convert back the u8 bytes to utf8

a random split into training and test sets

A proc-block takes a passage and a question as input and gives us BERT Encoding in form of input_ids, input_masks, segment_ids as output

It parses the text from a start logit to an end logit

Read an input from the environment.