welcome back king!

ritvik big thanks that you came back! I'm preparing as ML-eng to Yandex for last 6 months, and your channel is literature best resource. Diamond. You re doing really big job here.

Great to see you back!

u r the best, please go on for ever, one of the great, if you understand it then you you can explain it, but this is not the case in most of the cases, thanks

Love it!

Love that you're doing RNN tutorials when transformers are the meta these days. You should look into attention btw

Question: I find it interesting that in most examples it is the output of the RNN, not the hidden state, is the input to the dense layer. You would think it should be the embedding that is used to represent the sentence. I understand they are only different by a matrix multiplication. Still, I wonder why it is more often than not that people choose the output rather than the embedding.

You mentioned a while back if anyone wanted to here about listwise learn to rank to speak up. I do! I do! I have a problem I know how to tackle with pairwise. Don’t know if listwise is something doable for my problem because I just can’t figure it all out! If I knew a little more, I might know if I could!

Really Elegant. I was thinking about LSTM layers, Are they only used in sequence to sequence problem??

welcome back! how to apply SVM model to classify an alpha data, to realize the detection of driver’s sleepless? Thank you

In this example, does vectorized layer just make one hot vector?

please do a video on transformers and GPT

what i believe is our ml are not as smart is they should be that is because it is not done by people like you, we need to improve our tensor understanding, you can do it, just see throu the depth ...... thanks which you may not need, regard less, is it understood that is what matters!!

use pytorch reeeeeeeeeeeee