENT","links":[61]}],"properties":{"cnr_id":"comfy-core","ver":"0.3.46","Node name for S&R":"KSamplerAdvanced","widget_ue_connectable":{"add_noise":true,"noise_seed":true,"steps":true,"cfg":true,"sampler_name":true,"scheduler":true,"start_at_step":true,"end_at_step":true,"return_with_leftover_noise":true}},"widgets_values":["enable",100,"fixed",10,1,"res_2s","beta57",0,4,"disable",""],"color":"#223","bgcolor":"#335"},{"id":8,"type":"VAELoader","pos":[3099.625732421875,84.64628601074219],"size":[210,58],"flags":{},"order":3,"mode":0,"inputs":[{"localized_name":"vae_name","name":"vae_name","type":"COMBO","widget":{"name":"vae_name"},"link":null}],"outputs":[{"localized_name":"VAE","name":"VAE","type":"VAE","links":[9]}],"properties":{"cnr_id":"comfy-core","ver":"0.3.43","Node name for S&R":"VAELoader","widget_ue_connectable":{"vae_name":true}},"widgets_values":["wan_2.1_vae.safetensors"],"color":"#323","bgcolor":"#535"},{"id":9,"type":"VAEDecode","pos":[4427.91357421875,361.8782653808594],"size":[211.20114135742188,46],"flags":{},"order":14,"mode":0,"inputs":[{"localized_name":"samples","name":"samples","type":"LATENT","link":65},{"localized_name":"vae","name":"vae","type":"VAE","link":9}],"outputs":[{"localized_name":"IMAGE","name":"IMAGE","type":"IMAGE","links":[8]}],"properties":{"cnr_id":"comfy-core","ver":"0.3.43","Node name for S&R":"VAEDecode","widget_ue_connectable":{}},"widgets_values":[],"color":"#323","bgcolor":"#535"},{"id":10,"type":"SaveImage","pos":[4669.361328125,-330.9233703613281],"size":[788.8705444335938,772.3082275390625],"flags":{},"order":15,"mode":0,"inputs":[{"localized_name":"images","name":"images","type":"IMAGE","link":8},{"localized_name":"filename_prefix","name":"filename_prefix","type":"STRING","widget":{"name":"filename_prefix"},"link":null}],"outputs":[],"properties":{"cnr_id":"comfy-core","ver":"0.3.43","Node name for S&R":"SaveImage","widget_ue_connectable":{"filename_prefix":true}},"widgets_values":["ComfyUI"],"color":"#233","bgcolor":"#355"},{"id":71,"type":"LoraLoaderModelOnly","pos":[3346.9189453125,-28.68145751953125],"size":[270,82],"flags":{},"order":10,"mode":0,"inputs":[{"localized_name":"model","name":"model","type":"MODEL","link":128},{"localized_name":"lora_name","name":"lora_name","type":"COMBO","widget":{"name":"lora_name"},"link":null},{"localized_name":"strength_model","name":"strength_model","type":"FLOAT","widget":{"name":"strength_model"},"link":null}],"outputs":[{"localized_name":"MODEL","name":"MODEL","type":"MODEL","links":[129]}],"properties":{"cnr_id":"comfy-core","ver":"0.3.46","Node name for S&R":"LoraLoaderModelOnly","widget_ue_connectable":{}},"widgets_values":["WAN2.2_LowNoise_Instagirl_V2_Pruned.safetensors",1]},{"id":74,"type":"UnetLoaderGGUF","pos":[3042.22900390625,-197.35244750976562],"size":[270,58],"flags":{},"order":5,"mode":0,"inputs":[{"localized_name":"unet_name","name":"unet_name","type":"COMBO","widget":{"name":"unet_name"},"link":null}],"outputs":[{"localized_name":"MODEL","name":"MODEL","type":"MODEL","links":[136]}],"properties":{"cnr_id":"ComfyUI-GGUF","ver":"b3ec875a68d94b758914fd48d30571d953bb7a54","widget_ue_connectable":{},"Node name for