train.py

#   Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

import gym
import numpy as np
import os
import queue
import time
import threading
import parl
from atari_model import AtariModel
from atari_agent import AtariAgent
from parl.env.atari_wrappers import wrap_deepmind
from parl.utils import logger, summary, get_gpu_count
from parl.utils.scheduler import PiecewiseScheduler
from parl.utils.time_stat import TimeStat
from parl.utils.window_stat import WindowStat
from parl.utils import machine_info

from actor import Actor


class Learner(object):
    def __init__(self, config):
        self.config = config
        self.sample_data_queue = queue.Queue(
            maxsize=config['sample_queue_max_size'])

        #=========== Create Agent ==========
        env = gym.make(config['env_name'])
        env = wrap_deepmind(env, dim=config['env_dim'], obs_format='NCHW')
        obs_shape = env.observation_space.shape

        act_dim = env.action_space.n

        model = AtariModel(act_dim)
        algorithm = parl.algorithms.IMPALA(
            model,
            sample_batch_steps=self.config['sample_batch_steps'],
            gamma=self.config['gamma'],
            vf_loss_coeff=self.config['vf_loss_coeff'],
            clip_rho_threshold=self.config['clip_rho_threshold'],
            clip_pg_rho_threshold=self.config['clip_pg_rho_threshold'])
        self.agent = AtariAgent(algorithm, obs_shape, act_dim,
                                self.learn_data_provider)

        if machine_info.is_gpu_available():
            assert get_gpu_count() == 1, 'Only support training in single GPU,\
                    Please set environment variable: `export CUDA_VISIBLE_DEVICES=[GPU_ID_TO_USE]` .'

        self.cache_params = self.agent.get_weights()
        self.params_lock = threading.Lock()
        self.params_updated = False
        self.cache_params_sent_cnt = 0
        self.total_params_sync = 0

        #========== Learner ==========
        self.lr, self.entropy_coeff = None, None
        self.lr_scheduler = PiecewiseScheduler(config['lr_scheduler'])
        self.entropy_coeff_scheduler = PiecewiseScheduler(
            config['entropy_coeff_scheduler'])

        self.total_loss_stat = WindowStat(100)
        self.pi_loss_stat = WindowStat(100)
        self.vf_loss_stat = WindowStat(100)
        self.entropy_stat = WindowStat(100)
        self.kl_stat = WindowStat(100)
        self.learn_time_stat = TimeStat(100)
        self.start_time = None

        self.learn_thread = threading.Thread(target=self.run_learn)
        self.learn_thread.setDaemon(True)
        self.learn_thread.start()

        #========== Remote Actor ===========
        self.remote_count = 0

        self.batch_buffer = []
        self.remote_metrics_queue = queue.Queue()
        self.sample_total_steps = 0

        self.create_actors()

    def learn_data_provider(self):
        """ Data generator for fluid.layers.py_reader
        """
        while True:
            sample_data = self.sample_data_queue.get()
            self.sample_total_steps += sample_data['obs'].shape[0]
            self.batch_buffer.append(sample_data)

            buffer_size = sum(
                [data['obs'].shape[0] for data in self.batch_buffer])
            if buffer_size >= self.config['train_batch_size']:
                batch = {}
                for key in self.batch_buffer[0].keys():
                    batch[key] = np.concatenate(
                        [data[key] for data in self.batch_buffer])
                self.batch_buffer = []

                obs_np = batch['obs'].astype('float32')
                actions_np = batch['actions'].astype('int64')
                behaviour_logits_np = batch['behaviour_logits'].astype(
                    'float32')
                rewards_np = batch['rewards'].astype('float32')
                dones_np = batch['dones'].astype('float32')

                self.lr = self.lr_scheduler.step()
                self.entropy_coeff = self.entropy_coeff_scheduler.step()

                yield [
                    obs_np, actions_np, behaviour_logits_np, rewards_np,
                    dones_np,
                    np.float32(self.lr),
                    np.array([self.entropy_coeff], dtype='float32')
                ]

    def run_learn(self):
        """ Learn loop
        """
        while True:
            with self.learn_time_stat:
                total_loss, pi_loss, vf_loss, entropy, kl = self.agent.learn()

