ding.bonus.td3

ding.bonus.td3

TD3Agent

Overview

Class of agent for training, evaluation and deployment of Reinforcement learning algorithm Twin Delayed Deep Deterministic Policy Gradient(TD3). For more information about the system design of RL agent, please refer to https://di-engine-docs.readthedocs.io/en/latest/03_system/agent.html.

Interface: __init__, train, deploy, collect_data, batch_evaluate, best

supported_env_list = list(supported_env_cfg.keys()) class-attribute instance-attribute

Overview

List of supported envs.

Examples: >>> from ding.bonus.td3 import TD3Agent >>> print(TD3Agent.supported_env_list)

best property

Overview

Load the best model from the checkpoint directory, which is by default in folder exp_name/ckpt/eval.pth.tar. The return value is the agent with the best model.

Returns: - (:obj:TD3Agent): The agent with the best model. Examples: >>> agent = TD3Agent(env_id='LunarLanderContinuous-v2') >>> agent.train() >>> agent.best

.. note:: The best model is the model with the highest evaluation return. If this method is called, the current model will be replaced by the best model.

__init__(env_id=None, env=None, seed=0, exp_name=None, model=None, cfg=None, policy_state_dict=None)

Overview

Initialize agent for TD3 algorithm.

Arguments: - env_id (:obj:str): The environment id, which is a registered environment name in gym or gymnasium. If env_id is not specified, env_id in cfg.env must be specified. If env_id is specified, env_id in cfg.env will be ignored. env_id should be one of the supported envs, which can be found in supported_env_list. - env (:obj:BaseEnv): The environment instance for training and evaluation. If env is not specified, env_id`` or ``cfg.env.env_id`` must be specified. ``env_id`` or ``cfg.env.env_id`` will be used to create environment instance. If ``env`` is specified, ``env_id`` and ``cfg.env.env_id`` will be ignored. - seed (:obj:int): The random seed, which is set before running the program. Default to 0. - exp_name (:obj:str): The name of this experiment, which will be used to create the folder to save log data. Default to None. If not specified, the folder name will be ``env_id``-``algorithm``. - model (:obj:torch.nn.Module): The model of TD3 algorithm, which should be an instance of class :class:ding.model.ContinuousQAC. If not specified, a default model will be generated according to the configuration. - cfg (:obj:Union[EasyDict, dict]): The configuration of TD3 algorithm, which is a dict. Default to None. If not specified, the default configuration will be used. The default configuration can be found in ``ding/config/example/TD3/gym_lunarlander_v2.py``. - policy_state_dict (:obj:str`): The path of policy state dict saved by PyTorch a in local file. If specified, the policy will be loaded from this file. Default to None.

.. note:: An RL Agent Instance can be initialized in two basic ways. For example, we have an environment with id LunarLanderContinuous-v2 registered in gym, and we want to train an agent with TD3 algorithm with default configuration. Then we can initialize the agent in the following ways: >>> agent = TD3Agent(env_id='LunarLanderContinuous-v2') or, if we want can specify the env_id in the configuration: >>> cfg = {'env': {'env_id': 'LunarLanderContinuous-v2'}, 'policy': ...... } >>> agent = TD3Agent(cfg=cfg) There are also other arguments to specify the agent when initializing. For example, if we want to specify the environment instance: >>> env = CustomizedEnv('LunarLanderContinuous-v2') >>> agent = TD3Agent(cfg=cfg, env=env) or, if we want to specify the model: >>> model = ContinuousQAC(**cfg.policy.model) >>> agent = TD3Agent(cfg=cfg, model=model) or, if we want to reload the policy from a saved policy state dict: >>> agent = TD3Agent(cfg=cfg, policy_state_dict='LunarLanderContinuous-v2.pth.tar') Make sure that the configuration is consistent with the saved policy state dict.

train(step=int(10000000.0), collector_env_num=None, evaluator_env_num=None, n_iter_save_ckpt=1000, context=None, debug=False, wandb_sweep=False)

Overview

Train the agent with TD3 algorithm for step iterations with collector_env_num collector environments and evaluator_env_num evaluator environments. Information during training will be recorded and saved by wandb.

