* cherry-pick #5129 

* Add BERT. (#5125)

* Fix single data file in BERT pre-training. (#5127)

* add embedding example in embeddings.md

* add embedding.md link in toke_embedding readme

* fix hung

* add shuffle batch

* update

* reader_seed to shuffle_seed

* seed for shuffle batch

* add more English embedding name

* fix doc bug

* delete useless description

* add comments of TokenEmbedding

* add embedding model info

fix for pgl for the erniesage

* update __init__

* solve conflicts

* update couplet readme

* add transformer-xl for language model

* fix readme and change cross-entropy api

* change api back

fix erniesage for pgl 2.0 alpha 

* update __init__

* fix readme version info

* modify readme

* modify readme

* modify readme

* modify readme

* update codes

* update codes

* update codes

* update docs

* update docs

* update docs

* update docs

* add amp for ResNet200_vd

* change use_fp16 to use_amp

* fix script format

* change batch_size

* fix glue readme

* fix glue readme cased

* fix bleu bug

* fix bleu bug

* fix doc example

* minor fix

* fix squad dataset bug

* fix readme

* Add dependency to readme

* minor fix

* change run parameter

* fix readme

* Fix link error in PaddleNLP README.

* Fix link error in README_en.

* update codes

* update codes

* update codes

* update docs

* update docs

* add grad clip to pretrain, modify glue readme and improve glue score print

* add ernie-2.0 in run_glue.py and readme

fix url bug for word2vec

* add data.md

* backup code

* Add example for APIs.

* add readme.

* code backup.

* Modify the data.md

* fixed bug.

* Optimize the data.md