Commit
•
f16fef6
1
Parent(s):
ecfb95b
正确转换prompt (#1)
Browse files- 正确转换prompt (b75de531977fdea2e1fa26885437e25dbc4eae8c)
Co-authored-by: animelover <[email protected]>
app.py
CHANGED
@@ -8,76 +8,194 @@ from compel import Compel, ReturnedEmbeddingsType
|
|
8 |
|
9 |
import re
|
10 |
|
11 |
-
|
12 |
-
|
13 |
-
|
14 |
-
|
|
|
15 |
def parse_prompt_attention(text):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
16 |
res = []
|
17 |
-
|
18 |
-
|
19 |
-
|
20 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
21 |
return res
|
22 |
|
23 |
-
def prompt_attention_to_invoke_prompt(
|
24 |
-
|
25 |
-
for
|
26 |
-
|
27 |
-
|
28 |
-
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
40 |
|
41 |
def get_embed_new(prompt, pipeline, compel, only_convert_string=False, compel_process_sd=False):
|
|
|
42 |
if compel_process_sd:
|
43 |
return merge_embeds(tokenize_line(prompt, pipeline.tokenizer), compel)
|
44 |
else:
|
45 |
# fix bug weights conversion excessive emphasis
|
46 |
-
prompt = prompt.replace("((", "(").replace("))", ")")
|
47 |
|
48 |
# Convert to Compel
|
49 |
attention = parse_prompt_attention(prompt)
|
50 |
-
|
51 |
-
# 新增处理,当 attention 为空时
|
52 |
-
if not attention:
|
53 |
-
if only_convert_string:
|
54 |
-
return prompt
|
55 |
-
else:
|
56 |
-
conditioning, pooled = compel(prompt)
|
57 |
-
return conditioning, pooled
|
58 |
|
59 |
-
global_attention_chunks = []
|
60 |
-
# 下面的部分保持不变
|
61 |
for att in attention:
|
62 |
-
for
|
63 |
-
|
64 |
-
for
|
65 |
temp_dict = {
|
66 |
"weight": round(att[1], 2),
|
67 |
-
"
|
68 |
-
"prompt": f'{
|
69 |
}
|
70 |
-
|
71 |
|
72 |
max_tokens = pipeline.tokenizer.model_max_length - 2
|
73 |
-
|
74 |
current_list = []
|
75 |
current_length = 0
|
76 |
-
for item in
|
77 |
-
if current_length + item['
|
78 |
-
|
79 |
current_list = [[item['prompt'], item['weight']]]
|
80 |
-
current_length = item['
|
81 |
else:
|
82 |
if not current_list:
|
83 |
current_list.append([item['prompt'], item['weight']])
|
@@ -86,14 +204,19 @@ def get_embed_new(prompt, pipeline, compel, only_convert_string=False, compel_pr
|
|
86 |
current_list.append([item['prompt'], item['weight']])
|
87 |
else:
|
88 |
current_list[-1][0] += f" {item['prompt']}"
|
89 |
-
current_length += item['
|
90 |
if current_list:
|
91 |
-
|
92 |
|
93 |
if only_convert_string:
|
94 |
-
return ' '.join([prompt_attention_to_invoke_prompt(i) for i in
|
|
|
|
|
95 |
|
96 |
-
|
|
|
|
|
|
|
97 |
|
98 |
if not torch.cuda.is_available():
|
99 |
DESCRIPTION += "\n<p>你现在运行在CPU上 但是此项目只支持GPU.</p>"
|
@@ -138,24 +261,22 @@ def infer(
|
|
138 |
tokenizer=[pipe.tokenizer, pipe.tokenizer_2],
|
139 |
text_encoder=[pipe.text_encoder, pipe.text_encoder_2],
|
140 |
returned_embeddings_type=ReturnedEmbeddingsType.PENULTIMATE_HIDDEN_STATES_NON_NORMALIZED,
|
141 |
-
requires_pooled=[False, True]
|
|
|
142 |
)
|
143 |
# 在 infer 函数中调用 get_embed_new
|
144 |
-
|
145 |
-
|
146 |
-
|
147 |
-
|
148 |
-
|
149 |
-
else:
|
150 |
-
negative_conditioning = None
|
151 |
-
negative_pooled = None
|