Arguments: - step (:obj:int): The total training environment steps of all collector environments. Default to 1e7. - collector_env_num (:obj:int): The collector environment number. Default to None. If not specified, it will be set according to the configuration. - evaluator_env_num (:obj:int): The evaluator environment number. Default to None. If not specified, it will be set according to the configuration. - n_iter_save_ckpt (:obj:int): The frequency of saving checkpoint every training iteration. Default to 1000. - context (:obj:str): The multi-process context of the environment manager. Default to None. It can be specified as spawn, fork or forkserver. - debug (:obj:bool): Whether to use debug mode in the environment manager. Default to False. If set True, base environment manager will be used for easy debugging. Otherwise, subprocess environment manager will be used. - wandb_sweep (:obj:bool): Whether to use wandb sweep, which is a hyper-parameter optimization process for seeking the best configurations. Default to False. If True, the wandb sweep id will be used as the experiment name. Returns: - (:obj:TrainingReturn): The training result, of which the attributions are: - wandb_url (:obj:str): The weight & biases (wandb) project url of the trainning experiment.

deploy(enable_save_replay=False, concatenate_all_replay=False, replay_save_path=None, seed=None, debug=False)

Overview

Deploy the agent with TD3 algorithm by interacting with the environment, during which the replay video can be saved if enable_save_replay is True. The evaluation result will be returned.

Arguments: - enable_save_replay (:obj:bool): Whether to save the replay video. Default to False. - concatenate_all_replay (:obj:bool): Whether to concatenate all replay videos into one video. Default to False. If enable_save_replay is False, this argument will be ignored. If enable_save_replay is True and concatenate_all_replay is False, the replay video of each episode will be saved separately. - replay_save_path (:obj:str): The path to save the replay video. Default to None. If not specified, the video will be saved in exp_name/videos. - seed (:obj:Union[int, List]): The random seed, which is set before running the program. Default to None. If not specified, self.seed will be used. If seed is an integer, the agent will be deployed once. If seed is a list of integers, the agent will be deployed once for each seed in the list. - debug (:obj:bool): Whether to use debug mode in the environment manager. Default to False. If set True, base environment manager will be used for easy debugging. Otherwise, subprocess environment manager will be used. Returns: - (:obj:EvalReturn): The evaluation result, of which the attributions are: - eval_value (:obj:np.float32): The mean of evaluation return. - eval_value_std (:obj:np.float32): The standard deviation of evaluation return.

collect_data(env_num=8, save_data_path=None, n_sample=None, n_episode=None, context=None, debug=False)

Overview

Collect data with TD3 algorithm for n_episode episodes with env_num collector environments. The collected data will be saved in save_data_path if specified, otherwise it will be saved in exp_name/demo_data.

Arguments: - env_num (:obj:int): The number of collector environments. Default to 8. - save_data_path (:obj:str): The path to save the collected data. Default to None. If not specified, the data will be saved in exp_name/demo_data. - n_sample (:obj:int): The number of samples to collect. Default to None. If not specified, n_episode must be specified. - n_episode (:obj:int): The number of episodes to collect. Default to None. If not specified, n_sample must be specified. - context (:obj:str): The multi-process context of the environment manager. Default to None. It can be specified as spawn, fork or forkserver. - debug (:obj:bool): Whether to use debug mode in the environment manager. Default to False. If set True, base environment manager will be used for easy debugging. Otherwise, subprocess environment manager will be used.

batch_evaluate(env_num=4, n_evaluator_episode=4, context=None, debug=False)

Overview

Evaluate the agent with TD3 algorithm for n_evaluator_episode episodes with env_num evaluator environments. The evaluation result will be returned. The difference between methods batch_evaluate and deploy is that batch_evaluate will create multiple evaluator environments to evaluate the agent to get an average performance, while deploy will only create one evaluator environment to evaluate the agent and save the replay video.

Arguments: - env_num (:obj:int): The number of evaluator environments. Default to 4. - n_evaluator_episode (:obj:int): The number of episodes to evaluate. Default to 4. - context (:obj:str): The multi-process context of the environment manager. Default to None. It can be specified as spawn, fork or forkserver. - debug (:obj:bool): Whether to use debug mode in the environment manager. Default to False. If set True, base environment manager will be used for easy debugging. Otherwise, subprocess environment manager will be used. Returns: - (:obj:EvalReturn): The evaluation result, of which the attributions are: - eval_value (:obj:np.float32): The mean of evaluation return. - eval_value_std (:obj:np.float32): The standard deviation of evaluation return.

