Content
@
0 reply
0 recast
0 reaction
Daniel Lombraña
@teleyinex.eth
If you would have to start with a graph/vector DB which one would you use? A quick extra if it runs natively in AWS as a service.
1 reply
0 recast
3 reactions
Henry 🧾
@hengar.eth
Different use cases for graph db vs vector db (with some overlap, but primary needs could determine which way to lean). What do you need it for?
1 reply
0 recast
0 reaction
Garrett
@ghard
For vectors, I've found it's best to just use a vector plug in for an sql Db. Might run into issues at scale but for starting its the best
0 reply
0 recast
0 reaction
Sigrlami ⚛️❄️
@sigrlami
It highly depends on the task. Start with the pgvector extension for Postgres for validation. If you need to run ML tasks again depending on requirements, Redis if you don’t care about losing some data/path, Qdrant for GPU-heavy tasks. You can run everything on EC2 😉
0 reply
0 recast
0 reaction