Is there a difference between summary

Is there a difference between summary index and document summary index? Can anyone help me better understand these two

9 comments

@kapa.ai

Let me check the knowledge base for you. One moment! :books:

Sorry @Uzair, I seem to be having a bit of trouble. Don't worry though, I've pinged my support and they will try to fix me up! :technologist:

Perhaps try again in 15mins?

LLogan M

Summary index sends all nodes to the llm to refine an answer

Document summary index generates a summary for each document, and uses that summary to pick which document(s) to use for the query

UUzair

Thank you, I am having another problem.

UUzair

I made DocumentSummaryIndexLLMRetriever as shown

Plain Text

retriever = DocumentSummaryIndexLLMRetriever(
    index=doc_summary_index,
    choice_batch_size=10,
    choice_top_k=2,
)

retrieved_nodes = retriever.retrieve("Dé un resumen de los tratamientos para la fibrosis quística")
print(len(retrieved_nodes))

UUzair

but it is showing error

Plain Text

---------------------------------------------------------------------------
IndexError                                Traceback (most recent call last)
<ipython-input-35-3639e5dd6690> in <cell line: 1>()
----> 1 retrieved_nodes = retriever.retrieve("Dé un resumen de los tratamientos para la fibrosis quística")
      2 print(len(retrieved_nodes))

3 frames
/usr/local/lib/python3.10/dist-packages/llama_index/core/instrumentation/dispatcher.py in wrapper(func, instance, args, kwargs)
    272             )
    273             try:
--> 274                 result = func(*args, **kwargs)
    275             except BaseException as e:
    276                 self.event(SpanDropEvent(span_id=id_, err_str=str(e)))

/usr/local/lib/python3.10/dist-packages/llama_index/core/base/base_retriever.py in retrieve(self, str_or_query_bundle)
    242                 payload={EventPayload.QUERY_STR: query_bundle.query_str},
    243             ) as retrieve_event:
--> 244                 nodes = self._retrieve(query_bundle)
    245                 nodes = self._handle_recursive_retrieval(query_bundle, nodes)
    246                 retrieve_event.on_end(

/usr/local/lib/python3.10/dist-packages/llama_index/core/indices/document_summary/retrievers.py in _retrieve(self, query_bundle)
     96                 query_str=query_str,
     97             )
---> 98             raw_choices, relevances = self._parse_choice_select_answer_fn(
     99                 raw_response, len(summary_nodes)
    100             )

/usr/local/lib/python3.10/dist-packages/llama_index/core/indices/utils.py in default_parse_choice_select_answer_fn(answer, num_choices, raise_error)
    102                     "answer_num: <int>, answer_relevance: <float>"
    103                 )
--> 104         answer_num = int(line_tokens[0].split(":")[1].strip())
    105         if answer_num > num_choices:
    106             continue

IndexError: list index out of range

UUzair

Can you help me with this error? Why is it occuring and how to fix it?

LLogan M

The LLM is selecting which summary to use, and failing to pick in a way that llama-index can parse

The LLM retriever is a tad flakey this way, tbh I would just use the embedding retriever

Add a reply

Find answers from the community

Is there a difference between summary