This is the code for the paper Revisiting Fundamentals of Experience Replay
by
William Fedus, Prajit Ramachandran, Rishabh Agarwal, Yoshua Bengio, Hugo
Larochelle, Mark Rowland and Will Dabney
All of the commands below are run from the parent google_research
directory.
Start a virtualenv with these commands:
virtualenv -p python3 .
source ./bin/activate
Then install necessary packages:
pip install -r experience_replay/requirements.txt
To train the agent execute,
python -m experience_replay.train \
--gin_files=experience_replay/configs/dqn.gin \
--schedule=continuous_train_and_eval \
--base_dir=/tmp/experience_replay \
--gin_bindings=experience_replay.replay_memory.prioritized_replay_buffer.WrappedPrioritizedReplayBuffer.replay_capacity=1000000 \
--gin_bindings=ElephantDQNAgent.oldest_policy_in_buffer=250000 \
--gin_bindings="ElephantDQNAgent.replay_scheme='uniform'" \
--gin_bindings="atari_lib.create_atari_environment.game_name='Pong'"
These correspond to the default hyperparameters. The replay ratio may be
adjusted by changing the oldest_policy_in_buffer
.