Benchmarking Deep Reinforcement Learning for Continuous Control

Authors: Yan Duan, Xi Chen, Rein Houthooft, John Schulman, Pieter Abbeel

ICML 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We report novel findings based on the systematic evaluation of a range of implemented reinforcement learning algorithms.
Researcher Affiliation Collaboration University of California, Berkeley, Department of Electrical Engineering and Computer Sciences Ghent University i Minds, Department of Information Technology Open AI
Pseudocode No The paper describes algorithms but does not provide pseudocode or algorithm blocks.
Open Source Code Yes Both the benchmark and reference implementations are released at https://github.com/ rllab/rllab in order to facilitate experimental reproducibility and to encourage adoption by other researchers.
Open Datasets Yes The benchmark and reference implementations are available at https: //github.com/rllab/rllab, allowing for the development, implementation, and evaluation of new algorithms and tasks.
Dataset Splits No The paper mentions 'Hyperparameter Tuning' and selecting tasks for grid search but does not specify dataset splits (e.g., train/validation/test splits) for reproduction.
Hardware Specification No The paper does not specify the hardware used for experiments.
Software Dependencies No The paper mentions using Box2D (Catto, 2011) and Mu Jo Co (Todorov et al., 2012) physics simulators but does not provide specific version numbers for these or other software dependencies.
Experiment Setup Yes In this section, we elaborate on the experimental setup used to generate the results. For the DDPG algorithm, we used the hyperparametes reported in Lillicrap et al. (2015). For the other algorithms, we follow the approach in (Mnih et al., 2015), and we select two tasks in each category, on which a grid search of hyperparameters is performed. Each choice of hyperparameters is executed under five random seeds.