I want to implement LSTM
input_size: 10x512x7x7 (DataType: double)
output_size: 512x7x7 (DataType: double)
Do I need embedding layer, I heard embedding is used to convert word_to_number…
Could you explain more about embedding layer in LSTM…?