from typing import Union, Optional, List, Any, Tuple import os import torch import numpy as np from ditk import logging from functools import partial from tensorboardX import SummaryWriter from copy import deepcopy from ding.envs import get_vec_env_setting, create_env_manager from ding.worker import BaseLearner, InteractionSerialEvaluator, BaseSerialCommander, create_buffer, \ create_serial_collector from ding.config import read_config, compile_config from ding.policy import create_policy from ding.utils import set_pkg_seed from .utils import random_collect, mark_not_expert def serial_pipeline_r2d3( input_cfg: Union[str, Tuple[dict, dict]], expert_cfg: Union[str, Tuple[dict, dict]], seed: int = 0, env_setting: Optional[List[Any]] = None, model: Optional[torch.nn.Module] = None, expert_model: Optional[torch.nn.Module] = None, max_train_iter: Optional[int] = int(1e10), max_env_step: Optional[int] = int(1e10), ) -> 'Policy': # noqa """ Overview: Serial pipeline r2d3 entry: we create this serial pipeline in order to\ implement r2d3 in DI-engine. For now, we support the following envs\ Lunarlander, Pong, Qbert. The demonstration\ data come from the expert model. We use a well-trained model to \ generate demonstration data online Arguments: - input_cfg (:obj:`Union[str, Tuple[dict, dict]]`): Config in dict type. \ ``str`` type means config file path. \ ``Tuple[dict, dict]`` type means [user_config, create_cfg]. - seed (:obj:`int`): Random seed. - env_setting (:obj:`Optional[List[Any]]`): A list with 3 elements: \ ``BaseEnv`` subclass, collector env config, and evaluator env config. - model (:obj:`Optional[torch.nn.Module]`): Instance of torch.nn.Module. - expert_model (:obj:`Optional[torch.nn.Module]`): Instance of torch.nn.Module.\ The default model is DQN(**cfg.policy.model) - max_train_iter (:obj:`Optional[int]`): Maximum policy update iterations in training. - max_env_step (:obj:`Optional[int]`): Maximum collected environment interaction steps. Returns: - policy (:obj:`Policy`): Converged policy. """ if isinstance(input_cfg, str): cfg, create_cfg = read_config(input_cfg) expert_cfg, expert_create_cfg = read_config(expert_cfg) else: cfg, create_cfg = deepcopy(input_cfg) expert_cfg, expert_create_cfg = expert_cfg create_cfg.policy.type = create_cfg.policy.type + '_command' expert_create_cfg.policy.type = expert_create_cfg.policy.type + '_command' env_fn = None if env_setting is None else env_setting[0] cfg = compile_config(cfg, seed=seed, env=env_fn, auto=True, create_cfg=create_cfg, save_cfg=True) expert_cfg = compile_config( expert_cfg, seed=seed, env=env_fn, auto=True, create_cfg=expert_create_cfg, save_cfg=True ) # Create main components: env, policy if env_setting is None: env_fn, collector_env_cfg, evaluator_env_cfg = get_vec_env_setting(cfg.env) else: env_fn, collector_env_cfg, evaluator_env_cfg = env_setting collector_env = create_env_manager(cfg.env.manager, [partial(env_fn, cfg=c) for c in collector_env_cfg]) expert_collector_env = create_env_manager( expert_cfg.env.manager, [partial(env_fn, cfg=c) for c in collector_env_cfg] ) evaluator_env = create_env_manager(cfg.env.manager, [partial(env_fn, cfg=c) for c in evaluator_env_cfg]) expert_collector_env.seed(cfg.seed) collector_env.seed(cfg.seed) evaluator_env.seed(cfg.seed, dynamic_seed=False) expert_policy = create_policy(expert_cfg.policy, model=expert_model, enable_field=['collect', 'command']) set_pkg_seed(cfg.seed, use_cuda=cfg.policy.cuda) policy = create_policy(cfg.policy, model=model, enable_field=['learn', 'collect', 'eval', 'command']) expert_policy.collect_mode.load_state_dict(torch.load(expert_cfg.policy.collect.model_path, map_location='cpu')) # Create worker components: learner, collector, evaluator, replay buffer, commander. tb_logger = SummaryWriter(os.path.join('./{}/log/'.format(cfg.exp_name), 'serial')) learner = BaseLearner(cfg.policy.learn.learner, policy.learn_mode, tb_logger, exp_name=cfg.exp_name) collector = create_serial_collector( cfg.policy.collect.collector, env=collector_env, policy=policy.collect_mode, tb_logger=tb_logger, exp_name=cfg.exp_name ) expert_collector = create_serial_collector( expert_cfg.policy.collect.collector, env=expert_collector_env, policy=expert_policy.collect_mode, tb_logger=tb_logger, exp_name=expert_cfg.exp_name ) evaluator = InteractionSerialEvaluator( cfg.policy.eval.evaluator, evaluator_env, policy.eval_mode, tb_logger, exp_name=cfg.exp_name ) replay_buffer = create_buffer(cfg.policy.other.replay_buffer, tb_logger=tb_logger, exp_name=cfg.exp_name) commander = BaseSerialCommander( cfg.policy.other.commander, learner, collector, evaluator, replay_buffer, policy.command_mode ) expert_commander = BaseSerialCommander( expert_cfg.policy.other.commander, learner, expert_collector, evaluator, replay_buffer, expert_policy.command_mode ) # we create this to avoid the issue of eps, this is an issue due to the sample collector part. expert_collect_kwargs = expert_commander.step() if 'eps' in expert_collect_kwargs: expert_collect_kwargs['eps'] = -1 # ========== # Main loop # ========== # Learner's before_run