nodes.py 56.5 KB
Newer Older
comfyanonymous's avatar
comfyanonymous committed
1
2
3
4
5
import torch

import os
import sys
import json
6
import hashlib
7
import traceback
8
import math
9
import time
space-nuko's avatar
space-nuko committed
10
11
import struct
from io import BytesIO
comfyanonymous's avatar
comfyanonymous committed
12

13
from PIL import Image, ImageOps
comfyanonymous's avatar
comfyanonymous committed
14
15
from PIL.PngImagePlugin import PngInfo
import numpy as np
16
import safetensors.torch
comfyanonymous's avatar
comfyanonymous committed
17

sALTaccount's avatar
sALTaccount committed
18

comfyanonymous's avatar
comfyanonymous committed
19
sys.path.insert(0, os.path.join(os.path.dirname(os.path.realpath(__file__)), "comfy"))
comfyanonymous's avatar
comfyanonymous committed
20
21


22
import comfy.diffusers_load
comfyanonymous's avatar
comfyanonymous committed
23
import comfy.samplers
24
import comfy.sample
comfyanonymous's avatar
comfyanonymous committed
25
import comfy.sd
comfyanonymous's avatar
comfyanonymous committed
26
import comfy.utils
space-nuko's avatar
space-nuko committed
27
from comfy.taesd.taesd import TAESD
comfyanonymous's avatar
comfyanonymous committed
28

29
import comfy.clip_vision
30

31
import comfy.model_management
32
import importlib
comfyanonymous's avatar
comfyanonymous committed
33

34
import folder_paths
35

Dr.Lt.Data's avatar
Dr.Lt.Data committed
36

37
38
39
40
41
class LatentPreviewer:
    def decode_latent_to_preview(self, device, x0):
        pass


42
def before_node_execution():
43
    comfy.model_management.throw_exception_if_processing_interrupted()
44

45
def interrupt_processing(value=True):
46
    comfy.model_management.interrupt_current_processing(value)
47

48
MAX_RESOLUTION=8192
space-nuko's avatar
space-nuko committed
49
MAX_PREVIEW_RESOLUTION = 512
50

comfyanonymous's avatar
comfyanonymous committed
51
52
53
class CLIPTextEncode:
    @classmethod
    def INPUT_TYPES(s):
54
        return {"required": {"text": ("STRING", {"multiline": True}), "clip": ("CLIP", )}}
comfyanonymous's avatar
comfyanonymous committed
55
56
57
    RETURN_TYPES = ("CONDITIONING",)
    FUNCTION = "encode"

58
59
    CATEGORY = "conditioning"

comfyanonymous's avatar
comfyanonymous committed
60
    def encode(self, clip, text):
comfyanonymous's avatar
comfyanonymous committed
61
62
63
64
65
66
67
68
69
        return ([[clip.encode(text), {}]], )

class ConditioningCombine:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": {"conditioning_1": ("CONDITIONING", ), "conditioning_2": ("CONDITIONING", )}}
    RETURN_TYPES = ("CONDITIONING",)
    FUNCTION = "combine"

70
71
    CATEGORY = "conditioning"

comfyanonymous's avatar
comfyanonymous committed
72
73
74
    def combine(self, conditioning_1, conditioning_2):
        return (conditioning_1 + conditioning_2, )

FizzleDorf's avatar
FizzleDorf committed
75
76
77
class ConditioningAverage :
    @classmethod
    def INPUT_TYPES(s):
comfyanonymous's avatar
comfyanonymous committed
78
79
        return {"required": {"conditioning_to": ("CONDITIONING", ), "conditioning_from": ("CONDITIONING", ),
                              "conditioning_to_strength": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01})
FizzleDorf's avatar
FizzleDorf committed
80
81
82
83
84
85
                             }}
    RETURN_TYPES = ("CONDITIONING",)
    FUNCTION = "addWeighted"

    CATEGORY = "conditioning"

comfyanonymous's avatar
comfyanonymous committed
86
    def addWeighted(self, conditioning_to, conditioning_from, conditioning_to_strength):
FizzleDorf's avatar
FizzleDorf committed
87
        out = []
comfyanonymous's avatar
comfyanonymous committed
88
89
90
91
92
93
94
95
96
97
98
99
100
101

        if len(conditioning_from) > 1:
            print("Warning: ConditioningAverage conditioning_from contains more than 1 cond, only the first one will actually be applied to conditioning_to.")

        cond_from = conditioning_from[0][0]

        for i in range(len(conditioning_to)):
            t1 = conditioning_to[i][0]
            t0 = cond_from[:,:t1.shape[1]]
            if t0.shape[1] < t1.shape[1]:
                t0 = torch.cat([t0] + [torch.zeros((1, (t1.shape[1] - t0.shape[1]), t1.shape[2]))], dim=1)

            tw = torch.mul(t1, conditioning_to_strength) + torch.mul(t0, (1.0 - conditioning_to_strength))
            n = [tw, conditioning_to[i][1].copy()]
FizzleDorf's avatar
FizzleDorf committed
102
103
104
            out.append(n)
        return (out, )

comfyanonymous's avatar
comfyanonymous committed
105
106
107
108
class ConditioningSetArea:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": {"conditioning": ("CONDITIONING", ),
109
110
111
112
                              "width": ("INT", {"default": 64, "min": 64, "max": MAX_RESOLUTION, "step": 8}),
                              "height": ("INT", {"default": 64, "min": 64, "max": MAX_RESOLUTION, "step": 8}),
                              "x": ("INT", {"default": 0, "min": 0, "max": MAX_RESOLUTION, "step": 8}),
                              "y": ("INT", {"default": 0, "min": 0, "max": MAX_RESOLUTION, "step": 8}),
comfyanonymous's avatar
comfyanonymous committed
113
114
115
116
117
                              "strength": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 10.0, "step": 0.01}),
                             }}
    RETURN_TYPES = ("CONDITIONING",)
    FUNCTION = "append"

118
119
    CATEGORY = "conditioning"

120
    def append(self, conditioning, width, height, x, y, strength):
comfyanonymous's avatar
comfyanonymous committed
121
122
123
124
125
        c = []
        for t in conditioning:
            n = [t[0], t[1].copy()]
            n[1]['area'] = (height // 8, width // 8, y // 8, x // 8)
            n[1]['strength'] = strength
126
            n[1]['set_area_to_bounds'] = False
comfyanonymous's avatar
comfyanonymous committed
127
            c.append(n)
comfyanonymous's avatar
comfyanonymous committed
128
        return (c, )
comfyanonymous's avatar
comfyanonymous committed
129

Jacob Segal's avatar
Jacob Segal committed
130
131
132
133
134
135
class ConditioningSetMask:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": {"conditioning": ("CONDITIONING", ),
                              "mask": ("MASK", ),
                              "strength": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 10.0, "step": 0.01}),
136
                              "set_cond_area": (["default", "mask bounds"],),
Jacob Segal's avatar
Jacob Segal committed
137
138
139
140
141
142
                             }}
    RETURN_TYPES = ("CONDITIONING",)
    FUNCTION = "append"

    CATEGORY = "conditioning"

143
    def append(self, conditioning, mask, set_cond_area, strength):
Jacob Segal's avatar
Jacob Segal committed
144
        c = []
145
146
147
        set_area_to_bounds = False
        if set_cond_area != "default":
            set_area_to_bounds = True
Jacob Segal's avatar
Jacob Segal committed
148
149
150
151
152
153
        if len(mask.shape) < 3:
            mask = mask.unsqueeze(0)
        for t in conditioning:
            n = [t[0], t[1].copy()]
            _, h, w = mask.shape
            n[1]['mask'] = mask
Jacob Segal's avatar
Jacob Segal committed
154
            n[1]['set_area_to_bounds'] = set_area_to_bounds
155
            n[1]['mask_strength'] = strength
Jacob Segal's avatar
Jacob Segal committed
156
157
158
            c.append(n)
        return (c, )

comfyanonymous's avatar
comfyanonymous committed
159
160
161
162
163
164
165
class VAEDecode:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "samples": ("LATENT", ), "vae": ("VAE", )}}
    RETURN_TYPES = ("IMAGE",)
    FUNCTION = "decode"

166
167
    CATEGORY = "latent"

comfyanonymous's avatar
comfyanonymous committed
168
    def decode(self, vae, samples):
169
        return (vae.decode(samples["samples"]), )
comfyanonymous's avatar
comfyanonymous committed
170

171
172
173
174
175
176
177
178
179
180
181
182
class VAEDecodeTiled:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "samples": ("LATENT", ), "vae": ("VAE", )}}
    RETURN_TYPES = ("IMAGE",)
    FUNCTION = "decode"

    CATEGORY = "_for_testing"

    def decode(self, vae, samples):
        return (vae.decode_tiled(samples["samples"]), )

space-nuko's avatar
space-nuko committed
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
class TAESDDecode:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "samples": ("LATENT", ), "taesd": ("TAESD", )}}
    RETURN_TYPES = ("IMAGE",)
    FUNCTION = "decode"

    CATEGORY = "latent"

    def decode(self, taesd, samples):
        device = comfy.model_management.get_torch_device()
        # [B, C, H, W] -> [B, H, W, C]
        pixels = taesd.decoder(samples["samples"].to(device)).permute(0, 2, 3, 1).detach().clamp(0, 1)
        return (pixels, )

comfyanonymous's avatar
comfyanonymous committed
198
199
200
201
202
203
204
class VAEEncode:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "pixels": ("IMAGE", ), "vae": ("VAE", )}}
    RETURN_TYPES = ("LATENT",)
    FUNCTION = "encode"

205
206
    CATEGORY = "latent"

