Hugging Face's logo

Spaces:
Wmcs91
/
Versatile-Diffusion
Runtime error

App Files Community
Versatile-Diffusion
File size: 1,390 Bytes 
fb53ec8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
 
sd_base:
  symbol: sd
  find_unused_parameters: true

sd_autoencoder:
  type: autoencoderkl
  args:
    embed_dim: 4
    monitor: val/rec_loss
    ddconfig:
      double_z: true
      z_channels: 4
      resolution: 256
      in_channels: 3
      out_ch: 3
      ch: 128
      ch_mult: [1, 2, 4, 4]
      num_res_blocks: 2
      attn_resolutions: []
      dropout: 0.0
    lossconfig:
      target: torch.nn.Identity
  pth: pretrained/kl-f8.pth

sd_t2i:
  super_cfg: sd_base
  type: sd_t2i
  args:
    first_stage_config: MODEL(sd_autoencoder)
    cond_stage_config: MODEL(clip_text_frozen)
    unet_config: MODEL(openai_unet_sd)
    beta_linear_start: 0.00085
    beta_linear_end: 0.012
    num_timesteps_cond: 1
    timesteps: 1000
    scale_factor: 0.18215
    use_ema: true

sd_t2i_noema:
  super_cfg: sd
  args:
    use_ema: false

#####################
# sd with full clip #
#####################

sd_t2i_fullclip_backward_compatible:
  super_cfg: sd_t2i
  args:
    cond_stage_config: MODEL(clip_frozen_encode_text_noproj)

sd_t2i_fullclip_backward_compatible_noema:
  super_cfg: sd_t2i_noema
  args:
    cond_stage_config: MODEL(clip_frozen_encode_text_noproj)

sd_t2i_fullclip:
  super_cfg: sd_t2i
  args:
    cond_stage_config: MODEL(clip_frozen_encode_text)

sd_variation:
  super_cfg: sd_t2i
  type: sd_variation
  args:
    cond_stage_config: MODEL(clip_vision_frozen_justin)