actor_critic.py

"""
solving pendulum using actor-critic model
"""

import gym
import numpy as np 
from keras.models import Sequential, Model
from keras.layers import Dense, Dropout, Input
from keras.layers.merge import Add, Multiply
from keras.optimizers import Adam
import keras.backend as K

import tensorflow as tf

import random
from collections import deque

# determines how to assign values to each state, i.e. takes the state
# and action (two-input model) and determines the corresponding value
class ActorCritic:
	def __init__(self, env, sess):
		self.env  = env
		self.sess = sess

		self.learning_rate = 0.001
		self.epsilon = 1.0
		self.epsilon_decay = .995
		self.gamma = .95
		self.tau   = .125

		# ===================================================================== #
		#                               Actor Model                             #
		# Chain rule: find the gradient of chaging the actor network params in  #
		# getting closest to the final value network predictions, i.e. de/dA    #
		# Calculate de/dA as = de/dC * dC/dA, where e is error, C critic, A act #
		# ===================================================================== #

		self.memory = deque(maxlen=2000)
		self.actor_state_input, self.actor_model = self.create_actor_model()
		_, self.target_actor_model = self.create_actor_model()

		self.actor_critic_grad = tf.placeholder(tf.float32, 
			[None, self.env.action_space.shape[0]]) # where we will feed de/dC (from critic)
		
		actor_model_weights = self.actor_model.trainable_weights
		self.actor_grads = tf.gradients(self.actor_model.output, 
			actor_model_weights, -self.actor_critic_grad) # dC/dA (from actor)
		grads = zip(self.actor_grads, actor_model_weights)
		self.optimize = tf.train.AdamOptimizer(self.learning_rate).apply_gradients(grads)

		# ===================================================================== #
		#                              Critic Model                             #
		# ===================================================================== #		

		self.critic_state_input, self.critic_action_input, \
			self.critic_model = self.create_critic_model()
		_, _, self.target_critic_model = self.create_critic_model()

		self.critic_grads = tf.gradients(self.critic_model.output, 
			self.critic_action_input) # where we calcaulte de/dC for feeding above
		
		# Initialize for later gradient calculations
		self.sess.run(tf.initialize_all_variables())

	# ========================================================================= #
	#                              Model Definitions                            #
	# ========================================================================= #

	def create_actor_model(self):
		state_input = Input(shape=self.env.observation_space.shape)
		h1 = Dense(24, activation='relu')(state_input)
		h2 = Dense(48, activation='relu')(h1)
		h3 = Dense(24, activation='relu')(h2)
		output = Dense(self.env.action_space.shape[0], activation='relu')(h3)
		
		model = Model(input=state_input, output=output)
		adam  = Adam(lr=0.001)
		model.compile(loss="mse", optimizer=adam)
		return state_input, model

	def create_critic_model(self):
		state_input = Input(shape=self.env.observation_space.shape)
		state_h1 = Dense(24, activation='relu')(state_input)
		state_h2 = Dense(48)(state_h1)
		
		action_input = Input(shape=self.env.action_space.shape)
		action_h1    = Dense(48)(action_input)
		
		merged    = Add()([state_h2, action_h1])
		merged_h1 = Dense(24, activation='relu')(merged)
		output = Dense(1, activation='relu')(merged_h1)
		model  = Model(input=[state_input,action_input], output=output)
		
		adam  = Adam(lr=0.001)
		model.compile(loss="mse", optimizer=adam)
		return state_input, action_input, model