207
208
209
210
    @staticmethod
    def vae_encode_crop_pixels(pixels):
        x = (pixels.shape[1] // 8) * 8
        y = (pixels.shape[2] // 8) * 8
211
        if pixels.shape[1] != x or pixels.shape[2] != y:
212
213
214
215
            x_offset = (pixels.shape[1] % 8) // 2
            y_offset = (pixels.shape[2] % 8) // 2
            pixels = pixels[:, x_offset:x + x_offset, y_offset:y + y_offset, :]
        return pixels
216

217
218
219
    def encode(self, vae, pixels):
        pixels = self.vae_encode_crop_pixels(pixels)
        t = vae.encode(pixels[:,:,:,:3])
220
        return ({"samples":t}, )
comfyanonymous's avatar
comfyanonymous committed
221

comfyanonymous's avatar
comfyanonymous committed
222
223
224
225
226
227
228
229
230
231
class VAEEncodeTiled:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "pixels": ("IMAGE", ), "vae": ("VAE", )}}
    RETURN_TYPES = ("LATENT",)
    FUNCTION = "encode"

    CATEGORY = "_for_testing"

    def encode(self, vae, pixels):
232
        pixels = VAEEncode.vae_encode_crop_pixels(pixels)
comfyanonymous's avatar
comfyanonymous committed
233
234
        t = vae.encode_tiled(pixels[:,:,:,:3])
        return ({"samples":t}, )
235

236
237
238
class VAEEncodeForInpaint:
    @classmethod
    def INPUT_TYPES(s):
239
        return {"required": { "pixels": ("IMAGE", ), "vae": ("VAE", ), "mask": ("MASK", ), "grow_mask_by": ("INT", {"default": 6, "min": 0, "max": 64, "step": 1}),}}
240
241
242
243
244
    RETURN_TYPES = ("LATENT",)
    FUNCTION = "encode"

    CATEGORY = "latent/inpaint"

245
    def encode(self, vae, pixels, mask, grow_mask_by=6):
246
247
        x = (pixels.shape[1] // 8) * 8
        y = (pixels.shape[2] // 8) * 8
248
        mask = torch.nn.functional.interpolate(mask.reshape((-1, 1, mask.shape[-2], mask.shape[-1])), size=(pixels.shape[1], pixels.shape[2]), mode="bilinear")
249

250
        pixels = pixels.clone()
251
        if pixels.shape[1] != x or pixels.shape[2] != y:
252
253
254
255
            x_offset = (pixels.shape[1] % 8) // 2
            y_offset = (pixels.shape[2] % 8) // 2
            pixels = pixels[:,x_offset:x + x_offset, y_offset:y + y_offset,:]
            mask = mask[:,:,x_offset:x + x_offset, y_offset:y + y_offset]
256

257
        #grow mask by a few pixels to keep things seamless in latent space
258
259
260
261
262
263
264
265
        if grow_mask_by == 0:
            mask_erosion = mask
        else:
            kernel_tensor = torch.ones((1, 1, grow_mask_by, grow_mask_by))
            padding = math.ceil((grow_mask_by - 1) / 2)

            mask_erosion = torch.clamp(torch.nn.functional.conv2d(mask.round(), kernel_tensor, padding=padding), 0, 1)

266
        m = (1.0 - mask.round()).squeeze(1)
267
268
        for i in range(3):
            pixels[:,:,:,i] -= 0.5
269
            pixels[:,:,:,i] *= m
270
271
272
            pixels[:,:,:,i] += 0.5
        t = vae.encode(pixels)

273
        return ({"samples":t, "noise_mask": (mask_erosion[:,:,:x,:y].round())}, )
comfyanonymous's avatar
comfyanonymous committed
274

space-nuko's avatar
space-nuko committed
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
class TAESDEncode:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "pixels": ("IMAGE", ), "taesd": ("TAESD", )}}
    RETURN_TYPES = ("LATENT",)
    FUNCTION = "encode"

    CATEGORY = "latent"

    def encode(self, taesd, pixels):
        device = comfy.model_management.get_torch_device()
        # [B, H, W, C] -> [B, C, H, W]
        samples = taesd.encoder(pixels.permute(0, 3, 1, 2).to(device)).to(device)
        return ({"samples": samples}, )

290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
class TAESDPreviewerImpl(LatentPreviewer):
    def __init__(self, taesd):
        self.taesd = taesd

    def decode_latent_to_preview(self, device, x0):
        x_sample = self.taesd.decoder(x0.to(device))[0].detach()
        x_sample = self.taesd.unscale_latents(x_sample)  # returns value in [-2, 2]
        x_sample = x_sample * 0.5
        return x_sample

class TAESDPreviewer:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "taesd": ("TAESD", ), }}
    RETURN_TYPES = ("LATENT_PREVIEWER",)
    FUNCTION = "make_previewer"

    CATEGORY = "latent/previewer"

    def make_previewer(self, taesd):
        return (TAESDPreviewerImpl(taesd), )
Dr.Lt.Data's avatar
Dr.Lt.Data committed
311
312
313

class SaveLatent:
    def __init__(self):
314
        self.output_dir = folder_paths.get_output_directory()
Dr.Lt.Data's avatar
Dr.Lt.Data committed
315
316
317
318

    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "samples": ("LATENT", ),
319
                              "filename_prefix": ("STRING", {"default": "latents/ComfyUI"})},
Dr.Lt.Data's avatar
Dr.Lt.Data committed
320
321
322
323
324
325
326
327
328
329
                "hidden": {"prompt": "PROMPT", "extra_pnginfo": "EXTRA_PNGINFO"},
                }
    RETURN_TYPES = ()
    FUNCTION = "save"

    OUTPUT_NODE = True

    CATEGORY = "_for_testing"

    def save(self, samples, filename_prefix="ComfyUI", prompt=None, extra_pnginfo=None):
330
        full_output_folder, filename, counter, subfolder, filename_prefix = folder_paths.get_save_image_path(filename_prefix, self.output_dir)
Dr.Lt.Data's avatar
Dr.Lt.Data committed
331
332
333
334
335
336

        # support save metadata for latent sharing
        prompt_info = ""
        if prompt is not None:
            prompt_info = json.dumps(prompt)

337
        metadata = {"prompt": prompt_info}
Dr.Lt.Data's avatar
Dr.Lt.Data committed
338
339
340
341
342
343
344
        if extra_pnginfo is not None:
            for x in extra_pnginfo:
                metadata[x] = json.dumps(extra_pnginfo[x])

        file = f"{filename}_{counter:05}_.latent"
        file = os.path.join(full_output_folder, file)

345
346
347
348
        output = {}
        output["latent_tensor"] = samples["samples"]

        safetensors.torch.save_file(output, file, metadata=metadata)
Dr.Lt.Data's avatar
Dr.Lt.Data committed
349
350
351
352
353
354
355

        return {}


class LoadLatent:
    @classmethod
    def INPUT_TYPES(s):
356
357
        input_dir = folder_paths.get_input_directory()
        files = [f for f in os.listdir(input_dir) if os.path.isfile(os.path.join(input_dir, f)) and f.endswith(".latent")]
Dr.Lt.Data's avatar
Dr.Lt.Data committed
358
359
360
361
362
363
364
365
        return {"required": {"latent": [sorted(files), ]}, }

    CATEGORY = "_for_testing"

    RETURN_TYPES = ("LATENT", )
    FUNCTION = "load"

    def load(self, latent):
366
367
        latent_path = folder_paths.get_annotated_filepath(latent)
        latent = safetensors.torch.load_file(latent_path, device="cpu")
368
        samples = {"samples": latent["latent_tensor"].float()}
369
        return (samples, )
Dr.Lt.Data's avatar
Dr.Lt.Data committed
370

371
372
373
374
375
376
377
378
379
380
381
382
383
384
    @classmethod
    def IS_CHANGED(s, latent):
        image_path = folder_paths.get_annotated_filepath(latent)
        m = hashlib.sha256()
        with open(image_path, 'rb') as f:
            m.update(f.read())
        return m.digest().hex()

    @classmethod
    def VALIDATE_INPUTS(s, latent):
        if not folder_paths.exists_annotated_filepath(latent):
            return "Invalid latent file: {}".format(latent)
        return True

Dr.Lt.Data's avatar
Dr.Lt.Data committed
385

comfyanonymous's avatar
comfyanonymous committed
386
387
388
class CheckpointLoader:
    @classmethod
    def INPUT_TYPES(s):
389
390
        return {"required": { "config_name": (folder_paths.get_filename_list("configs"), ),
                              "ckpt_name": (folder_paths.get_filename_list("checkpoints"), )}}
comfyanonymous's avatar
comfyanonymous committed
391
392
393
    RETURN_TYPES = ("MODEL", "CLIP", "VAE")
    FUNCTION = "load_checkpoint"

394
    CATEGORY = "advanced/loaders"
395

comfyanonymous's avatar
comfyanonymous committed
396
    def load_checkpoint(self, config_name, ckpt_name, output_vae=True, output_clip=True):
397
398
        config_path = folder_paths.get_full_path("configs", config_name)
        ckpt_path = folder_paths.get_full_path("checkpoints", ckpt_name)
399
        return comfy.sd.load_checkpoint(config_path, ckpt_path, output_vae=True, output_clip=True, embedding_directory=folder_paths.get_folder_paths("embeddings"))
comfyanonymous's avatar
comfyanonymous committed
400

401
402
403
class CheckpointLoaderSimple:
    @classmethod
    def INPUT_TYPES(s):
404
        return {"required": { "ckpt_name": (folder_paths.get_filename_list("checkpoints"), ),
405
406
407
408
                             }}
    RETURN_TYPES = ("MODEL", "CLIP", "VAE")
    FUNCTION = "load_checkpoint"

