Is it possible to extract either the sql query or the sql result from the QueryPipeline?

[simontoftnielsen]

I want to return the actual result from the sql query to the user.
I am using the QueryPipeline that looks like this:

qp = QP(
modules={
"input": InputComponent(),
"table_retriever": obj_retriever,
"table_output_parser": table_parser_component,
"text2sql_prompt": text2sql_component,
"text2sql_llm": llm,
"sql_output_parser": sql_parser_component,
"sql_retriever": sql_retriever,
"response_synthesis_prompt": response_synthesis_prompt,
"response_synthesis_llm": llm,
},
verbose=True,
)

qp.add_chain(["input", "table_retriever", "table_output_parser"])
qp.add_link("input", "text2sql_prompt", dest_key="query_str")
qp.add_link("table_output_parser", "text2sql_prompt", dest_key="schema")
qp.add_chain(
["text2sql_prompt", "text2sql_llm", "sql_output_parser", "sql_retriever"]
)
qp.add_link(
"sql_output_parser", "response_synthesis_prompt", dest_key="sql_query"
)
qp.add_link(
"sql_retriever", "response_synthesis_prompt", dest_key="context_str"
)
qp.add_link("input", "response_synthesis_prompt", dest_key="query_str")
qp.add_link("response_synthesis_prompt", "response_synthesis_llm")

How ever the qp object does not return either the sql query that has been used or the result from the sql query.
Is it possible to somehow extract either of these from the pipeline?

[dosubot[bot]]

Hey there, @simontoftnielsen! Great to see you diving into another intriguing puzzle with LlamaIndex. How's everything going on your end?

Based on the code you've provided, it seems like you're trying to extract the SQL query or its result from the QueryPipeline (QP) object. However, the QP object itself does not directly handle SQL queries or their results. Instead, it orchestrates the execution of a series of modules (components) that you've added to the pipeline.

If you want to extract the SQL query or its result, you'll need to look into the specific modules that handle these tasks. In your case, these modules might be text2sql_llm, sql_output_parser, and sql_retriever.

Here's a brief explanation of how you might approach this:

1. Check the Implementation of Query Components: Look into the implementation of the modules that are responsible for handling the SQL query. For instance, text2sql_llm might be the module that generates the SQL query, sql_output_parser might parse the output of the SQL query, and sql_retriever might be the one executing the SQL query and retrieving the result. If these modules have methods or properties that expose the SQL query or its result, you can use them to extract the information you need.

2. Logging or Debugging: If the SQL query is being executed but you just want to know what it is for debugging or logging purposes, you could modify the module that executes the SQL query to log the query when it's executed. This would give you visibility into the SQL queries that are being run.

Remember, the QP object is just a framework for executing a series of tasks in a specific order. It doesn't know or care about the specifics of what each task is doing. So, if you want to extract specific information like an SQL query or its result, you'll need to get that from the specific task (module) that handles it.

I hope this helps! If you have any more questions or need further clarification, feel free to ask.

Sources

• llama-index-legacy/llama_index/legacy/query_pipeline/query.py
• llama-index-core/llama_index/core/query_pipeline/query.py

About Dosu

This response is meant to be useful and save you time. It is not meant to be a precise solution, but rather a starting point for your own research.

Help me be more useful! Please leave a 👍 if this is helpful and 👎 if it is irrelevant.

If you want to continue the conversation, start your reply with @dosu-bot.

[simontoftnielsen]

Yes. That is clear. But how do i access the modules?
I believe it would be the sql_parser_component or the sql_retriever I need to access.

The end goal is to have either the sql query or ideally the sql result saved as a variable that I can use later on