abstracts-search

abstracts-search is a project about indexing 200+ million academic publications into a single semantic search engine. The method behind it is to take the publicly-available abstracts in the OpenAlex dataset, generate embeddings using the stellaen1.5Bv5 model, then train a fast index using faiss. Furthermore, this can be repeated to keep the index synced with the quarterly snapshots of the OpenAlex dataset.

The project is split into four repositories:

• abstracts-search: An orchestration Makefile and the helper program oa_jsonl.
• abstracts-embeddings: The raw embeddings (released under CC0) as a Hugging Face Dataset.
• abstracts-index: The search interface as a Hugging Face Space (also released under CC0).
• abstracts-faiss: The trained index as a Hugging Face Dataset (also released under CC0).

Additionally, this project loads sidecar-search (originally spun off from this project), a set of CLI build tools for building sidecar indexes.

Running

An Internet connection is always needed. All data associated with the publications (titles, abstracts, authors, etc) is retrieved from OpenAlex and is not provided with this project. abstracts-faiss is only a trained index that maps query embeddings to directly to OpenAlex IDs, ranked by semantic similarity. These IDs are then used to query OpenAlex.

For running the search interface alone, without regular sync (last synced on March 30th, 2026 as of the time of writing), the only repo that needs to be cloned is abstracts-index:

git clone https://huggingface.co/spaces/colonelwatch/abstracts-index
cd abstracts-index
python3 -m venv .venv source .venv/bin/activate pip install -r requirements.txt
MODELNAME="NovaSearch/stellaen1.5Bv5" PROMPTNAME="s2pquery" TRUSTREMOTECODE=1 python3 app.py

Building

Building of abstracts-embeddings was originally done on a Google Cloud machine with four L4 GPUs over one week, with all subsequent building done on a machine with 16 GB of RAM (plus 16 GB swap), an RTX 3060 12GB, and a 2 TB scratch disk, so altogether this poses the minimum requirements for now. That initial build can be bypassed by using abstracts-embeddings as a base.

Ensure hf (Hugging Face CLI), jq, and mbuffer are available. Then, start with the following setup.

sudo nvidia-persistenced
git clone https://github.com/colonelwatch/abstracts-search cd abstracts-search
python3 -m venv .venv source .venv/bin/activate pip install -r requirements.txt

• To proceed by building from the abstracts-embeddings repository as a base, and run the following recovery commands.

hf download colonelwatch/abstracts-embeddings --repo-type dataset --local-dir abstracts-embeddings
make recover

• Otherwise, immediately proceed to the next step.

Finally, build the index by running the following make command.

SIDECARSEARCHMODEL="NovaSearch/stellaen1.5Bv5" SIDECARSEARCHTRUSTREMOTE_CODE=1 make BUILDFLAGS='-b 32' DUMPFLAGS='--shard-size 2097152 --row-group-size 65536' TRAINFLAGS='-N -c 65536'

Syncing

To sync, rerun the above make command.