def api_basic_agent_reset(game_type):
global basic_gs
basic_gs = TicTacToeGameState()
data = {"status": 200, "message": "Game State reseted", "game": "basic"}
return jsonify(data)
from agents import QuantumAgent, QuantumGroverAgent
from environments.tictactoe import TicTacToeGameState
from runners import run_to_the_end
if __name__ == "__main__":
gs = TicTacToeGameState()
agent0 = QuantumAgent()
agent1 = QuantumGroverAgent()
print(gs)
run_to_the_end([agent0, agent1], gs)
print(gs)
from agents import CommandLineAgent, DeepQLearningAgent, DDQNAgentWithPER
from environments.tictactoe import TicTacToeGameState
from runners import run_for_n_games_and_print_stats, run_step
if __name__ == "__main__":
gs = TicTacToeGameState()
agent0 = DeepQLearningAgent(action_space_size=gs.get_action_space_size(), neurons_per_hidden_layer=128,
hidden_layers=5)
agent1 = DDQNAgentWithPER(action_space_size=gs.get_action_space_size(), neurons_per_hidden_layer=128,
hidden_layers=5)
agent0.alpha = 0.1
agent0.epsilon = 0.005
agent1.alpha = 0.1
agent1.epsilon = 0.005
for i in range(100):
run_for_n_games_and_print_stats([agent0, agent1], gs, 1000)
agent0.epsilon = -1.0
agent1.epsilon = -1.0
run_for_n_games_and_print_stats([agent0, agent1], gs, 100)
gs_clone = gs.clone()
while not gs_clone.is_game_over():
run_step([agent0, CommandLineAgent()], gs_clone)
print(gs_clone)
gs_clone = gs.clone()
while not gs_clone.is_game_over():
run_step([CommandLineAgent(), agent1], gs_clone)
print(gs_clone)
from agents import CommandLineAgent, DeepQLearningAgent, PPOAgent, RandomAgent
from environments.tictactoe import TicTacToeGameState
from runners import run_for_n_games_and_print_stats, run_step
if __name__ == "__main__":
gs = TicTacToeGameState()
agent0 = PPOAgent(state_space_size=gs.get_vectorized_state().shape[0],
action_space_size=gs.get_action_space_size())
agent1 = RandomAgent()
for i in range(100):
run_for_n_games_and_print_stats([agent0, agent1], gs, 5000)
run_for_n_games_and_print_stats([agent0, agent1], gs, 100)
gs_clone = gs.clone()
while not gs_clone.is_game_over():
run_step([agent0, CommandLineAgent()], gs_clone)
print(gs_clone)
gs_clone = gs.clone()
while not gs_clone.is_game_over():
run_step([CommandLineAgent(), agent1], gs_clone)
print(gs_clone)
from agents import CommandLineAgent, DeepQLearningAgent
from environments.tictactoe import TicTacToeGameState
from runners import run_for_n_games_and_print_stats, run_step
if __name__ == "__main__":
gs = TicTacToeGameState()
agent0 = DeepQLearningAgent(action_space_size=gs.get_action_space_size())
agent1 = DeepQLearningAgent(action_space_size=gs.get_action_space_size())
agent0.alpha = 0.1
agent0.epsilon = 0.005
agent1.alpha = 0.1
agent1.epsilon = 0.005
for i in range(100):
run_for_n_games_and_print_stats([agent0, agent1], gs, 5)
agent0.epsilon = -1.0
agent1.epsilon = -1.0
run_for_n_games_and_print_stats([agent0, agent1], gs, 100)
gs_clone = gs.clone()
while not gs_clone.is_game_over():
run_step([agent0, CommandLineAgent()], gs_clone)
print(gs_clone)
gs_clone = gs.clone()
while not gs_clone.is_game_over():
run_step([CommandLineAgent(), agent1], gs_clone)
print(gs_clone)
from agents import RandomRolloutAgent, MOMCTSAgent, RandomAgent, HalfAlphaZeroAgent
from environments.tictactoe import TicTacToeGameState
from runners import run_to_the_end, run_for_n_games_and_print_stats
if __name__ == "__main__":
import tensorflow as tf
tf.compat.v1.disable_eager_execution()
gs = TicTacToeGameState()
agent0 = HalfAlphaZeroAgent(10, gs.get_action_space_size(), keep_memory=True)
agent1 = RandomAgent()
for _ in range(1000):
run_for_n_games_and_print_stats([agent0, agent1], gs, 100, shuffle_players=True)
from agents import RandomRolloutAgent, MOMCTSAgent, RandomAgent, ExpertApprenticeAgent
from environments.tictactoe import TicTacToeGameState
from runners import run_to_the_end, run_for_n_games_and_print_stats
if __name__ == "__main__":
import tensorflow as tf
tf.compat.v1.disable_eager_execution()
gs = TicTacToeGameState()
agent0 = ExpertApprenticeAgent(100, gs.get_action_space_size())
agent1 = RandomAgent()
for _ in range(1000):
run_for_n_games_and_print_stats([agent0, agent1], gs, 1000)
from agents import TabQLearningAgent, CommandLineAgent, RandomAgent
from environments.tictactoe import TicTacToeGameState
from runners import run_for_n_games_and_print_stats, run_step
if __name__ == "__main__":
gs = TicTacToeGameState()
agent0 = TabQLearningAgent()
agent1 = TabQLearningAgent()
agent0.alpha = 0.1
agent0.epsilon = 0.005
agent1.alpha = 0.1
agent1.epsilon = 0.005
for _ in range(100):
run_for_n_games_and_print_stats([agent0, agent1], gs, 5000)
agent0.epsilon = -1.0
agent1.epsilon = -1.0
run_for_n_games_and_print_stats([agent0, agent1], gs, 100)
gs_clone = gs.clone()
while not gs_clone.is_game_over():
run_step([agent0, CommandLineAgent()], gs_clone)
print(gs_clone)
gs_clone = gs.clone()
while not gs_clone.is_game_over():
run_step([CommandLineAgent(), agent1], gs_clone)
print(gs_clone)
from flask import Flask, render_template, jsonify, request
from runners import run_step
from agents import CommandLineAgent, QuantumAgent, WebAgent, QuantumGroverAgent
from environments.tictactoe import TicTacToeGameState
from runners import run_to_the_end
app = Flask(__name__, template_folder='static')
player_agent = WebAgent()
basic_agent = QuantumAgent()
grover_agent = QuantumGroverAgent()
basic_gs = TicTacToeGameState()
grover_gs = TicTacToeGameState()
# OLD SYNCHRONOUS CODE
# run_to_the_end([player_agent, basic_agent], basic_gs)
# run_to_the_end([player_agent, grover_agent], grover_gs)
######## FAST WEB ROOTING ##########
@app.route('/basic')
def basic_index():
data = {"title": "Basic Agent", "game": "basic"}
return render_template('html/index.html', data=data)
@app.route('/grover')
def grover_index():
data = {"title": "Grover Agent", "game": "grover"}
return render_template('html/index.html', data=data)