Self Attention with torch.nn.MultiheadAttention Module

[@SUDIPTODAS-r9l]

Got rid of the jargon , straight to the point , great tutorial

[@23232323rdurian]

Thank you! I've been trying to understand that math unsuccessfully for a long time.....seen lots of videos, but somehow yours explained best

[@mrdbourke]

Fantastic explanation, thank you very much!

[@saculzemog]

Very clear explanation. Well done

[@wolfisraging]

Loving it! Thanks a lot for the video!!!!!!

[@NONAME_G_R_I_D_]

All I needed tbh!! Thanks

[@figueraxiyana9411]

excellent, please keep uploading videos

[@zjp957]

Thank you for the explanation !

[@Alan-hs2wn]

love you, thank you so much

[@user-wr4yl7tx3w]

this was really helpful.

[@ahmedchaoukichami9345]

wow thank u so much good work

[@yimingxiao1033]

great explaination thanks a lot

[@yuchengli8009]

I have a question that why the different dimension of matrix can add together. For example, the 3x2 matrix add the bias 2x1?

[@yuchengli8009]

@@machinelearningwithpytorch hello there, such as @5:00

[@yuchengli8009]

yes, you are correct that 3x2 with 1x2

[@haneensuradi]

You do broadcasting

[@hx-vy1hn]

Thanks! Please add a Patreon account to help us fund your work.

[@JohnCena12355]

Nice video!

[@서로워]

Can you explain sparse attention? please please

[@ridwansalahudeen7621]

Excellent! You have a very sound comprehension of the module... How can I contact you?

[@rafaelgp9072]

Amazing

[@wishswiss]

thanks!

[@jaivalani4609]

What is called E here, last step is not understood , outW

[@cptmazi]

What?!! how do you add a 3x2 matrix to a 1x2 vector ?!!!

[@marcod6653]

it's a simple column-wise addition. So each column of the second matrix is added to all the elements in the same column of the first matrix