This is really well presented. One further topic for ppl to investigate if they’re curious is the latent spaces that the vectors are encoded with, the distance functions in those LS, and how traditional lookups might suffer.
This presentation took away the complexity at comprehending what vector search engine is. I was totally lost on the concept but got a clear understanding of what Ai search engine is general. Thanks for sharing 😘. With Weaviate, vector search engine at scale certainty is 💯✍🏼
I just found your channel and you are awesome! Love the whisper at the end (yup, it's the main reason I'm leaving this comment 😃 such great personality), well done!
Very well explained! However, usually the first step in a search engine is a candidate selection step, that reduces the number of candidates with a faster algorithm. Performing similarity matching between your query and EVERY item in the database is too heavy to compute in 1-2 seconds.
How do transformers understand abbreviations or bigram/trigrams on a character level of words? I never understand how this preprocessing step is done, I only learned about the old elmo method
This is really well presented. One further topic for ppl to investigate if they’re curious is the latent spaces that the vectors are encoded with, the distance functions in those LS, and how traditional lookups might suffer.
This presentation took away the complexity at comprehending what vector search engine is. I was totally lost on the concept but got a clear understanding of what Ai search engine is general. Thanks for sharing 😘. With Weaviate, vector search engine at scale certainty is 💯✍🏼
Thanks!
Super sleek explanation
I just found your channel and you are awesome! Love the whisper at the end (yup, it's the main reason I'm leaving this comment 😃 such great personality), well done!
Thank you for this content.
Very informative! Thanks!
Also: Congrats on the 7k subs! 🎉
🎉
Excellent presentation! 🙌🏻
Such a great subject to cover! Congrats on the sponsor as well! Super well explained as always!
Your channel is so underrated! I'm a PhD student and this is so helpful!!
Thanks! This makes me so happy being a PhD student too!
Your channel is so useful, as someone who's beginning to work in the world of AI
Thanks! Making useful things is our goal. 😊
That is a brilliant explanation! Love it!!
Very well explained! However, usually the first step in a search engine is a candidate selection step, that reduces the number of candidates with a faster algorithm. Performing similarity matching between your query and EVERY item in the database is too heavy to compute in 1-2 seconds.
This is super helpful and very intuitively explained, finally understand vector databases! :)
wonderful!
+1
Cool
do u have any example with Vector Geo data ? please make video on that that
How do transformers understand abbreviations or bigram/trigrams on a character level of words? I never understand how this preprocessing step is done, I only learned about the old elmo method
Please give a Python code to search vector database.