-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathCustomizedLocalSearch.py
265 lines (234 loc) · 9.26 KB
/
CustomizedLocalSearch.py
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
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
# Copyright (c) 2024 Microsoft Corporation.
# Licensed under the MIT License
"""LocalSearch implementation."""
import logging
import time
from collections.abc import AsyncGenerator
from typing import Any
import tiktoken
from graphrag.query.context_builder.builders import LocalContextBuilder
from graphrag.query.context_builder.conversation_history import (
ConversationHistory,
)
from graphrag.query.llm.base import BaseLLM, BaseLLMCallback
from graphrag.query.llm.text_utils import num_tokens
from graphrag.query.structured_search.base import SearchResult
from graphrag.query.structured_search.local_search.search import LocalSearch
from graphrag.query.structured_search.local_search.system_prompt import (
LOCAL_SEARCH_SYSTEM_PROMPT,
)
DEFAULT_LLM_PARAMS = {
"max_tokens": 1500,
"temperature": 0.0,
}
from multiModelsPictureProcess import get_content_by_mulit_model
log = logging.getLogger(__name__)
class MultiModelLocalSearch(LocalSearch):
"""Search orchestration for local search mode."""
def __init__(
self,
llm: BaseLLM,
context_builder: LocalContextBuilder,
token_encoder: tiktoken.Encoding | None = None,
system_prompt: str = LOCAL_SEARCH_SYSTEM_PROMPT,
response_type: str = "multiple paragraphs",
callbacks: list[BaseLLMCallback] | None = None,
llm_params: dict[str, Any] = DEFAULT_LLM_PARAMS,
context_builder_params: dict | None = None,
):
super().__init__(
llm=llm,
context_builder=context_builder,
token_encoder=token_encoder,
llm_params=llm_params,
context_builder_params=context_builder_params or {},
)
self.system_prompt = system_prompt
self.callbacks = callbacks
self.response_type = response_type
async def asearch(
self,
query: str,
conversation_history: ConversationHistory | None = None,
**kwargs,
) -> SearchResult:
"""Build local search context that fits a single context window and generate answer for the user query."""
start_time = time.time()
search_prompt = ""
context_text, context_records = self.context_builder.build_context(
query=query,
conversation_history=conversation_history,
**kwargs,
**self.context_builder_params,
)
log.info("GENERATE ANSWER: %s. QUERY: %s", start_time, query)
try:
search_prompt = self.system_prompt.format(
context_data=context_text, response_type=self.response_type
)
search_messages = [
{"role": "system", "content": search_prompt},
{"role": "user", "content": query},
]
response = await self.llm.agenerate(
messages=search_messages,
streaming=True,
callbacks=self.callbacks,
**self.llm_params,
)
return SearchResult(
response=response,
context_data=context_records,
context_text=context_text,
completion_time=time.time() - start_time,
llm_calls=1,
prompt_tokens=num_tokens(search_prompt, self.token_encoder),
)
except Exception:
log.exception("Exception in _asearch")
return SearchResult(
response="",
context_data=context_records,
context_text=context_text,
completion_time=time.time() - start_time,
llm_calls=1,
prompt_tokens=num_tokens(search_prompt, self.token_encoder),
)
async def asearchWithImage(
self,
query: str,
imageUrl: str,
conversation_history: ConversationHistory | None = None,
**kwargs,
) -> SearchResult:
"""Build local search context that fits a single context window and generate answer for the user query."""
start_time = time.time()
search_prompt = ""
imageContent = await get_content_by_mulit_model(imageUrl)
contextQuery = query + imageContent
context_text, context_records = self.context_builder.build_context(
query=contextQuery,
conversation_history=conversation_history,
**kwargs,
**self.context_builder_params,
)
log.info("GENERATE ANSWER: %s. QUERY: %s", start_time, query)
try:
search_prompt = self.system_prompt.format(
context_data=context_text, response_type=self.response_type
)
search_messages = [
{"role": "system", "content": search_prompt},
{"role": "user", "content": [
{
"type": "text",
"text": query
},
{
"type": "image_url",
"image_url": {
"url": imageUrl
}
}
] },
]
response = await self.llm.agenerate(
messages=search_messages,
streaming=True,
callbacks=self.callbacks,
**self.llm_params,
)
return SearchResult(
response=response,
context_data=context_records,
context_text=context_text,
completion_time=time.time() - start_time,
llm_calls=1,
prompt_tokens=num_tokens(search_prompt, self.token_encoder),
)
except Exception:
log.exception("Exception in _asearch")
return SearchResult(
response="",
context_data=context_records,
context_text=context_text,
completion_time=time.time() - start_time,
llm_calls=1,
prompt_tokens=num_tokens(search_prompt, self.token_encoder),
)
async def astream_search(
self,
query: str,
conversation_history: ConversationHistory | None = None,
) -> AsyncGenerator:
"""Build local search context that fits a single context window and generate answer for the user query."""
start_time = time.time()
context_text, context_records = self.context_builder.build_context(
query=query,
conversation_history=conversation_history,
**self.context_builder_params,
)
log.info("GENERATE ANSWER: %s. QUERY: %s", start_time, query)
search_prompt = self.system_prompt.format(
context_data=context_text, response_type=self.response_type
)
search_messages = [
{"role": "system", "content": search_prompt},
{"role": "user", "content": query},
]
# send context records first before sending the reduce response
yield context_records
async for response in self.llm.astream_generate( # type: ignore
messages=search_messages,
callbacks=self.callbacks,
**self.llm_params,
):
yield response
def search(
self,
query: str,
conversation_history: ConversationHistory | None = None,
**kwargs,
) -> SearchResult:
"""Build local search context that fits a single context window and generate answer for the user question."""
start_time = time.time()
search_prompt = ""
context_text, context_records = self.context_builder.build_context(
query=query,
conversation_history=conversation_history,
**kwargs,
**self.context_builder_params,
)
log.info("GENERATE ANSWER: %d. QUERY: %s", start_time, query)
try:
search_prompt = self.system_prompt.format(
context_data=context_text, response_type=self.response_type
)
search_messages = [
{"role": "system", "content": search_prompt},
{"role": "user", "content": query},
]
response = self.llm.generate(
messages=search_messages,
streaming=True,
callbacks=self.callbacks,
**self.llm_params,
)
return SearchResult(
response=response,
context_data=context_records,
context_text=context_text,
completion_time=time.time() - start_time,
llm_calls=1,
prompt_tokens=num_tokens(search_prompt, self.token_encoder),
)
except Exception:
log.exception("Exception in _map_response_single_batch")
return SearchResult(
response="",
context_data=context_records,
context_text=context_text,
completion_time=time.time() - start_time,
llm_calls=1,
prompt_tokens=num_tokens(search_prompt, self.token_encoder),
)