/Reinforcement-Learning-Assitant-Coach-System

Primary LanguageJupyter NotebookMIT LicenseMIT

AutoML algorithm search for different sparsity and latency

Dependencies

  • pyyaml == 5.3.0
  • torch == 1.8.0
  • torchvision == 0.9.0
  • tensorflow >=2.4.0

Pattern sets

  • You can modify any pruning patterns in the rl_input.py and admm.py, admm40.py, admm60.py.

How to run the experiment

cd AutoML
python rl_controller.py

-If you want to run the autoML on har dataset

cd AutoML/har_all
python rl_controller.py