Arknights: Endfield | Operators Collection
Intended to generate non-commercial fan works of Operators from the Arknights: Endfield video game.
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.
Version 3.0 (Latest)
This was re-trained with several attempts to reach the same level of quality as the full-finetune which included the same dataset:
The LoRA is not quite there, even when adjusting hyperparameters and extending the training time.
Little disappointed about this, but it may be an expected limitation of LoRA training.
Operators (game version 1.0):
female endministrator \(arknights\)
male endministrator \(arknights\)
perlica \(arknights\)
chen qianyu \(arknights\)
akekuri \(arknights\)
alesh \(arknights\)
antal \(arknights\)
arclight \(arknights\)
ardelia \(arknights\)
avywenna \(arknights\)
catcher \(arknights\)
da pan \(arknights\)
estella \(arknights\)
fluorite \(arknights\)
gilberta \(arknights\)
laevatain \(arknights\)
last rite \(arknights\)
lifeng \(arknights\)
snowshine \(arknights\)
wulfgard \(arknights\)
xaihi \(arknights\)
yvonne \(arknights\)
Operators (game version 1.1):
mi fu \(arknights\)
rossi \(arknights\)
tangtang \(arknights\)
zhuang fangyi \(arknights\)
Unreleased Operators (unconfirmed designs):
arcane \(arknights\)
camille \(arknights\)
feranmut proxy \(arknights\)Antagonists:
ardashir \(arknights\)
nefarith \(arknights\)
Works best in combination with NL if you name a character, then describe their basic appearance.
A vibrant and dynamic illustration of Yvonne from Arknights: Endfield, featuring her with long pink hair styled in twintails, pointy ears, and small horns, along with a playful tail...
To be fixed:
Pogranichnik had a mistake for his labelling
Ember did not seem to learn her outfit/features
Description
Trained on Anima Preview 3 Base
Dataset uses a mix of tags and natural language captions
Less prominent characters were not learned as well in this version
Training config:
# trained using diffusion-pipe commit 6e95020cad0b3cd7dcb93ce42b358669051bf6d2
output_dir = '/mnt/d/anima/training_output/arknights3'
dataset = 'dataset-anima.toml'
# training settings
epochs = 1000
# Per-resolution batch sizes
micro_batch_size_per_gpu = [[512, 64], [1024, 32], [1536, 16]]
pipeline_stages = 1
gradient_accumulation_steps = 1
gradient_clipping = 1
warmup_steps = 100
# misc settings
save_every_n_epochs = 1
#save_every_n_steps = 1000
#save_every_n_examples = 4096000
#checkpoint_every_n_epochs = 1
#checkpoint_every_n_minutes = 120
activation_checkpointing = true
#reentrant_activation_checkpointing = true
partition_method = 'parameters'
# partition_method = 'manual'
# partition_split = [10]
save_dtype = 'bfloat16'
caching_batch_size = 1
map_num_proc = 8
steps_per_print = 1
compile = true
[model]
type = 'anima'
transformer_path = '/mnt/c/workspace/ComfyUI_windows_portable_nvidia/ComfyUI_windows_portable/ComfyUI/models/diffusion_models/anima-preview3-base.safetensors'
vae_path = '/mnt/c/workspace/ComfyUI_windows_portable_nvidia/ComfyUI_windows_portable/ComfyUI/models/vae/qwen_image_vae.safetensors'
llm_path = '/mnt/c/workspace/ComfyUI_windows_portable_nvidia/ComfyUI_windows_portable/ComfyUI/models/text_encoders/qwen_3_06b_base.safetensors'
dtype = 'bfloat16'
#cache_text_embeddings = false
llm_adapter_lr = 2e-6
#timestep_sample_method = 'uniform'
#flux_shift = true
#multiscale_loss_weight = 0.5
sigmoid_scale = 1.3
[adapter]
type = 'lora'
rank = 32
dtype = 'bfloat16'
[optimizer]
type = 'adamw_optimi'
lr = 4e-5
betas = [0.9, 0.99]
weight_decay = 0.01
eps = 1e-8Dataset config:
resolutions = [512, 1024, 1536]
enable_ar_bucket = true
min_ar = 0.5
max_ar = 2.0
num_ar_buckets = 9
[[directory]]
path = '/mnt/d/training_data/images_arknights_captions'FAQ
Comments (8)
Do you make a new Version for Base later?
Currently training a full finetune of Base that includes this dataset, an update for https://civitai.red/models/2495369
@motimalu Hi, you updated your Basemodel, really good. Do you want to update this lora over the course of the month? Please let me know if I'm being too pushy.
Hey @VKilko no problem, yes I'll update this with an Anima base 1 version over the next week or so.
Maybe its already understood but just to re-iterate the Kirazuri (Anima) 3.0 model now also has been trained with the same dataset and can generate all the characters. For example: Rossi
@motimalu Thanks for the answer, have a nice day.
doesn't work on base1? or zhuang fangyi cannot be generated just by name
Hi, both should work on base 1.
v2.0 was not trained long enough for some characters to be learned as well as others, Zhuang Fangyi did not have much data and so isn't well represented.
v1.0 was trained on preview for much longer and has a stronger effect, but Zhuang Fangyi wasn't included in it.
when generating zhuang fangyi, no matter how it is described, there will always be two extra horns next to the ears. that was really strange
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.