https://github.com/scribd/brutus
Brute-force in-memory vector search
https://github.com/scribd/brutus
Last synced: 1 day ago
JSON representation
Brute-force in-memory vector search
- Host: GitHub
- URL: https://github.com/scribd/brutus
- Owner: scribd
- License: mit
- Created: 2025-04-18T21:33:09.000Z (12 months ago)
- Default Branch: main
- Last Pushed: 2026-03-20T20:43:49.000Z (16 days ago)
- Last Synced: 2026-03-21T11:40:39.866Z (15 days ago)
- Language: Rust
- Size: 2.42 MB
- Stars: 3
- Watchers: 4
- Forks: 1
- Open Issues: 1
-
Metadata Files:
- Readme: README.adoc
- License: LICENSE
- Code of conduct: CODE_OF_CONDUCT.md
Awesome Lists containing this project
README
ifdef::env-github[]
:tip-caption: :bulb:
:note-caption: :information_source:
:important-caption: :heavy_exclamation_mark:
:caution-caption: :fire:
:warning-caption: :warning:
endif::[]
:toc: macro
= Brutus
An API for performing brute force vector search backed by pre-computed data in
S3.
toc::[]
== Development
This project uses Rust, so builds and tests can be done with:
* `cargo build`
* `cargo test`
* _etc_
Optionally for hot-reloading of the service during development:
[source,bash]
----
cargo install cargo-watch catflap
----
=== Testing against localstack
[source,bash]
----
cargo test --feature integration
----
To run the hot-reloading webserver:
[source,bash]
----
./scripts/live-reload
----
You can then open link:http://localhost:5000[localhost:5000]
=== Data file schema
chunk_id:long
chunk_text:string
chunk_embedding:array
element:float
page:integer
doc_id:long
chunk_sequence:integer