jamesbriggs

8 ms·

31.
by jamesbriggs 5y ago
It will depend on your use-case, but primarily: (1) Pinecone uses dense vectors which can encode much more meaningful info, eg the actual 'semantic meaning' behind a sentence as we (people) would understand it, or the context in a
32.
by jamesbriggs 5y ago
Ideally I think it's useful to be familiar with Python and some NLP. Parts of the course can get reasonably in-depth, but we have tried to simplify every as much as possible and I think much of it can be followed without too much prere
33.
by jamesbriggs 5y ago
That's awesome, glad to hear it!
34.
by jamesbriggs 5y ago
Thanks!
35.
Show HN: I made a free course, 'NLP for Semantic Search'(pinecone.io)
239 points by jamesbriggs 5y ago|29 comments
36.
by jamesbriggs 5y ago
Hi all, I'm James, past data scientist, ML engineer, and now dev advocate at Pinecone. I've been dedicating much of the past few months to building a course on what I've found to be one of the most fascinating emerging techno
37.
by jamesbriggs 5y ago
Voronoi is super popular and typically is a component of solutions used in practice, we did write another article that covers what some of those might look like: https://www.pinecone.io/learn/composite-indexes/ TL
38.
by jamesbriggs 5y ago
I think Haystack is awesome but when the Faiss vector index starts getting too large their default 'Flat' index is too slow. I think a lot of people miss that they can fix that using custom Faiss indexes, which are set using `fais
39.
by jamesbriggs 5y ago
These models capture similarity on such a scale that they manage to learn patterns of language well, in terms of grammar they learn it indirectly - it is not manually coded, particularly the later models that we speak about. Our focus here
40.
by jamesbriggs 5y ago
A lot of websites scrape Medium without the authors having any idea, I assume it is another one of those
41.
by jamesbriggs 5y ago
Being from 2017 the article misses some of the coolest advances in semantic search, which is now pretty easy and lets you search in (almost) the same way they would ask a shop assistant when looking for something specific - "do you kno
42.
by jamesbriggs 5y ago
Also agree with this, it's almost like a marketing ploy - especially from OpenAI. They produce awesome stuff but things with GPT can get silly sometimes, like when they wouldn't release the larger versions because 'they were
43.
by jamesbriggs 5y ago
You also have plenty of comments which makes it much easier to follow than the Faiss code - I wish I had found this a few weeks ago, thanks for sharing.
44.
by jamesbriggs 5y ago
Yes your explanation is correct, if it hits an end block or overshoot the target it returns to the previous block in the next layer (not the start block), I submitted a change to the article and visual to make this clear - should be updated
45.
by jamesbriggs 5y ago
Yes I missed that, thanks for pointing it out!
46.
by jamesbriggs 5y ago
Yes that's why we wrote it! Other than in papers, finding info on HNSW is near-impossible so hopefully this will make a few lives easier. I came across this paper when researching for the article, I agree the link to brain networks is
47.
by jamesbriggs 5y ago
Hey I'm the author, feel free to ask me any questions!
48.
by jamesbriggs 5y ago
Yes as gk1 said, we're figuring out what can be said! I can't fully answer your questions beyond, having both indexes together helps us speed up the search (thanks to a lack of network/serialization/parsing overhead), an
49.
by jamesbriggs 5y ago
Hi all, I'm the author - I'll be here to answer any questions you have!
50.
by jamesbriggs 5y ago
Excalidraw.com for the style/graphs, then Adobe animate for the animation!
51.
by jamesbriggs 5y ago
Given the top 100 results produced by a flat index - how many of those are returned in the top 100 of our PQ/IVFPQ index. Although it's worth noting that this is at the lower end of performance for PQ, using the same Sift1M datase
52.
by jamesbriggs 5y ago
Excalidraw + Adobe Animate - I have never seen Excalidraw-animate before though so this could be new option!
53.
by jamesbriggs 5y ago
Welcome! Unfortunately faiss doesn't support sparse vectors. If you're using sparse vectors I would either not use faiss (there are better options like Elasticsearch) - or convert them into dense for use with faiss
54.
by jamesbriggs 5y ago
Hey all, I'm the author - feel free to ask me any questions about the article!