mabirck / modular_DeepRL

Attempt to implement A2C and PPO algorithm with modular properties of Maxout and LWTA. # UNFINISHED AND FAILED

modularReinforcementLearning

CONTINUOUS TASKS

#A2C python3 main.py --env-name "RoboschoolHumanoid-v1" "RoboschoolHumanoidFlagrun-v1" "RoboschoolHumanoidFlagrunHarder-v1" --num-stack 1 --num-frames 1000000 --act_func "maxout"

#PPO python3 main.py --env-name "RoboschoolHumanoid-v1" "RoboschoolHumanoidFlagrun-v1" "RoboschoolHumanoidFlagrunHarder-v1" --algo ppo --use-gae --vis-interval 1 --log-interval 1 --num-stack 1 --num-steps 2048 --num-processes 1 --lr 3e-4 --entropy-coef 0 --ppo-epoch 10 --num-mini-batch 32 --gamma 0.99 --tau 0.95 --num-frames 1000000 --act_func "tanh" --seed 22