Knowledge Base
Last updated
Last updated
The Knowledge Base feature in GenAI Stack is designed to centralize and manage your data effectively, ensuring it is easily accessible and queryable within your workflows. This component empowers you to integrate and utilize your data seamlessly across various GenAI Stack modules. Below is a step-by-step guide to creating your first Knowledge Base, along with detailed explanations of key configurations and an example to illustrate its integration into your overall workflow.
Steps to Create a Knowledge Base:
Access the Knowledge Base Section:
Navigate to the "Knowledge Base" option on the main page of your GenAI Stack site.
Click on the "Create" button to begin setting up your Knowledge Base.
Fill in the Required Details:
Name: Enter a name for your Knowledge Base. This should be descriptive enough to identify the data it contains.
Description: Provide a brief description of the Knowledge Base's purpose and contents. This helps in understanding the context and scope of the data.
Text Splitting: Specify how you want the text data to be split. This configuration determines how the text is chunked into smaller, manageable pieces. For example, you can split the text by sentences, paragraphs, or custom delimiters.
Embedding Details: Choose the embedding method to represent your data in vector form. Embeddings transform your textual data into numerical vectors that can be easily processed by machine learning models.
Upload Your Data:
You can upload files directly to populate your Knowledge Base. Supported file types include PDFs, Docs, and CSVs. Additionally, you can provide URLs to web pages, which will be scraped and included in your Knowledge Base.
Example: Suppose you are creating a Knowledge Base for AI research articles. You can upload a combination of PDFs of research papers, CSVs of bibliographic data, and URLs of relevant web pages.
Workflow Integration
After setting up your Knowledge Base, you can connect it to other components within the GenAI Stack to enhance your application's functionality. For instance, you can link the Knowledge Base to an Ensemble Retriever for efficient data retrieval and analysis.
Example:
Ensemble Retriever Integration: The screenshots below demonstrates how a Knowledge Base named "Chemistry Notes" can be connected to an Ensemble Retriever. This retriever can combine outputs from multiple retrieval models, such as the EnsembleRetriever and VectorStoreRetriever, to provide comprehensive and accurate results.
