def main(): parser = argparse.ArgumentParser() parser.add_argument('--bind-address', default='127.0.0.1') parser.add_argument('--port', '-p', type=int, default=5000) parser.add_argument('--pg-agent') parser.add_argument('--predict-agent') parser.add_argument('--q-agent') parser.add_argument('--ac-agent') args = parser.parse_args() bots = {'mcts': mcts.MCTSAgent(800, temperature=0.7)} if args.pg_agent: bots['pg'] = agent.load_policy_agent(h5py.File(args.pg_agent)) if args.predict_agent: bots['predict'] = agent.load_prediction_agent( h5py.File(args.predict_agent)) if args.q_agent: q_bot = rl.load_q_agent(h5py.File(args.q_agent)) q_bot.set_temperature(0.01) bots['q'] = q_bot if args.ac_agent: ac_bot = rl.load_ac_agent(h5py.File(args.ac_agent)) ac_bot.set_temperature(0.05) bots['ac'] = ac_bot web_app = httpfrontend.get_web_app(bots) web_app.run(host=args.bind_address, port=args.port, threaded=False)
def main(): parser = argparse.ArgumentParser() parser.add_argument('--bind-address', default='127.0.0.1') parser.add_argument('--port', '-p', type=int, default=5000) args = parser.parse_args() zero_agent = { 'zero': zero.load_zero_agent(h5py.File("bots/zeroagent_v1.hdf5", 'r')) } web_app = httpfrontend.get_web_app(zero_agent) web_app.run(host=args.bind_address, port=args.port, threaded=False)
def main(): samp = 1000 epo = 1 # tag::e2e_processor[] timestr = time.strftime("%Y%m%d-%H%M%S") model_h5filename = "./agents/deep_bot_" + timestr + "_s" + str( samp) + "e" + str(epo) + ".h5" go_board_rows, go_board_cols = 19, 19 nb_classes = go_board_rows * go_board_cols encoder = XPlaneEncoder((go_board_rows, go_board_cols)) data_dir = "data/" + str(encoder.num_planes) + "-planes" processor = GoDataProcessor(encoder=encoder.name(), data_directory=data_dir) X, y = processor.load_go_data(num_samples=samp) # end::e2e_processor[] # tag::e2e_model[] input_shape = (encoder.num_planes, go_board_rows, go_board_cols) model = Sequential() network_layers = large.layers(input_shape) for layer in network_layers: model.add(layer) try: with tf.device('/device:GPU:0'): model.add(Dense(nb_classes, activation='softmax')) model.compile(loss='categorical_crossentropy', optimizer='adadelta', metrics=['accuracy']) model.fit(X, y, batch_size=128, epochs=epo, verbose=1) # end::e2e_model[] # tag::e2e_agent[] deep_learning_bot = DeepLearningAgent(model, encoder) deep_learning_bot.serialize(h5py.File(model_h5filename, "w")) # end::e2e_agent[] # tag::e2e_load_agent[] model_file = h5py.File(model_h5filename, "r") bot_from_file = load_prediction_agent(model_file) web_app = get_web_app({'predict': bot_from_file}) web_app.run() # end::e2e_load_agent[] except RuntimeError as e: print(e)
def main(): workdir = '//home/nail//Code_Go//checkpoints//' os.chdir(workdir) bind_address = '127.0.0.1' port = 5000 predict_agent, pg_agent, q_agent, ac_agent = '', '', '', '' agent_type = input('Агент(pg/predict/q/ac = ').lower() if agent_type == 'pg': pg_agent = input( 'Введите имя файла для игры с ботом политика градиентов =') pg_agent = workdir + pg_agent + '.h5' if agent_type == 'predict': predict_agent = input( 'Введите имя файла для игры с ботом предсказания хода =') predict_agent = workdir + predict_agent + '.h5' if agent_type == 'q': q_agent = input( 'Введите имя файла для игры с ботом ценность действия =') q_agent = workdir + q_agent + '.h5' if agent_type == 'ac': ac_agent = input('Введите имя файла для игры с ботом актор-критик =') ac_agent = workdir + ac_agent + '.h5' bots = {'mcts': mcts.MCTSAgent(800, temperature=0.7)} if agent_type == 'pg': bots['pg'] = agent.load_policy_agent(h5py.File(pg_agent, 'r')) if agent_type == 'predict': bots['predict'] = agent.load_prediction_agent( h5py.File(predict_agent, 'r')) if agent_type == 'q': q_bot = rl.load_q_agent(h5py.File(q_agent, 'r')) q_bot.set_temperature(0.01) bots['q'] = q_bot if agent_type == 'ac': ac_bot = rl.load_ac_agent(h5py.File(ac_agent, 'r')) ac_bot.set_temperature(0.05) bots['ac'] = ac_bot web_app = httpfrontend.get_web_app(bots) web_app.run(host=bind_address, port=port, threaded=False)
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2' # Include the path to the local version of dlgo. dir_path = os.path.dirname(os.path.realpath(__file__)) sys.path.insert(0, dir_path) from dlgo.agent.predict import DeepLearningAgent, load_prediction_agent from dlgo.httpfrontend import get_web_app # load model model_file = h5py.File("betago/agents/predict_net.hdf5", "r") bot_from_file = load_prediction_agent(model_file) model_file.close() # build app app = get_web_app({'predict': bot_from_file}) # add favicon route @app.route("/favicon.ico") def favicon(): return send_from_directory(os.path.join(dir_path, "static"), "favicon.ico", mimetype="image/vnd.microsoft.icon") # add home page route @app.route('/') def index(): return redirect(url_for("static", filename="play_predict_19.html"))
def main(): bot = mcts.MCTSAgent(700, temperature=1.4) web_app = httpfrontend.get_web_app({'mcts': bot}) web_app.run()
def main(): bot = mcts.MCTSAgent(700, temperature=1.4) web_app = httpfrontend.get_web_app(bot, BOARD_SIZE) web_app.run()
import sys sys.path.append('../') from dlgo.agent import load_prediction_agent, load_policy_agent, AlphaGoMCTS from dlgo.rl import load_value_agent import h5py from dlgo import httpfrontend # 13.4 end fast_policy = load_prediction_agent( h5py.File('alphago_sl_policy_e20_2k.h5', 'r')) strong_policy = load_policy_agent(h5py.File('alphago_rl_policy_e20_2k.h5', 'r')) value = load_value_agent(h5py.File('alphago_value_e20_2k.h5', 'r')) alphago = AlphaGoMCTS(strong_policy, fast_policy, value) # HTTP FRONTEND bots = {} bots['predict'] = alphago web_app = httpfrontend.get_web_app(bots) web_app.run(threaded=False) # # TODO: implement GTP frontend