Hugging Face's logo

UmerHA
/
ConrolNetXS-SDXL-canny

Diffusers
Safetensors
Model card Files Community
UmerHA commited on
Commit
ecaef15
•
1 Parent(s): 4cd9b03

Upload 2 files

Browse files
Files changed (2) hide show
  1. config.json +81 -48
  2. diffusion_pytorch_model.safetensors +2 -2
config.json CHANGED
@@ -1,88 +1,96 @@
1
  {
2
  "_class_name": "ControlNetXSModel",
3
  "_diffusers_version": "0.22.0.dev0",
4
- "addition_embed_type": "text_time",
 
 
 
 
 
 
 
 
5
  "base_model_channel_sizes": {
6
- "dec": [
7
  [
8
- 2560,
9
- 1280
10
  ],
11
  [
12
- 2560,
13
- 1280
14
  ],
15
  [
16
- 1920,
17
- 1280
18
  ],
19
  [
20
- 1920,
21
- 640
22
  ],
23
  [
24
- 1280,
25
  640
26
  ],
27
  [
28
- 960,
29
  640
30
  ],
31
  [
32
- 960,
33
- 320
34
  ],
35
  [
36
  640,
37
- 320
38
  ],
39
  [
40
- 640,
41
- 320
42
  ]
43
  ],
44
- "enc": [
45
  [
46
- 4,
47
- 320
48
- ],
 
 
49
  [
50
- 320,
51
- 320
52
  ],
53
  [
54
- 320,
55
- 320
56
  ],
57
  [
58
- 320,
59
- 320
60
  ],
61
  [
62
- 320,
63
  640
64
  ],
65
  [
66
- 640,
67
  640
68
  ],
69
  [
70
- 640,
71
  640
72
  ],
73
  [
74
- 640,
75
- 1280
76
  ],
77
  [
78
- 1280,
79
- 1280
80
- ]
81
- ],
82
- "mid": [
83
  [
84
- 1280,
85
- 1280
86
  ]
87
  ]
88
  },
@@ -91,25 +99,48 @@
91
  64,
92
  128
93
  ],
94
- "control_attention_head_dim": 64,
95
- "control_model_ratio": 0.1,
96
- "control_scale": 0.95,
 
 
97
  "cross_attention_dim": 2048,
 
98
  "down_block_types": [
99
  "DownBlock2D",
100
  "CrossAttnDownBlock2D",
101
  "CrossAttnDownBlock2D"
102
  ],
103
- "global_pool_conditions": false,
104
- "hint_channels": 3,
 
 
 
 
 
105
  "learn_embedding": true,
106
- "model_channels": 320,
 
 
 
 
 
 
107
  "out_channels": 4,
108
- "time_control_scale": 1,
 
 
 
 
 
 
109
  "time_embedding_dim": 1280,
110
  "time_embedding_input_dim": 320,
 
 
 
111
  "transformer_layers_per_block": [
112
- 0,
113
  2,
114
  10
115
  ],
@@ -117,5 +148,7 @@
117
  "CrossAttnUpBlock2D",
118
  "CrossAttnUpBlock2D",
119
  "UpBlock2D"
120
- ]
 
 
121
  }
 
1
  {
2
  "_class_name": "ControlNetXSModel",
3
  "_diffusers_version": "0.22.0.dev0",
4
+ "_name_or_path": "/Users/umer/.cache/huggingface/hub/models--stabilityai--stable-diffusion-xl-base-1.0/snapshots/462165984030d82259a11f4367a4eed129e94a7b/unet",
5
+ "act_fn": "silu",
6
+ "addition_embed_type_num_heads": 64,
7
+ "attention_head_dim": [
8
+ 1,
9
+ 1,
10
+ 2
11
+ ],
12
+ "attention_type": "default",
13
  "base_model_channel_sizes": {
14
+ "down": [
15
  [
16
+ 4,
17
+ 320
18
  ],
19
  [
20
+ 320,
21
+ 320
22
  ],
23
  [
24
+ 320,
25
+ 320
26
  ],
27
  [
28
+ 320,
29
+ 320
30
  ],
31
  [
32
+ 320,
33
  640
34
  ],
35
  [
36
+ 640,
37
  640
38
  ],
39
  [
40
+ 640,
41
+ 640
42
  ],
43
  [
44
  640,
45
+ 1280
46
  ],
47
  [
48
+ 1280,
49
+ 1280
50
  ]
51
  ],
52
+ "mid": [
53
  [
54
+ 1280,
55
+ 1280
56
+ ]
57
+ ],
58
+ "up": [
59
  [
60
+ 2560,
61
+ 1280
62
  ],
63
  [
64
+ 2560,
65
+ 1280
66
  ],
67
  [
68
+ 1920,
69
+ 1280
70
  ],
71
  [
72
+ 1920,
73
  640
74
  ],
75
  [
76
+ 1280,
77
  640
78
  ],
79
  [
80
+ 960,
81
  640
82
  ],
83
  [
84
+ 960,
85
+ 320
86
  ],
87
  [
88
+ 640,
89
+ 320
90
+ ],
 
 
91
  [
92
+ 640,
93
+ 320
94
  ]
95
  ]
96
  },
 
99
  64,
100
  128
101
  ],
102
+ "center_input_sample": false,
103
+ "conditioning_channels": 3,
104
+ "controlnet_conditioning_channel_order": "rgb",
105
+ "conv_in_kernel": 3,
106
+ "conv_out_kernel": 3,
107
  "cross_attention_dim": 2048,
108
+ "cross_attention_norm": null,
109
  "down_block_types": [
110
  "DownBlock2D",
111
  "CrossAttnDownBlock2D",
112
  "CrossAttnDownBlock2D"
113
  ],
114
+ "downsample_padding": 1,
115
+ "dropout": 0.0,
116
+ "dual_cross_attention": false,
117
+ "encoder_hid_dim": null,
118
+ "encoder_hid_dim_type": null,
119
+ "in_channels": 4,
120
+ "layers_per_block": 2,
121
  "learn_embedding": true,
122
+ "mid_block_only_cross_attention": null,
123
+ "mid_block_scale_factor": 1,
124
+ "mid_block_type": "UNetMidBlock2DCrossAttn",
125
+ "norm_eps": 1e-05,
126
+ "norm_num_groups": 32,
127
+ "num_attention_heads": null,
128
+ "only_cross_attention": false,
129
  "out_channels": 4,
130
+ "resnet_out_scale_factor": 1.0,
131
+ "resnet_skip_time_act": false,
132
+ "resnet_time_scale_shift": "default",
133
+ "reverse_transformer_layers_per_block": null,
134
+ "sample_size": 128,
135
+ "time_cond_proj_dim": null,
136
+ "time_embedding_act_fn": null,
137
  "time_embedding_dim": 1280,
138
  "time_embedding_input_dim": 320,
139
+ "time_embedding_mix": 0.95,
140
+ "time_embedding_type": "positional",
141
+ "timestep_post_act": null,
142
  "transformer_layers_per_block": [
143
+ 1,
144
  2,
145
  10
146
  ],
 
148
  "CrossAttnUpBlock2D",
149
  "CrossAttnUpBlock2D",
150
  "UpBlock2D"
151
+ ],
152
+ "upcast_attention": null,
153
+ "use_linear_projection": true
154
  }
diffusion_pytorch_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2305c5457f836fcf2e310d44e3547d703e88ca28064bab2decdbbe688db6df9d
3
- size 169820668
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:de1984bcf1b697b3bb28b6c2b48ea5b427b675dbd88b7567c9e060aab1cca27e
3
+ size 169820488