def main():
#env_id = "CartPole-v1"
vix_env = trading_vix_env.trading_vix_env()
num_cpu = 20 # Number of processes to use
# Create the vectorized environment
env = SubprocVecEnv([make_env(vix_env, i) for i in range(num_cpu)])
# Create log dir
log_dir = './ppo_data'
os.makedirs(log_dir, exist_ok=True)
env = VecMonitor(env, log_dir)
callback = custom_call_back.CustomCallback(check_freq = 1000,log_dir = log_dir)
model = PPO('MlpPolicy', env, verbose=1,n_steps=500,batch_size = 10000)
model.learn(total_timesteps=2500000000,callback = callback)
def main():
# Create log dir
log_dir = './ddpg_data'
os.makedirs(log_dir, exist_ok=True)
vix_env = trading_vix_env.trading_vix_env()
env = Monitor(vix_env, log_dir)
# Create action noise because TD3 and DDPG use a deterministic policy
n_actions = env.action_space.shape[-1]
action_noise = NormalActionNoise(mean=np.zeros(n_actions), sigma=0.1 * np.ones(n_actions))
# Create the callback: check every 20000 steps
callback = custom_call_back.CustomCallback(check_freq = 20000,log_dir = log_dir)
# Create RL model
model = DDPG('MlpPolicy',env,action_noise = action_noise, verbose=2,batch_size = 10000)
# Train the agent
model.learn(total_timesteps=int(5e9), callback=callback)