Full Source Code

../ding/bonus/td3.py

1from typing import Optional, Union, List 2from ditk import logging 3from easydict import EasyDict 4import os 5import numpy as np 6import torch 7import treetensor.torch as ttorch 8from ding.framework import task, OnlineRLContext 9from ding.framework.middleware import CkptSaver, \ 10 wandb_online_logger, offline_data_saver, termination_checker, interaction_evaluator, StepCollector, data_pusher, \ 11 OffPolicyLearner, final_ctx_saver 12from ding.envs import BaseEnv 13from ding.envs import setup_ding_env_manager 14from ding.policy import TD3Policy 15from ding.utils import set_pkg_seed 16from ding.utils import get_env_fps, render 17from ding.config import save_config_py, compile_config 18from ding.model import ContinuousQAC 19from ding.data import DequeBuffer 20from ding.bonus.common import TrainingReturn, EvalReturn 21from ding.config.example.TD3 import supported_env_cfg 22from ding.config.example.TD3 import supported_env 23 24 25class TD3Agent: 26 """ 27 Overview: 28 Class of agent for training, evaluation and deployment of Reinforcement learning algorithm \ 29 Twin Delayed Deep Deterministic Policy Gradient(TD3). 30 For more information about the system design of RL agent, please refer to \ 31 <https://di-engine-docs.readthedocs.io/en/latest/03_system/agent.html>. 32 Interface: 33 ``__init__``, ``train``, ``deploy``, ``collect_data``, ``batch_evaluate``, ``best`` 34 """ 35 supported_env_list = list(supported_env_cfg.keys()) 36 """ 37 Overview: 38 List of supported envs. 39 Examples: 40 >>> from ding.bonus.td3 import TD3Agent 41 >>> print(TD3Agent.supported_env_list) 42 """ 43 44 def __init__( 45 self, 46 env_id: str = None, 47 env: BaseEnv = None, 48 seed: int = 0, 49 exp_name: str = None, 50 model: Optional[torch.nn.Module] = None, 51 cfg: Optional[Union[EasyDict, dict]] = None, 52 policy_state_dict: str = None, 53 ) -> None: 54 """ 55 Overview: 56 Initialize agent for TD3 algorithm. 57 Arguments: 58 - env_id (:obj:`str`): The environment id, which is a registered environment name in gym or gymnasium. \ 59 If ``env_id`` is not specified, ``env_id`` in ``cfg.env`` must be specified. \ 60 If ``env_id`` is specified, ``env_id`` in ``cfg.env`` will be ignored. \ 61 ``env_id`` should be one of the supported envs, which can be found in ``supported_env_list``. 62 - env (:obj:`BaseEnv`): The environment instance for training and evaluation. \ 63 If ``env`` is not specified, `env_id`` or ``cfg.env.env_id`` must be specified. \ 64 ``env_id`` or ``cfg.env.env_id`` will be used to create environment instance. \ 65 If ``env`` is specified, ``env_id`` and ``cfg.env.env_id`` will be ignored. 66 - seed (:obj:`int`): The random seed, which is set before running the program. \ 67 Default to 0. 68 - exp_name (:obj:`str`): The name of this experiment, which will be used to create the folder to save \ 69 log data. Default to None. If not specified, the folder name will be ``env_id``-``algorithm``. 70 - model (:obj:`torch.nn.Module`): The model of TD3 algorithm, which should be an instance of class \ 71 :class:`ding.model.ContinuousQAC`. \ 72 If not specified, a default model will be generated according to the configuration. 73 - cfg (:obj:Union[EasyDict, dict]): The configuration of TD3 algorithm, which is a dict. \ 74 Default to None. If not specified, the default configuration will be used. \ 75 The default configuration can be found in ``ding/config/example/TD3/gym_lunarlander_v2.py``. 76 - policy_state_dict (:obj:`str`): The path of policy state dict saved by PyTorch a in local file. \ 77 If specified, the policy will be loaded from this file. Default to None. 78 79 .. note:: 80 An RL Agent Instance can be initialized in two basic ways. \ 81 For example, we have an environment with id ``LunarLanderContinuous-v2`` registered in gym, \ 82 and we want to train an agent with TD3 algorithm with default configuration. \ 83 Then we can initialize the agent in the following ways: 84 >>> agent = TD3Agent(env_id='LunarLanderContinuous-v2') 85 or, if we want can specify the env_id in the configuration: 86 >>> cfg = {'env': {'env_id': 'LunarLanderContinuous-v2'}, 'policy': ...... } 87 >>> agent = TD3Agent(cfg=cfg) 88 There are also other arguments to specify the agent when initializing. 