sexual situations

physical violence

disturbing

male nudity

extremist

hanging

hate symbols

nazi party

emaciated bodies

female nudity

undressed

male underwear

female swimwear

genitals

female underwear

thick thighs

breasts out

strapless leotard

vore

bestiality

breast out

one breast out

huge breasts

gigantic breasts

huge butt

covered nipples

hair over breasts

no panties

sitting on face

anal

dildo riding

downblouse

oral

porn

futanari

hentai

nude

lingerie

sexy

nsfw

weapon violence

female swimwear or underwear

male swimwear or underwear

partial nudity

white supremacy

adult toys

graphic male nudity

illustrated explicit nudity

nudity

self injury

graphic female nudity

peeing

exposed female nipple

sexual activity

sexual intent

suggestive

revealing clothes

graphic violence or gore

corpses

wide hips

convenient censoring

oral invitation

blowjob

bukkake

cumshot

cumdrip

cum in pussy

cum on face

cum on hair

cum on body

cum on tongue

cum on hands

cum in mouth

cum on pussy

eat_cum

fellatio

implied fellatio

after fellatio

triple fellatio

autofellatio

pov fellatio

fucked silly

presenting ass

presenting pussy

spreading legs

spread legs

swimwear

This resulted from a bout of curiosity as some people claim Ia3 is the best thing since toasted bread. The result? It is an odd little thing emphasis in "Little" 218KB is comparable to a TI. You can compare against the <a target="_blank" rel="ugc" href="https://civitai.com/models/15051/iori-yoshizuki-is">LORA version</a> or the <a target="_blank" rel="ugc" href="https://civitai.com/models/145793/iori-yoshizuki-locon-version-is">LOCON version</a>.Image wise it looks as if you had pulled out all the influence of the model out of a LORA, In this particular case, as most of the dataset consisted of monochrome doujin and colored ones the resulting images in fact look rather like colored doujin images! That's terrible and incredible at the same time.Here's my analysis of pros and consPro:<ul><li>Very small</li><li>very quick to train with prodigy it requires 250~300 steps per epoch per concept for 8 epochs to give a decent result.</li><li>I suspect it works extremely well with mono source homogeneous datasets like screencaps without burning.</li><li>Seems to be great at retaining style.</li><li>Portability between models is not as bad as reported, It is inferior to LOCON but it is on par or slightly lower than LORAs.</li></ul>Cons:<ul><li>Feels slightly less responsive than LORA/LOCON to tags not in the dataset.</li><li>It is more sensitive to shortcomings of the dataset that normally get filled by the model(missing angles, low hair detail, fabric textures, low color, blurriness)</li><li>It won't express the model style strongly making it more neutral but also slightly flavorless.</li></ul>Conclusion:Ia3 seems to be perfect to make quick ok-ish LORA, for prototyping to quickly test the viability of a dataset and for beginners who don't care about things being "too exact" or don't care for their images to follow a particular style.I will add the dataset, training TOML file and an XYZ file comparing the results of 6 different models.<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/e9825fd0-cc98-4441-9934-b5787c2c30d5/width=525/e9825fd0-cc98-4441-9934-b5787c2c30d5.jpeg" />Btw got good results using: &lt;lora:yoshizuki_ioriV4:.8&gt; yoshizuki_iori,The only tagged outfit is:<ul><li>school_uniform_purple_shirt_blue_skirt_white_neckerchief_black_thighhighs</li></ul>

Trained using prodigy and type Ia3 using the following parameters:<pre><code>[general_args.args]
pretrained_model_name_or_path = "D:/wifediff/stable-diffusion-webui/models/Stable-diffusion/anything-v4.5-pruned.safetensors"
mixed_precision = "fp16"
seed = 80085
clip_skip = 2
xformers = true
max_data_loader_n_workers = 1
persistent_data_loader_workers = true
max_token_length = 225
prior_loss_weight = 1.0
max_train_epochs = 8
training_comment = "Trigger: yoshizuki_iori, School_uniform_Purple_shirt_Blue_skirt_White_neckerchief_Black_thighhighs"
cache_latents = true

[general_args.dataset_args]
resolution = 512
batch_size = 1

[network_args.args]
network_dim = 32
network_alpha = 16.0

[optimizer_args.args]
optimizer_type = "Prodigy"
lr_scheduler = "cosine"
learning_rate = 1.0
lr_scheduler_type = "LoraEasyCustomOptimizer.CustomOptimizers.CosineAnnealingWarmupRestarts"
lr_scheduler_num_cycles = 1
min_snr_gamma = 5

[saving_args.args]
output_dir = "D:/wifediff/lora/iorilora"
save_precision = "fp16"
save_model_as = "safetensors"
output_name = "yoshizuki_ioriV4"
tag_occurrence = true
save_toml = true
save_every_n_epochs = 1

[bucket_args.dataset_args]
enable_bucket = true
min_bucket_reso = 256
max_bucket_reso = 1024
bucket_reso_steps = 64

[network_args.args.network_args]
conv_dim = 32
conv_alpha = 16.0
algo = "ia3"

[optimizer_args.args.lr_scheduler_args]
min_lr = 1e-6
gamma = 0.9

[optimizer_args.args.optimizer_args]
weight_decay = "0.01"
decouple = "True"
use_bias_correction = "True"
safeguard_warmup = "True"
d_coef = "2"</code></pre>

v1.0

Iori Yoshizuki ( (IA)^3 experiment version ) - I''s