Big ANN @ NeurIPS’23 Practical Vector Search

olivia-moreira
Published  · 18 slides · 0 views
↓ Download
Big ANN @ NeurIPS’23 Practical Vector Search
1 / 1
Big ANN @ NeurIPS’23 Practical Vector Search - slide 1 of 18 Big ANN @ NeurIPS’23 Practical Vector Search - slide 2 of 18 Big ANN @ NeurIPS’23 Practical Vector Search - slide 3 of 18 Big ANN @ NeurIPS’23 Practical Vector Search - slide 4 of 18 Big ANN @ NeurIPS’23 Practical Vector Search - slide 5 of 18 Big ANN @ NeurIPS’23 Practical Vector Search - slide 6 of 18 Big ANN @ NeurIPS’23 Practical Vector Search - slide 7 of 18 Big ANN @ NeurIPS’23 Practical Vector Search - slide 8 of 18 Big ANN @ NeurIPS’23 Practical Vector Search - slide 9 of 18 Big ANN @ NeurIPS’23 Practical Vector Search - slide 10 of 18 Big ANN @ NeurIPS’23 Practical Vector Search - slide 11 of 18 Big ANN @ NeurIPS’23 Practical Vector Search - slide 12 of 18 Big ANN @ NeurIPS’23 Practical Vector Search - slide 13 of 18 Big ANN @ NeurIPS’23 Practical Vector Search - slide 14 of 18 Big ANN @ NeurIPS’23 Practical Vector Search - slide 15 of 18 Big ANN @ NeurIPS’23 Practical Vector Search - slide 16 of 18 Big ANN @ NeurIPS’23 Practical Vector Search - slide 17 of 18 Big ANN @ NeurIPS’23 Practical Vector Search - slide 18 of 18
Description
Big ANN NeurIPS23 Practical Vector Search Challenge Harsha Simhadri, Martin Aumüller, Dmitry Baranchuk, Matthijs Douze, Amir Ingber, Edo Liberty, Frank Liu , George Williams Microsoft, IT University of Copenhagen, Meta AI

Related Topics

Download this presentation From Below

"Big ANN @ NeurIPS’23 Practical Vector Search" is the property of its rightful owner. Permission is granted to download and print the materials on this website for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.

Presentation Transcript

01
Big ANN @ NeurIPS’23 Practical Vector Search Challenge Harsha Simhadri*, Martin Aumüller¤, Dmitry Baranchuk‡, Matthijs Douze†,
Amir Ingber*, Edo Liberty*, Frank Liu §, George Williams**
*Microsoft, ¤IT University of Copenhagen, †Meta AI Research, *Pinecone, ‡Yandex Labs, §Zilliz, **Ansible AI

Thanks to support from<br>
02
Why Approximate Nearest Neighbor Search? Semantic Search
Recommendation Systems
Security and threat analysis
…
RAG (Retrieval Augmented GenAI)
Ground the models in “facts”
Adapt to private and real-time data<br>
03
Many Algorithms. How to measure? Queries/sec(QPS)vs Recall@k
Latency, interactive & batch
Average Precision
DRAM and Disk Space
Power consumption
Indexing space and time ann-benchmarks.com; glove-100-angular<br>