Hugging Face's logo

CyberHarem
/
vigna_arknights

Text-to-Image
art
Not-For-All-Audiences
Model card Files Community
vigna_arknights
File size: 2,585 Bytes 
6d32cec
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
 
[ARB]
bucket_no_upscale = false
bucket_reso_steps = 64
enable_bucket = true
max_bucket_reso = 1584
min_bucket_reso = 327

[Attention]
mem_eff_attn = false
xformers = true

[Basics]
clip_skip = 2
max_train_epochs = 30
max_train_steps = 1073741824
pretrained_model_name_or_path = "******"
reg_data_dir = "******"
resolution = "720,720"
seed = 171927893
train_data_dir = "******"

[Cache_latents]
cache_latents = true
vae_batch_size = 1
cache_latents_to_disk = true

[Captions]
shuffle_caption = true
caption_extension = ".txt"
keep_tokens = 1
caption_dropout_rate = 0.05
caption_dropout_every_n_epochs = 0
caption_tag_dropout_rate = 0.0
max_token_length = 150
weighted_captions = false
token_warmup_min = 1
token_warmup_step = 0

[Data_augmentation]
color_aug = false
flip_aug = false
random_crop = false

[Dataset]
max_data_loader_n_workers = 8
persistent_data_loader_workers = true
dataset_repeats = 1

[Debugging]
debug_dataset = false

[Deprecated]
use_8bit_adam = false
use_lion_optimizer = false
learning_rate = 0.0002

[Further_improvement]
min_snr_gamma = 0
multires_noise_discount = 0.3
multires_noise_iterations = 6

[Huggingface]
save_state_to_huggingface = false
resume_from_huggingface = false
async_upload = false

[Logging]
logging_dir = "******"
log_with = "tensorboard"
log_prefix = "lora_"

[Lr_scheduler]
lr_scheduler_type = ""
lr_scheduler = "constant"
lr_warmup_steps = 0
lr_scheduler_num_cycles = 1
lr_scheduler_power = 1.0

[LyCORIS]
network_module = "lycoris.kohya"
network_args = [ "preset=attn-mlp", "algo=lora",]

[Network_setup]
dim_from_weights = false
network_alpha = 2
network_dim = 4
network_dropout = 0
network_train_text_encoder_only = false
network_train_unet_only = true
resume = false

[Optimizer]
gradient_accumulation_steps = 1
gradient_checkpointing = true
max_grad_norm = 1.0
optimizer_args = [ "weight_decay=0.1", "betas=0.9,0.99",]
optimizer_type = "AdamW8bit"
text_encoder_lr = 0.0006
train_batch_size = 8
unet_lr = 0.0006

[Others]
lowram = false
training_comment = "narugo1992's automated LoRA training, based on nebulae's config."

[Regularization]
prior_loss_weight = 1.0

[SDv2]
v2 = false
v_parameterization = false
scale_v_pred_loss_like_noise_pred = false

[Sampling_during_training]
sample_sampler = "ddim"

[Save]
output_dir = "******"
output_name = "vigna_arknights"
save_every_n_epochs = 2
save_every_n_steps = 1073741824
save_last_n_steps = 200
save_last_n_steps_state = 1
save_model_as = "safetensors"
save_precision = "fp16"
save_state = false

[Training_preciscion]
mixed_precision = "fp16"
full_fp16 = false
full_bf16 = false