152 |
|
153 |
# 在调用 pipe 时,使用新的参数名称(确保参数名称正确)
|
154 |
image = pipe(
|
155 |
-
prompt_embeds=conditioning,
|
156 |
-
pooled_prompt_embeds=pooled,
|
157 |
-
negative_prompt_embeds=
|
158 |
-
negative_pooled_prompt_embeds=
|
159 |
width=width,
|
160 |
height=height,
|
161 |
guidance_scale=guidance_scale,
|
|
|
8 |
|
9 |
import re
|
10 |
|
11 |
+
# =====================================
|
12 |
+
# Prompt weights
|
13 |
+
# =====================================
|
14 |
+
import torch
|
15 |
+
import re
|
16 |
def parse_prompt_attention(text):
|
17 |
+
re_attention = re.compile(r"""
|
18 |
+
\\\(|
|
19 |
+
\\\)|
|
20 |
+
\\\[|
|
21 |
+
\\]|
|
22 |
+
\\\\|
|
23 |
+
\\|
|
24 |
+
\(|
|
25 |
+
\[|
|
26 |
+
:([+-]?[.\d]+)\)|
|
27 |
+
\)|
|
28 |
+
]|
|
29 |
+
[^\\()\[\]:]+|
|
30 |
+
:
|
31 |
+
""", re.X)
|
32 |
+
|
33 |
res = []
|
34 |
+
round_brackets = []
|
35 |
+
square_brackets = []
|
36 |
+
|
37 |
+
round_bracket_multiplier = 1.1
|
38 |
+
square_bracket_multiplier = 1 / 1.1
|
39 |
+
|
40 |
+
def multiply_range(start_position, multiplier):
|
41 |
+
for p in range(start_position, len(res)):
|
42 |
+
res[p][1] *= multiplier
|
43 |
+
|
44 |
+
for m in re_attention.finditer(text):
|
45 |
+
text = m.group(0)
|
46 |
+
weight = m.group(1)
|
47 |
+
|
48 |
+
if text.startswith('\\'):
|
49 |
+
res.append([text[1:], 1.0])
|
50 |
+
elif text == '(':
|
51 |
+
round_brackets.append(len(res))
|
52 |
+
elif text == '[':
|
53 |
+
square_brackets.append(len(res))
|
54 |
+
elif weight is not None and len(round_brackets) > 0:
|
55 |
+
multiply_range(round_brackets.pop(), float(weight))
|
56 |
+
elif text == ')' and len(round_brackets) > 0:
|
57 |
+
multiply_range(round_brackets.pop(), round_bracket_multiplier)
|
58 |
+
elif text == ']' and len(square_brackets) > 0:
|
59 |
+
multiply_range(square_brackets.pop(), square_bracket_multiplier)
|
60 |
+
else:
|
61 |
+
parts = re.split(re.compile(r"\s*\bBREAK\b\s*", re.S), text)
|
62 |
+
for i, part in enumerate(parts):
|
63 |
+
if i > 0:
|
64 |
+
res.append(["BREAK", -1])
|
65 |
+
res.append([part, 1.0])
|
66 |
+
|
67 |
+
for pos in round_brackets:
|
68 |
+
multiply_range(pos, round_bracket_multiplier)
|
69 |
+
|
70 |
+
for pos in square_brackets:
|
71 |
+
multiply_range(pos, square_bracket_multiplier)
|
72 |
+
|
73 |
+
if len(res) == 0:
|
74 |
+
res = [["", 1.0]]
|
75 |
+
|
76 |
+
# merge runs of identical weights
|
77 |
+
i = 0
|
78 |
+
while i + 1 < len(res):
|
79 |
+
if res[i][1] == res[i + 1][1]:
|
80 |
+
res[i][0] += res[i + 1][0]
|
81 |
+
res.pop(i + 1)
|
82 |
+
else:
|
83 |
+
i += 1
|
84 |
+
|
85 |
return res
|
86 |
|
87 |
+
def prompt_attention_to_invoke_prompt(attention):
|
88 |
+
tokens = []
|
89 |
+
for text, weight in attention:
|
90 |
+
# Round weight to 2 decimal places
|
91 |
+
weight = round(weight, 2)
|
92 |
+
if weight == 1.0:
|
93 |
+
tokens.append(text)
|
94 |
+
elif weight < 1.0:
|
95 |
+
if weight < 0.8:
|
96 |
+
tokens.append(f"({text}){weight}")
|
97 |
+
else:
|
98 |
+
tokens.append(f"({text})-" + "-" * int((1.0 - weight) * 10))
