I'm struggling to get my head around how the output shape of an LSTM layer formed. How is the output unit value physically implemented in the layer?

For example, if I have an input shape of (128 x 6) with return_sequence=true andunits=10. How is the raw output of each LSTM unit converted into a (128x10) shaped output?


This is pretty accurate Understanding LSTM Weights

TL;DR, the output shape is 3D with Return sequences True, because output for each timestamp is returned and the output shape is 2D with Return sequences False, because output by last timestamp or LSTM unit is returned

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.