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
266
267
268
269
270
271
272
273
274
275
276 | class SubQuestionQueryEngine(BaseQueryEngine):
"""Sub question query engine.
A query engine that breaks down a complex query (e.g. compare and contrast) into
many sub questions and their target query engine for execution.
After executing all sub questions, all responses are gathered and sent to
response synthesizer to produce the final response.
Args:
question_gen (BaseQuestionGenerator): A module for generating sub questions
given a complex question and tools.
response_synthesizer (BaseSynthesizer): A response synthesizer for
generating the final response
query_engine_tools (Sequence[QueryEngineTool]): Tools to answer the
sub questions.
verbose (bool): whether to print intermediate questions and answers.
Defaults to True
use_async (bool): whether to execute the sub questions with asyncio.
Defaults to True
"""
def __init__(
self,
question_gen: BaseQuestionGenerator,
response_synthesizer: BaseSynthesizer,
query_engine_tools: Sequence[QueryEngineTool],
callback_manager: Optional[CallbackManager] = None,
verbose: bool = True,
use_async: bool = False,
) -> None:
self._question_gen = question_gen
self._response_synthesizer = response_synthesizer
self._metadatas = [x.metadata for x in query_engine_tools]
self._query_engines = {
tool.metadata.name: tool.query_engine for tool in query_engine_tools
}
self._verbose = verbose
self._use_async = use_async
super().__init__(callback_manager)
def _get_prompt_modules(self) -> PromptMixinType:
"""Get prompt sub-modules."""
return {
"question_gen": self._question_gen,
"response_synthesizer": self._response_synthesizer,
}
@classmethod
def from_defaults(
cls,
query_engine_tools: Sequence[QueryEngineTool],
llm: Optional[LLM] = None,
question_gen: Optional[BaseQuestionGenerator] = None,
response_synthesizer: Optional[BaseSynthesizer] = None,
verbose: bool = True,
use_async: bool = True,
) -> "SubQuestionQueryEngine":
callback_manager = Settings.callback_manager
if len(query_engine_tools) > 0:
callback_manager = query_engine_tools[0].query_engine.callback_manager
llm = llm or Settings.llm
if question_gen is None:
try:
from llama_index.question_gen.openai import (
OpenAIQuestionGenerator,
) # pants: no-infer-dep
# try to use OpenAI function calling based question generator.
# if incompatible, use general LLM question generator
question_gen = OpenAIQuestionGenerator.from_defaults(llm=llm)
except ImportError as e:
raise ImportError(
"`llama-index-question-gen-openai` package cannot be found. "
"Please install it by using `pip install `llama-index-question-gen-openai`"
)
except ValueError:
question_gen = LLMQuestionGenerator.from_defaults(llm=llm)
synth = response_synthesizer or get_response_synthesizer(
llm=llm,
callback_manager=callback_manager,
use_async=use_async,
)
return cls(
question_gen,
synth,
query_engine_tools,
callback_manager=callback_manager,
verbose=verbose,
use_async=use_async,
)
def _query(self, query_bundle: QueryBundle) -> RESPONSE_TYPE:
with self.callback_manager.event(
CBEventType.QUERY, payload={EventPayload.QUERY_STR: query_bundle.query_str}
) as query_event:
sub_questions = self._question_gen.generate(self._metadatas, query_bundle)
colors = get_color_mapping([str(i) for i in range(len(sub_questions))])
if self._verbose:
print_text(f"Generated {len(sub_questions)} sub questions.\n")
if self._use_async:
tasks = [
self._aquery_subq(sub_q, color=colors[str(ind)])
for ind, sub_q in enumerate(sub_questions)
]
qa_pairs_all = run_async_tasks(tasks)
qa_pairs_all = cast(List[Optional[SubQuestionAnswerPair]], qa_pairs_all)
