Vectors Generation

The generation of vectors needs to be used in conjunction with stores, which means that you need to configure stores before you can understand vectors.

Vectors are actually the result of embedding, which is the process of converting various types of data, such as text and images, into dense vector representations. This step is essential to facilitate efficient data processing and analysis within Casibase. With embedding, questions in chat and knowledge files in storage will be converted into vectors that will be used in the next step of knowledge search.

1. Refresh Vectors

The Refresh Vectors action is set as a button on each store data under the stores menu. In stores, since we will be setting up storage providers, it will provide us with a file tree for storing user files, so after configuring stores, save the configuration and return to the home page and you will see the file tree for the storage providers.

By clicking on the Refresh Vectors button for a particular stores, it will generate the corresponding vectors for all the files in the file tree for that stores by embedding them. The following figure shows the page and the operation.

2. View vectors

After that, we can view the specific vectors generated by that storages in the vector menu.

We can see that the files in the stores from the previous step of refreshing vectors have been converted into vectors to display here.

The edit page of my vectors shows specific information such as the name of the store, the name of the embedding model, the name of the file in which the embedding was performed, the file size, the dimension, the vectors data, and so on.