After training process, how to run dqn_stages with the save_model?
is0437if opened this issue · comments
For Turtlebot [ROS 1] Machine Learning: https://emanual.robotis.com/docs/en/platform/turtlebot3/machine_learning/
After training your stages, a model is saved in the save_model directory. For example: If I trained dqn_stage_4 for around 3000 episodes, it saves the trained model in "save_model" folder in several formats such as JSON and h5. Whenever trying to run the dqn_stage_4, it starts from the beginning. If I wanna test my trained model, how can I load the JSON or h5 saved model into code so that I can avoid training from the beginning every time running the code?
@ROBOTIS-David @ROBOTIS-Will
Hi is0437if, thank you for getting my attention to you.
I would like to transfer this issue to our TurtleBot3 issue section as this is a inquiry as to TurtleBot3.
@ROBOTIS-Will or @ashekim is going to help your issue, but you might as well fill the template so that a member in the team can give you a better solution to your inquiry.
Thank you :)
Thanks @ROBOTIS-David for forwarding my issue here! I am eagerly waiting for having a reply from @ashekim @ROBOTIS-Will