How to get started with LLMWare?
In this repository, we have small code snippets with a lot of annotation that describe key setup items to get started quickly and successfully with LLMWare:
-
Configuring a DB - for a fast start, you don't have to do anything except set the .active_db parameter to "sqlite" and all libraries will be written on a local embedded sqlite instance with no installation required.
from llmware.configs import LLMWareConfig LLMWareConfig().set_active_db("sqlite")
After writing content into your first library, you will find the sqlite db at the path defined here:
from llmware.configs import SQLiteConfig
SQLiteConfig().get_config("sqlite_db_folder_path")
-
Loading Sample Files - for an overview of the sample files available and how to pull down (integrated into many examples too.)
from llmware.setup import Setup Setup().load_sample_files()
-
Using the Model Catalog - this is the primary model discovery and loading class for LLMWare.
from llmware.models import ModelCatalog all_models = ModelCatalog().list_all_models() for model in all_models: print("model: ", model)
-
Working with Libraries - Libraries are the main organizing construct for text collections in LLMWare.
As you are getting started, we would also recommend that you check out the Fast Start examples and videos as well!