The procedure for increasing the token tolerance (token_index) and its impact

Hi team, we are currently in the same situation.
Do you plan to get back to increasing token tolerance?