Akiko Kamimura 神村秋子 | SISTERS The last day of summer
Version 3.0
Discord : https://discord.gg/YkqSnaXqEQ
Version 3.0 : Images : 200 From Anime
Sponsor : https://ko-fi.com/emaisuraai
prompt:
Simple White Background,
Orange Bikini, cleavage,bare shoulders, collarbone,navel,
<lora:Akiko_Kamimura_Sisters-KK77-V3:0.7>,
blonde hair,blue_eyes, Bangs, Long Hair,braid,
<lora:Oda_Non_Style-KK77-V2:0.3>,<lora:more_details:0.1>,
1 girl, 20yo,Young female,Beautiful long legs,Beautiful body,
Beautiful Nose,Beautiful character design, perfect eyes, perfect face,expressive eyes,perfect balance,
looking at viewer,(Focus on her face),closed mouth, (innocent_big_eyes:1.0),(Light_Smile:0.3),
official art,extremely detailed CG unity 8k wallpaper, perfect lighting,Colorful, Bright_Front_face_Lighting,White skin,
(masterpiece:1.0),(best_quality:1.0), ultra high res,4K,ultra-detailed,
photography, 8K, HDR, highres, absurdres:1.2, Kodak portra 400, film grain, blurry background, bokeh:1.2, lens flare, (vibrant_color:1.2),professional photograph,
(Beautiful,large_Breasts:1.4), (beautiful_face:1.5),(narrow_waist),
Negative prompt: EasyNegative, FastNegativeV2, bad-artist-anime, bad-hands-5, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, jpeg artifacts, signature, watermark, username, blurry, out of focus, censorship, Missing vagina, Blurry faces, Blank faces, bad face, Ugly, extra ear, amputee, missing hands, missing arms, missing legs, Extra fingers, 6 fingers, Extra feet, Missing nipples, ghost, futanari, Extra legs, Extra hands, panties,pants, (painting by bad-artist-anime:0.9), (painting by bad-artist:0.9), text, error, blurry, jpeg artifacts, cropped, normal quality, artist name, (worst quality, low quality:1.4),twisted_hands,fused_fingers,Face Shadow,NSFW,(worst quality:2), (low quality:2), (normal quality:2)
Steps: 48, Sampler: Euler a, CFG scale: 7, Seed: 2681472849, Size: 512x768, Model hash: 3e9211917c, Model: yesmix_v16Original, Denoising strength: 0.47, ControlNet 0: "Module: openpose_full, Model: control_v11p_sd15_openpose [cab727d4], Weight: 1, Resize Mode: Crop and Resize, Low Vram: False, Processor Res: 512, Guidance Start: 0, Guidance End: 1, Pixel Perfect: True, Control Mode: My prompt is more important, Save Detected Map: True", Hires upscale: 2, Hires steps: 18, Hires upscaler: R-ESRGAN 4x+ Anime6B, Lora hashes: "Akiko_Kamimura_Sisters-KK77-V3: a8d1a75871c2, Oda_Non_Style-KK77-V2: de8a898832ea, more_details: 3b8aa1d351ef", TI hashes: "easynegative: c74b4e810b03, FastNegativeV2: a7465e7cc2a2, bad-artist-anime: 5f7bea88750c, bad-hands-5: aa7651be154c", Version: v1.6.0
Time taken: 1 min. 31.0 sec.
Description
FAQ
Comments (7)
Your lora is very well done May I ask you what is your training process of the lora (making the dataset, number of images, number of steps... Etc) as I am working on a university project using lora so every bit of information will be very very useful to me, thank you in advance
You can download the training images that is data folder, right here,
https://civitai.com/api/download/training-data/80612
Thank You.
I post some my Kohya ss training step, to pixiv.
https://www.pixiv.net/artworks/108435224
And My Pc info also in that pixiv.
https://www.pixiv.net/artworks/108435224
This lora:
Folder 100_EAV P: 103 images found
Folder 100_EAV P: 10300 steps
max_train_steps = 5150
stop_text_encoder_training = 0
lr_warmup_steps = 515
accelerate launch --num_cpu_threads_per_process=2 "train_network.py" --pretrained_model_name_or_path="E:\Game 2023\AIWAIFU\SD15\stable-diffusion-webui\models\Stable-diffusion\yesmix_v16Original.safetensors" --train_data_dir="E:\Game 2023\AIWAIFU\kohya_ss20230512\kohya_ss\kohya_ss\YMAKELORA\Model\20230525\Akiko_Kamimura_SISTERS\2" --resolution=512,512 --output_dir="E:\Game 2023\AIWAIFU\kohya_ss20230512\kohya_ss\kohya_ss\YMAKELORA\Model\20230525\Akiko_Kamimura_SISTERS\2" --network_alpha="128" --save_model_as=safetensors --network_module=networks.lora --network_args "conv_dim=1" "conv_alpha=1" --text_encoder_lr=5e-05 --unet_lr=0.0001 --network_dim=32 --output_name="Akiko_Kamimura_SISTERS-KK77-V1" --lr_scheduler_num_cycles="1" --learning_rate="0.0001" --lr_scheduler="cosine" --lr_warmup_steps="515" --train_batch_size="2" --max_train_steps="5150" --save_every_n_epochs="1" --mixed_precision="fp16" --save_precision="fp16" --seed="1" --caption_extension=".txt" --cache_latents --optimizer_type="AdamW8bit" --max_data_loader_n_workers="1" --clip_skip=2 --bucket_reso_steps=64 --xformers --bucket_no_upscale
prepare tokenizer
Use DreamBooth method.
ignore directory without repeats / 繰り返し回数のないディレクトリを無視します: sample
prepare images.
found directory E:\Game 2023\AIWAIFU\kohya_ss20230512\kohya_ss\kohya_ss\YMAKELORA\Model\20230525\Akiko_Kamimura_SISTERS\2\100_EAV P contains 103 image files
10300 train images with repeating.
0 reg images.
no regularization images / 正則化画像が見つかりませんでした
[Dataset 0]
batch_size: 2
resolution: (512, 512)
enable_bucket: False
[Subset 0 of Dataset 0]
image_dir: "E:\Game 2023\AIWAIFU\kohya_ss20230512\kohya_ss\kohya_ss\YMAKELORA\Model\20230525\Akiko_Kamimura_SISTERS\2\100_EAV P"
image_count: 103
num_repeats: 100
shuffle_caption: False
keep_tokens: 0
caption_dropout_rate: 0.0
caption_dropout_every_n_epoches: 0
caption_tag_dropout_rate: 0.0
color_aug: False
flip_aug: False
face_crop_aug_range: None
random_crop: False
token_warmup_min: 1,
token_warmup_step: 0,
is_reg: False
class_tokens: EAV P
caption_extension: .txt
Time to output:
saving checkpoint:
E:\Game 2023\AIWAIFU\kohya_ss20230512\kohya_ss\kohya_ss\YMAKELORA\Model\20230525\Akiko_Kamimura_SISTERS\2\Akiko_Kamimura_SISTERS-KK77-V1.safetensors
model saved.
steps: 100%|███████████████████████████████████████████████████████████| 5150/5150 [23:10<00:00, 3.70it/s, loss=0.095]
I need Chika and Haruka too!