from typing import List, Dict, Any, Tuple from collections import namedtuple import copy import torch from torch.optim import AdamW from ding.torch_utils import Adam, to_device from ding.rl_utils import q_nstep_td_data, q_nstep_td_error, get_nstep_return_data, get_train_sample, \ dqfd_nstep_td_error, dqfd_nstep_td_data from ding.model import model_wrap from ding.utils import POLICY_REGISTRY from ding.utils.data import default_collate, default_decollate from .dqn import DQNPolicy from .common_utils import default_preprocess_learn from copy import deepcopy @POLICY_REGISTRY.register('dqfd') class DQFDPolicy(DQNPolicy): r""" Overview: Policy class of DQFD algorithm, extended by Double DQN/Dueling DQN/PER/multi-step TD. Config: == ==================== ======== ============== ======================================== ======================= ID Symbol Type Default Value Description Other(Shape) == ==================== ======== ============== ======================================== ======================= 1 ``type`` str dqn | RL policy register name, refer to | This arg is optional, | registry ``POLICY_REGISTRY`` | a placeholder 2 ``cuda`` bool False | Whether to use cuda for network | This arg can be diff- | erent from modes 3 ``on_policy`` bool False | Whether the RL algorithm is on-policy | or off-policy 4 ``priority`` bool True | Whether use priority(PER) | Priority sample, | update priority 5 | ``priority_IS`` bool True | Whether use Importance Sampling Weight | ``_weight`` | to correct biased update. If True, | priority must be True. 6 | ``discount_`` float 0.97, | Reward's future discount factor, aka. | May be 1 when sparse | ``factor`` [0.95, 0.999] | gamma | reward env 7 ``nstep`` int 10, | N-step reward discount sum for target [3, 5] | q_value estimation 8 | ``lambda1`` float 1 | multiplicative factor for n-step 9 | ``lambda2`` float 1 | multiplicative factor for the | supervised margin loss 10 | ``lambda3`` float 1e-5 | L2 loss 11 | ``margin_fn`` float 0.8 | margin function in JE, here we set | this as a constant 12 | ``per_train_`` int 10 | number of pertraining iterations | ``iter_k`` 13 | ``learn.update`` int 3 | How many updates(iterations) to train | This args can be vary | ``per_collect`` | after collector's one collection. Only | from envs. Bigger val | valid in serial training | means more off-policy 14 | ``learn.batch_`` int 64 | The number of samples of an iteration | ``size`` 15 | ``learn.learning`` float 0.001 | Gradient step length of an iteration. | ``_rate`` 16 | ``learn.target_`` int 100 | Frequency of target network update. | Hard(assign) update | ``update_freq`` 17 | ``learn.ignore_`` bool False | Whether ignore done for target value | Enable it for some | ``done`` | calculation. | fake termination env 18 ``collect.n_sample`` int [8, 128] | The number of training samples of a | It varies from | call of collector. | different envs 19 | ``collect.unroll`` int 1 | unroll length of an iteration | In RNN, unroll_len>1 | ``_len`` == ==================== ======== ============== ======================================== ======================= """ config = dict( type='dqfd', cuda=False, on_policy=False, priority=True, # (bool) Whether use Importance Sampling Weight to correct biased update. If True, priority must be True. priority_IS_weight=True, discount_factor=0.99, nstep=10, learn=dict( # multiplicative factor for each loss lambda1=1.0, # n-step return lambda2=1.0, # supervised loss lambda3=1e-5, # L2 # margin function in JE, here we implement this as a constant margin_function=0.8, # number of pertraining iterations per_train_iter_k=10, # How many updates(iterations) to train after collector's one collection. # Bigger "update_per_collect" means bigger off-policy. # collect data -> update policy-> collect data -> ... update_per_collect=3, batch_size=64, learning_rate=0.001, # ============================================================== # The following configs are algorithm-specific # ============================================================== # (int) Frequence of target network update. target_update_freq=100, # (bool) Whether ignore done(usually for max step termination env) ignore_done=False, ), # collect_mode config collect=dict( # (int) Only one of [n_sample, n_episode] should be set # n_sample=8, # (int) Cut trajectories into pieces with length "unroll_len". unroll_len=1, # The hyperparameter pho, the demo ratio, control the propotion of data\ # coming from expert demonstrations versus from the agent's own experience. pho=0.5, ), eval=dict(), # other config other=dict( # Epsilon greedy with decay. eps=dict( # (str) Decay type. Support ['exp', 'linear']. type='exp', start=0.95, end=0.1, # (int) Decay length(env step) decay=10000, ), replay_buffer=dict(replay_buffer_size=10000, ), ), ) def _init_learn(self) -> None: """ Overview: Learn mode init method. Called by ``self.