fix phase & format save_step output information

w2v infer bug

test=develop, bug fix for w2v, youtube_dnn'readme

* fix selected_gpus bug for local_cluster

* fix bugs about single machine multicards training for local_cluster

* fix python -m config.yaml -b backend.yaml

* re add selected_gpus_num config for single train mode

* fix bug for submit for paddlecloud k8s
Co-authored-by: Nwuzhihua <35824027+fuyinno4@users.noreply.github.com>

* update readme

* change readme

* change some fuzzy describe

* change

* add some notes
Co-authored-by: Ntangwei12 <tangwei12@baidu.com>
Co-authored-by: Nwuzhihua <35824027+fuyinno4@users.noreply.github.com>

[mmoe] fix phases problems

Gru4rec

add custom_read.md

Co-authored-by: Nwuzhihua <35824027+fuyinno4@users.noreply.github.com>

change simnet and dssm data