S&R":"UnetLoaderGGUF"},"widgets_values":["Wan2.2-T2V-A14B-LowNoise-Q8_0.gguf"]},{"id":73,"type":"UnetLoaderGGUF","pos":[3041.61181640625,-302.9368591308594],"size":[270,58],"flags":{},"order":4,"mode":0,"inputs":[{"localized_name":"unet_name","name":"unet_name","type":"COMBO","widget":{"name":"unet_name"},"link":null}],"outputs":[{"localized_name":"MODEL","name":"MODEL","type":"MODEL","links":[135]}],"properties":{"cnr_id":"ComfyUI-GGUF","ver":"b3ec875a68d94b758914fd48d30571d953bb7a54","widget_ue_connectable":{},"Node name for S&R":"UnetLoaderGGUF"},"widgets_values":["Wan2.2-T2V-A14B-HighNoise-Q8_0.gguf"]},{"id":3,"type":"CLIPTextEncode","pos":[3637.72607421875,-338.9566955566406],"size":[450,350],"flags":{},"order":8,"mode":0,"inputs":[{"localized_name":"clip","name":"clip","type":"CLIP","link":124},{"localized_name":"text","name":"text","type":"STRING","widget":{"name":"text"},"link":null}],"outputs":[{"localized_name":"CONDITIONING","name":"CONDITIONING","type":"CONDITIONING","links":[58,63]}],"title":"Positive Prompt","properties":{"cnr_id":"comfy-core","ver":"0.3.43","Node name for S&R":"CLIPTextEncode","widget_ue_connectable":{"text":true}},"widgets_values":[""],"color":"#2a363b","bgcolor":"#3f5159"}],"links":[[8,9,0,10,0,"IMAGE"],[9,8,0,9,1,"VAE"],[58,3,0,35,1,"CONDITIONING"],[59,4,0,35,2,"CONDITIONING"],[60,5,0,35,3,"LATENT"],[61,35,0,36,3,"LATENT"],[63,3,0,36,1,"CONDITIONING"],[64,4,0,36,2,"CONDITIONING"],[65,36,0,9,0,"LATENT"],[115,67,0,35,5,"INT"],[116,67,0,36,5,"INT"],[121,22,0,68,1,"CLIP"],[124,68,1,3,0,"CLIP"],[125,68,1,4,0,"CLIP"],[128,70,0,71,0,"MODEL"],[129,71,0,72,0,"MODEL"],[132,72,0,36,0,"MODEL"],[134,68,0,35,0,"MODEL"],[135,73,0,68,0,"MODEL"],[136,74,0,70,0,"MODEL"]],"groups":[],"config":{},"extra":{"ds":{"scale":0.7756848205605316,"offset":[-2564.324446129299,766.807977939455]},"frontendVersion":"1.23.4","ue_links":[],"links_added_by_ue":[],"VHS_latentpreview":true,"VHS_latentpreviewrate":0,"VHS_MetadataImage":true,"VHS_KeepIntermediate":true},"version":0.4}
hi, thanks for the lora ! :)
a question, what does it mean: "harmful content" ?
can i create nsfw (xxx or x) content with it or not?
edit: after generating some images now i understand your point haha! thanks anyway though!
Could you post a workflow using the native WAN 2.2 instead of using all the 2.1 loras that have been known to not be super compatible with the Wan 2.2 weights?
I'm assuming this is mainly a low noise model, what strength should one put for high noise? both 1?
Edit: I'm leaving it for the others with the same question, I'm dumb, download it and your questions will be answered
Hi, thanks for the update. Could you please upload the safetensor and the workflow separately? I'm using API to download lora's remotely and the output is a .zip file while its expecting a .safetensor. Thanks in advance
It's not possible, I would have done it that way otherwise
Does anyone have any recommendations on which Wan 2.2 model is best with 8gb VRAM?
Try out the different Q4 2.2 models, then work your way down to smaller sizes if it's too much. Q4 should work fine if you got enough system DRAM. I've use models over 16.5gb on my 12VRAM with optimized workflow and good amount of system RAM. I recommend at starting 64gb ram for your PC but should scrape by with 32gb. I go over 90gb system RAM on some workflows. There is also WANGP 2.2 one click easy installer on Pinokio UI which is great for low VRAM users.