|
99 |
+
else:
|
100 |
+
if weight < 1.3:
|
101 |
+
tokens.append(f"({text})" + "+" * int((weight - 1.0) * 10))
|
102 |
+
else:
|
103 |
+
tokens.append(f"({text}){weight}")
|
104 |
+
return "".join(tokens)
|
105 |
+
|
106 |
+
def concat_tensor(t):
|
107 |
+
t_list = torch.split(t, 1, dim=0)
|
108 |
+
t = torch.cat(t_list, dim=1)
|
109 |
+
return t
|
110 |
+
|
111 |
+
def merge_embeds(prompt_chanks, compel):
|
112 |
+
num_chanks = len(prompt_chanks)
|
113 |
+
if num_chanks != 0:
|
114 |
+
power_prompt = 1/(num_chanks*(num_chanks+1)//2)
|
115 |
+
prompt_embs = compel(prompt_chanks)
|
116 |
+
t_list = list(torch.split(prompt_embs, 1, dim=0))
|
117 |
+
for i in range(num_chanks):
|
118 |
+
t_list[-(i+1)] = t_list[-(i+1)] * ((i+1)*power_prompt)
|
119 |
+
prompt_emb = torch.stack(t_list, dim=0).sum(dim=0)
|
120 |
+
else:
|
121 |
+
prompt_emb = compel('')
|
122 |
+
return prompt_emb
|
123 |
+
|
124 |
+
def detokenize(chunk, actual_prompt):
|
125 |
+
chunk[-1] = chunk[-1].replace('</w>', '')
|
126 |
+
chanked_prompt = ''.join(chunk).strip()
|
127 |
+
while '</w>' in chanked_prompt:
|
128 |
+
if actual_prompt[chanked_prompt.find('</w>')] == ' ':
|
129 |
+
chanked_prompt = chanked_prompt.replace('</w>', ' ', 1)
|
130 |
+
else:
|
131 |
+
chanked_prompt = chanked_prompt.replace('</w>', '', 1)
|
132 |
+
actual_prompt = actual_prompt.replace(chanked_prompt,'')
|
133 |
+
return chanked_prompt.strip(), actual_prompt.strip()
|
134 |
+
|
135 |
+
def tokenize_line(line, tokenizer): # split into chunks
|
136 |
+
actual_prompt = line.lower().strip()
|
137 |
+
actual_tokens = tokenizer.tokenize(actual_prompt)
|
138 |
+
max_tokens = tokenizer.model_max_length - 2
|
139 |
+
comma_token = tokenizer.tokenize(',')[0]
|
140 |
+
|
141 |
+
chunks = []
|
142 |
+
chunk = []
|
143 |
+
for item in actual_tokens:
|
144 |
+
chunk.append(item)
|
145 |
+
if len(chunk) == max_tokens:
|
146 |
+
if chunk[-1] != comma_token:
|
147 |
+
for i in range(max_tokens-1, -1, -1):
|
148 |
+
if chunk[i] == comma_token:
|
149 |
+
actual_chunk, actual_prompt = detokenize(chunk[:i+1], actual_prompt)
|
150 |
+
chunks.append(actual_chunk)
|
151 |
+
chunk = chunk[i+1:]
|
152 |
+
break
|
153 |
+
else:
|
154 |
+
actual_chunk, actual_prompt = detokenize(chunk, actual_prompt)
|
155 |
+
chunks.append(actual_chunk)
|
156 |
+
chunk = []
|
157 |
+
else:
|
158 |
+
actual_chunk, actual_prompt = detokenize(chunk, actual_prompt)
|
159 |
+
chunks.append(actual_chunk)
|
160 |
+
chunk = []
|
161 |
+
if chunk:
|
162 |
+
actual_chunk, _ = detokenize(chunk, actual_prompt)
|
163 |
+
chunks.append(actual_chunk)
|
164 |
+
|
165 |
+
return chunks
|
166 |
|
167 |
def get_embed_new(prompt, pipeline, compel, only_convert_string=False, compel_process_sd=False):
|
168 |
+
|
169 |
if compel_process_sd:
|
170 |
return merge_embeds(tokenize_line(prompt, pipeline.tokenizer), compel)
|
171 |
else:
|
172 |
# fix bug weights conversion excessive emphasis
|
173 |
+
prompt = prompt.replace("((", "(").replace("))", ")").replace("\\", "\\\\\\")
