wereanimal_2.0
세부 정보
| 모델 유형 | LORA |
| 기본 모델 | SD 1.5 |
| 게시일 | 2023-04-05 |
| 학습된 단어 | werenaimal werecreature Humanoid weredog weresloth weredeer werebird werelizard wereshark werecat werewolf weretiger werefox wereAxolotl werelion weremouse wererabbit werefrog werepanthers wereSideburns weresheep wereraccoon werebear weresnake werehorse wereCaterpillar werepig wereLeopard |
이 버전에 대해
Increased species and resolution compared to version 1.0
When using common species such as dogs tiger, cat......, you only need to add the species name without necessarily using a trigger word.
However, using a trigger word can more accurately call up specific data such as gender, caterpillars......
There are the following species:
-- were_\(dog\), were_\(male-dog\), were_\(female-dog\),weredog, dog,
--were_\(bird\), were_\(male-bird\), were_\(female-bird\), werebird, bird,
-- were_\(lizard\), were_\(male-lizard\), were_\(female-lizard\),werelizard, lizard,
--were_\(shark\),Same as above.......
--were_\(cat\),Same as above.......
--were_\(wolf\),Same as above.......
--were_\(tiger\),Same as above.......
--were_\(fox\),Same as above.......
--were_\(salamander\),Same as above.......
--were_\(lion\),Same as above.......
--were_\(mouse\),Same as above.......
--were_\(rabbit\),Same as above.......
--were_\(frog\),Same as above.......
--were_\(ox\),Same as above.......
--were_\(Chimpanzees\),Same as above.......
--were_\(horse\),Same as above.......
--were_\(weasel\),Same as above.......
--were_\(Caterpillar\),Same as above.......
--were_\(deer\),Same as above.......
--were_\(pig\),Same as above.......
--were_\(sloth\),Same as above.......
load StableDiffusion checkpoint Folder 10_cat-female_BD_1: 72 images found Folder 10_cat-female_BD_1: 720 steps Folder 10_wolf_male_BD: 75 images found Folder 10_wolf_male_BD: 750 steps Folder 11_tiger_male_BD: 73 images found Folder 11_tiger_male_BD: 803 steps Folder 125_ox_femail_BD: 6 images found Folder 125_ox_femail_BD: 750 steps Folder 125_rabbit_male_BD: 6 images found Folder 125_rabbit_male_BD: 750 steps Folder 12_cat-male_BD1: 66 images found Folder 12_cat-male_BD1: 792 steps Folder 13_fox-female_BD: 58 images found Folder 13_fox-female_BD: 754 steps Folder 13_fox-male_BD: 58 images found Folder 13_fox-male_BD: 754 steps Folder 150_pig: 1 images found Folder 150_pig: 150 steps Folder 16_salamander_BD: 49 images found Folder 16_salamander_BD: 784 steps Folder 20_lion-male_BD: 38 images found Folder 20_lion-male_BD: 760 steps Folder 20_mouse_male_BD: 38 images found Folder 20_mouse_male_BD: 760 steps Folder 20_rabbit_female_BD: 40 images found Folder 20_rabbit_female_BD: 800 steps Folder 21_bird-female_BD_1: 36 images found Folder 21_bird-female_BD_1: 756 steps Folder 23_mouse_female_BD: 33 images found Folder 23_mouse_female_BD: 759 steps Folder 23_werefrog: 33 images found Folder 23_werefrog: 759 steps Folder 250_hippo-female_BD: 3 images found Folder 250_hippo-female_BD: 750 steps Folder 25_lizard-female_BD: 30 images found Folder 25_lizard-female_BD: 750 steps Folder 25_tiger_female_BD: 30 images found Folder 25_tiger_female_BD: 750 steps Folder 26_Leopard_male_BD: 29 images found Folder 26_Leopard_male_BD: 754 steps Folder 26_Sideburns_female_BD: 29 images found Folder 26_Sideburns_female_BD: 754 steps Folder 27_werejellyfish: 28 images found Folder 27_werejellyfish: 756 steps Folder 29_shark_female_BD: 26 images found Folder 29_shark_female_BD: 754 steps Folder 29_wolf_female_BD: 26 images found Folder 29_wolf_female_BD: 754 steps Folder 30_Leopard_female_BD: 25 images found