Skip to content

Running Evaluating a policy

Glen edited this page Dec 5, 2017 · 1 revision

Often you may want to see the result of learning a policy or how a policy is currently doing during trianing. This is easy. All you need to do is run simModel.py with the same settings file used for training.

Examples:

  1. Sim a 2d walker
python3 simModel.py learn/settings/terrainRLImitate/CACLA/A_CACLA_DeepNN.json
  1. Sim a 3d walker
python3 simModel.py learn/settings/terrainRLImitate3D/CACLA/A_CACLA_DeepNN.json
  1. Sim a 3D walker with footstep targets
python3 simModel.py learn/settings/terrainRLImitate3D/CACLA/CACLA_DeepNNTanH.json
Clone this wiki locally