            self.params_updated = True

            self.total_loss_stat.add(total_loss)
            self.pi_loss_stat.add(pi_loss)
            self.vf_loss_stat.add(vf_loss)
            self.entropy_stat.add(entropy)
            self.kl_stat.add(kl)

    def create_actors(self):
        """ Connect to the cluster and start sampling of the remote actor.
        """
        parl.connect(self.config['master_address'])

        logger.info('Waiting for {} remote actors to connect.'.format(
            self.config['actor_num']))

        for i in range(self.config['actor_num']):
            self.remote_count += 1
            logger.info('Remote actor count: {}'.format(self.remote_count))
            if self.start_time is None:
                self.start_time = time.time()

            remote_thread = threading.Thread(target=self.run_remote_sample)
            remote_thread.setDaemon(True)
            remote_thread.start()

    def run_remote_sample(self):
        """ Sample data from remote actor and update parameters of remote actor.
        """
        remote_actor = Actor(self.config)

        cnt = 0
        remote_actor.set_weights(self.cache_params)
        while True:
            batch = remote_actor.sample()
            self.sample_data_queue.put(batch)

            cnt += 1
            if cnt % self.config['get_remote_metrics_interval'] == 0:
                metrics = remote_actor.get_metrics()
                if metrics:
                    self.remote_metrics_queue.put(metrics)

            self.params_lock.acquire()

            if self.params_updated and self.cache_params_sent_cnt >= self.config[
                    'params_broadcast_interval']:
                self.params_updated = False
                self.cache_params = self.agent.get_weights()
                self.cache_params_sent_cnt = 0
            self.cache_params_sent_cnt += 1
            self.total_params_sync += 1

            self.params_lock.release()

            remote_actor.set_weights(self.cache_params)

    def log_metrics(self):
        """ Log metrics of learner and actors
        """
        if self.start_time is None:
            return

        metrics = []
        while True:
            try:
                metric = self.remote_metrics_queue.get_nowait()
                metrics.append(metric)
            except queue.Empty:
                break

        episode_rewards, episode_steps = [], []
        for x in metrics:
            episode_rewards.extend(x['episode_rewards'])
            episode_steps.extend(x['episode_steps'])
        max_episode_rewards, mean_episode_rewards, min_episode_rewards, \
                max_episode_steps, mean_episode_steps, min_episode_steps =\
                None, None, None, None, None, None
        if episode_rewards:
            mean_episode_rewards = np.mean(np.array(episode_rewards).flatten())
            max_episode_rewards = np.max(np.array(episode_rewards).flatten())
            min_episode_rewards = np.min(np.array(episode_rewards).flatten())

            mean_episode_steps = np.mean(np.array(episode_steps).flatten())
            max_episode_steps = np.max(np.array(episode_steps).flatten())
            min_episode_steps = np.min(np.array(episode_steps).flatten())

        metric = {
            'sample_steps': self.sample_total_steps,
            'max_episode_rewards': max_episode_rewards,
            'mean_episode_rewards': mean_episode_rewards,
            'min_episode_rewards': min_episode_rewards,
            'max_episode_steps': max_episode_steps,
            'mean_episode_steps': mean_episode_steps,
            'min_episode_steps': min_episode_steps,
            'sample_queue_size': self.sample_data_queue.qsize(),
            'total_params_sync': self.total_params_sync,
            'cache_params_sent_cnt': self.cache_params_sent_cnt,
            'total_loss': self.total_loss_stat.mean,
            'pi_loss': self.pi_loss_stat.mean,
            'vf_loss': self.vf_loss_stat.mean,
            'entropy': self.entropy_stat.mean,
            'kl': self.kl_stat.mean,
            'learn_time_s': self.learn_time_stat.mean,
            'elapsed_time_s': int(time.time() - self.start_time),
            'lr': self.lr,
            'entropy_coeff': self.entropy_coeff,
        }

        for key, value in metric.items():
            if value is not None:
                summary.add_scalar(key, value, self.sample_total_steps)

        logger.info(metric)


if __name__ == '__main__':
    from impala_config import config

    learner = Learner(config)
    assert config['log_metrics_interval_s'] > 0

    while True:
        time.sleep(config['log_metrics_interval_s'])

        learner.log_metrics()