|
174 |
|
175 |
# Convert to Compel
|
176 |
attention = parse_prompt_attention(prompt)
|
177 |
+
global_attention_chanks = []
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
178 |
|
|
|
|
|
179 |
for att in attention:
|
180 |
+
for chank in att[0].split(','):
|
181 |
+
temp_prompt_chanks = tokenize_line(chank, pipeline.tokenizer)
|
182 |
+
for small_chank in temp_prompt_chanks:
|
183 |
temp_dict = {
|
184 |
"weight": round(att[1], 2),
|
185 |
+
"lenght": len(pipeline.tokenizer.tokenize(f'{small_chank},')),
|
186 |
+
"prompt": f'{small_chank},'
|
187 |
}
|
188 |
+
global_attention_chanks.append(temp_dict)
|
189 |
|
190 |
max_tokens = pipeline.tokenizer.model_max_length - 2
|
191 |
+
global_prompt_chanks = []
|
192 |
current_list = []
|
193 |
current_length = 0
|
194 |
+
for item in global_attention_chanks:
|
195 |
+
if current_length + item['lenght'] > max_tokens:
|
196 |
+
global_prompt_chanks.append(current_list)
|
197 |
current_list = [[item['prompt'], item['weight']]]
|
198 |
+
current_length = item['lenght']
|
199 |
else:
|
200 |
if not current_list:
|
201 |
current_list.append([item['prompt'], item['weight']])
|
|
|
204 |
current_list.append([item['prompt'], item['weight']])
|
205 |
else:
|
206 |
current_list[-1][0] += f" {item['prompt']}"
|
207 |
+
current_length += item['lenght']
|
208 |
if current_list:
|
209 |
+
global_prompt_chanks.append(current_list)
|
210 |
|
211 |
if only_convert_string:
|
212 |
+
return ' '.join([prompt_attention_to_invoke_prompt(i) for i in global_prompt_chanks])
|
213 |
+
|
214 |
+
return merge_embeds([prompt_attention_to_invoke_prompt(i) for i in global_prompt_chanks], compel)
|
215 |
|
216 |
+
def add_comma_after_pattern_ti(text):
|
217 |
+
pattern = re.compile(r'\b\w+_\d+\b')
|
218 |
+
modified_text = pattern.sub(lambda x: x.group() + ',', text)
|
219 |
+
return modified_text
|
220 |
|
221 |
if not torch.cuda.is_available():
|
222 |
DESCRIPTION += "\n<p>你现在运行在CPU上 但是此项目只支持GPU.</p>"
|
|
|
261 |
tokenizer=[pipe.tokenizer, pipe.tokenizer_2],
|
262 |
text_encoder=[pipe.text_encoder, pipe.text_encoder_2],
|
263 |
returned_embeddings_type=ReturnedEmbeddingsType.PENULTIMATE_HIDDEN_STATES_NON_NORMALIZED,
|
264 |
+
requires_pooled=[False, True],
|
265 |
+
truncate_long_prompts=False
|
266 |
)
|
267 |
# 在 infer 函数中调用 get_embed_new
|
268 |
+
if not use_negative_prompt:
|
269 |
+
negative_prompt = ""
|
270 |
+
prompt = get_embed_new(prompt, pipe, compel, only_convert_string=True)
|
271 |
+
negative_prompt = get_embed_new(negative_prompt, pipe, compel, only_convert_string=True)
|
272 |
+
conditioning, pooled = compel([prompt, neg_prompt]) # 必须同时处理来保证长度相等
|
|
|
|
|
|
|
273 |
|
274 |
# 在调用 pipe 时,使用新的参数名称(确保参数名称正确)
|
275 |
image = pipe(
|
276 |
+
prompt_embeds=conditioning[0:1],
|
277 |
+
pooled_prompt_embeds=pooled[0:1],
|
278 |
+
negative_prompt_embeds=conditioning[1:2],
|
279 |
+
negative_pooled_prompt_embeds=pooled[1:2],
|
280 |
width=width,
|
281 |
height=height,
|
282 |
guidance_scale=guidance_scale,
|