The Vespa Engine

@vespaengine

The Vespa Engine on Twitter. Will post new blog posts from and important news about Vespa.

Vrijeme pridruživanja: rujan 2017.

Tweetovi

Blokirali ste korisnika/cu @vespaengine

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @vespaengine

  1. prije 4 sata

    Vespa improvements from January Among other things this includes new tensor functions needed to run BERT models. All of it already released and running in production of course.

    Poništi
  2. 21. sij
    Poništi
  3. proslijedio/la je Tweet
    18. pro 2019.

    One of many applications which uses in . Check out this presentation from AWS re:Invent 2019: How Verizon Media implemented push notification using Amazon DynamoDB. 150K/s push notifications, resolving of targets done by Vespa.

    Poništi
  4. 3. pro 2019.

    Getting started with machine-learned ranking using Vespa

    Poništi
  5. 28. stu 2019.

    If you have it's not too late to make yourself a state of the art and infinitely scalable e-commerce site in time for Black Friday. explains how in

    Poništi
  6. proslijedio/la je Tweet
    27. stu 2019.

    Working on new sample applications for this morning. A simple neural network trained with , exported in format and deployed to Vespa for serving. Text embeddings from Google's Universal Sentence Encoder. Work building on

    Poništi
  7. proslijedio/la je Tweet
    28. lis 2019.

    New / sample app which reproduces Google's paper on semantic retrieval for Q-A applications ( ) released today . 44% of 88K questions with the correct answer @ position 1 on SQuAD v1.1 train set

    Poništi
  8. 28. lis 2019.
    Poništi
  9. proslijedio/la je Tweet
    25. lis 2019.

    Comparing nearest neighbor ranking performance of versus using dense tensor dot product: Vespa 5x faster than Elastic on same hw, same data and queries. Try it out!

    Poništi
  10. 1. lis 2019.

    Newly released Vespa features: We added support for float tensors values, in addition to double. Obviously useful if you are memory bound, but we also measure up to 30% better ranking performance due to reduced memory bandwidth usage.

    Poništi
  11. 21. kol 2019.

    It's been a good summer for coding in the north 🌧️🌧️ We got some new features in Vespa released: BM25 ranking feature, searchable parents, tensor summary features and metric export

    Poništi
  12. proslijedio/la je Tweet
    19. srp 2019.

    Do you wish could be in your living room right now talking about Scalable Machine-Learned Model Serving? Not a problem! Follow this link to make your wish a reality:

    Poništi
  13. proslijedio/la je Tweet
    5. lip 2019.

    Applying machine learning in online applications requires solving the problem of model serving. At explains the problem and architectural solution, and shows how Vespa can be used to implement the solution. Read the full abstract here:

    Poništi
  14. 3. svi 2019.

    We're starting a new series of posts where we provide a complete and scalable Vespa application with a frontend for a popular use case. First out: Shopping sites!

    Poništi
  15. 24. tra 2019.

    Join us at in Dallas, TX, June 27.-29. where will talk about using Vespa to evaluate ML models over many data points with low latency.

    Poništi
  16. proslijedio/la je Tweet
    15. tra 2019.

    Applying machine learning in online applications requires solving the problem of model serving. At explains the problem and architectural solution, and shows how Vespa can be used to implement the solution. Read the full abstract here:

    Poništi
  17. proslijedio/la je Tweet
    10. tra 2019.

    With you can fine tune your term frequency contribution by custom occurrence tables & also built in rank-type for different type of text fields & . In addition the nativeRank of Vespa also includes proximity ranking.

    Poništi
  18. proslijedio/la je Tweet
    20. ožu 2019.

    Are you familiar with ? It is an service engine for processing data and making inferences at scale in . Learn with at

    Poništi
  19. 4. ožu 2019.
    Poništi
  20. 18. velj 2019.

    It turns out that when you open source 1.7M lines of code in 150 flat modules people will keep asking for directions. Today we're publishing a map to the code base.

    Poništi

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·