__init__``, initialize the optimizer, algorithm arguments, main \ and target model. """ self.lambda1 = self._cfg.learn.lambda1 # n-step return self.lambda2 = self._cfg.learn.lambda2 # supervised loss self.lambda3 = self._cfg.learn.lambda3 # L2 # margin function in JE, here we implement this as a constant self.margin_function = self._cfg.learn.margin_function self._priority = self._cfg.priority self._priority_IS_weight = self._cfg.priority_IS_weight # Optimizer # two optimizers: the performance of adamW is better than adam, so we recommend using the adamW. self._optimizer = AdamW(self._model.parameters(), lr=self._cfg.learn.learning_rate, weight_decay=self.lambda3) # self._optimizer = Adam(self._model.parameters(), lr=self._cfg.learn.learning_rate, weight_decay=self.lambda3) self._gamma = self._cfg.discount_factor self._nstep = self._cfg.nstep # use model_wrapper for specialized demands of different modes self._target_model = copy.deepcopy(self._model) self._target_model = model_wrap( self._target_model, wrapper_name='target', update_type='assign', update_kwargs={'freq': self._cfg.learn.target_update_freq} ) self._learn_model = model_wrap(self._model, wrapper_name='argmax_sample') self._learn_model.reset() self._target_model.reset() def _forward_learn(self, data: Dict[str, Any]) -> Dict[str, Any]: """ Overview: Forward computation graph of learn mode(updating policy). Arguments: - data (:obj:`Dict[str, Any]`): Dict type data, a batch of data for training, values are torch.Tensor or \ np.ndarray or dict/list combinations. Returns: - info_dict (:obj:`Dict[str, Any]`): Dict type data, a info dict indicated training result, which will be \ recorded in text log and tensorboard, values are python scalar or a list of scalars. ArgumentsKeys: - necessary: ``obs``, ``action``, ``reward``, ``next_obs``, ``done`` - optional: ``value_gamma``, ``IS`` ReturnsKeys: - necessary: ``cur_lr``, ``total_loss``, ``priority`` - optional: ``action_distribution`` """ data = default_preprocess_learn( data, use_priority=self._priority, use_priority_IS_weight=self._cfg.priority_IS_weight, ignore_done=self._cfg.learn.ignore_done, use_nstep=True ) data['done_1'] = data['done_1'].float() if self._cuda: data = to_device(data, self._device) # ==================== # Q-learning forward # ==================== self._learn_model.train() self._target_model.train() # Current q value (main model) q_value = self._learn_model.forward(data['obs'])['logit'] # Target q value with torch.no_grad(): target_q_value = self._target_model.forward(data['next_obs'])['logit'] target_q_value_one_step = self._target_model.forward(data['next_obs_1'])['logit'] # Max q value action (main model) target_q_action = self._learn_model.forward(data['next_obs'])['action'] target_q_action_one_step = self._learn_model.forward(data['next_obs_1'])['action'] # modify the tensor type to match the JE computation in dqfd_nstep_td_error is_expert = data['is_expert'].float() data_n = dqfd_nstep_td_data( q_value, target_q_value, data['action'], target_q_action, data['reward'], data['done'], data['done_1'], data['weight'], target_q_value_one_step, target_q_action_one_step, is_expert # set is_expert flag(expert 1, agent 0) ) value_gamma = data.get('value_gamma') loss, td_error_per_sample, loss_statistics = dqfd_nstep_td_error( data_n, self._gamma, self.lambda1, self.lambda2, self.margin_function, nstep=self._nstep, value_gamma=value_gamma ) # ==================== # Q-learning update # ==================== self._optimizer.zero_grad() loss.backward() if self._cfg.multi_gpu: self.sync_gradients(self._learn_model) self._optimizer.step() # ============= # after update # ============= self._target_model.update(self._learn_model.state_dict()) return { 'cur_lr': self._optimizer.defaults['lr'], 'total_loss': loss.item(), 'priority': td_error_per_sample.abs().tolist(), # Only discrete action satisfying len(data['action'])==1 can return this and draw histogram on tensorboard. # '[histogram]action_distribution': data['action'], } def _get_train_sample(self, data: List[Dict[str, Any]]) -> List[Dict[str, Any]]: """ Overview: For a given trajectory(transitions, a list of transition) data, process it into a list of sample that \ can be used for training directly. A train sample can be a processed transition(DQN with nstep TD) \ or some continuous transitions(DRQN). Arguments: - data (:obj:`List[Dict[str, Any]`): The trajectory data(a list of transition), each element is the same \ format as the return value of ``self._process_transition`` method. Returns: - samples (:obj:`dict`): The list of training samples. .. note:: We will vectorize ``process_transition`` and ``get_train_sample`` method in the following release version. \ And the user can customize the this data processing procecure by overriding this two methods and collector \ itself. """ data_1 = deepcopy(get_nstep_return_data(data, 1, gamma=self._gamma)) data = get_nstep_return_data( data, self._nstep, gamma=self._gamma ) # here we want to include one-step next observation for i in range(len(data)): data[i]['next_obs_1'] = data_1[i]['next_obs'] # concat the one-step next observation data[i]['done_1'] = data_1[i]['done'] return get_train_sample(data, self._unroll_len)