File size: 4,797 Bytes
4450790
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
133
134
135
136
137
138
139
140
141
142
"""
    This file is part of ComfyUI.
    Copyright (C) 2024 Stability AI

    This program is free software: you can redistribute it and/or modify
    it under the terms of the GNU General Public License as published by
    the Free Software Foundation, either version 3 of the License, or
    (at your option) any later version.

    This program is distributed in the hope that it will be useful,
    but WITHOUT ANY WARRANTY; without even the implied warranty of
    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
    GNU General Public License for more details.

    You should have received a copy of the GNU General Public License
    along with this program.  If not, see <https://www.gnu.org/licenses/>.
"""

import torch
import nodes
import comfy.utils


class StableCascade_EmptyLatentImage:
    def __init__(self, device="cpu"):
        self.device = device

    @classmethod
    def INPUT_TYPES(s):
        return {"required": {
            "width": ("INT", {"default": 1024, "min": 256, "max": nodes.MAX_RESOLUTION, "step": 8}),
            "height": ("INT", {"default": 1024, "min": 256, "max": nodes.MAX_RESOLUTION, "step": 8}),
            "compression": ("INT", {"default": 42, "min": 4, "max": 128, "step": 1}),
            "batch_size": ("INT", {"default": 1, "min": 1, "max": 4096})
        }}
    RETURN_TYPES = ("LATENT", "LATENT")
    RETURN_NAMES = ("stage_c", "stage_b")
    FUNCTION = "generate"

    CATEGORY = "latent/stable_cascade"

    def generate(self, width, height, compression, batch_size=1):
        c_latent = torch.zeros([batch_size, 16, height // compression, width // compression])
        b_latent = torch.zeros([batch_size, 4, height // 4, width // 4])
        return ({
            "samples": c_latent,
        }, {
            "samples": b_latent,
        })

class StableCascade_StageC_VAEEncode:
    def __init__(self, device="cpu"):
        self.device = device

    @classmethod
    def INPUT_TYPES(s):
        return {"required": {
            "image": ("IMAGE",),
            "vae": ("VAE", ),
            "compression": ("INT", {"default": 42, "min": 4, "max": 128, "step": 1}),
        }}
    RETURN_TYPES = ("LATENT", "LATENT")
    RETURN_NAMES = ("stage_c", "stage_b")
    FUNCTION = "generate"

    CATEGORY = "latent/stable_cascade"

    def generate(self, image, vae, compression):
        width = image.shape[-2]
        height = image.shape[-3]
        out_width = (width // compression) * vae.downscale_ratio
        out_height = (height // compression) * vae.downscale_ratio

        s = comfy.utils.common_upscale(image.movedim(-1,1), out_width, out_height, "bicubic", "center").movedim(1,-1)

        c_latent = vae.encode(s[:,:,:,:3])
        b_latent = torch.zeros([c_latent.shape[0], 4, (height // 8) * 2, (width // 8) * 2])
        return ({
            "samples": c_latent,
        }, {
            "samples": b_latent,
        })

class StableCascade_StageB_Conditioning:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "conditioning": ("CONDITIONING",),
                              "stage_c": ("LATENT",),
                             }}
    RETURN_TYPES = ("CONDITIONING",)

    FUNCTION = "set_prior"

    CATEGORY = "conditioning/stable_cascade"

    def set_prior(self, conditioning, stage_c):
        c = []
        for t in conditioning:
            d = t[1].copy()
            d['stable_cascade_prior'] = stage_c['samples']
            n = [t[0], d]
            c.append(n)
        return (c, )

class StableCascade_SuperResolutionControlnet:
    def __init__(self, device="cpu"):
        self.device = device

    @classmethod
    def INPUT_TYPES(s):
        return {"required": {
            "image": ("IMAGE",),
            "vae": ("VAE", ),
        }}
    RETURN_TYPES = ("IMAGE", "LATENT", "LATENT")
    RETURN_NAMES = ("controlnet_input", "stage_c", "stage_b")
    FUNCTION = "generate"

    EXPERIMENTAL = True
    CATEGORY = "_for_testing/stable_cascade"

    def generate(self, image, vae):
        width = image.shape[-2]
        height = image.shape[-3]
        batch_size = image.shape[0]
        controlnet_input = vae.encode(image[:,:,:,:3]).movedim(1, -1)

        c_latent = torch.zeros([batch_size, 16, height // 16, width // 16])
        b_latent = torch.zeros([batch_size, 4, height // 2, width // 2])
        return (controlnet_input, {
            "samples": c_latent,
        }, {
            "samples": b_latent,
        })

NODE_CLASS_MAPPINGS = {
    "StableCascade_EmptyLatentImage": StableCascade_EmptyLatentImage,
    "StableCascade_StageB_Conditioning": StableCascade_StageB_Conditioning,
    "StableCascade_StageC_VAEEncode": StableCascade_StageC_VAEEncode,
    "StableCascade_SuperResolutionControlnet": StableCascade_SuperResolutionControlnet,
}