Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[RLlib] Fix PPOTorchPolicy producing float metrics when not using critic #27980

Merged
merged 4 commits into from
Aug 22, 2022
Merged

[RLlib] Fix PPOTorchPolicy producing float metrics when not using critic #27980

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
6b50794
initial
ArturNiederfahrenhorst Aug 18, 2022
06b838f
Model no variance with explained variance = 1
ArturNiederfahrenhorst Aug 18, 2022
69b81d9
unify with TF
ArturNiederfahrenhorst Aug 18, 2022
4648a09
sven's comments
ArturNiederfahrenhorst Aug 18, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 4 additions & 2 deletions rllib/algorithms/ppo/ppo_torch_policy.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -148,8 +148,10 @@ def reduce_mean_valid(t):
mean_vf_loss = reduce_mean_valid(vf_loss_clipped)
# Ignore the value function.
else:
value_fn_out = 0
vf_loss_clipped = mean_vf_loss = 0.0
value_fn_out = torch.tensor(0.0).to(mean_policy_loss.device)
vf_loss_clipped = mean_vf_loss = torch.tensor(0.0).to(
mean_policy_loss.device
)

total_loss = reduce_mean_valid(
-surrogate_loss
Expand Down
3 changes: 3 additions & 0 deletions rllib/utils/torch_utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -211,6 +211,9 @@ def explained_variance(y: TensorType, pred: TensorType) -> TensorType:
The explained variance given a pair of labels and predictions.
"""
y_var = torch.var(y, dim=[0])
if y_var == 0.0:
# Model case in which y does not vary with explained variance of -1
return torch.tensor(-1.0).to(pred.device)
diff_var = torch.var(y - pred, dim=[0])
min_ = torch.tensor([-1.0]).to(pred.device)
return torch.max(min_, 1 - (diff_var / y_var))[0]
Expand Down