	# ========================================================================= #
	#                               Model Training                              #
	# ========================================================================= #

	def remember(self, cur_state, action, reward, new_state, done):
		self.memory.append([cur_state, action, reward, new_state, done])

	def _train_actor(self, samples):
		for sample in samples:
			cur_state, action, reward, new_state, _ = sample
			predicted_action = self.actor_model.predict(cur_state)
			grads = self.sess.run(self.critic_grads, feed_dict={
				self.critic_state_input:  cur_state,
				self.critic_action_input: predicted_action
			})[0]

			self.sess.run(self.optimize, feed_dict={
				self.actor_state_input: cur_state,
				self.actor_critic_grad: grads
			})
            
	def _train_critic(self, samples):
		for sample in samples:
			cur_state, action, reward, new_state, done = sample
			if not done:
				target_action = self.target_actor_model.predict(new_state)
				future_reward = self.target_critic_model.predict(
					[new_state, target_action])[0][0]
				reward += self.gamma * future_reward
			self.critic_model.fit([cur_state, action], reward, verbose=0)
		
	def train(self):
		batch_size = 32
		if len(self.memory) < batch_size:
			return

		rewards = []
		samples = random.sample(self.memory, batch_size)
		self._train_critic(samples)
		self._train_actor(samples)

	# ========================================================================= #
	#                         Target Model Updating                             #
	# ========================================================================= #

	def _update_actor_target(self):
		actor_model_weights  = self.actor_model.get_weights()
		actor_target_weights = self.target_critic_model.get_weights()
		
		for i in range(len(actor_target_weights)):
			actor_target_weights[i] = actor_model_weights[i]
		self.target_critic_model.set_weights(actor_target_weights)

	def _update_critic_target(self):
		critic_model_weights  = self.critic_model.get_weights()
		critic_target_weights = self.critic_target_model.get_weights()
		
		for i in range(len(critic_target_weights)):
			critic_target_weights[i] = critic_model_weights[i]
		self.critic_target_model.set_weights(critic_target_weights)		

	def update_target(self):
		self._update_actor_target()
		self._update_critic_target()

	# ========================================================================= #
	#                              Model Predictions                            #
	# ========================================================================= #

	def act(self, cur_state):
		self.epsilon *= self.epsilon_decay
		if np.random.random() < self.epsilon:
			return self.env.action_space.sample()
		return self.actor_model.predict(cur_state)

def main():
	sess = tf.Session()
	K.set_session(sess)
	env = gym.make("Pendulum-v0")
	actor_critic = ActorCritic(env, sess)

	num_trials = 10000
	trial_len  = 500

	cur_state = env.reset()
	action = env.action_space.sample()
	while True:
		env.render()
		cur_state = cur_state.reshape((1, env.observation_space.shape[0]))
		action = actor_critic.act(cur_state)
		action = action.reshape((1, env.action_space.shape[0]))

		new_state, reward, done, _ = env.step(action)
		new_state = new_state.reshape((1, env.observation_space.shape[0]))

		actor_critic.remember(cur_state, action, reward, new_state, done)
		actor_critic.train()

		cur_state = new_state

if __name__ == "__main__":
	main()

I think there is a mistake in line 153.
Instead of self.critic_target_model.get_weights() it should be self.target_critic_model.get_weights()

Thanks Nikunj !

…

----- Mail original ----- De: "Nikunj" <notifications@github.com> À: "yashpatel5400" <yashpatel5400@noreply.github.com> Cc: "Alex Boulangé" <aboul@free.fr>, "Manual" <manual@noreply.github.com> Envoyé: Jeudi 23 Mai 2019 15:35:11 Objet: Re: yashpatel5400/actor_critic.py Yes, Alex, the script is running for me. The update functions (both) are actually unutilized in the script, so not posing an issue. The results, however, can be improved considerably by some modifications to the code, for example, using the update actor and critic functions and maybe by playing a bit with the hyperparameters (number of layers in the actor/critic model, etc.). Also, the script runs forever right now (while True: ) though there is mention of num_trials and trial_len . Try using these variables to break the learning at your desired behaviour of policy. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub , or mute the thread .

Hi Nikunj, finaly imo this is not a mistake
( I'm working on a fork, I'll share it soon, with tau implementation)

Why isn't update_target(in line 159) executed in the main function? I think that if this is not done, the target isn't updated.
I'm glad if you tell me the reason.

Why the target is not updated at the end, also the training main loop seems to be wrong.I think there should be a double for loop for the training step. Also, I think the tau value is not used for target.

The _update_actor_target() and _update_critic_target() function has not been called anyhow in the main loop. Is is an error ?? Please review the code.