409
    CATEGORY = "loaders"
410

411
    def load_checkpoint(self, ckpt_name, output_vae=True, output_clip=True):
412
        ckpt_path = folder_paths.get_full_path("checkpoints", ckpt_name)
413
        out = comfy.sd.load_checkpoint_guess_config(ckpt_path, output_vae=True, output_clip=True, embedding_directory=folder_paths.get_folder_paths("embeddings"))
414
415
        return out

sALTaccount's avatar
sALTaccount committed
416
417
418
class DiffusersLoader:
    @classmethod
    def INPUT_TYPES(cls):
419
        paths = []
sALTaccount's avatar
sALTaccount committed
420
        for search_path in folder_paths.get_folder_paths("diffusers"):
421
            if os.path.exists(search_path):
422
423
424
425
                for root, subdir, files in os.walk(search_path, followlinks=True):
                    if "model_index.json" in files:
                        paths.append(os.path.relpath(root, start=search_path))

426
        return {"required": {"model_path": (paths,), }}
sALTaccount's avatar
sALTaccount committed
427
428
429
    RETURN_TYPES = ("MODEL", "CLIP", "VAE")
    FUNCTION = "load_checkpoint"

430
    CATEGORY = "advanced/loaders"
sALTaccount's avatar
sALTaccount committed
431
432

    def load_checkpoint(self, model_path, output_vae=True, output_clip=True):
sALTaccount's avatar
sALTaccount committed
433
434
        for search_path in folder_paths.get_folder_paths("diffusers"):
            if os.path.exists(search_path):
435
436
437
                path = os.path.join(search_path, model_path)
                if os.path.exists(path):
                    model_path = path
sALTaccount's avatar
sALTaccount committed
438
                    break
439

440
        return comfy.diffusers_load.load_diffusers(model_path, fp16=comfy.model_management.should_use_fp16(), output_vae=output_vae, output_clip=output_clip, embedding_directory=folder_paths.get_folder_paths("embeddings"))
sALTaccount's avatar
sALTaccount committed
441
442


443
444
445
446
447
448
449
450
class unCLIPCheckpointLoader:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "ckpt_name": (folder_paths.get_filename_list("checkpoints"), ),
                             }}
    RETURN_TYPES = ("MODEL", "CLIP", "VAE", "CLIP_VISION")
    FUNCTION = "load_checkpoint"

451
    CATEGORY = "loaders"
452
453
454
455
456
457

    def load_checkpoint(self, ckpt_name, output_vae=True, output_clip=True):
        ckpt_path = folder_paths.get_full_path("checkpoints", ckpt_name)
        out = comfy.sd.load_checkpoint_guess_config(ckpt_path, output_vae=True, output_clip=True, output_clipvision=True, embedding_directory=folder_paths.get_folder_paths("embeddings"))
        return out

comfyanonymous's avatar
comfyanonymous committed
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
class CLIPSetLastLayer:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "clip": ("CLIP", ),
                              "stop_at_clip_layer": ("INT", {"default": -1, "min": -24, "max": -1, "step": 1}),
                              }}
    RETURN_TYPES = ("CLIP",)
    FUNCTION = "set_last_layer"

    CATEGORY = "conditioning"

    def set_last_layer(self, clip, stop_at_clip_layer):
        clip = clip.clone()
        clip.clip_layer(stop_at_clip_layer)
        return (clip,)

474
475
476
477
478
class LoraLoader:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "model": ("MODEL",),
                              "clip": ("CLIP", ),
479
                              "lora_name": (folder_paths.get_filename_list("loras"), ),
480
481
                              "strength_model": ("FLOAT", {"default": 1.0, "min": -10.0, "max": 10.0, "step": 0.01}),
                              "strength_clip": ("FLOAT", {"default": 1.0, "min": -10.0, "max": 10.0, "step": 0.01}),
482
483
484
485
486
487
488
                              }}
    RETURN_TYPES = ("MODEL", "CLIP")
    FUNCTION = "load_lora"

    CATEGORY = "loaders"

    def load_lora(self, model, clip, lora_name, strength_model, strength_clip):
489
490
491
        if strength_model == 0 and strength_clip == 0:
            return (model, clip)

492
        lora_path = folder_paths.get_full_path("loras", lora_name)
493
494
495
        model_lora, clip_lora = comfy.sd.load_lora_for_models(model, clip, lora_path, strength_model, strength_clip)
        return (model_lora, clip_lora)

496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
class TomePatchModel:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "model": ("MODEL",),
                              "ratio": ("FLOAT", {"default": 0.3, "min": 0.0, "max": 1.0, "step": 0.01}),
                              }}
    RETURN_TYPES = ("MODEL",)
    FUNCTION = "patch"

    CATEGORY = "_for_testing"

    def patch(self, model, ratio):
        m = model.clone()
        m.set_model_tomesd(ratio)
        return (m, )

comfyanonymous's avatar
comfyanonymous committed
512
513
514
class VAELoader:
    @classmethod
    def INPUT_TYPES(s):
515
        return {"required": { "vae_name": (folder_paths.get_filename_list("vae"), )}}
comfyanonymous's avatar
comfyanonymous committed
516
517
518
    RETURN_TYPES = ("VAE",)
    FUNCTION = "load_vae"

519
520
    CATEGORY = "loaders"

comfyanonymous's avatar
comfyanonymous committed
521
522
    #TODO: scale factor?
    def load_vae(self, vae_name):
523
        vae_path = folder_paths.get_full_path("vae", vae_name)
comfyanonymous's avatar
comfyanonymous committed
524
525
526
        vae = comfy.sd.VAE(ckpt_path=vae_path)
        return (vae,)

space-nuko's avatar
space-nuko committed
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
class TAESDLoader:
    @classmethod
    def INPUT_TYPES(s):
        model_list = folder_paths.get_filename_list("taesd")
        return {"required": {
            "encoder_name": (model_list, { "default": "taesd_encoder.pth" }),
            "decoder_name": (model_list, { "default": "taesd_decoder.pth" })
        }}
    RETURN_TYPES = ("TAESD",)
    FUNCTION = "load_taesd"

    CATEGORY = "loaders"

    def load_taesd(self, encoder_name, decoder_name):
        device = comfy.model_management.get_torch_device()
        encoder_path = folder_paths.get_full_path("taesd", encoder_name)
        decoder_path = folder_paths.get_full_path("taesd", decoder_name)
        taesd = TAESD(encoder_path, decoder_path).to(device)
        return (taesd,)

comfyanonymous's avatar
comfyanonymous committed
547
548
549
class ControlNetLoader:
    @classmethod
    def INPUT_TYPES(s):
550
        return {"required": { "control_net_name": (folder_paths.get_filename_list("controlnet"), )}}
comfyanonymous's avatar
comfyanonymous committed
551
552
553
554
555
556
557

    RETURN_TYPES = ("CONTROL_NET",)
    FUNCTION = "load_controlnet"

    CATEGORY = "loaders"

    def load_controlnet(self, control_net_name):
558
        controlnet_path = folder_paths.get_full_path("controlnet", control_net_name)
comfyanonymous's avatar
comfyanonymous committed
559
560
561
        controlnet = comfy.sd.load_controlnet(controlnet_path)
        return (controlnet,)

562
563
564
565
class DiffControlNetLoader:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "model": ("MODEL",),
566
                              "control_net_name": (folder_paths.get_filename_list("controlnet"), )}}
567
568
569
570
571
572
573

    RETURN_TYPES = ("CONTROL_NET",)
    FUNCTION = "load_controlnet"

    CATEGORY = "loaders"

    def load_controlnet(self, model, control_net_name):
574
        controlnet_path = folder_paths.get_full_path("controlnet", control_net_name)
575
576
577
        controlnet = comfy.sd.load_controlnet(controlnet_path, model)
        return (controlnet,)

comfyanonymous's avatar
comfyanonymous committed
578
579
580
581

class ControlNetApply:
    @classmethod
    def INPUT_TYPES(s):
582
583
584
585
586
        return {"required": {"conditioning": ("CONDITIONING", ),
                             "control_net": ("CONTROL_NET", ),
                             "image": ("IMAGE", ),
                             "strength": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 10.0, "step": 0.01})
                             }}
comfyanonymous's avatar
comfyanonymous committed
587
588
589
590
591
    RETURN_TYPES = ("CONDITIONING",)
    FUNCTION = "apply_controlnet"

    CATEGORY = "conditioning"

592
    def apply_controlnet(self, conditioning, control_net, image, strength):
593
594
595
        if strength == 0:
            return (conditioning, )

comfyanonymous's avatar
comfyanonymous committed
596
597
598
599
        c = []
        control_hint = image.movedim(-1,1)
        for t in conditioning:
            n = [t[0], t[1].copy()]
comfyanonymous's avatar
comfyanonymous committed
600
601
602
603
            c_net = control_net.copy().set_cond_hint(control_hint, strength)
            if 'control' in t[1]:
                c_net.set_previous_controlnet(t[1]['control'])
            n[1]['control'] = c_net
comfyanonymous's avatar
comfyanonymous committed
604
605
606
            c.append(n)
        return (c, )

607
608
609
class CLIPLoader:
    @classmethod
    def INPUT_TYPES(s):
610
        return {"required": { "clip_name": (folder_paths.get_filename_list("clip"), ),
611
612
613
614
615
616
                             }}
    RETURN_TYPES = ("CLIP",)
    FUNCTION = "load_clip"

    CATEGORY = "loaders"

617
    def load_clip(self, clip_name):
618
        clip_path = folder_paths.get_full_path("clip", clip_name)
comfyanonymous's avatar
comfyanonymous committed
619
        clip = comfy.sd.load_clip(ckpt_path=clip_path, embedding_directory=folder_paths.get_folder_paths("embeddings"))
620
621
        return (clip,)

