Hugging Face's logo

glif-loradex-trainer
/
swapagrawal14_flux_dev_Indian_anime_fusion_v1

Text-to-Image
Diffusers
template:sd-lora
flux
lora
Model card Files Community
swapagrawal14_flux_dev_Indian_anime_fusion_v1
File size: 2,324 Bytes 
feb3983
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
 
config:
  name: flux_dev_Indian_anime_fusion_v1
  process:
  - datasets:
    - cache_latents_to_disk: true
      caption_dropout_rate: 0.2
      caption_ext: txt
      folder_path: /root/lorahub/flux_dev_Indian_anime_fusion_v1/dataset
      resolution:
      - 512
      - 768
      - 1024
      shuffle_tokens: false
      token_dropout_rate: 0.01
    device: cuda:0
    model:
      is_flux: true
      name_or_path: black-forest-labs/FLUX.1-dev
      quantize: true
      text_encoder_bits: 8
    network:
      linear: 42
      linear_alpha: 42
      transformer_only: true
      type: lora
    performance_log_every: 500
    sample:
      height: 1024
      neg: ''
      prompts:
      - a woman selling ice cream in dairy farm  [trigger]
      - a woman sitting on sofa and taking care of her pet [trigger]
      - a man doing farming in hot summer with bull [trigger]
      - worshiping the cow is the purity in Hinduism  [trigger]
      - children celebrating diwali in light, crackers and lightning  [trigger]
      - a businessman got first prize in his field, smiling and thanking on stage
        [trigger]
      sample_every: 500
      sample_steps: 25
      sampler: flowmatch
      seed: 42
      walk_seed: true
      width: 1024
    save:
      dtype: float16
      max_step_saves_to_keep: 3
      save_every: 500
      save_format: diffusers
    train:
      batch_size: 1
      dtype: bf16
      ema_config:
        ema_decay: 0.99
        use_ema: true
      gradient_accumulation_steps: 1
      gradient_checkpointing: true
      linear_timesteps: true
      loss_type: mse
      lr: 0.0002
      noise_scheduler: flowmatch
      optimizer: adamw8bit
      reg_weight: 1.0
      steps: 3000
      target_noise_multiplier: 1.0
      train_text_encoder: false
      train_unet: true
    training_folder: /root/lorahub
    trigger_word: ' indo_swaping_chan '
    type: sd_trainer
job: extension
meta:
  description: This LoRa model is designed to capture the fusion of Indian cultural
    elements with anime aesthetics. It emphasizes intricate details in traditional
    attire, jewelry, and patterns, while maintaining the expressive, stylized look
    of anime art. The result is a unique blend that highlights Indian heritage through
    a colorful, high-quality anime style with depth and character.