CodeInstructionToCodeGenerator is an operator that takes a natural language instruction and uses an LLM to generate a corresponding code snippet. This is often used as a step in a 'self-instruct' style data synthesis pipeline for code.

__init__ function

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

Parameter	Type	Default	Description
llm_serving	LLMServingABC	Required	Large language model serving instance for inference and generation.
prompt_template	PromptABC, str, None	None	Prompt template object. If None, uses CodeInstructionToCodeGeneratorPrompt. If str, uses DiyCodePrompt.

Prompt Template Descriptions

Prompt Template Name	Primary Use	Applicable Scenarios	Feature Description

run function

def run(self, storage: DataFlowStorage, input_key: str = "instruction", output_key: str = "generated_code")

Name	Type	Default	Description
storage	DataFlowStorage	Required	DataFlow storage instance for reading and writing data.
input_key	str	"instruction"	Input column name, corresponding to the natural language instruction field.
output_key	str	"generated_code"	Output column name, corresponding to the generated code field.

🧠 Example Usage

🧾 Default Output Format (Output Format)

Field	Type	Description
instruction	str	The input natural language instruction.
generated_code	str	The code snippet generated by the model.

Example Input:

{
    "instruction": "Write a Python function to calculate the factorial of a number."
}

Example Output:

{
    "instruction": "Write a Python function to calculate the factorial of a number.",
    "generated_code": "def factorial(n):\n    if n == 0:\n        return 1\n    else:\n        return n * factorial(n-1)"
}