622
623
624
class CLIPVisionLoader:
    @classmethod
    def INPUT_TYPES(s):
625
        return {"required": { "clip_name": (folder_paths.get_filename_list("clip_vision"), ),
626
627
628
629
630
631
632
                             }}
    RETURN_TYPES = ("CLIP_VISION",)
    FUNCTION = "load_clip"

    CATEGORY = "loaders"

    def load_clip(self, clip_name):
633
        clip_path = folder_paths.get_full_path("clip_vision", clip_name)
634
        clip_vision = comfy.clip_vision.load(clip_path)
635
636
637
638
639
640
641
642
        return (clip_vision,)

class CLIPVisionEncode:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "clip_vision": ("CLIP_VISION",),
                              "image": ("IMAGE",)
                             }}
comfyanonymous's avatar
comfyanonymous committed
643
    RETURN_TYPES = ("CLIP_VISION_OUTPUT",)
644
645
    FUNCTION = "encode"

646
    CATEGORY = "conditioning"
647
648
649
650
651
652
653
654

    def encode(self, clip_vision, image):
        output = clip_vision.encode_image(image)
        return (output,)

class StyleModelLoader:
    @classmethod
    def INPUT_TYPES(s):
655
        return {"required": { "style_model_name": (folder_paths.get_filename_list("style_models"), )}}
656
657
658
659
660
661
662

    RETURN_TYPES = ("STYLE_MODEL",)
    FUNCTION = "load_style_model"

    CATEGORY = "loaders"

    def load_style_model(self, style_model_name):
663
        style_model_path = folder_paths.get_full_path("style_models", style_model_name)
664
665
666
667
668
669
670
        style_model = comfy.sd.load_style_model(style_model_path)
        return (style_model,)


class StyleModelApply:
    @classmethod
    def INPUT_TYPES(s):
671
672
673
        return {"required": {"conditioning": ("CONDITIONING", ),
                             "style_model": ("STYLE_MODEL", ),
                             "clip_vision_output": ("CLIP_VISION_OUTPUT", ),
674
675
676
677
                             }}
    RETURN_TYPES = ("CONDITIONING",)
    FUNCTION = "apply_stylemodel"

comfyanonymous's avatar
comfyanonymous committed
678
    CATEGORY = "conditioning/style_model"
679

680
681
    def apply_stylemodel(self, clip_vision_output, style_model, conditioning):
        cond = style_model.get_cond(clip_vision_output)
682
        c = []
683
684
        for t in conditioning:
            n = [torch.cat((t[0], cond), dim=1), t[1].copy()]
685
686
687
            c.append(n)
        return (c, )

688
689
690
691
692
693
class unCLIPConditioning:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": {"conditioning": ("CONDITIONING", ),
                             "clip_vision_output": ("CLIP_VISION_OUTPUT", ),
                             "strength": ("FLOAT", {"default": 1.0, "min": -10.0, "max": 10.0, "step": 0.01}),
694
                             "noise_augmentation": ("FLOAT", {"default": 0.0, "min": 0.0, "max": 1.0, "step": 0.01}),
695
696
697
698
                             }}
    RETURN_TYPES = ("CONDITIONING",)
    FUNCTION = "apply_adm"

699
    CATEGORY = "conditioning"
700

701
    def apply_adm(self, conditioning, clip_vision_output, strength, noise_augmentation):
702
703
704
        if strength == 0:
            return (conditioning, )

705
706
707
        c = []
        for t in conditioning:
            o = t[1].copy()
708
            x = (clip_vision_output, strength, noise_augmentation)
709
710
711
712
713
714
715
716
            if "adm" in o:
                o["adm"] = o["adm"][:] + [x]
            else:
                o["adm"] = [x]
            n = [t[0], o]
            c.append(n)
        return (c, )

717
718
719
720
721
722
723
724
class GLIGENLoader:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "gligen_name": (folder_paths.get_filename_list("gligen"), )}}

    RETURN_TYPES = ("GLIGEN",)
    FUNCTION = "load_gligen"

comfyanonymous's avatar
comfyanonymous committed
725
    CATEGORY = "loaders"
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746

    def load_gligen(self, gligen_name):
        gligen_path = folder_paths.get_full_path("gligen", gligen_name)
        gligen = comfy.sd.load_gligen(gligen_path)
        return (gligen,)

class GLIGENTextBoxApply:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": {"conditioning_to": ("CONDITIONING", ),
                              "clip": ("CLIP", ),
                              "gligen_textbox_model": ("GLIGEN", ),
                              "text": ("STRING", {"multiline": True}),
                              "width": ("INT", {"default": 64, "min": 8, "max": MAX_RESOLUTION, "step": 8}),
                              "height": ("INT", {"default": 64, "min": 8, "max": MAX_RESOLUTION, "step": 8}),
                              "x": ("INT", {"default": 0, "min": 0, "max": MAX_RESOLUTION, "step": 8}),
                              "y": ("INT", {"default": 0, "min": 0, "max": MAX_RESOLUTION, "step": 8}),
                             }}
    RETURN_TYPES = ("CONDITIONING",)
    FUNCTION = "append"

comfyanonymous's avatar
comfyanonymous committed
747
    CATEGORY = "conditioning/gligen"
748
749
750
751
752
753
754
755
756
757
758
759
760
761

    def append(self, conditioning_to, clip, gligen_textbox_model, text, width, height, x, y):
        c = []
        cond, cond_pooled = clip.encode_from_tokens(clip.tokenize(text), return_pooled=True)
        for t in conditioning_to:
            n = [t[0], t[1].copy()]
            position_params = [(cond_pooled, height // 8, width // 8, y // 8, x // 8)]
            prev = []
            if "gligen" in n[1]:
                prev = n[1]['gligen'][2]

            n[1]['gligen'] = ("position", gligen_textbox_model, prev + position_params)
            c.append(n)
        return (c, )
762

comfyanonymous's avatar
comfyanonymous committed
763
764
765
766
767
768
class EmptyLatentImage:
    def __init__(self, device="cpu"):
        self.device = device

    @classmethod
    def INPUT_TYPES(s):
769
770
        return {"required": { "width": ("INT", {"default": 512, "min": 64, "max": MAX_RESOLUTION, "step": 8}),
                              "height": ("INT", {"default": 512, "min": 64, "max": MAX_RESOLUTION, "step": 8}),
comfyanonymous's avatar
comfyanonymous committed
771
772
773
774
                              "batch_size": ("INT", {"default": 1, "min": 1, "max": 64})}}
    RETURN_TYPES = ("LATENT",)
    FUNCTION = "generate"

775
776
    CATEGORY = "latent"

comfyanonymous's avatar
comfyanonymous committed
777
778
    def generate(self, width, height, batch_size=1):
        latent = torch.zeros([batch_size, 4, height // 8, width // 8])
779
        return ({"samples":latent}, )
comfyanonymous's avatar
comfyanonymous committed
780

comfyanonymous's avatar
comfyanonymous committed
781

782
783
784
785
786
class LatentFromBatch:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "samples": ("LATENT",),
                              "batch_index": ("INT", {"default": 0, "min": 0, "max": 63}),
787
                              "length": ("INT", {"default": 1, "min": 1, "max": 64}),
788
789
                              }}
    RETURN_TYPES = ("LATENT",)
790
    FUNCTION = "frombatch"
791

792
    CATEGORY = "latent/batch"
793

794
    def frombatch(self, samples, batch_index, length):
795
796
797
        s = samples.copy()
        s_in = samples["samples"]
        batch_index = min(s_in.shape[0] - 1, batch_index)
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
        length = min(s_in.shape[0] - batch_index, length)
        s["samples"] = s_in[batch_index:batch_index + length].clone()
        if "noise_mask" in samples:
            masks = samples["noise_mask"]
            if masks.shape[0] == 1:
                s["noise_mask"] = masks.clone()
            else:
                if masks.shape[0] < s_in.shape[0]:
                    masks = masks.repeat(math.ceil(s_in.shape[0] / masks.shape[0]), 1, 1, 1)[:s_in.shape[0]]
                s["noise_mask"] = masks[batch_index:batch_index + length].clone()
        if "batch_index" not in s:
            s["batch_index"] = [x for x in range(batch_index, batch_index+length)]
        else:
            s["batch_index"] = samples["batch_index"][batch_index:batch_index + length]
        return (s,)
    
class RepeatLatentBatch:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "samples": ("LATENT",),
                              "amount": ("INT", {"default": 1, "min": 1, "max": 64}),
                              }}
    RETURN_TYPES = ("LATENT",)
    FUNCTION = "repeat"

    CATEGORY = "latent/batch"

    def repeat(self, samples, amount):
        s = samples.copy()
        s_in = samples["samples"]
        
        s["samples"] = s_in.repeat((amount, 1,1,1))
        if "noise_mask" in samples and samples["noise_mask"].shape[0] > 1:
            masks = samples["noise_mask"]
            if masks.shape[0] < s_in.shape[0]:
                masks = masks.repeat(math.ceil(s_in.shape[0] / masks.shape[0]), 1, 1, 1)[:s_in.shape[0]]
            s["noise_mask"] = samples["noise_mask"].repeat((amount, 1,1,1))
        if "batch_index" in s:
            offset = max(s["batch_index"]) - min(s["batch_index"]) + 1
            s["batch_index"] = s["batch_index"] + [x + (i * offset) for i in range(1, amount) for x in s["batch_index"]]
838
        return (s,)
comfyanonymous's avatar
comfyanonymous committed
839

comfyanonymous's avatar
comfyanonymous committed
840
class LatentUpscale:
841
    upscale_methods = ["nearest-exact", "bilinear", "area", "bislerp"]
842
    crop_methods = ["disabled", "center"]
comfyanonymous's avatar
comfyanonymous committed
843
844
845
846

