* solve prediction problems.

* fix docs.

* update dynabert

* update

* update

* update

* add transformer_pruner to autocompress

* add auto

* update transformer pruner

* ptq-hpo support eval function

* update

* update

* update

* update docs

* add comment

* fix patterns

* add an interface to predict the latency of compressed model

* just for rerun

* delete temporary model files
Co-authored-by: Nceci3 <ceci3@users.noreply.github.com>

* fix a pruner's bug

* support predict pruned model

* support predict pruned model

* support sparse model

* add interface for quick pruning

* add hardware list for latencyPredictor

* add ptq data-free method