def sac_policy_loss(
self,
log_probs: torch.Tensor,
q1p_outs: Dict[str, torch.Tensor],
loss_masks: torch.Tensor,
discrete: bool,
) -> torch.Tensor:
_ent_coef = torch.exp(self._log_ent_coef)
mean_q1 = torch.mean(torch.stack(list(q1p_outs.values())), axis=0)
if not discrete:
mean_q1 = mean_q1.unsqueeze(1)
batch_policy_loss = torch.mean(_ent_coef * log_probs - mean_q1, dim=1)
policy_loss = ModelUtils.masked_mean(batch_policy_loss, loss_masks)
else:
action_probs = log_probs.exp()
branched_per_action_ent = ModelUtils.break_into_branches(
log_probs * action_probs, self.act_size
)
branched_q_term = ModelUtils.break_into_branches(
mean_q1 * action_probs, self.act_size
)
branched_policy_loss = torch.stack(
[
torch.sum(_ent_coef[i] * _lp - _qt, dim=1, keepdim=True)
for i, (_lp, _qt) in enumerate(
zip(branched_per_action_ent, branched_q_term)
)
],
dim=1,
)
batch_policy_loss = torch.squeeze(branched_policy_loss)
policy_loss = ModelUtils.masked_mean(batch_policy_loss, loss_masks)
return policy_loss
def sac_entropy_loss(
self, log_probs: torch.Tensor, loss_masks: torch.Tensor, discrete: bool
) -> torch.Tensor:
if not discrete:
with torch.no_grad():
target_current_diff = torch.sum(log_probs + self.target_entropy, dim=1)
entropy_loss = -1 * ModelUtils.masked_mean(
self._log_ent_coef * target_current_diff, loss_masks
)
else:
with torch.no_grad():
branched_per_action_ent = ModelUtils.break_into_branches(
log_probs * log_probs.exp(), self.act_size
)
target_current_diff_branched = torch.stack(
[
torch.sum(_lp, axis=1, keepdim=True) + _te
for _lp, _te in zip(
branched_per_action_ent, self.target_entropy
)
],
axis=1,
)
target_current_diff = torch.squeeze(
target_current_diff_branched, axis=2
)
entropy_loss = -1 * ModelUtils.masked_mean(
torch.mean(self._log_ent_coef * target_current_diff, axis=1), loss_masks
)
return entropy_loss
def sac_entropy_loss(
self, log_probs: ActionLogProbs, loss_masks: torch.Tensor
) -> torch.Tensor:
_cont_ent_coef, _disc_ent_coef = (
self._log_ent_coef.continuous,
self._log_ent_coef.discrete,
)
entropy_loss = 0
if self._action_spec.discrete_size > 0:
with torch.no_grad():
# Break continuous into separate branch
disc_log_probs = log_probs.all_discrete_tensor
branched_per_action_ent = ModelUtils.break_into_branches(
disc_log_probs * disc_log_probs.exp(),
self._action_spec.discrete_branches,
)
target_current_diff_branched = torch.stack(
[
torch.sum(_lp, axis=1, keepdim=True) + _te
for _lp, _te in zip(
branched_per_action_ent, self.target_entropy.discrete
)
],
axis=1,
)
target_current_diff = torch.squeeze(
target_current_diff_branched, axis=2
)
entropy_loss += -1 * ModelUtils.masked_mean(
torch.mean(_disc_ent_coef * target_current_diff, axis=1), loss_masks
)
if self._action_spec.continuous_size > 0:
with torch.no_grad():
cont_log_probs = log_probs.continuous_tensor
target_current_diff = torch.sum(
cont_log_probs + self.target_entropy.continuous, dim=1
)
# We update all the _cont_ent_coef as one block
entropy_loss += -1 * ModelUtils.masked_mean(
_cont_ent_coef * target_current_diff, loss_masks
)
return entropy_loss