I have two csv files. One with around 700 columns is properly loaded in Tensorflow Input Pipelines, but the second one with 700 same and around 100 extra columns is not loaded p
