spaCy and Explosion: past, present & future

Transcript

1. spaCy and Explosion: past, present & future Matthew Honnibal
 Ines

   Montani Explosion

2. 1982 - 2014 Before spaCy

3. 1982 - 2014 Before spaCy Some time before 2005

4. 1982 - 2014 Before spaCy PhD 2005-2010 (Sydney) Some time

   before 2005

5. 1982 - 2014 Before spaCy PhD 2005-2010 (Sydney) ACL 2009

   (Singapore) Some time before 2005

6. July 2014 First commit on spaCy

7. Early 2015 spaCy is first released

8. 2015 First collaborations • demos & visualizers like displaCy •

   first concepts of a modern approach to NLP annotation tools

9. 2015 First collaborations “Baskets” concept • demos & visualizers like

   displaCy • first concepts of a modern approach to NLP annotation tools

10. 2015 First collaborations “Baskets” concept Binary annotation tool concept •

    demos & visualizers like displaCy • first concepts of a modern approach to NLP annotation tools

11. Early 2016 First highlights: sense2vec

12. Early 2016 First highlights: German model • first non-English model

    • non-projective dependencies • developed by Wolfgang Seeker

13. Late 2016 Explosion • new company for AI developer tools

    • bootstrapped through consulting 
 for the first 6 months • funded through software sales 
 since 2017 • 100% independent and profitable

14. Late 2016 Explosion • new company for AI developer tools

    • bootstrapped through consulting 
 for the first 6 months • funded through software sales 
 since 2017 • 100% independent and profitable Our bets about NLP • NLP won’t just be a cloud API • number of developers will increase • annotation is better in-house

15. 2017 spaCy v2.0 • shift to deep learning • smaller

    and updatable models • custom pipeline components • custom extension attributes • built-in text classification • built-in displaCy visualizers • many other improvements Thinc, spaCy’s machine learning library

16. July 2017 neuralcoref by Hugging Face Community package for coreference


    resolution with spaCy

17. Late 2017 Prodigy • first commercial product • modern annotation

    tool • fully scriptable in Python

18. Late 2017 Prodigy • first commercial product • modern annotation

    tool • fully scriptable in Python

19. Late 2017 Prodigy • first commercial product • modern annotation

    tool • fully scriptable in Python users 2,000+ 250+ companies incl.

20. July 2018 10,000 stars on GitHub

21. , Early 2019 spaCy v2.1 • transfer learning and pretraining

    • 2-3 times faster tokenization • enhanced match pattern API • built-in rule-based NER • many other improvements

22. , Early 2019 spaCy v2.1 • transfer learning and pretraining

    • 2-3 times faster tokenization • enhanced match pattern API • built-in rule-based NER • many other improvements Transfer learning • better models with less data – huge win! • how to adapt for spaCy without bigger (and slower) models? • spacy pretrain is a pretty cool compromise

23. April 2019 Free interactive online course course.spacy.io

24. So what’s next 
 for spaCy?

25. July 2019 Explosion team Matthew Ines Montani Honnibal

26. July 2019 Explosion team Matthew Ines Montani Justin DuJardin Honnibal

27. July 2019 Explosion team Matthew Ines Montani Sebastián Ramírez Guadalupe

    Romero Giannis Daras Justin DuJardin Van Landeghem Sofie Honnibal

28. Today spaCy IRL • 2 sold-out corporate training days •

    200+ conference attendees

29. What’s next? spaCy v3.0 • morphological features • entity linking

    • non-entity span tagging • static analysis of processing pipeline and its components

30. What’s next? spaCy v3.0 Vision for spaCy • focus on

    data structures and pipeline • build support for new tasks even if we don’t have a model • make sure it’s easy to BYO model • keep shipping good defaults • morphological features • entity linking • non-entity span tagging • static analysis of processing pipeline and its components

31. What’s next? spaCy v3.0 Vision for spaCy • focus on

    data structures and pipeline • build support for new tasks even if we don’t have a model • make sure it's easy to BYO model • keep shipping good defaults What’s out-of-scope? • anything generative: summarization, machine translation, etc. • multi-modal: audio, video, etc. • research assistance: plenty of good frameworks for developing novel techniques • morphological features • entity linking • non-entity span tagging • static analysis of processing pipeline and its components

32. , What’s next? spaCy ecosystem in your cloud • whole

    systems, not just libraries • programmable, extensible cluster • running under your control • automated setup, good defaults • full data privacy – we don’t want your data!

33. , What’s next? spaCy ecosystem in your cloud • whole

    systems, not just libraries • programmable, extensible cluster • running under your control • automated setup, good defaults • full data privacy – we don’t want your data! Processing with Dask

34. , What’s next? spaCy ecosystem in your cloud • whole

    systems, not just libraries • programmable, extensible cluster • running under your control • automated setup, good defaults • full data privacy – we don’t want your data! Prodigy Scale Processing with Dask

35. commits 10,000+ contributors 300+ 60+ extension packages 13,500+ GitHub stars

    2,800+ closed issues 80+ releases

36. Thank you!