25 lines
676 B
Python
25 lines
676 B
Python
"""
|
|
Machine learning utilities for Catan:
|
|
- Agents (random, policy-based)
|
|
- Observation/action encoders
|
|
- Trainers for reinforcement learning and evolutionary strategies
|
|
"""
|
|
|
|
from .agents import Agent, PolicyAgent, RandomAgent, finalize_action
|
|
from .encoding import encode_action, encode_observation
|
|
from .trainers import EvolutionStrategyTrainer, ReinforcementLearningTrainer
|
|
from .selfplay import PPOConfig, SelfPlayPPOTrainer
|
|
|
|
__all__ = [
|
|
"Agent",
|
|
"PolicyAgent",
|
|
"RandomAgent",
|
|
"encode_action",
|
|
"encode_observation",
|
|
"ReinforcementLearningTrainer",
|
|
"EvolutionStrategyTrainer",
|
|
"finalize_action",
|
|
"SelfPlayPPOTrainer",
|
|
"PPOConfig",
|
|
]
|