@31:39 On this. I think, there is the last linear layer project the data into the number of videos to do the softmax. The weights of that layer associated with each video is the vector for each video. Intuitively, if the user vector has large dot product with this video vector, it will have large logit for the softmax thus most probably a match.
sorry for my audio quality I had the nice mic set up and was talking into it the whole time but zoom was set to receive audio input from my earpods....🤦♂️
You put the candidate retrieval and ranking model in the same machine(For example, using SM) Under the SM, user_id -> invoke ANN(db) to get candidates(a bunch of item_ids) -> invoke FS with item_id and user_id to get features separately -> invoke ranking model -> return a bunch of items with score in the sorted manner descendingly. Everything should be done within 200 ms p99
Also, how can we expose our candidate generation and ranking services via generic APIs, so other users can mix-and-match as required? We’ll want to consider these in the long-term roadmap. I'm wondering sh
Thanks for this wonderful talk! There is one point though which I would like to clear. At 14:50, when you talk about the request coming from a user, the user's browser history items is also seen to get the candidate sets. At that point of time, is the present item that a user is currently looking at is also being seen as the input?
Hi Eugene, Thanks for the great video. One question has been troubling me is that for recommendation engine why we can't simply use a GNN to generate user and item embeddings and then use a similarity method such as cosine or dot product to rank items vis a vis a classical two tower model. For all the user, item meta data and other user-item implicit interactions (click, purchase etc.) and other contextual ranking signals embeddings can be generated. These embeddings can be concatenated and then do a dot product with item to rank and serve online. Do you see any challenges in this. Pls advise on priority as I am preparing for an int. Thanks in advance.
this is a singaporean channel! nice to see singapore high quality youtube content!!!!!!!!!!!!
Really insightful. Thank you very much for putting the time and effort on the presentation. I really appreciated and learned from the video
This was an amazing presentation and there's a reason it's your most popular video now. Thank you.
Great content, no distractions, to the point. Thanks a lot.
IT WAS SO COOL AND INSIGHTFUL! MANY THANKS!
A minor correction: Skipgram already uses Negative Sampling @MLOps
Very great talk, lots of great explanations and diagrams all-in-one! Thanks for sharing!
Great talk! Thanks for sharing!
@31:39 On this. I think, there is the last linear layer project the data into the number of videos to do the softmax. The weights of that layer associated with each video is the vector for each video. Intuitively, if the user vector has large dot product with this video vector, it will have large logit for the softmax thus most probably a match.
Great details and examples, Eugene.
Great! Thanks for sharing
Thanks for watching
very interesting talk! thanks for sharing.
sorry for my audio quality I had the nice mic set up and was talking into it the whole time but zoom was set to receive audio input from my earpods....🤦♂️
Great Content Man
Great overview.
Great talk
You put the candidate retrieval and ranking model in the same machine(For example, using SM)
Under the SM,
user_id -> invoke ANN(db) to get candidates(a bunch of item_ids) -> invoke FS with item_id and user_id to get features separately -> invoke ranking model -> return a bunch of items with score in the sorted manner descendingly.
Everything should be done within 200 ms p99
Also, how can we expose our candidate generation and ranking services via generic APIs, so other users can mix-and-match as required? We’ll want to consider these in the long-term roadmap.
I'm wondering sh
Great session !!
This is gold!
Pretty clear and interesting!
Awesome interview
Eugene is a legend, great interview!
yes he is!
Sum of user scores for CFI2I and SWINGI2I should be at the nominator, please correct me if I am wrong.
Great content!
Thanks for this wonderful talk! There is one point though which I would like to clear. At 14:50, when you talk about the request coming from a user, the user's browser history items is also seen to get the candidate sets. At that point of time, is the present item that a user is currently looking at is also being seen as the input?
Which type of databases can be used for storing vetted content and ranking done through Deep Learning? Any video/article which recommends databases?
TPP = The Personalization Platform?
🎉🎉🎉❤❤❤
🔥
Is it possible to download this information from some resources?
Is it possible to get the slides? Great talk.
yess! eugeneyan.com/speaking/mlops-community-recsys/
Works better with .75 speed
how do I poo poo it lol what does it mean
Join us at our first in-person conference on June 25 all about AI Quality: www.aiqualityconference.com/
Very interesting talk! I like how questions are answered based on evidence/use-case i.e. how real time recommendation is a bit of an overkill.
Hi Eugene, Thanks for the great video. One question has been troubling me is that for recommendation engine why we can't simply use a GNN to generate user and item embeddings and then use a similarity method such as cosine or dot product to rank items vis a vis a classical two tower model. For all the user, item meta data and other user-item implicit interactions (click, purchase etc.) and other contextual ranking signals embeddings can be generated. These embeddings can be concatenated and then do a dot product with item to rank and serve online. Do you see any challenges in this. Pls advise on priority as I am preparing for an int. Thanks in advance.