89 For example, if we want to specify the environment instance: 90 >>> env = CustomizedEnv('LunarLanderContinuous-v2') 91 >>> agent = TD3Agent(cfg=cfg, env=env) 92 or, if we want to specify the model: 93 >>> model = ContinuousQAC(**cfg.policy.model) 94 >>> agent = TD3Agent(cfg=cfg, model=model) 95 or, if we want to reload the policy from a saved policy state dict: 96 >>> agent = TD3Agent(cfg=cfg, policy_state_dict='LunarLanderContinuous-v2.pth.tar') 97 Make sure that the configuration is consistent with the saved policy state dict. 98 """ 99 100 assert env_id is not None or cfg is not None, "Please specify env_id or cfg." 101 102 if cfg is not None and not isinstance(cfg, EasyDict): 103 cfg = EasyDict(cfg) 104 105 if env_id is not None: 106 assert env_id in TD3Agent.supported_env_list, "Please use supported envs: {}".format( 107 TD3Agent.supported_env_list 108 ) 109 if cfg is None: 110 cfg = supported_env_cfg[env_id] 111 else: 112 assert cfg.env.env_id == env_id, "env_id in cfg should be the same as env_id in args." 113 else: 114 assert hasattr(cfg.env, "env_id"), "Please specify env_id in cfg." 115 assert cfg.env.env_id in TD3Agent.supported_env_list, "Please use supported envs: {}".format( 116 TD3Agent.supported_env_list 117 ) 118 default_policy_config = EasyDict({"policy": TD3Policy.default_config()}) 119 default_policy_config.update(cfg) 120 cfg = default_policy_config 121 122 if exp_name is not None: 123 cfg.exp_name = exp_name 124 self.cfg = compile_config(cfg, policy=TD3Policy) 125 self.exp_name = self.cfg.exp_name 126 if env is None: 127 self.env = supported_env[cfg.env.env_id](cfg=cfg.env) 128 else: 129 assert isinstance(env, BaseEnv), "Please use BaseEnv as env data type." 130 self.env = env 131 132 logging.getLogger().setLevel(logging.INFO) 133 self.seed = seed 134 set_pkg_seed(self.seed, use_cuda=self.cfg.policy.cuda) 135 if not os.path.exists(self.exp_name): 136 os.makedirs(self.exp_name) 137 save_config_py(self.cfg, os.path.join(self.exp_name, 'policy_config.py')) 138 if model is None: 139 model = ContinuousQAC(**self.cfg.policy.model) 140 self.buffer_ = DequeBuffer(size=self.cfg.policy.other.replay_buffer.replay_buffer_size) 141 self.policy = TD3Policy(self.cfg.policy, model=model) 142 if policy_state_dict is not None: 143 self.policy.learn_mode.load_state_dict(policy_state_dict) 144 self.checkpoint_save_dir = os.path.join(self.exp_name, "ckpt") 145 146 def train( 147 self, 148 step: int = int(1e7), 149 collector_env_num: int = None, 150 evaluator_env_num: int = None, 151 n_iter_save_ckpt: int = 1000, 152 context: Optional[str] = None, 153 debug: bool = False, 154 wandb_sweep: bool = False, 155 ) -> TrainingReturn: 156 """ 157 Overview: 158 Train the agent with TD3 algorithm for ``step`` iterations with ``collector_env_num`` collector \ 159 environments and ``evaluator_env_num`` evaluator environments. Information during training will be \ 160 recorded and saved by wandb. 161 Arguments: 162 - step (:obj:`int`): The total training environment steps of all collector environments. Default to 1e7. 163 - collector_env_num (:obj:`int`): The collector environment number. Default to None. \ 164 If not specified, it will be set according to the configuration. 165 - evaluator_env_num (:obj:`int`): The evaluator environment number. Default to None. \ 166 If not specified, it will be set according to the configuration. 167 - n_iter_save_ckpt (:obj:`int`): The frequency of saving checkpoint every training iteration. \ 168 Default to 1000. 169 - context (:obj:`str`): The multi-process context of the environment manager. Default to None. \ 170 It can be specified as ``spawn``, ``fork`` or ``forkserver``. 