This is a demonstration usage of late chunking via a Jina embedding model and Elastic vector database capabilities.
https://joeywhelan.github.io/late-chunking/
- Jupyter notebook
- Builds an Elastic Serverless deployment via Terraform
- Indexes two data sets and then compares late chunking performance on each
- Deletes the entire deployment via Terraform
- uv
- terraform
- Elastic Cloud account and API key
- Jina API key
- Python
- Edit the terraform.tfvars.sample and rename to terraform.tfvars
- Create a Python virtual environment
- Execute notebook