Issues
- 0
Research and Develop Visualization Tools for Analyzing the Performance of the Self-play system.
#21 opened by gaoyuankidult - 0
Modify BrainParameters Python class to include new information regarding whether a brain is the "main" brain or "opponen" brain
#7 opened by Pete-York - 0
Modify / Create External Brain type to include whether a brain is the "main" or "opponent" brain
#9 opened by Pete-York - 0
Create scope string variable in C# side, pass it to Python BrainParameters class. In SelfPlayTrainerController replace models selectively
#11 opened by Pete-York - 0
Create a new SelfPlayTrainerController which inherits from TrainerController and adds the self-play functionality
#2 opened by Pete-York - 2
- 0
Create a tensorflow tf.train.Saver for each self-play opponent brain so they save only their own scope. This will be used for save / restore operations
#3 opened by Pete-York - 0
- 0
Refactor TrainerController so that start_learning function becomes more modular
#1 opened by Pete-York - 0
Create environment.
#10 opened by Pete-York - 0
Include Opponent Policy Change Interval (OPCI) and delta inside trainer_config.yaml
#8 opened by Pete-York - 0