Folder 30_Leopard_female_BD: 750 steps Folder 33_deer-male_BD: 23 images found Folder 33_deer-male_BD: 759 steps Folder 33_sheep_male_BD: 23 images found Folder 33_sheep_male_BD: 759 steps Folder 357_sloth_BD: 2 images found Folder 357_sloth_BD: 714 steps Folder 375_Leopard_female_panthers_BD: 2 images found Folder 375_Leopard_female_panthers_BD: 750 steps Folder 40_lion-female_BD: 19 images found Folder 40_lion-female_BD: 760 steps Folder 44_raccoon_male_BD: 17 images found Folder 44_raccoon_male_BD: 748 steps Folder 50_bear_BD1: 15 images found Folder 50_bear_BD1: 750 steps Folder 50_raccoon_female_BD: 15 images found Folder 50_raccoon_female_BD: 750 steps Folder 50_sheep_female_BD: 16 images found Folder 50_sheep_female_BD: 800 steps Folder 58_Sideburns_male_BD: 13 images found Folder 58_Sideburns_male_BD: 754 steps Folder 68_snake_BD: 11 images found Folder 68_snake_BD: 748 steps Folder 69_Leopard_male_panthers_BD: 11 images found Folder 69_Leopard_male_panthers_BD: 759 steps Folder 75_ox_male_BD: 10 images found Folder 75_ox_male_BD: 750 steps Folder 7_dog-male_BD: 124 images found Folder 7_dog-male_BD: 868 steps Folder 84_Chimpanzees-male_BD: 9 images found Folder 84_Chimpanzees-male_BD: 756 steps Folder 84_horse-female_BD: 9 images found Folder 84_horse-female_BD: 756 steps Folder 84_weasel_female_BD: 9 images found Folder 84_weasel_female_BD: 756 steps Folder 84_wereCaterpillar: 9 images found Folder 84_wereCaterpillar: 756 steps Folder 8_bird-male_BD_1: 103 images found Folder 8_bird-male_BD_1: 824 steps Folder 94_deer-female_BD: 8 images found Folder 94_deer-female_BD: 752 steps Folder 9_dog-female_BD: 90 images found Folder 9_dog-female_BD: 810 steps Folder 9_lizard-male_BD: 86 images found Folder 9_lizard-male_BD: 774 steps Folder 9_shark_male_BD: 89 images found Folder 9_shark_male_BD: 801 steps max_train_steps = 108081 stop_text_encoder_training = 0 lr_warmup_steps = 0Here is the trainer command as a reference. It will not be executed:
accelerate launch –num_cpu_threads_per_process=2 “train_network.py” –pretrained_model_name_or_path=“D:/ai/stable-diffusion-webui/models/Stable-diffusion/v1-5-pruned.safetensors” –train_data_dir=“D:\ai\test\image\100_wereanimal” –resolution=768,768 –output_dir=“D:/ai/test/model” –logging_dir=“D:/ai/test/log” –network_alpha=“128” –save_model_as=safetensors –network_module=networks.lora –text_encoder_lr=5e-5 –unet_lr=0.0001 –network_dim=200 –output_name=“7_wereanimal_1v5_768_0001BD5e_NR128NA128off” –lr_scheduler_num_cycles=“3” –learning_rate=“0.0001” –lr_scheduler=“constant” –train_batch_size=“1” –max_train_steps=“108081” –save_every_n_epochs=“1” –mixed_precision=“bf16” –save_precision=“bf16” –seed=“1234” –caption_extension=”.txt” –cache_latents –optimizer_type=“AdamW8bit” –max_data_loader_n_workers=“1” –clip_skip=2 –bucket_reso_steps=64 –xformers –bucket_no_upscale
모델 설명
It seems that the first version: wereanimal_1.0 has been removed from the website.
This is a LORA for animal humanoid .
Setting the weight to 0.2-0.8 will result in better performance.
Using the following Checkpoints model will result in more stable performance:
https://civitai.com/models/11718/dungeons-n-waifus-new-version-22
https://civitai.com/models/6250/dosmix
https://civitai.com/models/6925/realdosmix
https://civitai.com/models/7371/rev-animated
-You can also incorporate the Rola module of any real animal to achieve a more realistic effect.
-Using the pose controller can significantly improve the success rate.
It would be great if you could share your works using this module with me!