Lightrag ingestion speed

[themindfactory]

Just downloaded and installed LR Server I also connected it to a PSQL server, I very patiently waited for it to ingest a few hunder docs, 30kB each, I have settings at 1024 chunks using openai for embedding, and allow 8 at a time, and it is very slow almost 2 minutes per doc... I have a huge project in mind and this is just way to slow... however I do not want to give up due to the fact that once I got those 200 in, the output was amazing in mix mode! And its all local on a very decent machine 12 core, 32GB mem...