171 - debug (:obj:`bool`): Whether to use debug mode in the environment manager. Default to False. \ 172 If set True, base environment manager will be used for easy debugging. Otherwise, \ 173 subprocess environment manager will be used. 174 - wandb_sweep (:obj:`bool`): Whether to use wandb sweep, \ 175 which is a hyper-parameter optimization process for seeking the best configurations. \ 176 Default to False. If True, the wandb sweep id will be used as the experiment name. 177 Returns: 178 - (:obj:`TrainingReturn`): The training result, of which the attributions are: 179 - wandb_url (:obj:`str`): The weight & biases (wandb) project url of the trainning experiment. 180 """ 181 182 if debug: 183 logging.getLogger().setLevel(logging.DEBUG) 184 logging.debug(self.policy._model) 185 # define env and policy 186 collector_env_num = collector_env_num if collector_env_num else self.cfg.env.collector_env_num 187 evaluator_env_num = evaluator_env_num if evaluator_env_num else self.cfg.env.evaluator_env_num 188 collector_env = setup_ding_env_manager(self.env, collector_env_num, context, debug, 'collector') 189 evaluator_env = setup_ding_env_manager(self.env, evaluator_env_num, context, debug, 'evaluator') 190 191 with task.start(ctx=OnlineRLContext()): 192 task.use( 193 interaction_evaluator( 194 self.cfg, 195 self.policy.eval_mode, 196 evaluator_env, 197 render=self.cfg.policy.eval.render if hasattr(self.cfg.policy.eval, "render") else False 198 ) 199 ) 200 task.use(CkptSaver(policy=self.policy, save_dir=self.checkpoint_save_dir, train_freq=n_iter_save_ckpt)) 201 task.use( 202 StepCollector( 203 self.cfg, 204 self.policy.collect_mode, 205 collector_env, 206 random_collect_size=self.cfg.policy.random_collect_size 207 if hasattr(self.cfg.policy, 'random_collect_size') else 0, 208 ) 209 ) 210 task.use(data_pusher(self.cfg, self.buffer_)) 211 task.use(OffPolicyLearner(self.cfg, self.policy.learn_mode, self.buffer_)) 212 task.use( 213 wandb_online_logger( 214 metric_list=self.policy._monitor_vars_learn(), 215 model=self.policy._model, 216 anonymous=True, 217 project_name=self.exp_name, 218 wandb_sweep=wandb_sweep, 219 ) 220 ) 221 task.use(termination_checker(max_env_step=step)) 222 task.use(final_ctx_saver(name=self.exp_name)) 223 task.run() 224 225 return TrainingReturn(wandb_url=task.ctx.wandb_url) 226 227 def deploy( 228 self, 229 enable_save_replay: bool = False, 230 concatenate_all_replay: bool = False, 231 replay_save_path: str = None, 232 seed: Optional[Union[int, List]] = None, 233 debug: bool = False 234 ) -> EvalReturn: 235 """ 236 Overview: 237 Deploy the agent with TD3 algorithm by interacting with the environment, during which the replay video \ 238 can be saved if ``enable_save_replay`` is True. The evaluation result will be returned. 239 Arguments: 240 - enable_save_replay (:obj:`bool`): Whether to save the replay video. Default to False. 241 - concatenate_all_replay (:obj:`bool`): Whether to concatenate all replay videos into one video. \ 242 Default to False. If ``enable_save_replay`` is False, this argument will be ignored. \ 243 If ``enable_save_replay`` is True and ``concatenate_all_replay`` is False, \ 244 the replay video of each episode will be saved separately. 245 - replay_save_path (:obj:`str`): The path to save the replay video. Default to None. \ 246 If not specified, the video will be saved in ``exp_name/videos``. 247 - seed (:obj:`Union[int, List]`): The random seed, which is set before running the program. \ 248 Default to None. If not specified, ``self.seed`` will be used. \ 249 If ``seed`` is an integer, the agent will be deployed once. \ 250 If ``seed`` is a list of integers, the agent will be deployed once for each seed in the list. 251 - debug (:obj:`bool`): Whether to use debug mode in the environment manager. Default to False. \ 252 If set True, base environment manager will be used for easy debugging. Otherwise, \ 253 subprocess environment manager will be used. 254 Returns: 255 - (:obj:`EvalReturn`): The evaluation result, of which the attributions are: 256 - eval_value (:obj:`np.float32`): The mean of evaluation return. 