Switch to branch another_sparse_branch_ppo to run the experiments on our model. The code in master is used to run experiments on the vanilla maml-trpo model.
To avoid any conflict with your existing Python setup, and to keep this project self-contained, it is suggested to work in a virtual environment with virtualenv
. To install virtualenv
:
pip install --upgrade virtualenv
Create a virtual environment, activate it and install the requirements in requirements.txt
.
virtualenv venv
source venv/bin/activate
pip install -r requirements.txt
You can use the main.py
script in order to run reinforcement learning experiments. This script was tested with Python 3.5.
python main.py --env-name HalfCheetahRandVelEnv-v1 --fast-batch-size 20 --meta-batch-size 40 --output-folder hcv-1 --num-workers 16 --embed-size 32 --exp-lr 7e-4 --baseline-type nn --nonlinearity tanh --num-layers-pre 1 --hidden-size 64 --seed 0
A huge part of this implementation is borrowed from the MAML implementation of tristandeleu/pytorch-maml-rl in Pytorch. A huge thanks to them for open-sourcing their implementation.