📘 Overview

The Text2SQLCoTGenerator is an operator that generates Chain-of-Thought (CoT) reasoning for converting natural language questions into SQL queries based on a given database schema. It leverages a large language model (LLM) to produce a detailed reasoning process that leads to the final SQL query, enhancing the model's transparency and accuracy for complex text-to-SQL tasks.

__init__

def __init__(self, 
             llm_serving: LLMServingABC, 
             database_manager: DatabaseManager,
             prompt_template=None
             )

Parameter	Type	Default	Description
llm_serving	LLMServingABC	Required	Large Language Model serving instance, used for executing inference and generation.
database_manager	DatabaseManager	Required	Manages database connections, schema retrieval, and query execution.
prompt_template	PromptABC	None	Prompt template object for building the input. Defaults to Text2SQLCotGeneratorPrompt if not provided.

Prompt Template Descriptions

Prompt Template Name	Primary Use	Applicable Scenarios	Feature Description

run

def run(self, 
        storage: DataFlowStorage, 
        input_sql_key: str = "SQL",
        input_question_key: str = "question",
        input_db_id_key: str = "db_id",
        input_evidence_key: str = "evidence",
        output_cot_key: str = "cot_reasoning"
        )

Name	Type	Default	Description
storage	DataFlowStorage	Required	DataFlow storage instance responsible for reading and writing data.
input_sql_key	str	"SQL"	Input column name for the gold SQL query.
input_question_key	str	"question"	Input column name for the natural language question.
input_db_id_key	str	"db_id"	Input column name for the database identifier.
input_evidence_key	str	"evidence"	Input column name for the evidence text.
output_cot_key	str	"cot_reasoning"	Output column name for the generated Chain-of-Thought reasoning.

🧠 Example Usage