Yuzuriha (Hell’s Paradise)
Yuzuriha (Hell\'s Paradise)
- TypeLORA
- ModelSDXL 1.0
- Trained WordsYuzuriha \(Jigokuraku\),1girl,japanese clothes ,purple hair,pink eyes,ninja,ponytail,no bra
- TagLORA1girl,solo, anime character female hells paradise jigokuraku yuzuriha
This is my first Lora so it might not be the best but i’ve tested it on multiple anime XL models such as
animeChangefulXL_v10ReleasedCandidate
I found that i got the best results with the lora weighted from .4 to .6 and if i wanted her normal outfit to show .6 to .9 seemed good but i used Kohya_SS Gui to train this lora and since i didn’t know exactly which settings would be good for a character lora i went with one of the XL presets and I’ll post that below as well as uploading the exact dataset i used with the caption files included so maybe someone who has much more experience then i do can improve this lora if they feel like it.
{
“LoRA_type”: “Standard”,
“LyCORIS_preset”: “full”,
“adaptive_noise_scale”: 0,
“additional_parameters”: “”,
“block_alphas”: “”,
“block_dims”: “”,
“block_lr_zero_threshold”: “”,
“bucket_no_upscale”: true,
“bucket_reso_steps”: 64,
“cache_latents”: true,
“cache_latents_to_disk”: true,
“caption_dropout_every_n_epochs”: 0.0,
“caption_dropout_rate”: 0,
“caption_extension”: “.txt”,
“clip_skip”: “1”,
“color_aug”: false,
“constrain”: 0.0,
“conv_alpha”: 1,
“conv_block_alphas”: “”,
“conv_block_dims”: “”,
“conv_dim”: 1,
“debiased_estimation_loss”: false,
“decompose_both”: false,
“dim_from_weights”: true,
“down_lr_weight”: “”,
“enable_bucket”: true,
“epoch”: 35,
“factor”: -1,
“flip_aug”: true,
“full_bf16”: false,
“full_fp16”: false,
“gradient_accumulation_steps”: 1,
“gradient_checkpointing”: true,
“keep_tokens”: 0,
“learning_rate”: 1.0,
“logging_dir”: “/notebooks/y4train/log”,
“lora_network_weights”: “/notebooks/y4train/model/yuzuriha_a-step00002000.safetensors”,
“lr_scheduler”: “cosine”,
“lr_scheduler_args”: “”,
“lr_scheduler_num_cycles”: “”,
“lr_scheduler_power”: “”,
“lr_warmup”: 5,
“max_bucket_reso”: 2048,
“max_data_loader_n_workers”: “8”,
“max_grad_norm”: 1,
“max_resolution”: “1024,1024”,
“max_timestep”: 1000,
“max_token_length”: “225”,
“max_train_epochs”: “”,
“max_train_steps”: “”,
“mem_eff_attn”: false,
“mid_lr_weight”: “”,
“min_bucket_reso”: 256,
“min_snr_gamma”: 5,
“min_timestep”: 0,
“mixed_precision”: “bf16”,
“model_list”: “custom”,
“module_dropout”: 0,
“multires_noise_discount”: 0,
“multires_noise_iterations”: 0,
“network_alpha”: 256,
“network_dim”: 256,
“network_dropout”: 0,
“no_token_padding”: false,
“noise_offset”: 0.0357,
“noise_offset_type”: “Original”,
“num_cpu_threads_per_process”: 2,
“optimizer”: “DAdaptAdam”,
“optimizer_args”: “\”decouple=True\” \”weight_decay=0.2\” \”betas=0.9,0.99\” \”growth_rate=1.02\””,
“output_dir”: “/notebooks/y4train/model”,
“output_name”: “yuzuriha_a”,
“persistent_data_loader_workers”: true,
“pretrained_model_name_or_path”: “stabilityai/stable-diffusion-xl-base-1.0”,
“prior_loss_weight”: 1.0,
“random_crop”: false,
“rank_dropout”: 0,
“rank_dropout_scale”: false,
“reg_data_dir”: “”,
“rescaled”: false,
“resume”: “/notebooks/y4train/model/yuzuriha_a-step00002000-state”,
“sample_every_n_epochs”: 0,
“sample_every_n_steps”: 50,
“sample_prompts”: “masterpiece, best quality, award-winning illustration, yuzuriha (jigokuraku), 1girl, narrow waist, hips, bent over, from side, bare shoulders, alluring thighs, thick ass, ass focus, looking to the side at viewer, leaning forward, seductive pose, arched back, bending over, dynamic_pose, solo, purple hair, looking at viewer, bandages, pink eyes, ninja, no bra, medium breasts, hair between eyes, ponytail, bandaged arm, cleavage, japanese clothes, purple hair –n aidxlv05_neg,(3d,realistic:1.5),worst_quality,low_quality,bad_anatomy,blurry,lowres,negativeXL_D –w 1024 –h 1024 –d 1337 –l 9 –s 30”,
“sample_sampler”: “euler_a”,
“save_every_n_epochs”: 0,
“save_every_n_steps”: 250,
“save_last_n_steps”: 0,
“save_last_n_steps_state”: 0,
“save_model_as”: “safetensors”,
“save_precision”: “bf16”,
“save_state”: true,
“scale_v_pred_loss_like_noise_pred”: false,
“scale_weight_norms”: 0,
“sdxl”: true,
“sdxl_cache_text_encoder_outputs”: false,
“sdxl_no_half_vae”: true,
“seed”: “1337”,
“shuffle_caption”: false,
“stop_text_encoder_training_pct”: 0,
“text_encoder_lr”: 1.0,
“train_batch_size”: 6,
“train_data_dir”: “/notebooks/y4train/img”,
“train_norm”: false,
“train_on_input”: true,
“training_comment”: “”,
“unet_lr”: 1.0,
“unit”: 1,
“up_lr_weight”: “”,
“use_cp”: false,
“use_scalar”: false,
“use_tucker”: false,
“use_wandb”: false,
“v2”: false,
“v_parameterization”: false,
“v_pred_like_loss”: 0,
“vae”: “”,
“vae_batch_size”: 0,
“wandb_api_key”: “”,
“weighted_captions”: false,
“xformers”: “sdpa”
}
One thing i noticed though was the size of the lora was 1.7gb which is way to big for a character lora so i used kohya_ss gui’s lora tools to change the 256 dim into a 32 dim, tested before and after and both seemed the same which resulted in a 170mb instead of almost 2gb