Skip to content

Latest commit

 

History

History
27 lines (18 loc) · 805 Bytes

README.md

File metadata and controls

27 lines (18 loc) · 805 Bytes
Raw

Stochastic Lower Bound Optimization

This is the TensorFlow implementation for the paper Algorithmic Framework for Model-based Deep Reinforcement Learning with Theoretical Guarantees. A PyTorch version will be released later.

Requirements

  1. rllab (commit number b3a2899)
  2. TensorFlow (== 1.9)
  3. NumPy (>= 1.14.5)
  4. Python 3.6

Run

Before running, please make sure that rllab and baselines are available

python main.py -c configs/algos/slbo.yml configs/envs/half_cheetah.yml -s log_dir=/tmp

If you want to change hyper-parameters, you can either modify a corresponding yml file or change it temporarily by appending model.hidden_sizes='[1000,1000]' in the command line.

License

See LICENSE for additional details.