Mornye (Wuthering Waves) | 莫宁《鸣潮》
Intended to generate non-commercial fan works of the character Mornye from the Wuthering Waves video game.
Trained on a small mixed NL and tags dataset, was interested to see the control NL tagging might have over transparency concepts like the characters see-through legs.
ℹ️ LoRA work best when applied to the base models on which they are trained. Please read the About This Version on the appropriate base models and workflow/training information.
Trigger:
mornye \(wuthering waves\)Works best in combination with NL if you name a character, then describe their basic appearance.
An artwork of Mornye from Wuthering Waves with long, silvery-white hair, red eyes. She wears a black dress with a high collar and decorative jewelry, partially covered by a loose white jacket that drapes open...Description
Trained on Anima Preview
Assume that any lora trained on the preview version won't work well on the final version.
Trained the mixed tags and natural language dataset.
Tags are dropped at 0.3, so that the trigger should function alone to express the outfit.
Used diffusion-pipe - fork by @bluvoll
Config:
# dataset-anima.toml
# Resolution settings.
resolutions = [1024]
# Aspect ratio bucketing settings
enable_ar_bucket = true
min_ar = 0.5
max_ar = 2.0
num_ar_buckets = 7
[[directory]] # IMAGES
# Path to the directory containing images and their corresponding caption files.
path = '/mnt/d/training_data/images'
num_repeats = 1
resolutions = [1024]
# config.toml
# Change these paths
output_dir = '/mnt/d/anima/training_output'
dataset = 'dataset-anima.toml'
# training settings
epochs = 60
micro_batch_size_per_gpu = 6
pipeline_stages = 1
gradient_accumulation_steps = 1
gradient_clipping = 1.0
warmup_steps = 10
train_llm_adapter = true
# eval settings
eval_every_n_epochs = 1
eval_before_first_step = true
eval_micro_batch_size_per_gpu = 1
eval_gradient_accumulation_steps = 1
# misc settings
save_every_n_epochs = 2
checkpoint_every_n_minutes = 120
activation_checkpointing = true
partition_method = 'parameters'
save_dtype = 'bfloat16'
caching_batch_size = 1
steps_per_print = 1
[model]
type = 'anima'
transformer_path = '/mnt/c/workspace/models/diffusion_models/anima-preview.safetensors'
vae_path = '/mnt/c/workspace/models/vae/qwen_image_vae.safetensors'
qwen_path = '../qwen0.6/Qwen3-0.6B/'
dtype = 'bfloat16'
timestep_sample_method = 'logit_normal'
sigmoid_scale = 1.0
shift = 3.0
# Caption Processing Options
cache_text_embeddings = false
# NOTE: Requires cache_text_embeddings = false to work!
# For cached embeddings, use cache_shuffle_num in your dataset config instead.
shuffle_tags = true
tag_delimiter = ', '
keep_first_n_tags = 5
shuffle_keep_first_n = 5
tag_dropout_percent = 0.3
protected_tags_file = './protected_tags.txt'
nl_shuffle_sentences = false
nl_keep_first_sentence = true
# 'tags' 'nl' 'mixed'
caption_mode = 'mixed'
debug_caption_processing = true
debug_caption_interval = 100
[adapter]
type = 'lora'
rank = 32
dtype = 'bfloat16'
# AdamW from the optimi library is a good default since it automatically uses Kahan summation when training bfloat16 weights.
[optimizer]
type = 'adamw_optimi'
lr = 5e-5
betas = [0.9, 0.99]
weight_decay = 0.01
eps = 1e-8