Ver NOOB V-PRED V2 EQ
Trained with improved parameters based on NOOB V-PRED EQ VAE exprimental version: https://civarchive.com/models/1858821/noobai-v-pred-10-with-eq-vae
Ver NOOB V-PRED 1.0
Trained based on NOOB V-PRED V1.0 with OneTrainer. See about this version for training details. Changed trigger words to prevent unwanted shower head or waterdrop.
Pros:
Better response to trigger words. There is no need to put the weight of cumextremely high to get the desired results in this version.
Cons:
Still you need some trials or weight changing to get what you want.
Very stiff gestures. It is difficult to get views with differences.
Distortion still exists
Ver NOOB Experimental
Trained based on NOOB E-PRED V1.0. Different trigger words set due to the improvement knowledge of model. For example, cumbath is not needed now because the model can do its generation pretty well.
Updated
Used PonyXL to train this one. Updated dataset to achieve better performance. It enhanced this concept on PonyXL and related models.
Adding a concept: cum_through
Introduction
This LyCORIS is used to enhance the concept of the scenes with massive cum like cumbath and cumshower. In NAI and other hentai models, it is possible to do that, but hard to get the results you wanted. So I made this.
How to use it
I recommand using it in weight 0.7-0.8. Higher weight may give you overbaked results. Works well with other tatoo loras.
This locon has two trigger words. One is cumbath, use it when you need your character partically submerged in cum. One is cumshower, use it when you need the cum over the body.
Training details:
Traingset is of about 400 images. They are mirrored before training. Half of them are chosen as regularization images. Total steps are about 15000.
I divided the training set into two parts tagged with two different trigger words.
Regularization = true
resolution=768
batch_size=1
epoch=10
network_dim=32
network_alpha=32
clip_skip=2
Using AdamW8bits as optimizer:
lr="1e-4"
unet_lr="1e-4"
text_encoder_lr="1e-5"
Locon parameters:
conv_dim=4
conv_alpha=4
Description
Used PonyXL to train this one. Updated dataset to achieve better performance. It enhanced this concept on PonyXL and related models.
Adding a concept: cum_through
Parameters:
{ "LoRA_type": "LyCORIS/LoCon", "LyCORIS_preset": "full", "adaptive_noise_scale": 0, "additional_parameters": "", "block_alphas": "", "block_dims": "", "block_lr_zero_threshold": "", "bucket_no_upscale": true, "bucket_reso_steps": 128, "cache_latents": false, "cache_latents_to_disk": false, "caption_dropout_every_n_epochs": 0.0, "caption_dropout_rate": 0, "caption_extension": ".txt", "clip_skip": 2, "color_aug": false, "constrain": 0.0, "conv_alpha": 4, "conv_block_alphas": "", "conv_block_dims": "", "conv_dim": 8, "debiased_estimation_loss": false, "decompose_both": false, "dim_from_weights": false, "down_lr_weight": "", "enable_bucket": true, "epoch": 10, "factor": -1, "flip_aug": false, "fp8_base": false, "full_bf16": true, "full_fp16": false, "gpu_ids": "", "gradient_accumulation_steps": "1", "gradient_checkpointing": true, "keep_tokens": 1, "learning_rate": 1.0, "log_tracker_config": "", "log_tracker_name": "", "logging_dir": "", "lora_network_weights": "", "lr_scheduler": "cosine", "lr_scheduler_args": "", "lr_scheduler_num_cycles": "", "lr_scheduler_power": "", "lr_warmup": 0, "max_bucket_reso": 4096, "max_data_loader_n_workers": "0", "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": 1024, "min_snr_gamma": 0, "min_timestep": 0, "mixed_precision": "bf16", "model_list": "custom", "module_dropout": 0, "multi_gpu": false, "multires_noise_discount": 0.1, "multires_noise_iterations": 6, "network_alpha": 8, "network_dim": 16, "network_dropout": 0, "noise_offset": 0, "noise_offset_type": "Multires", "num_cpu_threads_per_process": 2, "num_machines": 1, "num_processes": 1, "optimizer": "Prodigy", "optimizer_args": "weight_decay=0.01 d_coef=1 use_bias_correction=True safeguard_warmup=False betas=0.9,0.99 decouple=True ", "output_dir": "/root/autodl-tmp/output", "output_name": "cumflood-ponyXL", "persistent_data_loader_workers": false, "pretrained_model_name_or_path": "/root/kohya_ss/models/ponyDiffusionV6XL.safetensors", "prior_loss_weight": 1.0, "random_crop": false, "rank_dropout": 0, "rank_dropout_scale": false, "reg_data_dir": "/root/autodl-tmp/reg_cumflood_XL", "rescaled": false, "resume": "", "sample_every_n_epochs": 1, "sample_every_n_steps": 0, "sample_prompts": "score_9, score_8_up, score_7_up, anime BREAK, cumshower, cumbath, bathtub, futanari, gigantic_penis, veiny penis, veiny testicles, erection, 1girl, solo, nude, excessive_cum, ejaculation, projectile_cum --n source_pony, source_furry, source_cartoon, greyscale, text, artist name, signature, timestamp, --w 1024 --h 1024", "sample_sampler": "euler_a", "save_every_n_epochs": 1, "save_every_n_steps": 0, "save_last_n_steps": 0, "save_last_n_steps_state": 0, "save_model_as": "safetensors", "save_precision": "bf16", "save_state": false, "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": "31337", "shuffle_caption": false, "stop_text_encoder_training_pct": 0, "text_encoder_lr": 1.0, "train_batch_size": 2, "train_data_dir": "/root/autodl-tmp/cumflood_XL", "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": "/root/kohya_ss/models/sdxl_vae_fp16fix.safetensors", "vae_batch_size": 0, "wandb_api_key": "", "wandb_run_name": "", "weighted_captions": false, "xformers": "xformers" }
FAQ
Comments (3)
what is cum_through?
A tag describes cum overflow from mouth. Check it in rule34 you will get a clearer picture.
can someone make a SD1.5 and SDXL cumtribute lora model