Retrievers
Prompt Retriever Node
The Prompt Retriever node is designed to store prompt templates with associated names and descriptions. These stored prompts can later be queried and used by a MultiPromptChain. This node is particularly useful for organizing and managing multiple prompts that can be dynamically selected based on the context or requirements of a conversation or task.
Node Details
- Name: promptRetriever
- Type: PromptRetriever
- Category: Retrievers
- Version: 1.0
Input Parameters
-
Prompt Name
- Type: string
- Description: A unique identifier for the prompt template
- Example: “physics-qa”
-
Prompt Description
- Type: string
- Description: A brief explanation of what the prompt does and when it should be used
- Example: “Good for answering questions about physics”
-
Prompt System Message
- Type: string
- Description: The actual prompt template or system message that guides the AI’s behavior
- Example: “You are a very smart physics professor. You are great at answering questions about physics in a concise and easy to understand manner. When you don’t know the answer to a question you admit that you don’t know.”
Output
The node initializes and returns a PromptRetriever object, which encapsulates the provided prompt information (name, description, and system message).Usage
This node is typically used as part of a larger system where multiple specialized prompts are needed. By storing prompts with metadata, it allows for:- Organized Prompt Management: Keeping track of multiple prompts for different purposes.
- Dynamic Prompt Selection: Enabling systems to choose the most appropriate prompt based on the current context or user query.
- Improved Maintainability: Centralizing prompt storage and making it easier to update or modify prompts without changing the underlying code.
Integration
The Prompt Retriever is often used in conjunction with a MultiPromptChain, which can dynamically select and use the most appropriate prompt based on the input or context. This allows for creating more flexible and adaptive AI systems that can handle a wide range of queries or tasks by selecting the most suitable prompt template.Example Use Case
In a multi-purpose AI assistant, you might have several Prompt Retriever nodes set up:- One for physics questions
- One for literature analysis
- One for coding help
- One for general conversation
Previous
Reciprocal Rank Fusion RetrieverThe Reciprocal Rank Fusion (RRF) Retriever is a specialized retriever that uses the Reciprocal Rank Fusion algorithm to re-rank search results obtained from multiple query generations. This node enhances the retrieval process by generating synthetic queries and combining their results for improved relevance.
Next