hi, in the notebook of C5W3A1, I think there a couple of things wrong in section 2.1:
- in the architecture, the plot in the left panel shows the outputs, but we have
y<1>twice. - The explanation reads
The post-attention LSTM at time 't' only takes the hidden state ๐ โจ๐กโฉ and cell state ๐โจ๐กโฉ as input.but it should actually be๐ โจ๐ก-1โฉand๐โจ๐ก-1โฉ.
Hope that helps!
Thanks