mirror of
https://github.com/Unstructured-IO/unstructured.git
synced 2025-07-09 18:15:55 +00:00
e052c2a9b2
* pgvector requirements * first pass on pgvector notebook and sql alchemy file * created code for loading vectors into db * added query for embedding distance * updates to pgvector notebook * update function with time decay * update pgvector notebook to use example code * remove old create table script * add readme for pgvector * update example to use get_date()
18 lines
965 B
Markdown
18 lines
965 B
Markdown
# Loading `unstructured` outputs into Postgres with `pgvector`
|
|
|
|
The following example shows how to load `unstructured` output into Postgres with the
|
|
`pgvector` extension installed. Combining the similarity search functionality of
|
|
`pgvector` with the traditional RDBMS capabilities of Postgres allow users to performing
|
|
similarity searches that are conditioned on metadata or biased toward more recent documents.
|
|
Use cases include document discovery and more sophisticated retrieval augmented generation
|
|
for LLMs.
|
|
The [`langchain` docs](https://docs.langchain.com/docs/components/memory/) have more information
|
|
about retrieval augmented generation.
|
|
|
|
## Running the example
|
|
1. Install [Postgres](https://www.postgresql.org/docs/15/tutorial-install.html).
|
|
1. Install [`pgvector`](https://github.com/pgvector/pgvector)
|
|
1. Run `pip install -r requirements.txt` to install the Python dependencies.
|
|
1. Run `jupyter-notebook to start.
|
|
1. Run the `pgvector.ipynb` notebook.
|