他サイトに転載・再配布しないで下さい。営利目的も厳禁です。
请勿转载或转发至其他网站。 商业目的也被严格禁止。
다른 사이트에 전재하지 마십시오.상업 이용 엄금입니다.
Please do not reprint or redistribute it to other sites. Commercial purposes are also strictly prohibited.
Wan 2.1 ComfyUI repackaged
Hunyuan Video Comfy-Diffusion-BF16
SD 1.5 BRA v6.0
Wan2.1/Hunyuan Video Lora Trainning with GUI in Windows
Description
parameters
masterpiece,beautiful,extremely detailed,finely detailed,exquisite,(ultra high res),(8k UHD best quality), (realistic, photo-realistic:1.37), depth_of,blurry_background, photo_inset,Dramatic lighting,reference_inset,photo_background,photorealistic,looking_at_viewer,1girl,detailed skin,shiny skin,kanna<lora:kanna hashimoto:1>
Negative prompt: (((nude))),(((naked))),worst quality, low quality:1.2,painting,sketches,logo, watermark, text,(worst quality:2),(low quality:2),(normal quality:2),lowers,normal quality,((monochrome)),((grayscale)),skin spots,acnes,skin blemishes,age spot,nsfw,ugly face,breasts out,(((esotropia))),(((squint))),hat,(((cleavage))),hide the chest,park,
Steps: 20, Sampler: DPM++ 2M Karras, CFG scale: 7, Seed: 3578797827
FAQ
Comments (4)
可以给一下训练参数吗
我是参考YouTube开始玩LoRA的，所以对参数等不太了解。 我会发布它以防万一。
--train_batch_size=2
--max_train_epochs=10
--optimizer_type=AdamW8bit
--learning_rate=1e-4
--network_dim=128
--network_alpha=64
@shimesaba_EX 感谢老哥，你的lora制作的非常好
@PATREONaiiiiii
非常感谢。 我的 LoRA 是由运气组成的。