- Weaviate Newsletter
- Posts
- Weaviate newsletter - Week 17
Weaviate newsletter - Week 17
Dear Weaviate follower,
This week I want to share a new podcast and some small updates... Enjoy :)
Weaviate Updates
π€« Soon to come: Support Filtering by Timestamp and Batch-Delete by Query.
Weaviate Podcast
π Weaviate Podcast #14: Yi-Lin Sung, Jaemin Cho and Mohit Bansal, a research team from UNC, present their incredible work on VL-Adapter
Other Weaviate News
πΉ Interesting session about Open Source Databases with Laura Ham and more on dev.to.
π Curious about this weeks hot topic benchmarking? Read this thread from Nils Reimers and unravel.
π Meet us on the Knowledge Graph Conference next week in New York.
Have a great week,
Donna
A helpful explainer from the creator of SBERT himself!
It appears research and understanding of Vector DBs and embeddings is still nascent even among leading NLP organizations... π
Hard problems are the best ones to solve!! Let's go @weaviate_io πππ
β Rajko RadovanoviΔ (@rajko_rad)
5:42 PM β’ Apr 23, 2022
An ASR system I helped build just transcribed "vector search" as "better search". I did not do anything, it is all AI.
β Dmitry Kan (@DmitryKan)
4:47 PM β’ Apr 9, 2022
Since benchmarking random d=128 vectors is the big topic today: efConstr=512, ef=256, m=64 leads to 93% Recall@10 in @WeAviate. Mean latency of ~12ms with p99 of ~14ms. W/ 30 CPUs that's ~2550 QPS including object retrieval from disk, HTTP request, client-side parsing, etc.
β Etienne Dilocker (@etiennedi)
2:37 PM β’ Apr 22, 2022