Please, help me
Prompt outputs failed validation: KSamplerAdvanced: - Value not in list: sampler_name: 'res_2s' not in (list of length 40) - Value not in list: scheduler: 'beta57' not in ['simple', 'sgm_uniform', 'karras', 'exponential', 'ddim_uniform', 'beta', 'normal', 'linear_quadratic', 'kl_optimal'] KSamplerAdvanced: - Value not in list: sampler_name: 'res_2s' not in (list of length 40) - Value not in list: scheduler: 'beta57' not in ['simple', 'sgm_uniform', 'karras', 'exponential', 'ddim_uniform', 'beta', 'normal', 'linear_quadratic', 'kl_optimal']
hardnon thanks
hardnon hey, do you know if its possible to install these files if you have the pinokio version of comfyui? i can not find a venv folder anywhere.
ducky66 No idea, I just fly by the seat of my pants. I don't know more than just what allows me to get things working.
ducky66 oh you have to make a venv folder.... its a virtual environment you have to create through your Terminal.
In the comfyui root folder I just run each line:
right click in ComfyUI -> Open in terminal
py -3.12 -m venv venv
cd venv/Scripts
./Activate.ps1
cd ../..
This would go crazy on flux, hope you can do it someday
Please stop polluting this space with broken workflows that don't work. Your discord automatically bans people who ask about the broken workflow or missing models. What is a l3n0v0.safetensors? This sort of behavior should be banned from this community.
It is referring to this lora:
https://civitai.com/models/1662740/i-dunno-how-to-call-this-lora-ultrareal
TesterFranz That helps now. However it doesn't solve the underlying problem.
wow the audacity to complain about something like this that is offered literally for free to you. what did you contribute? anything at all? of course your profile is completely empty. your ban already seems more than justified. what kind of low and useless human waste you are.
engineer There is no underlying problem. I loaded the workflow after downloading ALL of the loras listed and it worked perfectly.
theloraprodigy I build a table. I give it to you for free. You set up a meal on it. The table collapses. Would you not at least raise an eyebrow and ask me to improve my carpentry skills? Heck, I would want to know.
He literally links the lenovo model in the lora's description.
Lol are people illiterate now? Can you not take 2 mins to read the description first before complaining?
Workflow works fine for me... don't know why you're having issues
@engineer wow, just wow. that analogy was jaw-dropping. NO, I WOULD NOT ask you to improve your carpenter skills. I would even hesitate to tell you it broke. I would just be glad you actually took the time to build a table and even more so, GIVE IT TO ME FOR FREE. I would not raise a single hair on any brow and I would just be grateful for your efforts and try to fix it myself if I wanted to keep it. How the hell are you reasoning? Disgraceful thinking.
Are you able to upload the LORAs separately? I use a Civitdownloader with ComfyUI because my upload speed isn't very good, and I use runpod so I have to upload these which takes hours.
you know you can just download the models directly in runpod right?
click on download, once download starts on your browser, copy the download link from downloads tab of your browser and then got to runpod terminal , cd into loras folder then run the below command with the download link updated with yours, update the link in quote with what you have copied from browser, my link wont work, it has a expiry token, so you have to copy the link which you get from the browser -
wget -O insta.zip "https://civitai-delivery-worker-prod.5ac0637cfd0766c97916cefa3764fbdf.r2.cloudflarestorage.com/model/7053464/instagirlv2202B.hgkp.zip?X-Amz-Expires=86400&response-content-disposition=attachment%3B%20filename%3D%22Instagirlv2%20%2B%20workflow.zip%22&X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=e01358d793ad6966166af8b3064953ad/20250808/us-east-1/s3/aws4_request&X-Amz-Date=20250808T121239Z&X-Amz-SignedHeaders=host&X-Amz-Signature=babb3ad4d2c9ad889fe95715498a289401abcc5237707b8914f84c5d7bcb2715"
To upload a LoRa much faster to Jupyter in Runpod:
Go to terminal
pip install gdown
cd /workspace/ComfyUI/models/loras
gdown --fuzzy PASTE-LORA-GOOGLE-DRIVE-LINK
Is it possible to just make standard gens ?
Nice loras. Unusual sampler and scheduler - res_2s and beta57. Where can I get them?
have you found it. I am struggling to install it
marqs89 hey, i have comfyui installed from pinokio and i do not have a venv folder, so im lost how to install this. i even downloaded the "portable version" and tried installing it there and when following instructions on the page your linking, it gives error, i ran the command in the custom nodes folder and it appeared to work, but the samplers are still not showing up.
You are the real deal mate. Great lora!
trying to run the example comfy workflow. everything works, only it seems like output resolution is lower compared to examples. any ideas?
