-
-
Notifications
You must be signed in to change notification settings - Fork 295
Expand file tree
/
Copy pathWuerstchenBaseDataLoader.py
More file actions
162 lines (132 loc) · 8 KB
/
WuerstchenBaseDataLoader.py
File metadata and controls
162 lines (132 loc) · 8 KB
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
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
import os
from modules.dataLoader.BaseDataLoader import BaseDataLoader
from modules.dataLoader.mixin.DataLoaderText2ImageMixin import DataLoaderText2ImageMixin
from modules.dataLoader.wuerstchen.EncodeWuerstchenEffnet import EncodeWuerstchenEffnet
from modules.model.BaseModel import BaseModel
from modules.model.WuerstchenModel import WuerstchenModel
from modules.modelSetup.BaseModelSetup import BaseModelSetup
from modules.modelSetup.BaseWuerstchenSetup import BaseWuerstchenSetup
from modules.util import factory
from modules.util.config.TrainConfig import TrainConfig
from modules.util.enum.ModelType import ModelType
from modules.util.torch_util import torch_gc
from modules.util.TrainProgress import TrainProgress
from mgds.pipelineModules.DecodeTokens import DecodeTokens
from mgds.pipelineModules.EncodeClipText import EncodeClipText
from mgds.pipelineModules.MapData import MapData
from mgds.pipelineModules.NormalizeImageChannels import NormalizeImageChannels
from mgds.pipelineModules.SaveImage import SaveImage
from mgds.pipelineModules.SaveText import SaveText
from mgds.pipelineModules.ScaleImage import ScaleImage
from mgds.pipelineModules.Tokenize import Tokenize
class WuerstchenBaseDataLoader(
BaseDataLoader,
DataLoaderText2ImageMixin,
):
def _preparation_modules(self, config: TrainConfig, model: WuerstchenModel):
downscale_image = ScaleImage(in_name='image', out_name='image', factor=0.75)
normalize_image = NormalizeImageChannels(image_in_name='image', image_out_name='image', mean=(0.485, 0.456, 0.406), std=(0.229, 0.224, 0.225))
encode_image = EncodeWuerstchenEffnet(in_name='image', out_name='latent_image', effnet_encoder=model.effnet_encoder, autocast_contexts=[model.autocast_context, model.effnet_encoder_autocast_context], dtype=model.effnet_encoder_train_dtype.torch_dtype())
downscale_mask = ScaleImage(in_name='mask', out_name='latent_mask', factor=0.75 * 0.03125) # *0.75 / 32.0
add_embeddings_to_prompt = MapData(in_name='prompt', out_name='prompt', map_fn=model.add_prior_text_encoder_embeddings_to_prompt)
tokenize_prompt = Tokenize(in_name='prompt', tokens_out_name='tokens', mask_out_name='tokens_mask', tokenizer=model.prior_tokenizer, max_token_length=model.prior_tokenizer.model_max_length)
if model.model_type.is_wuerstchen_v2():
encode_prompt = EncodeClipText(in_name='tokens', tokens_attention_mask_in_name='tokens_mask', hidden_state_out_name='text_encoder_hidden_state', pooled_out_name=None, add_layer_norm=True,
text_encoder=model.prior_text_encoder, hidden_state_output_index=-1, autocast_contexts=[model.autocast_context], dtype=model.train_dtype.torch_dtype())
elif model.model_type.is_stable_cascade():
encode_prompt = EncodeClipText(in_name='tokens', tokens_attention_mask_in_name='tokens_mask', hidden_state_out_name='text_encoder_hidden_state', pooled_out_name='pooled_text_encoder_output', add_layer_norm=False,
text_encoder=model.prior_text_encoder, hidden_state_output_index=-1, autocast_contexts=[model.autocast_context], dtype=model.train_dtype.torch_dtype())
modules = [downscale_image, normalize_image, encode_image]
if config.masked_training or config.model_type.has_mask_input():
modules.append(downscale_mask)
modules += [add_embeddings_to_prompt, tokenize_prompt]
if not config.train_text_encoder_or_embedding():
modules.append(encode_prompt)
return modules
