pipeline_tag: text-to-image
license: other
license_name: faipl-1.0-sd
license_link: LICENSE
decoder:
- Disty0/sotediffusion-wuerstchen3-alpha1-decoder
SoteDiffusion Wuerstchen3
Anime finetune of Würstchen V3.
Currently is in early state in training.
No commercial use thanks to StabilityAI.
Release Notes
Did major cleanup on the dataset in this release.
Changed the training parameters and started from a fresh state.
Switch to FairAI license. (Still no commercial use.)
UI Guide
SD.Next
Switch to the dev branch:
git checkout dev
Go to Models -> Huggingface and type Disty0/sotediffusion-wuerstchen3-alpha1-decoder into the model name and press download.
Load Disty0/sotediffusion-wuerstchen3-alpha1-decoder after the download process is complete.
Parameters:
Sampler: Default
Steps: 30 or 40
Secondary Steps: 10
CFG: 8
Secondary CFG: 1 or 1.2
ComfyUI
Please refer to CivitAI: https://civitai.com/models/353284
Code Example
pip install diffusers
import torch
from diffusers import StableCascadeCombinedPipeline
device = "cuda"
dtype = torch.bfloat16
model = "Disty0/sotediffusion-wuerstchen3-alpha1-decoder"
pipe = StableCascadeCombinedPipeline.from_pretrained(model, torch_dtype=dtype)
# send everything to the gpu:
pipe = pipe.to(device, dtype=dtype)
pipe.prior_pipe = pipe.prior_pipe.to(device, dtype=dtype)
# or enable model offload to save vram:
# pipe.enable_model_cpu_offload()
prompt = "extremely aesthetic, best quality, newest, general, 1girl, solo, looking at viewer, blush, slight smile, cat ears, long hair, dress, bare shoulders, cherry blossoms, flowers, petals, vegetation, wind,"
negative_prompt = "very displeasing, worst quality, oldest, monochrome, sketch, loli, child,"
output = pipe(
width=1024,
height=1536,
prompt=prompt,
negative_prompt=negative_prompt,
decoder_guidance_scale=1.2,
prior_guidance_scale=8.0,
prior_num_inference_steps=40,
output_type="pil",
num_inference_steps=10
).images[0]
## do something with the output image
Training Status:
GPU used for training: 1x AMD RX 7900 XTX 24GB
GPU Hours: 100
| dataset name | training done | remaining |
|---|---|---|
| newest | 003 | 228 |
| recent | 003 | 169 |
| mid | 003 | 121 |
| early | 003 | 067 |
| oldest | 003 | 017 |
| pixiv | 003 | 039 |
| visual novel cg | 003 | 025 |
| anime wallpaper | 003 | 010 |
| Total | 32 | 682 |
Note: chunks starts from 0 and there are 8000 images per chunk
Dataset:
GPU used for captioning: 1x Intel ARC A770 16GB
GPU Hours: 350
Model used for captioning: SmilingWolf/wd-swinv2-tagger-v3
Command:
python /mnt/DataSSD/AI/Apps/kohya_ss/sd-scripts/finetune/tag_images_by_wd14_tagger.py --model_dir "/mnt/DataSSD/AI/models/wd14_tagger_model" --repo_id "SmilingWolf/wd-swinv2-tagger-v3" --recursive --remove_underscore --use_rating_tags --character_tags_first --character_tag_expand --append_tags --onnx --caption_separator ", " --general_threshold 0.35 --character_threshold 0.50 --batch_size 4 --caption_extension ".txt" ./
| dataset name | total images | total chunk |
|---|---|---|
| newest | 1.848.331 | 232 |
| recent | 1.380.630 | 173 |
| mid | 993.227 | 125 |
| early | 566.152 | 071 |
| oldest | 160.397 | 021 |
| pixiv | 343.614 | 043 |
| visual novel cg | 231.358 | 029 |
| anime wallpaper | 104.790 | 014 |
| Total | 5.628.499 | 708 |
Note:
- Smallest size is 1280x600 | 768.000 pixels
- Deduped based on image similarity using czkawka-zli
Tags:
aesthetic tags, quality tags, date tags, custom tags, rating tags, character tags, rest of the tags
Date:
| tag | date |
|---|---|
| newest | 2022 to 2024 |
| recent | 2019 to 2021 |
| mid | 2015 to 2018 |
| early | 2011 to 2014 |
| oldest | 2005 to 2010 |
Aesthetic Tags:
Model used: shadowlilac/aesthetic-shadow-v2
| score greater than | tag |
|---|---|
| 0.90 | extremely aesthetic |
| 0.80 | very aesthetic |
| 0.70 | aesthetic |
| 0.50 | slightly aesthetic |
| 0.40 | not displeasing |
| 0.30 | not aesthetic |
| 0.20 | slightly displeasing |
| 0.10 | displeasing |
| rest of them | very displeasing |
Quality Tags:
Model used: https://huggingface.co/hakurei/waifu-diffusion-v1-4/blob/main/models/aes-B32-v0.pth
