* fix hung

* add shuffle batch

* update

* reader_seed to shuffle_seed

* seed for shuffle batch

* add the benchmark glue for some task

* fix the qnil result for the bert benchmark

* change the from the logger to framework logger

* fix translation dataset and simplify transformer reader

Co-authored-by: Nwawltor <fanzeyang0904@hotmail.com>

* update the readme for the benchmark

* udpate README file for the bert benchmark

* uddate the readme format for the benchmark bert
Co-authored-by: Nwawltor <fanzeyang0904@hotmail.com>
Co-authored-by: Nfangzeyang <fangzeyang@baidu.com>

* add custom op gelu

* refine run_pretrain

* add the support of amp for the fleet train
* remove the attribute for the amp

Co-authored-by: Nfangzeyang <fangzeyang@baidu.com>