You can change the output resolution in the workflow
is this video model?
robinhud5738936 what is a video if not a series of images? yes, models trained on images work for video.
@playtime_ai when I'm using your workflow it just generates black images, no errors at all...
This looks great! Side note; Uploading a LoRA trained on WAN and presumably a dataset you don't own, then slapping a restrictive license on is not legally sound.
Only one negative thing about this Lora. It tries to add a nose ring way to often, but then it's mostly only a part of a nose ring
try adding a negative prompt for nose ring
is this trained on photos or videos? ty
Mostly likely photos I imagine. Gooning for high quality videos of girls on insta-sluts is rather rare. Hopefully OP will make a Pinterest/tiktok girls Wan lmao. I've made a flux Pinterest girls but I wouldn't release that lol
great one, where do I find the sampler and the scheduler
marqs89 thanks mate. it is taking lots of time for 81 frames, it is been 30 mins on a6000 48GB with sampler euler and scheduler simple, is it gonna be better with the configured sampler and scheduler?
robinhud5738936 I did not try videos with this workflow, only pics. Will try it later and give you feedback! The res sampler will likely take a bit longer but it's worth it! But if you generate 81 frames with such a big resolution and 10 steps it will likely take very long yes. You will likely have to either reduce resolution, steps or frames. Or all of it ;)
robinhud5738936 so i did 81 frames, 480x720, 10 steps, res_2s sampler -> 450 seconds on a 4090. Quality obviously not very good with that resolution. Wan 2.2 is awesome but using two models takes time sadly. Oh btw, 81 frames is not 5 seconds bc the videos are 24 FPS so if you want 5 seconds you have to do 24*seconds+1=121 frames, which takes even longer of course :D
marqs89 14b fps is 16 - 5b fps is 24
dasfajhi i use the 14b and if i use 81fps i only get like 3 seconds?
EDIT: Forget it, i'm just a dummy :D
You are correct!
marqs89 you can use the Film VFI node at the end of the generation process to interpolate frames, taking it from the native 16fps to 32fps without too much additional compute.
how do i train a charchter lora for this ?
How do I even run this LoRA as of now both replicate and Fal don't provide a Wan2.2 lora model to run
Great LoRA.
What's up with the license?
This is a derivative of Wan.
Do you own the rights to the dataset used? if not, your license is as good as solid food for my dentureless grandma
Nobody will know how you will use the model, those licenses are useless anyway.
I'm assuming it's just in case some scammy company uses it or something
It's a LoRa. Saying that they can't license it because it was trained on another model is like saying that you can't sell the tires you hand crafted and built yourself to fit a Ford because you don't own Ford motor company.
wyldhunt You're right that creating something compatible with another product doesn't automatically restrict its sale. But the LoRA situation isn't quite like crafting tires--it’s more like modifying a car engine using proprietary parts and then trying to resell the whole thing under your own terms.
LoRAs are trained on base models like Wan 2.2, which themselves may have licensing restrictions. If the base model or its training data isn't owned or fully licensed by the LoRA creator, then slapping a restrictive license on the derivative work can be legally shaky. It's not just about compatibility--it's about derivative rights and the provenance of the data used.
The right-to-repair analogy works better when you're modifying something you own. But in this case, if the LoRA was trained on a model or dataset the creator doesn't have rights to, then asserting licensing control over the output is like selling a remix of a song you don’t own the rights to--regardless of how much you tweaked the bassline.
So while the spirit of open innovation is admirable, licensing in the AI space hinges on data ownership, model provenance, and derivative rights. That’s why some folks are raising eyebrows about the restrictive license here.
AdaptiveVision Depends. If an investigation happens somewhere and ask for proofs, or forget to hide the meta, they can just paste the image/video in comyfui and see what you used :P
Once they they will scan stuffs with AI and see the fingerprint of everything used haha
yes.wan2.2 got the license of internet videos than give you wan2.2
Alright123 All you need to do is save the image without metadata.
@knigitz I can't agree. You base it off of the model to match the tensor shapes. In your analogy, that would be like saying that you made something to fit a Ford V8 engine. There is nothing proprietary about the tensors or the shape of the LoRa. The LoRa is its own model, with its own data, using a tensor pattern that is compatible with a specific model. How could someone claim ownership because their model uses the pattern that the LoRa was designed to fit?