* update nccl2 dist train tips

* update

* update

* update

* update

* follow comments

* add anakin docker

* add

* add doc

*  delete docker

* Update index_anakin_ch.rst

* Update index_anakin_ch.rst

* Update index_anakin_ch.rst

* Update index_anakin_ch.rst

* Update index_anakin.rst

* Update anakin_parser_design_ch.md

* Update and rename anakin_parser_design_ch.md to anakin_parser_design.md

* Delete index_anakin_ch.rst

* Update anakin_parser_design.md

* update

* update

* Update anakin_parser_design.md

*  fix format

*  fix format

* fix format

* Delete .DS_Store

* Delete .DS_Store

* Delete menu.json

* Update anakin_parser_design.md

* change fluid to paddle

* Update anakin_arm_benchmark.md

* Update anakin_gpu_benchmark.md

* fix .md to .html

* fix .md to .html

* fix .md to .html

* update anakin docs

* update anakin docs

* Delete menu.json

* Delete int8_design_anakin.md

* update_en_structure

* remove doc/fluid/api/.DS_Store

* add windows compile, install guide support

* adjust the format

* correct the format
test=develop

* fix the style
test=develop

* fix issue
test=develop

* disable the windows cpp-reference doc temporarily
test=develop

* restore the inference doc
test=develop

* add async_training.rst

* add cpu_train_best_practice.rst

* clean code

* add doc for large_scale_sparse_feature_training

* add index for distribute train api guide

* follow_comments

* fix_link_format

* fix_link_format

* follow_comments_large_scale

* follow_comments

fix https://github.com/PaddlePaddle/FluidDoc/pull/400#discussion_r237335307

Add sparse update API guide.