Search results

Allow PPO to train in multi-agent games (more than just 1 opponent)

agents

enhancementNew feature or requestNew feature or request

ppo

training

Status: Open.

#51 In roeey777/Splendor-AI;

· roeey777 opened on Dec 24, 2024

DQN - add a DQN/DDQN/Rainbow agent

agents

dqn

enhancementNew feature or requestNew feature or request

Status: Open.

#27 In roeey777/Splendor-AI;

· roeey777 opened on Oct 12, 2024

Engine - support agent nicknames to shorten execution arguments

enhancementNew feature or requestNew feature or request

Status: Open.

#12 In roeey777/Splendor-AI;

· roeey777 opened on Sep 30, 2024

Genetic - use different representations (maybe ensambles or polynomial regression or even a MLP)

agents

enhancementNew feature or requestNew feature or request

genetic-algorithm

Status: Open.

#11 In roeey777/Splendor-AI;

· roeey777 opened on Sep 30, 2024

PPO2 - add a PPO2 agent

agents

enhancementNew feature or requestNew feature or request

ppo

Status: Open.

#7 In roeey777/Splendor-AI;

· roeey777 opened on Sep 30, 2024

PPO - log metrices using tensorboard instead/alongside of csv file

agents

enhancementNew feature or requestNew feature or request

logging

ppo

training

Status: Open.

#4 In roeey777/Splendor-AI;

· roeey777 opened on Sep 30, 2024

PPO - Use multi-processing & vectorized environments to speed-up the learning process

agents

enhancementNew feature or requestNew feature or request

ppo

training

Status: Open.

#2 In roeey777/Splendor-AI;

· roeey777 opened on Sep 30, 2024