rapirent / OpenAI-gym-MountainCar-v0

OpenAI-gym-MountainCar-v0

simple q-learning

files & directories

• fig/*: the figures of each experiment
• console_log/*: the console output of each experiment
• hw4.py

requirements

• python3.6
• tensorflow
• more details are writed as a requirements.txt file

usage

• run an experiments
  • use MODEL_NAME to specify size of model name
  • use TRAIN_EPISODE to specify train times
  • use STEP_LIMIT to specify the maximum possible action step in each episode
  • use STATE_NUM to specify the state number size of Q-matrix
  • use LEARNING_RATE to specify the learning rate

$ python3 hw4.py [-h] [-n MODEL_NAME] [-e TRAIN_EPISODE] [-l STEP_LIMIT] [-s STATE_NUM] [-lr LEARNING_RATE]

• You can use shell script to execute the models

$ bash command.sh

detail and report

You can refer to jupyter notebook report.ipynb

LICNESE

MIT