else:
qa_pairs_all = [
self._query_subq(sub_q, color=colors[str(ind)])
for ind, sub_q in enumerate(sub_questions)
]
# filter out sub questions that failed
qa_pairs: List[SubQuestionAnswerPair] = list(filter(None, qa_pairs_all))
nodes = [self._construct_node(pair) for pair in qa_pairs]
source_nodes = [node for qa_pair in qa_pairs for node in qa_pair.sources]
response = self._response_synthesizer.synthesize(
query=query_bundle,
nodes=nodes,
additional_source_nodes=source_nodes,
)
query_event.on_end(payload={EventPayload.RESPONSE: response})
return response
async def _aquery(self, query_bundle: QueryBundle) -> RESPONSE_TYPE:
with self.callback_manager.event(
CBEventType.QUERY, payload={EventPayload.QUERY_STR: query_bundle.query_str}
) as query_event:
sub_questions = await self._question_gen.agenerate(
self._metadatas, query_bundle
)
colors = get_color_mapping([str(i) for i in range(len(sub_questions))])
if self._verbose:
print_text(f"Generated {len(sub_questions)} sub questions.\n")
tasks = [
self._aquery_subq(sub_q, color=colors[str(ind)])
for ind, sub_q in enumerate(sub_questions)
]
qa_pairs_all = await asyncio.gather(*tasks)
qa_pairs_all = cast(List[Optional[SubQuestionAnswerPair]], qa_pairs_all)
# filter out sub questions that failed
qa_pairs: List[SubQuestionAnswerPair] = list(filter(None, qa_pairs_all))
nodes = [self._construct_node(pair) for pair in qa_pairs]
source_nodes = [node for qa_pair in qa_pairs for node in qa_pair.sources]
response = await self._response_synthesizer.asynthesize(
query=query_bundle,
nodes=nodes,
additional_source_nodes=source_nodes,
)
query_event.on_end(payload={EventPayload.RESPONSE: response})
return response
def _construct_node(self, qa_pair: SubQuestionAnswerPair) -> NodeWithScore:
node_text = (
f"Sub question: {qa_pair.sub_q.sub_question}\nResponse: {qa_pair.answer}"
)
return NodeWithScore(node=TextNode(text=node_text))
async def _aquery_subq(
self, sub_q: SubQuestion, color: Optional[str] = None
) -> Optional[SubQuestionAnswerPair]:
try:
with self.callback_manager.event(
CBEventType.SUB_QUESTION,
payload={EventPayload.SUB_QUESTION: SubQuestionAnswerPair(sub_q=sub_q)},
) as event:
question = sub_q.sub_question
query_engine = self._query_engines[sub_q.tool_name]
if self._verbose:
print_text(f"[{sub_q.tool_name}] Q: {question}\n", color=color)
response = await query_engine.aquery(question)
response_text = str(response)
if self._verbose:
print_text(f"[{sub_q.tool_name}] A: {response_text}\n", color=color)
qa_pair = SubQuestionAnswerPair(
sub_q=sub_q, answer=response_text, sources=response.source_nodes
)
event.on_end(payload={EventPayload.SUB_QUESTION: qa_pair})
return qa_pair
except ValueError:
logger.warning(f"[{sub_q.tool_name}] Failed to run {question}")
return None
def _query_subq(
self, sub_q: SubQuestion, color: Optional[str] = None
) -> Optional[SubQuestionAnswerPair]:
try:
with self.callback_manager.event(
CBEventType.SUB_QUESTION,
payload={EventPayload.SUB_QUESTION: SubQuestionAnswerPair(sub_q=sub_q)},
) as event:
question = sub_q.sub_question
query_engine = self._query_engines[sub_q.tool_name]
if self._verbose:
print_text(f"[{sub_q.tool_name}] Q: {question}\n", color=color)
response = query_engine.query(question)
response_text = str(response)
if self._verbose:
print_text(f"[{sub_q.tool_name}] A: {response_text}\n", color=color)
qa_pair = SubQuestionAnswerPair(
sub_q=sub_q, answer=response_text, sources=response.source_nodes
)
event.on_end(payload={EventPayload.SUB_QUESTION: qa_pair})
return qa_pair
except ValueError:
logger.warning(f"[{sub_q.tool_name}] Failed to run {question}")
return None
|