search

harsha-simhadri / big-ann-benchmarks

Framework for evaluating ANNS algorithms on billion scale datasets.
https://big-ann-benchmarks.com
MIT License
341 stars 114 forks source link

pinecone's sparse algorithm #264

Closed ingberam closed 8 months ago

ingberam commented 9 months ago

pinecone's sparse algorithm Expected results on the standard Azure VM: 8k QPS (public), 7.5k QPS (private) for recall=0.9 (for sparse-full). @harsha-simhadri can you run and update the leaderboard plots? Thanks. running time < 10min.

harsha-simhadri commented 8 months ago

pinecone_smips,pinecone_smips,sparse-full,10,8152.069513110283,0.0,-1.0,14008096.0,0.0,0.0,sparse,0.8926074498567335 pinecone_smips,pinecone_smips,sparse-full,10,8083.498770890919,0.0,-1.0,14008096.0,0.0,0.0,sparse,0.9019340974212036 pinecone_smips,pinecone_smips,sparse-full,10,8008.758767543293,0.0,-1.0,14008096.0,0.0,0.0,sparse,0.909541547277937 pinecone_smips,pinecone_smips,sparse-full,10,7899.129073983437,0.0,-1.0,14008096.0,0.0,0.0,sparse,0.9141833810888252 pinecone_smips,pinecone_smips,sparse-full,10,7808.109720157484,0.0,-1.0,14008096.0,0.0,0.0,sparse,0.917593123209169 pinecone_smips,pinecone_smips,sparse-full,10,7705.006165059365,0.0,-1.0,14008096.0,0.0,0.0,sparse,0.9208166189111747 pinecone_smips,pinecone_smips,sparse-full,10,7602.261838576096,0.0,-1.0,14008096.0,0.0,0.0,sparse,0.9236532951289398 pinecone_smips,pinecone_smips,sparse-full,10,7518.281790453895,0.0,-1.0,14008096.0,0.0,0.0,sparse,0.9251432664756447