TenaOS — Qdrant collection snapshots

Pre-built Qdrant collection snapshots so a new TenaOS operator can stand up a working knowledge base in one download instead of running the full offline indexing pipeline.

File Size Collection What's in it
who_msf_guidelines.snapshot ~449 MB who_msf_guidelines Embedded chunks of WHO + MSF clinical guidelines
ciel_concepts.snapshot ~327 MB ciel_concepts Embedded CIEL concept names + descriptions

Both collections are encoded with google/embeddinggemma-300m dense vectors + BM25 sparse vectors. The same two collections are queried by the kb-guidelines (:4276) and kb-ciel (:4277) daemons inside the TenaOS Docker image.

Restore on a fresh Qdrant

hf download beza4588/tenaos-qdrant-snapshots --local-dir ./snapshots --repo-type dataset

for c in who_msf_guidelines ciel_concepts; do
  curl -fsS -X POST "http://localhost:6333/collections/${c}/snapshots/upload" \
       -H "Content-Type: multipart/form-data" \
       -F "snapshot=@./snapshots/${c}.snapshot"
done

The TenaOS bootstrap script (scripts/fetch-models.sh) performs this restore automatically on first container boot.

Checksums

File SHA-256
who_msf_guidelines.snapshot 5ce04462aac136968878c449c038fb6f173818da199f5e2e64b1dc8f35877c96
ciel_concepts.snapshot 154db5e8761c689d7d628a46c57319edbebe2b60216acb0327f636521e950cc3

License

Apache 2.0 for the packaging. The underlying corpora retain their original licenses (WHO SMART Guidelines, MSF Clinical Guidelines, CIEL).

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support