| score greater than | tag |
|---|---|
| 0.980 | best quality |
| 0.900 | high quality |
| 0.750 | great quality |
| 0.500 | medium quality |
| 0.250 | normal quality |
| 0.125 | bad quality |
| 0.025 | low quality |
| rest of them | worst quality |
Rating Tags
- general
- sensitive
- nsfw
- explicit nsfw
Custom Tags:
| dataset name | custom tag |
|---|---|
| image boards | date, |
| pixiv | art by Display_Name, |
| visual novel cg | Full_VN_Name (short_3_letter_name), visual novel cg, |
| anime wallpaper | date, anime wallpaper, |
Training Params:
Software used: Kohya SD-Scripts with Stable Cascade branch
Base model: Disty0/sote-diffusion-cascade-alpha0
Command:
LD_PRELOAD=/usr/lib/libtcmalloc.so.4 accelerate launch --mixed_precision fp16 --num_cpu_threads_per_process 1 stable_cascade_train_stage_c.py \
--mixed_precision fp16 \
--save_precision fp16 \
--full_fp16 \
--sdpa \
--gradient_checkpointing \
--train_text_encoder \
--resolution "1024,1024" \
--train_batch_size 2 \
--gradient_accumulation_steps 8 \
--learning_rate 1e-5 \
--learning_rate_te1 1e-5 \
--lr_scheduler constant_with_warmup \
--lr_warmup_steps 100 \
--optimizer_type adafactor \
--optimizer_args "scale_parameter=False" "relative_step=False" "warmup_init=False" \
--max_grad_norm 0 \
--token_warmup_min 1 \
--token_warmup_step 0 \
--shuffle_caption \
--caption_separator ", " \
--caption_dropout_rate 0 \
--caption_tag_dropout_rate 0 \
--caption_dropout_every_n_epochs 0 \
--dataset_repeats 1 \
--save_state \
--save_every_n_steps 256 \
--sample_every_n_steps 64 \
--max_token_length 225 \
--max_train_epochs 1 \
--caption_extension ".txt" \
--max_data_loader_n_workers 2 \
--persistent_data_loader_workers \
--enable_bucket \
--min_bucket_reso 256 \
--max_bucket_reso 4096 \
--bucket_reso_steps 64 \
--bucket_no_upscale \
--log_with tensorboard \
--output_name sotediffusion-wr3_3b \
--train_data_dir /mnt/DataSSD/AI/anime_image_dataset/combined/combined-0004/0005 \
--in_json /mnt/DataSSD/AI/anime_image_dataset/combined/combined-0004/0005.json \
--output_dir /mnt/DataSSD/AI/SoteDiffusion/Wuerstchen3/sotediffusion-wr3_3b-4/0005 \
--logging_dir /mnt/DataSSD/AI/SoteDiffusion/Wuerstchen3/sotediffusion-wr3_3b-4/0005/logs \
--resume /mnt/DataSSD/AI/SoteDiffusion/Wuerstchen3/sotediffusion-wr3_3b-4/0004/sotediffusion-wr3_3b-state \
--stage_c_checkpoint_path /mnt/DataSSD/AI/SoteDiffusion/Wuerstchen3/sotediffusion-wr3_3b-4/0004/sotediffusion-wr3_3b.safetensors \
--text_model_checkpoint_path /mnt/DataSSD/AI/SoteDiffusion/Wuerstchen3/sotediffusion-wr3_3b-4/0004/sotediffusion-wr3_3b_text_model.safetensors \
--effnet_checkpoint_path /mnt/DataSSD/AI/models/wuerstchen3/effnet_encoder.safetensors \
--previewer_checkpoint_path /mnt/DataSSD/AI/models/wuerstchen3/previewer.safetensors \
--sample_prompts /mnt/DataSSD/AI/SoteDiffusion/Wuerstchen3/config/sotediffusion-prompt.txt
Limitations and Bias
Bias
- This model is intended for anime illustrations.
Realistic capabilites are not tested at all. - Still underbaked.
Limitations
- Can fall back to realistic.
Add "realistic" tag to the negatives when this happens. - Far shot eyes are can bad.
- Anatomy and hands can bad.
License
(This part is copied directly from Animagine V3.1 and modified.)
SoteDiffusion models falls under Fair AI Public License 1.0-SD license, which is compatible with Stable Diffusion models’ license. Key points:
- Modification Sharing: If you modify SoteDiffusion models, you must share both your changes and the original license.
- Source Code Accessibility: If your modified version is network-accessible, provide a way (like a download link) for others to get the source code. This applies to derived models too.
- Distribution Terms: Any distribution must be under this license or another with similar rules.
- Compliance: Non-compliance must be fixed within 30 days to avoid license termination, emphasizing transparency and adherence to open-source values.
Notes: Anything not covered by Fair AI license is inherited from Stability AI Non-Commercial license which is named as LICENSE_INHERIT. Meaning, still no commercial use of any kind.