def _cache_modules(self, config: TrainConfig, model: WuerstchenModel, model_setup: BaseWuerstchenSetup):
image_split_names = [
'latent_image',
'original_resolution', 'crop_offset',
]
if config.masked_training or config.model_type.has_mask_input():
image_split_names.append('latent_mask')
image_aggregate_names = ['crop_resolution', 'image_path']
text_split_names = ['tokens', 'tokens_mask', 'text_encoder_hidden_state']
if model.model_type.is_stable_cascade():
text_split_names.append('pooled_text_encoder_output')
sort_names = text_split_names + image_aggregate_names + image_split_names + [
'prompt', 'concept'
]
def before_cache_image_fun():
model.to(self.temp_device)
model.effnet_encoder_to(self.train_device)
model.eval()
torch_gc()
return self._cache_modules_from_names(
model, model_setup,
image_split_names=image_split_names,
image_aggregate_names=image_aggregate_names,
text_split_names=text_split_names,
sort_names=sort_names,
config=config,
text_caching=not config.train_text_encoder_or_embedding(),
before_cache_image_fun=before_cache_image_fun
)
def _output_modules(self, config: TrainConfig, model: WuerstchenModel, model_setup: BaseWuerstchenSetup, is_validation: bool = False):
output_names = [
'image_path', 'latent_image',
'prompt',
'tokens',
'tokens_mask',
'original_resolution', 'crop_resolution', 'crop_offset',
]
if config.masked_training or config.model_type.has_mask_input():
output_names.append('latent_mask')
if not config.train_text_encoder_or_embedding():
output_names.append('text_encoder_hidden_state')
if model.model_type.is_stable_cascade():
output_names.append('pooled_text_encoder_output')
def before_cache_image_fun():
model.to(self.temp_device)
model.effnet_encoder_to(self.train_device)
model.eval()
torch_gc()
return self._output_modules_from_out_names(
model, model_setup,
output_names=output_names,
config=config,
before_cache_image_fun=before_cache_image_fun,
autocast_context=[model.autocast_context],
train_dtype=model.train_dtype,
is_validation=is_validation,
)
def _debug_modules(self, config: TrainConfig, model: WuerstchenModel):
debug_dir = os.path.join(config.debug_dir, "dataloader")
upscale_mask = ScaleImage(in_name='latent_mask', out_name='decoded_mask', factor=1.0/0.75)
decode_prompt = DecodeTokens(in_name='tokens', out_name='decoded_prompt', tokenizer=model.prior_tokenizer)
# SaveImage(image_in_name='latent_mask', original_path_in_name='image_path', path=debug_dir, in_range_min=0, in_range_max=1),
save_mask = SaveImage(image_in_name='decoded_mask', original_path_in_name='image_path', path=debug_dir, in_range_min=0, in_range_max=1)
save_prompt = SaveText(text_in_name='decoded_prompt', original_path_in_name='image_path', path=debug_dir)
# These modules don't really work, since they are inserted after a sorting operation that does not include this data
# SaveImage(image_in_name='mask', original_path_in_name='image_path', path=debug_dir, in_range_min=0, in_range_max=1),
# SaveImage(image_in_name='image', original_path_in_name='image_path', path=debug_dir, in_range_min=-1, in_range_max=1),
modules = []
if config.masked_training or config.model_type.has_mask_input():
modules += [upscale_mask, save_mask]
modules += [decode_prompt, save_prompt]
return modules
def _create_dataset(
self,
config: TrainConfig,
model: BaseModel,
model_setup: BaseModelSetup,
train_progress: TrainProgress,
is_validation: bool = False,
):
return DataLoaderText2ImageMixin._create_dataset(self,
config, model, model_setup, train_progress, is_validation,
aspect_bucketing_quantization=128,
supports_inpainting=False,
)
factory.register(BaseDataLoader, WuerstchenBaseDataLoader, ModelType.WUERSTCHEN_2)
factory.register(BaseDataLoader, WuerstchenBaseDataLoader, ModelType.STABLE_CASCADE_1)