Open fsh2017 opened 3 years ago
Mean-pooling is used.
Sorry, I am not familiar with the use of mean pooling. On the S2VT data set, In the process of extracting 2D features, the shape of NPY is (28,1000), and the shape of H5 is (1536,). How to do mean pooling? Thank you for your answer!
Sorry for my late reply, it seems that you are using the output of the last layer (i.e, the layer for classification), while the feature should come from one layer before.
I want to know what's going on in the process of converting data from NPY to H5PY? They don't have the same shape.