Below is a list of keywords used during the training of this model:
[33] Grimmjow
[33] breasts
[33] large_breasts
[33] blue_hair
[32] 1girl
[27] blue_eyes
[26] blush
[25] long_hair
[25] solo
[23] looking_at_viewer
[19] v-shaped_eyebrows
[19] bangs
[18] collarbone
[17] open_mouth
[17] navel
[16] teeth
[16] hair_between_eyes
[13] nipples
[13] cleavage
[12] angry
[12] upper_body
[11] shiny
[11] smile
[10] nude
[10] bare_shoulders
Description
FAQ
Comments (3)
This looks amazing! Could you please share how exactly this LoRA was trained? What rank, learning rate, and number of epochs did you use? How many repeats and images were included? And which program was used? I've already failed seven LoRAs, so I'd be really grateful for any information and your response!
uploaded dataset, config follows:
[[subsets]]
num_repeats = 4
caption_extension = ".txt"
shuffle_caption = true
flip_aug = false
is_reg = false
image_dir = "F:/Preprocess/Reprocess/Grimmjow"
keep_tokens = 1
[noise_args]
[sample_args]
[logging_args]
[general_args.args]
pretrained_model_name_or_path = "F:/AI/Training/Training_Model/ponyDiffusionV6XL_v6StartWithThisOne.safetensors"
mixed_precision = "fp16"
seed = 80085
max_data_loader_n_workers = 1
persistent_data_loader_workers = true
max_token_length = 225
prior_loss_weight = 1.0
sdxl = true
xformers = true
cache_latents = true
max_train_epochs = 10
no_half_vae = true
vae = "F:/AI/Training/Training_Model/sdxl_vae.safetensors"
[general_args.dataset_args]
resolution = 1024
batch_size = 2
[network_args.args]
network_dim = 32
network_alpha = 16.0
min_timestep = 0
max_timestep = 1000
network_train_unet_only = true
[optimizer_args.args]
optimizer_type = "Prodigy"
lr_scheduler = "cosine"
learning_rate = 1.0
max_grad_norm = 1.0
[saving_args.args]
output_dir = "F:/AI/stable-diffusion-webui-forge/webui/models/Lora"
save_precision = "fp16"
save_model_as = "safetensors"
tag_occurrence = true
save_toml = true
output_name = "Grimmjow_V3"
[bucket_args.dataset_args]
enable_bucket = true
min_bucket_reso = 256
max_bucket_reso = 2048
bucket_reso_steps = 256
[optimizer_args.args.optimizer_args]
weight_decay = "0.1"
betas = "0.9,0.99"
decouple = "True"
@Synthdark8 Thanks so much! 😊