rlworkgroup / garage

50%

master: 64%

LAST BUILD ON BRANCH fix_vvpg_v

branch: fix_vvpg_v

CHANGE BRANCH

x

Reset

• fix_vvpg_v
• 2018.10
• Automated_Benchmarking_normalized
• GautamDev
• IndexedSlices_warning
• add_baseline
• add_changelog
• add_cnn
• add_cnn_progress
• add_dict_support
• add_erwr
• add_init
• add_param
• add_tnpg
• add_vanilla_loss
• add_wrappers
• add_wrappers_progress
• automated-benchmarking
• backport-415
• backport-418
• backport-426
• backport-434
• backport-434-446
• backport-544
• backport-pr-418
• backport_622_201810
• backup_exp_runner
• bc
• behavioral_cloning
• behavioral_cloning_ag
• benchmark
• box2d-install
• box2d_install_linux
• bugs_ryan
• bump_gym
• changelog_2018101
• check_modules
• clean_mujoco_conv_import
• codecov
• copy_actions
• dead_code
• decouple_sampler_from_algo
• deterministicMLP
• dm_comtrol-mujoco_1.5
• dqn
• dqn_2
• dqn_benchmark
• dqn_debug
• dqn_testing
• dqn_tf
• dyn_rand_nested
• epilson_greedy
• epsilon_greedy
• eric-torch
• example_shebang
• exp_runner
• fix-install
• fix-logging
• fix-scripts
• fix_copy_mjkey
• fix_dm_control
• fix_docker_ci
• fix_failed_tests
• fix_flake8_exclude
• fix_gcc_osx
• fix_lasagne_import
• fix_lbfgs
• fix_mujoco_import
• fix_off_policy_sampler
• fix_pickled_openai
• fix_pol_ent
• fix_pol_re
• fix_pol_std
• fix_precommit
• fix_replay_buffer
• fix_sampler
• fix_setup_osx
• fix_sleeping_proc
• fix_test_quirks
• fix_travis_tests
• fix_vpg_ddpg
• flake8-refactor
• flaky_tests
• formater
• garage_docker
• garage_experiment
• garage_model
• ignore_egg_info
• ignore_flake8_D107
• kill_stub
• latest_reps
• logger_decouple_tensorboard
• make_test
• maml
• master
• meta_learning
• mjpro200
• model_benchmark
• model_fix_pickle
• models
• models_embed
• models_gautam
• models_new
• models_progress
• models_ryan
• move_dm_tests
• move_nb_utils
• multitask_env
• new-logger
• new_dm_control
• newer-logger
• no_docker_opengl
• not_scripts
• ppo-test
• pr420
• pre_commit-flake8
• product_to_tuple
• randomized_env_test
• refactor_imports
• release-2018.10
• remove-most-envs
• remove_gym_dupes
• remove_horizon
• remove_layer_ddpg_test
• remove_layers
• remove_layers_lstm
• remove_mjkey_check
• remove_sawyer
• remove_theano
• rename_macos
• reps
• revert-430-backport-418
• setup_py
• tb_images
• td3
• test_bug
• test_bug_wip
• test_keras
• test_keras_local
• tf112
• theano-gpu
• torch
• torch-rlkit
• tsan
• uint8_support
• uint8_test
• unused-deps
• update-contributing
• update-docs
• update_gitignore
• v2018.10
• v2018.10.0
• v2018.10.1
• viskit_flake8

Build # 2845

Build Type

push

travis-ci-com

Committed by CatherineSue

Commit Message

Add miscellaneous changes

 - Make the replay buffer base class garage.replay_buffer.ReplayBuffer
abstract. Rename the regular replay buffer to SimpleReplayBuffer.
 - Delete enums and pass ReplayBuffer class into algorithm. Now use can
construct a replay buffer by their own.
 - Move `configure_dims` and `dims_to_shapes` into replay buffer
constructor. Delete `env.reset()` in configure_dims. Notice the reason
that we need to use `env.step()` because it's the only way to get
env_info in gym.GoalEnv. And the env_info is used to compute reward in
`env.compute_reward`. But the source code in gym shows that the env_info
is not actually used when computing reward. So we can not store info in
transitions.
 - Set default value for optimizer params in DDPG.
 - Remove commented code.

WIP:
 - Redesign reuse in ContinuousMLPPolicy.
 - Redesign OffPolicyBatchPolopt.

Run Details

6541 of 13192 relevant lines covered (49.58%)

0.5 hits per line