def train_test_agent():
M = 10
env = GraphSamplingEnv(max_samples=M)
num_train_graphs = 10
agent = BaseAgent(env=env)
agent.learn(num_train_graphs)
agent.test()
def run(args):
M = 5
env = GraphSamplingEnv(max_samples=M)
agent = BaseAgent(env=env)
now = datetime.now()
logger.configure(
dir=f"./results/fixed_env/{now.strftime(TIMESTAMP_FORMAT)}")
agent.learn()
agent.test()
def run(args):
M = 3
env = GraphSamplingEnv(max_samples=M)
agent = BaseAgent(
env=env,
gamma=args["gamma"],
learning_rate=args["learning_rate"],
replay_buffer_size=args["replay_buffer_size"],
exploration_schedule_steps=args["exploration_schedule_steps"],
exploration_initial_prob=args["exploration_initial_prob"],
exploration_final_prob=args["exploration_final_prob"],
random_walk_sampling_args=SAMPLING_ARGS)
now = datetime.now()
logger.configure(dir=LOGDIR + f"{now.strftime(TIMESTAMP_FORMAT)}")
agent.learn()
agent.test()
from envs import GraphSamplingEnv
from agents import BaseAgent
# def train_test_agent():
print ("here")
M = 10
env = GraphSamplingEnv(max_samples=M)
num_train_graphs = 10
agent = BaseAgent(env=env)
agent.learn()#num_train_graphs)
agent.test()
# if __name__ == "__main__":
# train_test_agent()