LlamaIndex Hands-On Examples Using Synthetic Data.

I’m convinced that the most effective way to learn is through hands-on experiments. Engaging with actual code allows you to grasp how things function and assess the strengths and weaknesses of various options and approaches.

The examples will be published as a series of Python notebooks. We’ll cover topics related to using LLMs with LlamaIndex, including:

• Loading Data: Processing and ingesting data.
• Indexing: Creating data structures for LLM querying.
• Querying: Making prompt calls to an LLM.
• Evaluation: To improve you have to continuously be benchmarking and measuring the performance of your solutions.

These examples draw inspiration from the LlamaIndex Bottoms-Up Development video series.

The goal of the examples in this folder is to utilize the bubl-ai environment (container + library) to experiment with ideas and simultaneously learn LlamaIndex.

The synthetic dataset employed in these examples is the Williams Family Tree:

• bubl-ai post describing the dataset
• Code that creates the dataset
• Available on our Hugging Face site