Why can't I find H36m_preprocess.py, doesn't this dataset need this step?

[stm233]

For human3.6M, you need to follow the processing step by officially way and then you will get the input data for our merge steo.