ChatPromptTemplate

The ChatPromptTemplate class provides utilities for managing prompts, prompt templates, and conversation memory for AI interactions. It helps structure and format prompts for various AI tasks and maintain conversation context.

datapizza.modules.prompt.ChatPromptTemplate

Bases: Prompt

It takes as input a Memory, Chunks, Prompt and creates a Memory with all existing messages + the user's qry + function_call_retrieval + chunks retrieval. args: user_prompt_template: str # The user prompt jinja template retrieval_prompt_template: str # The retrieval prompt jinja template

format

format(
    memory=None,
    chunks=None,
    user_prompt="",
    retrieval_query="",
)

Creates a new memory object that includes: - Existing memory messages - User's query - Function call retrieval results - Chunks retrieval results

Parameters:

Name	Type	Description	Default
memory	Memory | None	The memory object to add the new messages to.	None
chunks	list[Chunk] | None	The chunks to add to the memory.	None
user_prompt	str	The user's query.	''
retrieval_query	str	The query to search the vectorstore for.	''

Returns:

Type	Description
Memory	A new memory object with the new messages.

Overview

The ChatPromptTemplate module provides utilities for managing prompts and prompt templates in AI pipelines.

from datapizza.modules.prompt import prompt

Features:

• Prompt template management and formatting
• Context-aware prompt construction
• Integration with memory systems for conversation history
• Structured prompt formatting for different AI tasks

Usage Examples

Basic Prompt Management

import uuid

from datapizza.modules.prompt import ChatPromptTemplate
from datapizza.type import Chunk

# Create structured prompts for different tasks
system_prompt = ChatPromptTemplate(
    user_prompt_template="You are helping with data analysis tasks, this is the user prompt: {{ user_prompt }}",
    retrieval_prompt_template="Retrieved content:\n{% for chunk in chunks %}{{ chunk.text }}\n{% endfor %}"
)

print(system_prompt.format(user_prompt="Hello, how are you?", chunks=[Chunk(id=str(uuid.uuid4()), text="This is a chunk"), Chunk(id=str(uuid.uuid4()), text="This is another chunk")]))