* added experience replay and n-step

* implementing distributional q value

* added distributional q-value

* added overview in qac_dist and d4pg

* derived D4PG from DDPG

* fixed a bug when action shape >1

* benchmark D4PG mujoco + minor fixs

-entry for DDPG mujoco
-entry for D4PG mujoco
-config for D4PG mujoco
-fixed style D4PG code
-unittests for QAC distributional

* formatted code

* minor updates (read description)

-added d4pg seria_entry test
-updated comments in QACDIST
-added d4pg in commander register
-added q_value in d4pg return dict
-added priority update in d4pg entry
-added assertion in QACDIST

* feature(zym): add offlineRL algo Discrete CQL.

* feature(zym): add offlineRL algo Discrete CQL; add hdf5 dataset for offlineRL.

* style(nyz): add ctools.pysc2 import in smac env

* feature(nyz): add smac docker build(enable docker, smac docker)

* fix(nyz): fix if condition syntax in deploy(enable docker, smac docker)

* fix(nyz): fix if condition syntax in deploy(enable docker, smac docker)

* fix(nyz): remove cache layer in smac docker(enable docker, smac docker)

* feature(nyz): use self-hosted runner in docker smac deploy(enable docker, smac docker)

* feature(nyz): build smac docker manually(enable docker, smac docker)

* feature(nyz): use docker buildx as default tool in smac and add SC2Map in setup(enable docker, smac docker)

* feature(nyz): add __init__.py in smac env maps(enable docker, smac docker)

* feature(nyz): add docker_mujoco build and upgrada numpy version to 1.20.0(enable docker)

* fix(nyz): fix numpy version compatibility bug and add -y option in apt-get(enable docker)

* fix(nyz): add libosmesa6-dev in Dockerfile.env(enable docker)

* fix(nyz): add permanent env variable about mujoco(enable docker)

* fix(nyz): change sh source to .(enable docker)

* fix(nyz): set env variable in bashrc(enable docker)

* fix(nyz): fix pip typo(enable docker)

* fix(nyz): add env in dockerfile(enable docker)

* add wqmix

* update annotation

* reformate

* update annotation

* update config

* fix annotation

* update as review

* fix as review

* add 5m6m MMM MMM2 config

* reformate

* fix(pu): fix rnn reset bug and add unittest

* fix(pu): fix rnn reset bug in centrally-weighted wqmix

* style(pu): yapf format and let WQMIXPolicy extend QMIXPolicy

* fix(pu): fix wqmix policy extend bug

* test(pu): add unittest test_wqmix

* fix(pu): fix mixer key bug in particle config

* feature(pu): add cooperative_navigation_wqmix_config

* style(pu): yapf format

* test(pu): change nn.Identity() to nn.Sequential()

* fix(pu): fix unittest bug in test_wqmix

* feature(nyz): move atari_py to ale-py and polish standard docker build(enable docker)

* fix(nyz): fix atari env import bug(enable docker)

* feature(nyz): add autorom install in docker(enable docker)

* feature(nyz): split base and env docker build(enable docker)

* fix(nyz): fix docker env source image bug(enable docker)

* add md ppo

* add doc string

fix(xjx): fix the catch statments that will never succeed in test networks; fix silence method (#71)

* style(nyz): add docker deploy workflow(enable docker)

* style(nyz): fix docker push info(enable docker)

* style(nyz): modify org name and image default name rule(enable docker)

* style(nyz): change default version to date(enable docker)

* fix(hansbug): try support spawn backend

* fix(hansbug): try fix the xxxxing problem in interaction spawn support && reformat the code style

* fix(nyz): disable silence decorator for spawn context interaction test
Co-authored-by: HansBug <killog@126.com>

* feature(wyh):mappo and ippo win rate and time

* feature(wyh):mappo and ippo epymarl win rate and time

* feature(wyh):smac epymarl commit id
Co-authored-by: N卫昱宏 <SENSETIME\weiyuhong@cn0214000504l.domain.sensetime.com>

* fix(wyh): plot function

* fix(wyh): plot function pytest

* fix(wyh):plot function modify comments

* feature(wyh):plot style
Co-authored-by: Nweiyuhong <weiyuhong@sensetime.com>

* start implementing bsuite env

* add bsuite env

* Implemented

* removed unused file

* added cartpole_swing environment

* Update test_bsuite_env.py

* added env in readme and in setup.py

* Create bsuite.png

* update md_dqn

* update offpolicy ppo

* add rainbow md policy

* format code

* del ppo; leave to future updates

* add doc string; fix rainbow returns

* test rnd

* fix mz config

* fix config

* fix(pu): fix r2d2

* fix(puyuan): fix r2d2

* feature(puyuan): add minigrid r2d2 config

* polish minigrid config

* modified as review

* fix(pu): fix bugffor compatibility

* polish(pu): add annotations and polish slice operation

* style(pu): run format.sh

* style(pu): correct yapf format

* fix(pu): fix config

* fix(pu): fix done slice bug and lstm reset bug

* style(pu): format config

* polish(pu): polish config params for cartpole, lunarlander and minigrid

* polish(pu): polish minigrid config params

* Update r2d2.py

* polish(pu): polish rnn reset problem

* fix(pu): fix merge error

* polish(pu): polish cartpole config

* polish(nyz): polish cartpole r2d2 config for faster convergence

* test(nyz): enable r2d2 algotest
Co-authored-by: Nniuyazhe <niuyazhe@sensetime.com>

* fix_formatted_config_bug_eval

* fix(wyh):add config pytest

* feature(nyz): add trueskill as league metric, naive elo calculator, fix game_env info bug

* fix(nyz): fix league player mutate bug

* fix(nyz): fix league unittest bug

* feature(nyz): add elo ranking in league metric env

* polish(nyz): modify fixed eval policy and trueskill init

* add_scheduler_module

* fix_change_range_and_factor

* cooldown_counter_bug_fix

* add_div_mode

* code_format_fixed

* fix_pr_bug

* add_unnitest_module

* add_patience_test

* polish(nyz): polish scheduler design and fix league mode scheduler bug

* fix(nyz): fix merge test_metric.py bug
Co-authored-by: Nniuyazhe <niuyazhe@sensetime.com>
Co-authored-by: N李可 <like2@CN0014008466M.local>

* fix_mappo_bug_masknan_and_dict_cannot_unsqueeze

* squeeze_bug

* test dijob

* test: wait for dijob Succeeded phase, and read coordinator logs

* test: update wait condition

* ci: update algo_test.yaml and flake check

* test: move kubernetes package to where it will be used

* feat: add k8s launcher

* feat: install kubectl when install k3d

* feat: add orchestrator launcher and a test case

* ci: install kubernetes related package and cli

* style: format code

* style: flake check code

* test k8s launcher

* ci: change back to unit test

* feat: delete cert manager when delete orchestrator

* style: flake8 check

* feat: merge k8s-launcher with k8s-helper

1. merge k8s-launcher with k8s-helper
2. move kubernetes package import to where it will be used
3. hack/install-k8s-tools.sh -> ding/scripts/install-k8s-tools.sh