Hugging Face's logo

williamberman
/
consistency-decoder

Diffusers
Safetensors
Model card Files
xet
Community
1
williamberman commited on
Commit
39f0fcf
·
1 Parent(s): 42028c0

vae not pipeline

Browse files
Files changed (5) hide show
  1. config.json +91 -0
  2. unet/diffusion_pytorch_model.safetensors → diffusion_pytorch_model.safetensors +2 -2
  3. model_index.json +0 -12
  4. scheduler/scheduler_config.json +0 -6
  5. unet/config.json +0 -45
config.json ADDED
@@ -0,0 +1,91 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_class_name": "ConsistencyDecoderVae",
3
+ "_diffusers_version": "0.23.0.dev0",
4
+ "block_out_channels": [
5
+ 128,
6
+ 256,
7
+ 512,
8
+ 512
9
+ ],
10
+ "decoder_args": {
11
+ "_use_default_values": [
12
+ "downsample_type",
13
+ "flip_sin_to_cos",
14
+ "mid_block_scale_factor",
15
+ "num_class_embeds",
16
+ "attention_head_dim",
17
+ "sample_size",
18
+ "act_fn",
19
+ "freq_shift",
20
+ "class_embed_type",
21
+ "attn_norm_num_groups",
22
+ "downsample_padding",
23
+ "upsample_type",
24
+ "dropout",
25
+ "center_input_sample"
26
+ ],
27
+ "act_fn": "silu",
28
+ "add_attention": false,
29
+ "attention_head_dim": 8,
30
+ "attn_norm_num_groups": null,
31
+ "block_out_channels": [
32
+ 320,
33
+ 640,
34
+ 1024,
35
+ 1024
36
+ ],
37
+ "center_input_sample": false,
38
+ "class_embed_type": null,
39
+ "down_block_types": [
40
+ "ResnetDownsampleBlock2D",
41
+ "ResnetDownsampleBlock2D",
42
+ "ResnetDownsampleBlock2D",
43
+ "ResnetDownsampleBlock2D"
44
+ ],
45
+ "downsample_padding": 1,
46
+ "downsample_type": "conv",
47
+ "dropout": 0.0,
48
+ "flip_sin_to_cos": true,
49
+ "freq_shift": 0,
50
+ "in_channels": 7,
51
+ "layers_per_block": 3,
52
+ "mid_block_scale_factor": 1,
53
+ "norm_eps": 1e-05,
54
+ "norm_num_groups": 32,
55
+ "num_class_embeds": null,
56
+ "num_train_timesteps": 1024,
57
+ "out_channels": 6,
58
+ "resnet_time_scale_shift": "scale_shift",
59
+ "sample_size": null,
60
+ "time_embedding_type": "learned",
61
+ "up_block_types": [
62
+ "ResnetUpsampleBlock2D",
63
+ "ResnetUpsampleBlock2D",
64
+ "ResnetUpsampleBlock2D",
65
+ "ResnetUpsampleBlock2D"
66
+ ],
67
+ "upsample_type": "conv"
68
+ },
69
+ "encoder_args": {
70
+ "act_fn": "silu",
71
+ "block_out_channels": [
72
+ 128,
73
+ 256,
74
+ 512,
75
+ 512
76
+ ],
77
+ "double_z": true,
78
+ "down_block_types": [
79
+ "DownEncoderBlock2D",
80
+ "DownEncoderBlock2D",
81
+ "DownEncoderBlock2D",
82
+ "DownEncoderBlock2D"
83
+ ],
84
+ "in_channels": 3,
85
+ "layers_per_block": 2,
86
+ "norm_num_groups": 32,
87
+ "out_channels": 4
88
+ },
89
+ "latent_channels": 4,
90
+ "scaling_factor": 0.18215
91
+ }
unet/diffusion_pytorch_model.safetensors → diffusion_pytorch_model.safetensors RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:08ea9dbc43a94fbf2dd45006af9e842451c8d8b51ae97f4001ba9ce3e7b3f2e8
3
- size 2485155472
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:df6386a06f741c7dfaaa4709b6db39aca345a406d8413dc2c4dbd48a0dce1cfb
3
+ size 1310944884
model_index.json DELETED
@@ -1,12 +0,0 @@
1
- {
2
- "_class_name": "ConsistencyModelPipeline",
3
- "_diffusers_version": "0.23.0.dev0",
4
- "scheduler": [
5
- "diffusers",
6
- "ConsistencyDecoderScheduler"
7
- ],
8
- "unet": [
9
- "diffusers",
10
- "UNet2DModel"
11
- ]
12
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
scheduler/scheduler_config.json DELETED
@@ -1,6 +0,0 @@
1
- {
2
- "_class_name": "ConsistencyDecoderScheduler",
3
- "_diffusers_version": "0.23.0.dev0",
4
- "num_train_timesteps": 1024,
5
- "sigma_data": 0.5
6
- }
 
 
 
 
 
 
 
unet/config.json DELETED
@@ -1,45 +0,0 @@
1
- {
2
- "_class_name": "UNet2DModel",
3
- "_diffusers_version": "0.23.0.dev0",
4
- "act_fn": "silu",
5
- "add_attention": false,
6
- "attention_head_dim": 8,
7
- "attn_norm_num_groups": null,
8
- "block_out_channels": [
9
- 320,
10
- 640,
11
- 1024,
12
- 1024
13
- ],
14
- "center_input_sample": false,
15
- "class_embed_type": null,
16
- "down_block_types": [
17
- "ResnetDownsampleBlock2D",
18
- "ResnetDownsampleBlock2D",
19
- "ResnetDownsampleBlock2D",
20
- "ResnetDownsampleBlock2D"
21
- ],
22
- "downsample_padding": 1,
23
- "downsample_type": "conv",
24
- "dropout": 0.0,
25
- "flip_sin_to_cos": true,
26
- "freq_shift": 0,
27
- "in_channels": 7,
28
- "layers_per_block": 3,
29
- "mid_block_scale_factor": 1,
30
- "norm_eps": 1e-05,
31
- "norm_num_groups": 32,
32
- "num_class_embeds": null,
33
- "num_train_timesteps": 1024,
34
- "out_channels": 6,
35
- "resnet_time_scale_shift": "scale_shift",
36
- "sample_size": null,
37
- "time_embedding_type": "learned",
38
- "up_block_types": [
39
- "ResnetUpsampleBlock2D",
40
- "ResnetUpsampleBlock2D",
41
- "ResnetUpsampleBlock2D",
42
- "ResnetUpsampleBlock2D"
43
- ],
44
- "upsample_type": "conv"
45
- }