Open timbektu opened 1 year ago
…st (batch_size) dimension, and the code breaks when using batch_size =1. pytorch lacks a functionality squeeze multiple dims at once, so using reshape() instead. patched the bug in this commit.
…st (batch_size) dimension, and the code breaks when using batch_size =1. pytorch lacks a functionality squeeze multiple dims at once, so using reshape() instead. patched the bug in this commit.