    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "samples": ("LATENT",), "upscale_method": (s.upscale_methods,),
847
848
                              "width": ("INT", {"default": 512, "min": 64, "max": MAX_RESOLUTION, "step": 8}),
                              "height": ("INT", {"default": 512, "min": 64, "max": MAX_RESOLUTION, "step": 8}),
849
                              "crop": (s.crop_methods,)}}
comfyanonymous's avatar
comfyanonymous committed
850
851
852
    RETURN_TYPES = ("LATENT",)
    FUNCTION = "upscale"

853
854
    CATEGORY = "latent"

855
    def upscale(self, samples, upscale_method, width, height, crop):
856
        s = samples.copy()
comfyanonymous's avatar
comfyanonymous committed
857
        s["samples"] = comfy.utils.common_upscale(samples["samples"], width // 8, height // 8, upscale_method, crop)
comfyanonymous's avatar
comfyanonymous committed
858
859
        return (s,)

comfyanonymous's avatar
comfyanonymous committed
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
class LatentUpscaleBy:
    upscale_methods = ["nearest-exact", "bilinear", "area", "bislerp"]

    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "samples": ("LATENT",), "upscale_method": (s.upscale_methods,),
                              "scale_by": ("FLOAT", {"default": 1.5, "min": 0.01, "max": 8.0, "step": 0.01}),}}
    RETURN_TYPES = ("LATENT",)
    FUNCTION = "upscale"

    CATEGORY = "latent"

    def upscale(self, samples, upscale_method, scale_by):
        s = samples.copy()
        width = round(samples["samples"].shape[3] * scale_by)
        height = round(samples["samples"].shape[2] * scale_by)
        s["samples"] = comfy.utils.common_upscale(samples["samples"], width, height, upscale_method, "disabled")
        return (s,)

comfyanonymous's avatar
comfyanonymous committed
879
880
881
882
883
884
885
886
887
class LatentRotate:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "samples": ("LATENT",),
                              "rotation": (["none", "90 degrees", "180 degrees", "270 degrees"],),
                              }}
    RETURN_TYPES = ("LATENT",)
    FUNCTION = "rotate"

comfyanonymous's avatar
comfyanonymous committed
888
    CATEGORY = "latent/transform"
comfyanonymous's avatar
comfyanonymous committed
889
890

    def rotate(self, samples, rotation):
891
        s = samples.copy()
comfyanonymous's avatar
comfyanonymous committed
892
893
894
895
896
897
898
899
        rotate_by = 0
        if rotation.startswith("90"):
            rotate_by = 1
        elif rotation.startswith("180"):
            rotate_by = 2
        elif rotation.startswith("270"):
            rotate_by = 3

900
        s["samples"] = torch.rot90(samples["samples"], k=rotate_by, dims=[3, 2])
comfyanonymous's avatar
comfyanonymous committed
901
        return (s,)
comfyanonymous's avatar
comfyanonymous committed
902
903
904
905
906
907
908
909
910
911

class LatentFlip:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "samples": ("LATENT",),
                              "flip_method": (["x-axis: vertically", "y-axis: horizontally"],),
                              }}
    RETURN_TYPES = ("LATENT",)
    FUNCTION = "flip"

comfyanonymous's avatar
comfyanonymous committed
912
    CATEGORY = "latent/transform"
comfyanonymous's avatar
comfyanonymous committed
913
914

    def flip(self, samples, flip_method):
915
        s = samples.copy()
comfyanonymous's avatar
comfyanonymous committed
916
        if flip_method.startswith("x"):
917
            s["samples"] = torch.flip(samples["samples"], dims=[2])
comfyanonymous's avatar
comfyanonymous committed
918
        elif flip_method.startswith("y"):
919
            s["samples"] = torch.flip(samples["samples"], dims=[3])
comfyanonymous's avatar
comfyanonymous committed
920
921

        return (s,)
comfyanonymous's avatar
comfyanonymous committed
922
923
924
925

class LatentComposite:
    @classmethod
    def INPUT_TYPES(s):
comfyanonymous's avatar
comfyanonymous committed
926
927
928
929
930
931
        return {"required": { "samples_to": ("LATENT",),
                              "samples_from": ("LATENT",),
                              "x": ("INT", {"default": 0, "min": 0, "max": MAX_RESOLUTION, "step": 8}),
                              "y": ("INT", {"default": 0, "min": 0, "max": MAX_RESOLUTION, "step": 8}),
                              "feather": ("INT", {"default": 0, "min": 0, "max": MAX_RESOLUTION, "step": 8}),
                              }}
comfyanonymous's avatar
comfyanonymous committed
932
933
934
935
936
    RETURN_TYPES = ("LATENT",)
    FUNCTION = "composite"

    CATEGORY = "latent"

comfyanonymous's avatar
comfyanonymous committed
937
938
939
    def composite(self, samples_to, samples_from, x, y, composite_method="normal", feather=0):
        x =  x // 8
        y = y // 8
940
        feather = feather // 8
comfyanonymous's avatar
comfyanonymous committed
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
        samples_out = samples_to.copy()
        s = samples_to["samples"].clone()
        samples_to = samples_to["samples"]
        samples_from = samples_from["samples"]
        if feather == 0:
            s[:,:,y:y+samples_from.shape[2],x:x+samples_from.shape[3]] = samples_from[:,:,:samples_to.shape[2] - y, :samples_to.shape[3] - x]
        else:
            samples_from = samples_from[:,:,:samples_to.shape[2] - y, :samples_to.shape[3] - x]
            mask = torch.ones_like(samples_from)
            for t in range(feather):
                if y != 0:
                    mask[:,:,t:1+t,:] *= ((1.0/feather) * (t + 1))

                if y + samples_from.shape[2] < samples_to.shape[2]:
                    mask[:,:,mask.shape[2] -1 -t: mask.shape[2]-t,:] *= ((1.0/feather) * (t + 1))
                if x != 0:
                    mask[:,:,:,t:1+t] *= ((1.0/feather) * (t + 1))
                if x + samples_from.shape[3] < samples_to.shape[3]:
                    mask[:,:,:,mask.shape[3]- 1 - t: mask.shape[3]- t] *= ((1.0/feather) * (t + 1))
            rev_mask = torch.ones_like(mask) - mask
            s[:,:,y:y+samples_from.shape[2],x:x+samples_from.shape[3]] = samples_from[:,:,:samples_to.shape[2] - y, :samples_to.shape[3] - x] * mask + s[:,:,y:y+samples_from.shape[2],x:x+samples_from.shape[3]] * rev_mask
        samples_out["samples"] = s
        return (samples_out,)
comfyanonymous's avatar
comfyanonymous committed
964

comfyanonymous's avatar
comfyanonymous committed
965
966
967
968
class LatentCrop:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "samples": ("LATENT",),
969
970
                              "width": ("INT", {"default": 512, "min": 64, "max": MAX_RESOLUTION, "step": 8}),
                              "height": ("INT", {"default": 512, "min": 64, "max": MAX_RESOLUTION, "step": 8}),
971
972
                              "x": ("INT", {"default": 0, "min": 0, "max": MAX_RESOLUTION, "step": 8}),
                              "y": ("INT", {"default": 0, "min": 0, "max": MAX_RESOLUTION, "step": 8}),
comfyanonymous's avatar
comfyanonymous committed
973
974
975
976
                              }}
    RETURN_TYPES = ("LATENT",)
    FUNCTION = "crop"

comfyanonymous's avatar
comfyanonymous committed
977
    CATEGORY = "latent/transform"
comfyanonymous's avatar
comfyanonymous committed
978
979

    def crop(self, samples, width, height, x, y):
980
981
        s = samples.copy()
        samples = samples['samples']
comfyanonymous's avatar
comfyanonymous committed
982
983
984
985
986
987
988
989
990
991
992
993
994
        x =  x // 8
        y = y // 8

        #enfonce minimum size of 64
        if x > (samples.shape[3] - 8):
            x = samples.shape[3] - 8
        if y > (samples.shape[2] - 8):
            y = samples.shape[2] - 8

        new_height = height // 8
        new_width = width // 8
        to_x = new_width + x
        to_y = new_height + y
995
        s['samples'] = samples[:,:,y:to_y, x:to_x]
comfyanonymous's avatar
comfyanonymous committed
996
997
        return (s,)

998
999
1000
1001
1002
1003
1004
1005
1006
class SetLatentNoiseMask:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "samples": ("LATENT",),
                              "mask": ("MASK",),
                              }}
    RETURN_TYPES = ("LATENT",)
    FUNCTION = "set_mask"

1007
    CATEGORY = "latent/inpaint"
1008
1009
1010

    def set_mask(self, samples, mask):
        s = samples.copy()
1011
        s["noise_mask"] = mask.reshape((-1, 1, mask.shape[-2], mask.shape[-1]))
1012
1013
        return (s,)

space-nuko's avatar
space-nuko committed
1014

1015
1016
def decode_latent_to_preview_image(previewer, device, preview_format, x0):
    x_sample = previewer.decode_latent_to_preview(device, x0)
space-nuko's avatar
space-nuko committed
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041

    x_sample = torch.clamp((x_sample + 1.0) / 2.0, min=0.0, max=1.0)
    x_sample = 255. * np.moveaxis(x_sample.cpu().numpy(), 0, 2)
    x_sample = x_sample.astype(np.uint8)

    preview_image = Image.fromarray(x_sample)

    if preview_image.size[0] > MAX_PREVIEW_RESOLUTION or preview_image.size[1] > MAX_PREVIEW_RESOLUTION:
        preview_image.thumbnail((MAX_PREVIEW_RESOLUTION, MAX_PREVIEW_RESOLUTION), Image.ANTIALIAS)

    preview_type = 1
    if preview_format == "JPEG":
        preview_type = 1
    elif preview_format == "PNG":
        preview_type = 2

    bytesIO = BytesIO()
    header = struct.pack(">I", preview_type)
    bytesIO.write(header)
    preview_image.save(bytesIO, format=preview_format)
    preview_bytes = bytesIO.getvalue()

    return preview_bytes


1042
def common_ksampler(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent, denoise=1.0, disable_noise=False, start_step=None, last_step=None, force_full_denoise=False, previewer=None):
1043
    device = comfy.model_management.get_torch_device()
1044
    latent_image = latent["samples"]
1045

comfyanonymous's avatar
comfyanonymous committed
1046
1047
1048
    if disable_noise:
        noise = torch.zeros(latent_image.size(), dtype=latent_image.dtype, layout=latent_image.layout, device="cpu")
    else:
1049
1050
        batch_inds = latent["batch_index"] if "batch_index" in latent else None
        noise = comfy.sample.prepare_noise(latent_image, seed, batch_inds)
comfyanonymous's avatar
comfyanonymous committed
1051