hook. learner.call_hook('before_run') if expert_cfg.policy.learn.expert_replay_buffer_size != 0: # for ablation study expert_buffer = create_buffer(expert_cfg.policy.other.replay_buffer, tb_logger=tb_logger, exp_name=cfg.exp_name) expert_data = expert_collector.collect( n_sample=expert_cfg.policy.learn.expert_replay_buffer_size, train_iter=learner.train_iter, policy_kwargs=expert_collect_kwargs ) for i in range(len(expert_data)): # set is_expert flag(expert 1, agent 0) # expert_data[i]['is_expert'] = 1 # for transition-based alg. expert_data[i]['is_expert'] = [1] * expert_cfg.policy.collect.unroll_len # for rnn/sequence-based alg. expert_buffer.push(expert_data, cur_collector_envstep=0) for _ in range(cfg.policy.learn.per_train_iter_k): # pretrain if evaluator.should_eval(learner.train_iter): stop, reward = evaluator.eval(learner.save_checkpoint, learner.train_iter, collector.envstep) if stop: break # Learn policy from collected data # Expert_learner will train ``update_per_collect == 1`` times in one iteration. train_data = expert_buffer.sample(learner.policy.get_attribute('batch_size'), learner.train_iter) learner.train(train_data, collector.envstep) if learner.policy.get_attribute('priority'): expert_buffer.update(learner.priority_info) learner.priority_info = {} # Accumulate plenty of data at the beginning of training. if cfg.policy.get('random_collect_size', 0) > 0: random_collect( cfg.policy, policy, collector, collector_env, commander, replay_buffer, postprocess_data_fn=mark_not_expert ) while True: collect_kwargs = commander.step() # Evaluate policy performance if evaluator.should_eval(learner.train_iter): stop, reward = evaluator.eval(learner.save_checkpoint, learner.train_iter, collector.envstep) if stop: break # Collect data by default config n_sample/n_episode new_data = collector.collect(train_iter=learner.train_iter, policy_kwargs=collect_kwargs) for i in range(len(new_data)): # set is_expert flag(expert 1, agent 0) new_data[i]['is_expert'] = [0] * expert_cfg.policy.collect.unroll_len replay_buffer.push(new_data, cur_collector_envstep=collector.envstep) # Learn policy from collected data for i in range(cfg.policy.learn.update_per_collect): if expert_cfg.policy.learn.expert_replay_buffer_size != 0: # Learner will train ``update_per_collect`` times in one iteration. # The hyperparameter pho, the demo ratio, control the propotion of data coming\ # from expert demonstrations versus from the agent's own experience. expert_batch_size = int( np.float32(np.random.rand(learner.policy.get_attribute('batch_size')) < cfg.policy.collect.pho ).sum() ) agent_batch_size = (learner.policy.get_attribute('batch_size')) - expert_batch_size train_data_agent = replay_buffer.sample(agent_batch_size, learner.train_iter) train_data_expert = expert_buffer.sample(expert_batch_size, learner.train_iter) if train_data_agent is None: # It is possible that replay buffer's data count is too few to train ``update_per_collect`` times logging.warning( "Replay buffer's data can only train for {} steps. ".format(i) + "You can modify data collect config, e.g. increasing n_sample, n_episode." ) break train_data = train_data_agent + train_data_expert learner.train(train_data, collector.envstep) if learner.policy.get_attribute('priority'): # When collector, set replay_buffer_idx and replay_unique_id for each data item, priority = 1.\ # When learner, assign priority for each data item according their loss learner.priority_info_agent = deepcopy(learner.priority_info) learner.priority_info_expert = deepcopy(learner.priority_info) learner.priority_info_agent['priority'] = learner.priority_info['priority'][0:agent_batch_size] learner.priority_info_agent['replay_buffer_idx'] = learner.priority_info['replay_buffer_idx'][ 0:agent_batch_size] learner.priority_info_agent['replay_unique_id'] = learner.priority_info['replay_unique_id'][ 0:agent_batch_size] learner.priority_info_expert['priority'] = learner.priority_info['priority'][agent_batch_size:] learner.priority_info_expert['replay_buffer_idx'] = learner.priority_info['replay_buffer_idx'][ agent_batch_size:] learner.priority_info_expert['replay_unique_id'] = learner.priority_info['replay_unique_id'][ agent_batch_size:] # Expert data and demo data update their priority separately. replay_buffer.update(learner.priority_info_agent) expert_buffer.update(learner.priority_info_expert) else: # Learner will train ``update_per_collect`` times in one iteration. train_data = replay_buffer.sample(learner.policy.get_attribute('batch_size'), learner.train_iter) if train_data is None: # It is possible that replay buffer's data count is too few to train ``update_per_collect`` times logging.warning( "Replay buffer's data can only train for {} steps. ".format(i) + "You can modify data collect config, e.g. increasing n_sample, n_episode." ) break learner.train(train_data, collector.envstep) if learner.policy.get_attribute('priority'): replay_buffer.update(learner.priority_info) if collector.envstep >= max_env_step or learner.train_iter >= max_train_iter: break # Learner's after_run hook. learner.call_hook('after_run') return policy