Skip to content

Retriever

RetrieverQueryEngine #

Bases: BaseQueryEngine

Retriever query engine.

Parameters:

Name Type Description Default
retriever BaseRetriever

A retriever object.

 required
response_synthesizer Optional[BaseSynthesizer]

A BaseSynthesizer object.

 None
callback_manager Optional[CallbackManager]

A callback manager.

 None
Source code in llama-index-core/llama_index/core/query_engine/retriever_query_engine.py 
 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
class RetrieverQueryEngine(BaseQueryEngine):
    """Retriever query engine.

    Args:
        retriever (BaseRetriever): A retriever object.
        response_synthesizer (Optional[BaseSynthesizer]): A BaseSynthesizer
            object.
        callback_manager (Optional[CallbackManager]): A callback manager.
    """

    def __init__(
        self,
        retriever: BaseRetriever,
        response_synthesizer: Optional[BaseSynthesizer] = None,
        node_postprocessors: Optional[List[BaseNodePostprocessor]] = None,
        callback_manager: Optional[CallbackManager] = None,
    ) -> None:
        self._retriever = retriever
        self._response_synthesizer = response_synthesizer or get_response_synthesizer(
            llm=Settings.llm,
            callback_manager=callback_manager or Settings.callback_manager,
        )

        self._node_postprocessors = node_postprocessors or []
        callback_manager = (
            callback_manager or self._response_synthesizer.callback_manager
        )
        for node_postprocessor in self._node_postprocessors:
            node_postprocessor.callback_manager = callback_manager
        super().__init__(callback_manager=callback_manager)

    def _get_prompt_modules(self) -> PromptMixinType:
        """Get prompt sub-modules."""
        return {"response_synthesizer": self._response_synthesizer}

    @classmethod
    def from_args(
        cls,
        retriever: BaseRetriever,
        llm: Optional[LLM] = None,
        response_synthesizer: Optional[BaseSynthesizer] = None,
        node_postprocessors: Optional[List[BaseNodePostprocessor]] = None,
        # response synthesizer args
        response_mode: ResponseMode = ResponseMode.COMPACT,
        text_qa_template: Optional[BasePromptTemplate] = None,
        refine_template: Optional[BasePromptTemplate] = None,
        summary_template: Optional[BasePromptTemplate] = None,
        simple_template: Optional[BasePromptTemplate] = None,
        output_cls: Optional[BaseModel] = None,
        use_async: bool = False,
        streaming: bool = False,
        **kwargs: Any,
    ) -> "RetrieverQueryEngine":
        """Initialize a RetrieverQueryEngine object.".

        Args:
            retriever (BaseRetriever): A retriever object.
            node_postprocessors (Optional[List[BaseNodePostprocessor]]): A list of
                node postprocessors.
            verbose (bool): Whether to print out debug info.
            response_mode (ResponseMode): A ResponseMode object.
            text_qa_template (Optional[BasePromptTemplate]): A BasePromptTemplate
                object.
            refine_template (Optional[BasePromptTemplate]): A BasePromptTemplate object.
            simple_template (Optional[BasePromptTemplate]): A BasePromptTemplate object.

            use_async (bool): Whether to use async.
            streaming (bool): Whether to use streaming.
            optimizer (Optional[BaseTokenUsageOptimizer]): A BaseTokenUsageOptimizer
                object.

        """
        llm = llm or Settings.llm

        response_synthesizer = response_synthesizer or get_response_synthesizer(
            llm=llm,
            text_qa_template=text_qa_template,
            refine_template=refine_template,
            summary_template=summary_template,
            simple_template=simple_template,
            response_mode=response_mode,
            output_cls=output_cls,
            use_async=use_async,
            streaming=streaming,
        )

        callback_manager = Settings.callback_manager

        return cls(
            retriever=retriever,
            response_synthesizer=response_synthesizer,
            callback_manager=callback_manager,
            node_postprocessors=node_postprocessors,
        )

    def _apply_node_postprocessors(
        self, nodes: List[NodeWithScore], query_bundle: QueryBundle
    ) -> List[NodeWithScore]:
        for node_postprocessor in self._node_postprocessors:
            nodes = node_postprocessor.postprocess_nodes(
                nodes, query_bundle=query_bundle
            )
        return nodes

    def retrieve(self, query_bundle: QueryBundle) -> List[NodeWithScore]:
        nodes = self._retriever.retrieve(query_bundle)
        return self._apply_node_postprocessors(nodes, query_bundle=query_bundle)

    async def aretrieve(self, query_bundle: QueryBundle) -> List[NodeWithScore]:
        nodes = await self._retriever.aretrieve(query_bundle)
        return self._apply_node_postprocessors(nodes, query_bundle=query_bundle)

    def with_retriever(self, retriever: BaseRetriever) -> "RetrieverQueryEngine":
        return RetrieverQueryEngine(
            retriever=retriever,
            response_synthesizer=self._response_synthesizer,
            callback_manager=self.callback_manager,
            node_postprocessors=self._node_postprocessors,
        )

    def synthesize(
        self,
        query_bundle: QueryBundle,
        nodes: List[NodeWithScore],
        additional_source_nodes: Optional[Sequence[NodeWithScore]] = None,
    ) -> RESPONSE_TYPE:
        return self._response_synthesizer.synthesize(
            query=query_bundle,
            nodes=nodes,
            additional_source_nodes=additional_source_nodes,
        )

    async def asynthesize(
        self,
        query_bundle: QueryBundle,
        nodes: List[NodeWithScore],
        additional_source_nodes: Optional[Sequence[NodeWithScore]] = None,
    ) -> RESPONSE_TYPE:
        return await self._response_synthesizer.asynthesize(
            query=query_bundle,
            nodes=nodes,
            additional_source_nodes=additional_source_nodes,
        )