1052
    noise_mask = None
1053
    if "noise_mask" in latent:
1054
        noise_mask = latent["noise_mask"]
comfyanonymous's avatar
comfyanonymous committed
1055

space-nuko's avatar
space-nuko committed
1056
1057
1058
1059
    preview_format = "JPEG"
    if preview_format not in ["JPEG", "PNG"]:
        preview_format = "JPEG"

1060
    pbar = comfy.utils.ProgressBar(steps)
1061
    def callback(step, x0, x, total_steps):
space-nuko's avatar
space-nuko committed
1062
        preview_bytes = None
1063
1064
        if previewer:
            preview_bytes = decode_latent_to_preview_image(previewer, device, preview_format, x0)
space-nuko's avatar
space-nuko committed
1065
        pbar.update_absolute(step + 1, total_steps, preview_bytes)
1066

1067
1068
    samples = comfy.sample.sample(model, noise, steps, cfg, sampler_name, scheduler, positive, negative, latent_image,
                                  denoise=denoise, disable_noise=disable_noise, start_step=start_step, last_step=last_step,
1069
                                  force_full_denoise=force_full_denoise, noise_mask=noise_mask, callback=callback)
1070
1071
1072
    out = latent.copy()
    out["samples"] = samples
    return (out, )
comfyanonymous's avatar
comfyanonymous committed
1073

comfyanonymous's avatar
comfyanonymous committed
1074
1075
1076
class KSampler:
    @classmethod
    def INPUT_TYPES(s):
comfyanonymous's avatar
comfyanonymous committed
1077
        return {"required":
comfyanonymous's avatar
comfyanonymous committed
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
                    {"model": ("MODEL",),
                    "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}),
                    "steps": ("INT", {"default": 20, "min": 1, "max": 10000}),
                    "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0}),
                    "sampler_name": (comfy.samplers.KSampler.SAMPLERS, ),
                    "scheduler": (comfy.samplers.KSampler.SCHEDULERS, ),
                    "positive": ("CONDITIONING", ),
                    "negative": ("CONDITIONING", ),
                    "latent_image": ("LATENT", ),
                    "denoise": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01}),
space-nuko's avatar
space-nuko committed
1088
1089
                     },
                "optional": {
1090
                    "previewer": ("LATENT_PREVIEWER",)
space-nuko's avatar
space-nuko committed
1091
                }}
comfyanonymous's avatar
comfyanonymous committed
1092
1093
1094
1095

    RETURN_TYPES = ("LATENT",)
    FUNCTION = "sample"

1096
1097
    CATEGORY = "sampling"

1098
1099
    def sample(self, model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise=1.0, previewer=None):
        return common_ksampler(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise=denoise, previewer=previewer)
comfyanonymous's avatar
comfyanonymous committed
1100

comfyanonymous's avatar
comfyanonymous committed
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
class KSamplerAdvanced:
    @classmethod
    def INPUT_TYPES(s):
        return {"required":
                    {"model": ("MODEL",),
                    "add_noise": (["enable", "disable"], ),
                    "noise_seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}),
                    "steps": ("INT", {"default": 20, "min": 1, "max": 10000}),
                    "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0}),
                    "sampler_name": (comfy.samplers.KSampler.SAMPLERS, ),
                    "scheduler": (comfy.samplers.KSampler.SCHEDULERS, ),
                    "positive": ("CONDITIONING", ),
                    "negative": ("CONDITIONING", ),
                    "latent_image": ("LATENT", ),
                    "start_at_step": ("INT", {"default": 0, "min": 0, "max": 10000}),
                    "end_at_step": ("INT", {"default": 10000, "min": 0, "max": 10000}),
                    "return_with_leftover_noise": (["disable", "enable"], ),
space-nuko's avatar
space-nuko committed
1118
1119
                     },
                "optional": {
1120
                    "previewer": ("LATENT_PREVIEWER",)
space-nuko's avatar
space-nuko committed
1121
                }}
comfyanonymous's avatar
comfyanonymous committed
1122
1123
1124
1125
1126

    RETURN_TYPES = ("LATENT",)
    FUNCTION = "sample"

    CATEGORY = "sampling"
comfyanonymous's avatar
comfyanonymous committed
1127

1128
    def sample(self, model, add_noise, noise_seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, start_at_step, end_at_step, return_with_leftover_noise, denoise=1.0, previewer=None):
comfyanonymous's avatar
comfyanonymous committed
1129
1130
1131
1132
1133
1134
        force_full_denoise = True
        if return_with_leftover_noise == "enable":
            force_full_denoise = False
        disable_noise = False
        if add_noise == "disable":
            disable_noise = True
1135
        return common_ksampler(model, noise_seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise=denoise, disable_noise=disable_noise, start_step=start_at_step, last_step=end_at_step, force_full_denoise=force_full_denoise, previewer=previewer)
comfyanonymous's avatar
comfyanonymous committed
1136
1137
1138

class SaveImage:
    def __init__(self):
1139
        self.output_dir = folder_paths.get_output_directory()
m957ymj75urz's avatar
m957ymj75urz committed
1140
        self.type = "output"
comfyanonymous's avatar
comfyanonymous committed
1141
1142
1143
1144

    @classmethod
    def INPUT_TYPES(s):
        return {"required": 
1145
                    {"images": ("IMAGE", ),
pythongosssss's avatar
tidy  
pythongosssss committed
1146
                     "filename_prefix": ("STRING", {"default": "ComfyUI"})},
pythongosssss's avatar
pythongosssss committed
1147
                "hidden": {"prompt": "PROMPT", "extra_pnginfo": "EXTRA_PNGINFO"},
comfyanonymous's avatar
comfyanonymous committed
1148
1149
1150
1151
1152
1153
1154
                }

    RETURN_TYPES = ()
    FUNCTION = "save_images"

    OUTPUT_NODE = True

1155
1156
    CATEGORY = "image"

pythongosssss's avatar
tidy  
pythongosssss committed
1157
    def save_images(self, images, filename_prefix="ComfyUI", prompt=None, extra_pnginfo=None):
1158
        full_output_folder, filename, counter, subfolder, filename_prefix = folder_paths.get_save_image_path(filename_prefix, self.output_dir, images[0].shape[1], images[0].shape[0])
m957ymj75urz's avatar
m957ymj75urz committed
1159
        results = list()
comfyanonymous's avatar
comfyanonymous committed
1160
1161
        for image in images:
            i = 255. * image.cpu().numpy()
1162
            img = Image.fromarray(np.clip(i, 0, 255).astype(np.uint8))
comfyanonymous's avatar
comfyanonymous committed
1163
1164
1165
1166
1167
1168
            metadata = PngInfo()
            if prompt is not None:
                metadata.add_text("prompt", json.dumps(prompt))
            if extra_pnginfo is not None:
                for x in extra_pnginfo:
                    metadata.add_text(x, json.dumps(extra_pnginfo[x]))
1169

1170
            file = f"{filename}_{counter:05}_.png"
1171
            img.save(os.path.join(full_output_folder, file), pnginfo=metadata, compress_level=4)
m957ymj75urz's avatar
m957ymj75urz committed
1172
1173
1174
1175
            results.append({
                "filename": file,
                "subfolder": subfolder,
                "type": self.type
Gavroche CryptoRUSH's avatar
Gavroche CryptoRUSH committed
1176
            })
1177
            counter += 1
comfyanonymous's avatar
comfyanonymous committed
1178

m957ymj75urz's avatar
m957ymj75urz committed
1179
        return { "ui": { "images": results } }
comfyanonymous's avatar
comfyanonymous committed
1180

pythongosssss's avatar
pythongosssss committed
1181
1182
class PreviewImage(SaveImage):
    def __init__(self):
1183
        self.output_dir = folder_paths.get_temp_directory()
m957ymj75urz's avatar
m957ymj75urz committed
1184
        self.type = "temp"
pythongosssss's avatar
pythongosssss committed
1185
1186
1187

    @classmethod
    def INPUT_TYPES(s):
1188
        return {"required":
pythongosssss's avatar
pythongosssss committed
1189
1190
1191
                    {"images": ("IMAGE", ), },
                "hidden": {"prompt": "PROMPT", "extra_pnginfo": "EXTRA_PNGINFO"},
                }
1192

1193
1194
1195
class LoadImage:
    @classmethod
    def INPUT_TYPES(s):
1196
        input_dir = folder_paths.get_input_directory()
comfyanonymous's avatar
comfyanonymous committed
1197
        files = [f for f in os.listdir(input_dir) if os.path.isfile(os.path.join(input_dir, f))]
1198
        return {"required":
comfyanonymous's avatar
comfyanonymous committed
1199
                    {"image": (sorted(files), )},
1200
                }
1201
1202

