/DY2P

This is the official repository for DY2P.

Primary LanguagePython

XXX (DY2P)

This repository is the official implementation of DY2P. Our implementation is based on SAC+AE.

Requirements

Required dependencies of this repo can be installed by running:

conda env create -f environment.yml

Then you can activate the environment by running:

source activate py3.6

Instructions

To train a DY2P agent on the cartpole swingup task from images, you can run:

python train.py \
            --domain_name cartpole  \
            --task_name swingup  \
            --action_repeat 8 \
            --save_tb \
            --seed 0 \
            --eval_freq 1250 \
            --batch_size 512 \
            --pre_transform_image_size 84 \
            --image_size 84 \
            --cody_lr 1e-4 \
            --results_dir ./logs \
            --time_step 2 \
            --omega_cody_loss 0.01 \
            --fc_output_logits True \
            --kl_use_target True \

or you can run the script for all six tasks:

bash train.sh