257 - eval_value_std (:obj:`np.float32`): The standard deviation of evaluation return. 258 """ 259 260 if debug: 261 logging.getLogger().setLevel(logging.DEBUG) 262 # define env and policy 263 env = self.env.clone(caller='evaluator') 264 265 if seed is not None and isinstance(seed, int): 266 seeds = [seed] 267 elif seed is not None and isinstance(seed, list): 268 seeds = seed 269 else: 270 seeds = [self.seed] 271 272 returns = [] 273 images = [] 274 if enable_save_replay: 275 replay_save_path = os.path.join(self.exp_name, 'videos') if replay_save_path is None else replay_save_path 276 env.enable_save_replay(replay_path=replay_save_path) 277 else: 278 logging.warning('No video would be generated during the deploy.') 279 if concatenate_all_replay: 280 logging.warning('concatenate_all_replay is set to False because enable_save_replay is False.') 281 concatenate_all_replay = False 282 283 def single_env_forward_wrapper(forward_fn, cuda=True): 284 285 def _forward(obs): 286 # unsqueeze means add batch dim, i.e. (O, ) -> (1, O) 287 obs = ttorch.as_tensor(obs).unsqueeze(0) 288 if cuda and torch.cuda.is_available(): 289 obs = obs.cuda() 290 action = forward_fn(obs, mode='compute_actor')["action"] 291 # squeeze means delete batch dim, i.e. (1, A) -> (A, ) 292 action = action.squeeze(0).detach().cpu().numpy() 293 return action 294 295 return _forward 296 297 forward_fn = single_env_forward_wrapper(self.policy._model, self.cfg.policy.cuda) 298 299 # reset first to make sure the env is in the initial state 300 # env will be reset again in the main loop 301 env.reset() 302 303 for seed in seeds: 304 env.seed(seed, dynamic_seed=False) 305 return_ = 0. 306 step = 0 307 obs = env.reset() 308 images.append(render(env)[None]) if concatenate_all_replay else None 309 while True: 310 action = forward_fn(obs) 311 obs, rew, done, info = env.step(action) 312 images.append(render(env)[None]) if concatenate_all_replay else None 313 return_ += rew 314 step += 1 315 if done: 316 break 317 logging.info(f'DQN deploy is finished, final episode return with {step} steps is: {return_}') 318 returns.append(return_) 319 320 env.close() 321 322 if concatenate_all_replay: 323 images = np.concatenate(images, axis=0) 324 import imageio 325 imageio.mimwrite(os.path.join(replay_save_path, 'deploy.mp4'), images, fps=get_env_fps(env)) 326 327 return EvalReturn(eval_value=np.mean(returns), eval_value_std=np.std(returns)) 328 329 def collect_data( 330 self, 331 env_num: int = 8, 332 save_data_path: Optional[str] = None, 333 n_sample: Optional[int] = None, 334 n_episode: Optional[int] = None, 335 context: Optional[str] = None, 336 debug: bool = False 337 ) -> None: 338 """ 339 Overview: 340 Collect data with TD3 algorithm for ``n_episode`` episodes with ``env_num`` collector environments. \ 341 The collected data will be saved in ``save_data_path`` if specified, otherwise it will be saved in \ 342 ``exp_name/demo_data``. 343 Arguments: 344 - env_num (:obj:`int`): The number of collector environments. Default to 8. 345 - save_data_path (:obj:`str`): The path to save the collected data. Default to None. \ 346 If not specified, the data will be saved in ``exp_name/demo_data``. 347 - n_sample (:obj:`int`): The number of samples to collect. Default to None. \ 348 If not specified, ``n_episode`` must be specified. 349 - n_episode (:obj:`int`): The number of episodes to collect. Default to None. \ 350 If not specified, ``n_sample`` must be specified. 351 - context (:obj:`str`): The multi-process context of the environment manager. Default to None. \ 352 It can be specified as ``spawn``, ``fork`` or ``forkserver``. 