raghavanone

@ArthurZucker I have questions for some of the comments, request for more clarification .

@NielsRogge @ArthurZucker All the feedbacks has been incorporated.

@amyeroberts All the comments have been addressed, The failing test are unrelated to this PR. Let me know if I need to anything to fix them.

@ArthurZucker I wanted to work on this issue, I did little more digging and found out that this issue (difference in input_ids by the tokenizer) happens when is not followed...

Okay, Let me dig further in this direction.

@gante I would like to pick up this. Any pointers on where to start ?