    @dispatcher.span
    def _query(self, query_bundle: QueryBundle) -> RESPONSE_TYPE:
        """Answer a query."""
        with self.callback_manager.event(
            CBEventType.QUERY, payload={EventPayload.QUERY_STR: query_bundle.query_str}
        ) as query_event:
            nodes = self.retrieve(query_bundle)
            response = self._response_synthesizer.synthesize(
                query=query_bundle,
                nodes=nodes,
            )
            query_event.on_end(payload={EventPayload.RESPONSE: response})

        return response

    @dispatcher.span
    async def _aquery(self, query_bundle: QueryBundle) -> RESPONSE_TYPE:
        """Answer a query."""
        with self.callback_manager.event(
            CBEventType.QUERY, payload={EventPayload.QUERY_STR: query_bundle.query_str}
        ) as query_event:
            nodes = await self.aretrieve(query_bundle)

            response = await self._response_synthesizer.asynthesize(
                query=query_bundle,
                nodes=nodes,
            )

            query_event.on_end(payload={EventPayload.RESPONSE: response})

        return response

    @property
    def retriever(self) -> BaseRetriever:
        """Get the retriever object."""
        return self._retriever

retriever property #

retriever: BaseRetriever

Get the retriever object.

from_args classmethod #

from_args(retriever: BaseRetriever, llm: Optional[LLM] = None, response_synthesizer: Optional[BaseSynthesizer] = None, node_postprocessors: Optional[List[BaseNodePostprocessor]] = None, response_mode: ResponseMode = ResponseMode.COMPACT, text_qa_template: Optional[BasePromptTemplate] = None, refine_template: Optional[BasePromptTemplate] = None, summary_template: Optional[BasePromptTemplate] = None, simple_template: Optional[BasePromptTemplate] = None, output_cls: Optional[BaseModel] = None, use_async: bool = False, streaming: bool = False, **kwargs: Any) -> RetrieverQueryEngine

Initialize a RetrieverQueryEngine object.".

Parameters:

Name Type Description Default
retriever BaseRetriever

A retriever object.

 required
node_postprocessors Optional[List[BaseNodePostprocessor]]

A list of node postprocessors.

 None
verbose bool

Whether to print out debug info.

 required
response_mode ResponseMode

A ResponseMode object.

 COMPACT
text_qa_template Optional[BasePromptTemplate]

A BasePromptTemplate object.

 None
refine_template Optional[BasePromptTemplate]

A BasePromptTemplate object.

 None
simple_template Optional[BasePromptTemplate]

A BasePromptTemplate object.

 None
use_async bool

Whether to use async.

 False
streaming bool

Whether to use streaming.

 False
optimizer Optional[BaseTokenUsageOptimizer]

A BaseTokenUsageOptimizer object.

 required
Source code in llama-index-core/llama_index/core/query_engine/retriever_query_engine.py 
 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
@classmethod
def from_args(
    cls,
    retriever: BaseRetriever,
    llm: Optional[LLM] = None,
    response_synthesizer: Optional[BaseSynthesizer] = None,
    node_postprocessors: Optional[List[BaseNodePostprocessor]] = None,
    # response synthesizer args
    response_mode: ResponseMode = ResponseMode.COMPACT,
    text_qa_template: Optional[BasePromptTemplate] = None,
    refine_template: Optional[BasePromptTemplate] = None,
    summary_template: Optional[BasePromptTemplate] = None,
    simple_template: Optional[BasePromptTemplate] = None,
    output_cls: Optional[BaseModel] = None,
    use_async: bool = False,
    streaming: bool = False,
    **kwargs: Any,
) -> "RetrieverQueryEngine":
    """Initialize a RetrieverQueryEngine object.".

    Args:
        retriever (BaseRetriever): A retriever object.
        node_postprocessors (Optional[List[BaseNodePostprocessor]]): A list of
            node postprocessors.
        verbose (bool): Whether to print out debug info.
        response_mode (ResponseMode): A ResponseMode object.
        text_qa_template (Optional[BasePromptTemplate]): A BasePromptTemplate
            object.
        refine_template (Optional[BasePromptTemplate]): A BasePromptTemplate object.
        simple_template (Optional[BasePromptTemplate]): A BasePromptTemplate object.

        use_async (bool): Whether to use async.
        streaming (bool): Whether to use streaming.
        optimizer (Optional[BaseTokenUsageOptimizer]): A BaseTokenUsageOptimizer
            object.

    """
    llm = llm or Settings.llm

    response_synthesizer = response_synthesizer or get_response_synthesizer(
        llm=llm,
        text_qa_template=text_qa_template,
        refine_template=refine_template,
        summary_template=summary_template,
        simple_template=simple_template,
        response_mode=response_mode,
        output_cls=output_cls,
        use_async=use_async,
        streaming=streaming,
    )

    callback_manager = Settings.callback_manager

    return cls(
        retriever=retriever,
        response_synthesizer=response_synthesizer,
        callback_manager=callback_manager,
        node_postprocessors=node_postprocessors,
    )