* add VisualDL as summary logging backend

* yapf

* copyright

* using visualdl as the default logging tool

* only support linux at the moment

* yapf

* yapf

* remove dependency on six

* unittest

* yapf

* replace tensorboard with summary to support VDL in the future

* unittest

* rename keys for record

* yapf

* fix a number of bug

* update comments

* paddle license

* revert the change on client_not_init_test.py

* fix timeout bug & unnitest

* add unit test for potential bugs

* yapf

* unit test

* logger

* logger#2

* fix job's bug that will cause dead-lock

* fix bug

* add missing files

* solve hanging

* fix comments; exit job completely

* hanging problem

* hanging problem#2

* hanging_program#3

* CMAKE

* remove unused variables

* add tensorboardX

* yapf

* Code License

* add test for import parl

* fix error of check_style

* change pip source

* refine naming

* add IMPALA algorithm and some common utils

* update README.md

* refactor files structure of impala algorithm; seperate numpy utils from utils

* add hyper parameter scheduler module; add entropy and lr scheduler in impala

* clip reward in atari wrapper instead of learner side; fix codestyle

* add benchmark result of impala; refine code of impala example; add obs_format in atari_wrappers

* Update README.md

* sync paras in program, fix deepcopy bug, python3 compatibility

* refactor code, add plutil directory, clean import order

* remove old comment

* refine comment

* fix codestyle

* cache sync program, add gputils module, refine model_base unittest

* fix codestyle

* refine sync params cache

* add fetch_value module

* redesign basic class in PARL

* code style fixed

* update yaml's version

* update yaml's version & update code to fix style problem

* add debug message for  function

* delete test code

* rename function: has_fun -> has_func

* prelimary implementations of ComputationTask, Algorithm and Model classes

* remove "model_func" from the args of an algorithm

* a clean clone() function for Algorithm and Model

* add use_next_value as a input to learn()

* further re-structure

* added Feedforward and RLAlgorithm classes

* maxid -> argmax

* discrete_distribution -> category_distribution

* category -> categorical

* revisions