353 - debug (:obj:`bool`): Whether to use debug mode in the environment manager. Default to False. \ 354 If set True, base environment manager will be used for easy debugging. Otherwise, \ 355 subprocess environment manager will be used. 356 """ 357 358 if debug: 359 logging.getLogger().setLevel(logging.DEBUG) 360 if n_episode is not None: 361 raise NotImplementedError 362 # define env and policy 363 env_num = env_num if env_num else self.cfg.env.collector_env_num 364 env = setup_ding_env_manager(self.env, env_num, context, debug, 'collector') 365 366 if save_data_path is None: 367 save_data_path = os.path.join(self.exp_name, 'demo_data') 368 369 # main execution task 370 with task.start(ctx=OnlineRLContext()): 371 task.use( 372 StepCollector( 373 self.cfg, self.policy.collect_mode, env, random_collect_size=self.cfg.policy.random_collect_size 374 ) 375 ) 376 task.use(offline_data_saver(save_data_path, data_type='hdf5')) 377 task.run(max_step=1) 378 logging.info( 379 f'TD3 collecting is finished, more than {n_sample} samples are collected and saved in `{save_data_path}`' 380 ) 381 382 def batch_evaluate( 383 self, 384 env_num: int = 4, 385 n_evaluator_episode: int = 4, 386 context: Optional[str] = None, 387 debug: bool = False 388 ) -> EvalReturn: 389 """ 390 Overview: 391 Evaluate the agent with TD3 algorithm for ``n_evaluator_episode`` episodes with ``env_num`` evaluator \ 392 environments. The evaluation result will be returned. 393 The difference between methods ``batch_evaluate`` and ``deploy`` is that ``batch_evaluate`` will create \ 394 multiple evaluator environments to evaluate the agent to get an average performance, while ``deploy`` \ 395 will only create one evaluator environment to evaluate the agent and save the replay video. 396 Arguments: 397 - env_num (:obj:`int`): The number of evaluator environments. Default to 4. 398 - n_evaluator_episode (:obj:`int`): The number of episodes to evaluate. Default to 4. 399 - context (:obj:`str`): The multi-process context of the environment manager. Default to None. \ 400 It can be specified as ``spawn``, ``fork`` or ``forkserver``. 401 - debug (:obj:`bool`): Whether to use debug mode in the environment manager. Default to False. \ 402 If set True, base environment manager will be used for easy debugging. Otherwise, \ 403 subprocess environment manager will be used. 404 Returns: 405 - (:obj:`EvalReturn`): The evaluation result, of which the attributions are: 406 - eval_value (:obj:`np.float32`): The mean of evaluation return. 407 - eval_value_std (:obj:`np.float32`): The standard deviation of evaluation return. 408 """ 409 410 if debug: 411 logging.getLogger().setLevel(logging.DEBUG) 412 # define env and policy 413 env_num = env_num if env_num else self.cfg.env.evaluator_env_num 414 env = setup_ding_env_manager(self.env, env_num, context, debug, 'evaluator') 415 416 # reset first to make sure the env is in the initial state 417 # env will be reset again in the main loop 418 env.launch() 419 env.reset() 420 421 evaluate_cfg = self.cfg 422 evaluate_cfg.env.n_evaluator_episode = n_evaluator_episode 423 424 # main execution task 425 with task.start(ctx=OnlineRLContext()): 426 task.use(interaction_evaluator(self.cfg, self.policy.eval_mode, env)) 427 task.run(max_step=1) 428 429 return EvalReturn(eval_value=task.ctx.eval_value, eval_value_std=task.ctx.eval_value_std) 430 431 @property 432 def best(self) -> 'TD3Agent': 433 """ 434 Overview: 435 Load the best model from the checkpoint directory, \ 436 which is by default in folder ``exp_name/ckpt/eval.pth.tar``. \ 437 The return value is the agent with the best model. 438 Returns: 439 - (:obj:`TD3Agent`): The agent with the best model. 440 Examples: 441 >>> agent = TD3Agent(env_id='LunarLanderContinuous-v2') 442 >>> agent.train() 443 >>> agent.best 444 445 .. note:: 446 The best model is the model with the highest evaluation return. If this method is called, the current \ 447 model will be replaced by the best model. 448 """ 449 450 best_model_file_path = os.path.join(self.checkpoint_save_dir, "eval.pth.tar") 451 # Load best model if it exists 452 if os.path.exists(best_model_file_path): 453 policy_state_dict = torch.load(best_model_file_path, map_location=torch.device("cpu")) 454 self.policy.learn_mode.load_state_dict(policy_state_dict) 455 return self