    CATEGORY = "image"
1203

1204
    RETURN_TYPES = ("IMAGE", "MASK")
1205
1206
    FUNCTION = "load_image"
    def load_image(self, image):
1207
        image_path = folder_paths.get_annotated_filepath(image)
1208
        i = Image.open(image_path)
1209
        i = ImageOps.exif_transpose(i)
1210
        image = i.convert("RGB")
1211
        image = np.array(image).astype(np.float32) / 255.0
1212
        image = torch.from_numpy(image)[None,]
1213
1214
1215
1216
1217
1218
        if 'A' in i.getbands():
            mask = np.array(i.getchannel('A')).astype(np.float32) / 255.0
            mask = 1. - torch.from_numpy(mask)
        else:
            mask = torch.zeros((64,64), dtype=torch.float32, device="cpu")
        return (image, mask)
1219

1220
1221
    @classmethod
    def IS_CHANGED(s, image):
1222
        image_path = folder_paths.get_annotated_filepath(image)
1223
1224
1225
1226
        m = hashlib.sha256()
        with open(image_path, 'rb') as f:
            m.update(f.read())
        return m.digest().hex()
pythongosssss's avatar
pythongosssss committed
1227

1228
1229
1230
1231
1232
1233
1234
    @classmethod
    def VALIDATE_INPUTS(s, image):
        if not folder_paths.exists_annotated_filepath(image):
            return "Invalid image file: {}".format(image)

        return True

1235
class LoadImageMask:
1236
    _color_channels = ["alpha", "red", "green", "blue"]
1237
1238
    @classmethod
    def INPUT_TYPES(s):
1239
        input_dir = folder_paths.get_input_directory()
comfyanonymous's avatar
comfyanonymous committed
1240
        files = [f for f in os.listdir(input_dir) if os.path.isfile(os.path.join(input_dir, f))]
1241
        return {"required":
comfyanonymous's avatar
comfyanonymous committed
1242
                    {"image": (sorted(files), ),
1243
                     "channel": (s._color_channels, ), }
1244
1245
                }

1246
    CATEGORY = "mask"
1247
1248
1249
1250

    RETURN_TYPES = ("MASK",)
    FUNCTION = "load_image"
    def load_image(self, image, channel):
1251
        image_path = folder_paths.get_annotated_filepath(image)
1252
        i = Image.open(image_path)
1253
        i = ImageOps.exif_transpose(i)
1254
1255
        if i.getbands() != ("R", "G", "B", "A"):
            i = i.convert("RGBA")
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
        mask = None
        c = channel[0].upper()
        if c in i.getbands():
            mask = np.array(i.getchannel(c)).astype(np.float32) / 255.0
            mask = torch.from_numpy(mask)
            if c == 'A':
                mask = 1. - mask
        else:
            mask = torch.zeros((64,64), dtype=torch.float32, device="cpu")
        return (mask,)

    @classmethod
    def IS_CHANGED(s, image, channel):
1269
        image_path = folder_paths.get_annotated_filepath(image)
1270
1271
1272
1273
        m = hashlib.sha256()
        with open(image_path, 'rb') as f:
            m.update(f.read())
        return m.digest().hex()
pythongosssss's avatar
pythongosssss committed
1274

1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
    @classmethod
    def VALIDATE_INPUTS(s, image, channel):
        if not folder_paths.exists_annotated_filepath(image):
            return "Invalid image file: {}".format(image)

        if channel not in s._color_channels:
            return "Invalid color channel: {}".format(channel)

        return True

comfyanonymous's avatar
comfyanonymous committed
1285
1286
1287
1288
1289
1290
1291
class ImageScale:
    upscale_methods = ["nearest-exact", "bilinear", "area"]
    crop_methods = ["disabled", "center"]

    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "image": ("IMAGE",), "upscale_method": (s.upscale_methods,),
1292
1293
                              "width": ("INT", {"default": 512, "min": 1, "max": MAX_RESOLUTION, "step": 1}),
                              "height": ("INT", {"default": 512, "min": 1, "max": MAX_RESOLUTION, "step": 1}),
comfyanonymous's avatar
comfyanonymous committed
1294
1295
1296
1297
                              "crop": (s.crop_methods,)}}
    RETURN_TYPES = ("IMAGE",)
    FUNCTION = "upscale"

1298
    CATEGORY = "image/upscaling"
1299

comfyanonymous's avatar
comfyanonymous committed
1300
1301
    def upscale(self, image, upscale_method, width, height, crop):
        samples = image.movedim(-1,1)
comfyanonymous's avatar
comfyanonymous committed
1302
        s = comfy.utils.common_upscale(samples, width, height, upscale_method, crop)
comfyanonymous's avatar
comfyanonymous committed
1303
1304
        s = s.movedim(1,-1)
        return (s,)
comfyanonymous's avatar
comfyanonymous committed
1305

1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
class ImageInvert:

    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "image": ("IMAGE",)}}

    RETURN_TYPES = ("IMAGE",)
    FUNCTION = "invert"

    CATEGORY = "image"

    def invert(self, image):
        s = 1.0 - image
        return (s,)


Guo Y.K's avatar
Guo Y.K committed
1322
1323
1324
1325
1326
1327
1328
class ImagePadForOutpaint:

    @classmethod
    def INPUT_TYPES(s):
        return {
            "required": {
                "image": ("IMAGE",),
1329
1330
1331
1332
                "left": ("INT", {"default": 0, "min": 0, "max": MAX_RESOLUTION, "step": 8}),
                "top": ("INT", {"default": 0, "min": 0, "max": MAX_RESOLUTION, "step": 8}),
                "right": ("INT", {"default": 0, "min": 0, "max": MAX_RESOLUTION, "step": 8}),
                "bottom": ("INT", {"default": 0, "min": 0, "max": MAX_RESOLUTION, "step": 8}),
1333
                "feathering": ("INT", {"default": 40, "min": 0, "max": MAX_RESOLUTION, "step": 1}),
Guo Y.K's avatar
Guo Y.K committed
1334
1335
1336
1337
1338
1339
1340
1341
            }
        }

    RETURN_TYPES = ("IMAGE", "MASK")
    FUNCTION = "expand_image"

    CATEGORY = "image"

1342
    def expand_image(self, image, left, top, right, bottom, feathering):
Guo Y.K's avatar
Guo Y.K committed
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
        d1, d2, d3, d4 = image.size()

        new_image = torch.zeros(
            (d1, d2 + top + bottom, d3 + left + right, d4),
            dtype=torch.float32,
        )
        new_image[:, top:top + d2, left:left + d3, :] = image

        mask = torch.ones(
            (d2 + top + bottom, d3 + left + right),
            dtype=torch.float32,
        )
1355

1356
1357
1358
1359
1360
        t = torch.zeros(
            (d2, d3),
            dtype=torch.float32
        )

1361
        if feathering > 0 and feathering * 2 < d2 and feathering * 2 < d3:
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380

            for i in range(d2):
                for j in range(d3):
                    dt = i if top != 0 else d2
                    db = d2 - i if bottom != 0 else d2

                    dl = j if left != 0 else d3
                    dr = d3 - j if right != 0 else d3

                    d = min(dt, db, dl, dr)

                    if d >= feathering:
                        continue

                    v = (feathering - d) / feathering

                    t[i, j] = v * v

        mask[top:top + d2, left:left + d3] = t
1381

Guo Y.K's avatar
Guo Y.K committed
1382
1383
1384
        return (new_image, mask)


comfyanonymous's avatar
comfyanonymous committed
1385
1386
NODE_CLASS_MAPPINGS = {
    "KSampler": KSampler,
1387
    "CheckpointLoaderSimple": CheckpointLoaderSimple,
comfyanonymous's avatar
comfyanonymous committed
1388
    "CLIPTextEncode": CLIPTextEncode,
comfyanonymous's avatar
comfyanonymous committed
1389
    "CLIPSetLastLayer": CLIPSetLastLayer,
comfyanonymous's avatar
comfyanonymous committed
1390
1391
    "VAEDecode": VAEDecode,
    "VAEEncode": VAEEncode,
1392
    "VAEEncodeForInpaint": VAEEncodeForInpaint,
comfyanonymous's avatar
comfyanonymous committed
1393
    "VAELoader": VAELoader,
space-nuko's avatar
space-nuko committed
1394
1395
    "TAESDDecode": TAESDDecode,
    "TAESDEncode": TAESDEncode,
1396
    "TAESDPreviewer": TAESDPreviewer,
space-nuko's avatar
space-nuko committed
1397
    "TAESDLoader": TAESDLoader,
comfyanonymous's avatar
comfyanonymous committed
1398
1399
    "EmptyLatentImage": EmptyLatentImage,
    "LatentUpscale": LatentUpscale,
comfyanonymous's avatar
comfyanonymous committed
1400
    "LatentUpscaleBy": LatentUpscaleBy,
1401
    "LatentFromBatch": LatentFromBatch,
1402
    "RepeatLatentBatch": RepeatLatentBatch,
comfyanonymous's avatar
comfyanonymous committed
1403
    "SaveImage": SaveImage,
pythongosssss's avatar
pythongosssss committed
1404
    "PreviewImage": PreviewImage,
comfyanonymous's avatar
comfyanonymous committed
1405
    "LoadImage": LoadImage,
1406
    "LoadImageMask": LoadImageMask,
comfyanonymous's avatar
comfyanonymous committed
1407
    "ImageScale": ImageScale,
1408
    "ImageInvert": ImageInvert,
Guo Y.K's avatar
Guo Y.K committed
1409
    "ImagePadForOutpaint": ImagePadForOutpaint,
FizzleDorf's avatar
FizzleDorf committed
1410
    "ConditioningAverage ": ConditioningAverage ,
comfyanonymous's avatar
comfyanonymous committed
1411
1412
    "ConditioningCombine": ConditioningCombine,
    "ConditioningSetArea": ConditioningSetArea,
Jacob Segal's avatar
Jacob Segal committed
1413
    "ConditioningSetMask": ConditioningSetMask,
comfyanonymous's avatar
comfyanonymous committed
1414
    "KSamplerAdvanced": KSamplerAdvanced,
1415
    "SetLatentNoiseMask": SetLatentNoiseMask,
comfyanonymous's avatar
comfyanonymous committed
1416
    "LatentComposite": LatentComposite,
comfyanonymous's avatar
comfyanonymous committed
1417
    "LatentRotate": LatentRotate,
comfyanonymous's avatar
comfyanonymous committed
1418
    "LatentFlip": LatentFlip,
comfyanonymous's avatar
comfyanonymous committed
1419
    "LatentCrop": LatentCrop,
1420
    "LoraLoader": LoraLoader,
1421
    "CLIPLoader": CLIPLoader,
1422
    "CLIPVisionEncode": CLIPVisionEncode,
1423
    "StyleModelApply": StyleModelApply,
1424
    "unCLIPConditioning": unCLIPConditioning,
comfyanonymous's avatar
comfyanonymous committed
1425
1426
    "ControlNetApply": ControlNetApply,
    "ControlNetLoader": ControlNetLoader,
1427
    "DiffControlNetLoader": DiffControlNetLoader,
comfyanonymous's avatar
comfyanonymous committed
1428
1429
    "StyleModelLoader": StyleModelLoader,
    "CLIPVisionLoader": CLIPVisionLoader,
1430
    "VAEDecodeTiled": VAEDecodeTiled,
comfyanonymous's avatar
comfyanonymous committed
1431
    "VAEEncodeTiled": VAEEncodeTiled,
1432
    "TomePatchModel": TomePatchModel,
1433
    "unCLIPCheckpointLoader": unCLIPCheckpointLoader,
1434
1435
1436
    "GLIGENLoader": GLIGENLoader,
    "GLIGENTextBoxApply": GLIGENTextBoxApply,

