MerchantryTidbits

rag-retrieval / library

Batched

Symptoms: too many sequential embed calls; provider rate limits on unbatched traffic.

Use it when

  • batch embedding requests to cut API cost
  • dynamic batching for inference callables

What it solves

Not the fit when

  • not a model host

Install

pip install batched

Invoke

Wrap embed or generate callables with batched.aio.BatchProcessor to cut cost-per-call under load.

Alternatives

No reviewed alternatives recorded yet.