import torch
import numpy as np
from torch import nn
from copy import deepcopy
import torch.nn.functional as F
from tianshou.data import Batch
from tianshou.policy import PGPolicy
[docs]class PPOPolicy(PGPolicy):
r"""Implementation of Proximal Policy Optimization. arXiv:1707.06347
:param torch.nn.Module actor: the actor network following the rules in
:class:`~tianshou.policy.BasePolicy`. (s -> logits)
:param torch.nn.Module critic: the critic network. (s -> V(s))
:param torch.optim.Optimizer optim: the optimizer for actor and critic
network.
:param torch.distributions.Distribution dist_fn: for computing the action.
:param float discount_factor: in [0, 1], defaults to 0.99.
:param float max_grad_norm: clipping gradients in back propagation,
defaults to ``None``.
:param float eps_clip: :math:`\epsilon` in :math:`L_{CLIP}` in the original
paper, defaults to 0.2.
:param float vf_coef: weight for value loss, defaults to 0.5.
:param float ent_coef: weight for entropy loss, defaults to 0.01.
:param action_range: the action range (minimum, maximum).
:type action_range: [float, float]
"""
def __init__(self, actor, critic, optim, dist_fn,
discount_factor=0.99,
max_grad_norm=.5,
eps_clip=.2,
vf_coef=.5,
ent_coef=.0,
action_range=None,
**kwargs):
super().__init__(None, None, dist_fn, discount_factor)
self._max_grad_norm = max_grad_norm
self._eps_clip = eps_clip
self._w_vf = vf_coef
self._w_ent = ent_coef
self._range = action_range
self.actor, self.actor_old = actor, deepcopy(actor)
self.actor_old.eval()
self.critic, self.critic_old = critic, deepcopy(critic)
self.critic_old.eval()
self.optim = optim
[docs] def train(self):
"""Set the module in training mode, except for the target network."""
self.training = True
self.actor.train()
self.critic.train()
[docs] def eval(self):
"""Set the module in evaluation mode, except for the target network."""
self.training = False
self.actor.eval()
self.critic.eval()
[docs] def __call__(self, batch, state=None, model='actor', **kwargs):
"""Compute action over the given batch data.
:return: A :class:`~tianshou.data.Batch` which has 4 keys:
* ``act`` the action.
* ``logits`` the network's raw output.
* ``dist`` the action distribution.
* ``state`` the hidden state.
More information can be found at
:meth:`~tianshou.policy.BasePolicy.__call__`.
"""
model = getattr(self, model)
logits, h = model(batch.obs, state=state, info=batch.info)
if isinstance(logits, tuple):
dist = self.dist_fn(*logits)
else:
dist = self.dist_fn(logits)
act = dist.sample()
if self._range:
act = act.clamp(self._range[0], self._range[1])
return Batch(logits=logits, act=act, state=h, dist=dist)
[docs] def sync_weight(self):
"""Synchronize the weight for the target network."""
self.actor_old.load_state_dict(self.actor.state_dict())
self.critic_old.load_state_dict(self.critic.state_dict())
[docs] def learn(self, batch, batch_size=None, repeat=1, **kwargs):
losses, clip_losses, vf_losses, ent_losses = [], [], [], []
r = batch.returns
batch.returns = (r - r.mean()) / (r.std() + self._eps)
batch.act = torch.tensor(batch.act)
batch.returns = torch.tensor(batch.returns)[:, None]
for _ in range(repeat):
for b in batch.split(batch_size):
vs_old, vs__old = self.critic_old(np.concatenate([
b.obs, b.obs_next])).split(b.obs.shape[0])
dist = self(b).dist
dist_old = self(b, model='actor_old').dist
target_v = b.returns.to(vs__old.device) + self._gamma * vs__old
adv = (target_v - vs_old).detach()
a = b.act.to(adv.device)
ratio = torch.exp(dist.log_prob(a) - dist_old.log_prob(a))
surr1 = ratio * adv
surr2 = ratio.clamp(
1. - self._eps_clip, 1. + self._eps_clip) * adv
clip_loss = -torch.min(surr1, surr2).mean()
clip_losses.append(clip_loss.item())
vf_loss = F.smooth_l1_loss(self.critic(b.obs), target_v)
vf_losses.append(vf_loss.item())
e_loss = dist.entropy().mean()
ent_losses.append(e_loss.item())
loss = clip_loss + self._w_vf * vf_loss - self._w_ent * e_loss
losses.append(loss.item())
self.optim.zero_grad()
loss.backward()
nn.utils.clip_grad_norm_(list(
self.actor.parameters()) + list(self.critic.parameters()),
self._max_grad_norm)
self.optim.step()
self.sync_weight()
return {
'loss': losses,
'loss/clip': clip_losses,
'loss/vf': vf_losses,
'loss/ent': ent_losses,
}