1437
    "CheckpointLoader": CheckpointLoader,
sALTaccount's avatar
sALTaccount committed
1438
    "DiffusersLoader": DiffusersLoader,
Dr.Lt.Data's avatar
Dr.Lt.Data committed
1439
1440
1441

    "LoadLatent": LoadLatent,
    "SaveLatent": SaveLatent
comfyanonymous's avatar
comfyanonymous committed
1442
1443
}

City's avatar
City committed
1444
1445
1446
1447
1448
NODE_DISPLAY_NAME_MAPPINGS = {
    # Sampling
    "KSampler": "KSampler",
    "KSamplerAdvanced": "KSampler (Advanced)",
    # Loaders
comfyanonymous's avatar
Rename.  
comfyanonymous committed
1449
1450
    "CheckpointLoader": "Load Checkpoint (With Config)",
    "CheckpointLoaderSimple": "Load Checkpoint",
City's avatar
City committed
1451
    "VAELoader": "Load VAE",
space-nuko's avatar
space-nuko committed
1452
    "TAESDLoader": "Load TAESD",
1453
    "TAESDPreviewer": "TAESD Previewer",
City's avatar
City committed
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
    "LoraLoader": "Load LoRA",
    "CLIPLoader": "Load CLIP",
    "ControlNetLoader": "Load ControlNet Model",
    "DiffControlNetLoader": "Load ControlNet Model (diff)",
    "StyleModelLoader": "Load Style Model",
    "CLIPVisionLoader": "Load CLIP Vision",
    "UpscaleModelLoader": "Load Upscale Model",
    # Conditioning
    "CLIPVisionEncode": "CLIP Vision Encode",
    "StyleModelApply": "Apply Style Model",
    "CLIPTextEncode": "CLIP Text Encode (Prompt)",
    "CLIPSetLastLayer": "CLIP Set Last Layer",
    "ConditioningCombine": "Conditioning (Combine)",
FizzleDorf's avatar
FizzleDorf committed
1467
    "ConditioningAverage ": "Conditioning (Average)",
City's avatar
City committed
1468
    "ConditioningSetArea": "Conditioning (Set Area)",
Jacob Segal's avatar
Jacob Segal committed
1469
    "ConditioningSetMask": "Conditioning (Set Mask)",
City's avatar
City committed
1470
1471
1472
1473
1474
1475
    "ControlNetApply": "Apply ControlNet",
    # Latent
    "VAEEncodeForInpaint": "VAE Encode (for Inpainting)",
    "SetLatentNoiseMask": "Set Latent Noise Mask",
    "VAEDecode": "VAE Decode",
    "VAEEncode": "VAE Encode",
space-nuko's avatar
space-nuko committed
1476
1477
    "TAESDDecode": "TAESD Decode",
    "TAESDEncode": "TAESD Encode",
City's avatar
City committed
1478
1479
1480
1481
1482
    "LatentRotate": "Rotate Latent",
    "LatentFlip": "Flip Latent",
    "LatentCrop": "Crop Latent",
    "EmptyLatentImage": "Empty Latent Image",
    "LatentUpscale": "Upscale Latent",
comfyanonymous's avatar
comfyanonymous committed
1483
    "LatentUpscaleBy": "Upscale Latent By",
City's avatar
City committed
1484
    "LatentComposite": "Latent Composite",
1485
1486
    "LatentFromBatch" : "Latent From Batch",
    "RepeatLatentBatch": "Repeat Latent Batch",
City's avatar
City committed
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
    # Image
    "SaveImage": "Save Image",
    "PreviewImage": "Preview Image",
    "LoadImage": "Load Image",
    "LoadImageMask": "Load Image (as Mask)",
    "ImageScale": "Upscale Image",
    "ImageUpscaleWithModel": "Upscale Image (using Model)",
    "ImageInvert": "Invert Image",
    "ImagePadForOutpaint": "Pad Image for Outpainting",
    # _for_testing
    "VAEDecodeTiled": "VAE Decode (Tiled)",
    "VAEEncodeTiled": "VAE Encode (Tiled)",
}

1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
def load_custom_node(module_path):
    module_name = os.path.basename(module_path)
    if os.path.isfile(module_path):
        sp = os.path.splitext(module_path)
        module_name = sp[0]
    try:
        if os.path.isfile(module_path):
            module_spec = importlib.util.spec_from_file_location(module_name, module_path)
        else:
            module_spec = importlib.util.spec_from_file_location(module_name, os.path.join(module_path, "__init__.py"))
        module = importlib.util.module_from_spec(module_spec)
        sys.modules[module_name] = module
        module_spec.loader.exec_module(module)
        if hasattr(module, "NODE_CLASS_MAPPINGS") and getattr(module, "NODE_CLASS_MAPPINGS") is not None:
            NODE_CLASS_MAPPINGS.update(module.NODE_CLASS_MAPPINGS)
1516
1517
            if hasattr(module, "NODE_DISPLAY_NAME_MAPPINGS") and getattr(module, "NODE_DISPLAY_NAME_MAPPINGS") is not None:
                NODE_DISPLAY_NAME_MAPPINGS.update(module.NODE_DISPLAY_NAME_MAPPINGS)
1518
            return True
1519
1520
        else:
            print(f"Skip {module_path} module for custom nodes due to the lack of NODE_CLASS_MAPPINGS.")
1521
            return False
1522
1523
1524
    except Exception as e:
        print(traceback.format_exc())
        print(f"Cannot import {module_path} module for custom nodes:", e)
1525
        return False
1526

Hacker 17082006's avatar
Hacker 17082006 committed
1527
def load_custom_nodes():
1528
    node_paths = folder_paths.get_folder_paths("custom_nodes")
1529
    node_import_times = []
1530
1531
1532
1533
1534
1535
1536
1537
    for custom_node_path in node_paths:
        possible_modules = os.listdir(custom_node_path)
        if "__pycache__" in possible_modules:
            possible_modules.remove("__pycache__")

        for possible_module in possible_modules:
            module_path = os.path.join(custom_node_path, possible_module)
            if os.path.isfile(module_path) and os.path.splitext(module_path)[1] != ".py": continue
1538
            if module_path.endswith(".disabled"): continue
1539
            time_before = time.perf_counter()
1540
            success = load_custom_node(module_path)
1541
            node_import_times.append((time.perf_counter() - time_before, module_path, success))
1542

1543
    if len(node_import_times) > 0:
comfyanonymous's avatar
comfyanonymous committed
1544
        print("\nImport times for custom nodes:")
1545
        for n in sorted(node_import_times):
1546
1547
1548
1549
1550
            if n[2]:
                import_message = ""
            else:
                import_message = " (IMPORT FAILED)"
            print("{:6.1f} seconds{}:".format(n[0], import_message), n[1])
1551
        print()
1552

1553
def init_custom_nodes():
1554
    load_custom_node(os.path.join(os.path.join(os.path.dirname(os.path.realpath(__file__)), "comfy_extras"), "nodes_hypernetwork.py"))
1555
1556
    load_custom_node(os.path.join(os.path.join(os.path.dirname(os.path.realpath(__file__)), "comfy_extras"), "nodes_upscale_model.py"))
    load_custom_node(os.path.join(os.path.join(os.path.dirname(os.path.realpath(__file__)), "comfy_extras"), "nodes_post_processing.py"))
1557
    load_custom_node(os.path.join(os.path.join(os.path.dirname(os.path.realpath(__file__)), "comfy_extras"), "nodes_mask.py"))
1558
    load_custom_node(os.path.join(os.path.join(os.path.dirname(os.path.realpath(__file__)), "comfy_extras"), "nodes_rebatch.py"))
1559
    load_custom_nodes()