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

Fix the initialization function of "GradSampleModuleFastGradientClipping" #675

Closed
wants to merge 1 commit into from
Closed

Fix the initialization function of "GradSampleModuleFastGradientClipping" #675

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
9 changes: 4 additions & 5 deletions opacus/grad_sample/grad_sample_module.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -105,10 +105,9 @@ def __init__(
``[K, batch_size, ...]``
loss_reduction: Indicates if the loss reduction (for aggregating the gradients)
is a sum or a mean operation. Can take values "sum" or "mean"
strict: If set to ``True``, the input module will be validated to check that
``GradSampleModule`` has grad sampler functions for all submodules of
the input module (i.e. if it knows how to calculate per sample gradients)
for all model parameters. If set to ``False``, per sample gradients will
strict: If set to ``True``, the input module will be validated to make sure that none of its submodules includes buffers,
which is not currently supported by Opacus.
If set to ``False``, per sample gradients will
be computed on "best effort" basis - they will be available where
possible and set to None otherwise. This is not recommended, because
some unsupported modules (e.g. BatchNorm) affect other parameters and
Expand All @@ -120,7 +119,7 @@ def __init__(
Raises:
NotImplementedError
If ``strict`` is set to ``True`` and module ``m`` (or any of its
submodules) doesn't have a registered grad sampler function.
submodules) includes a buffer.
"""
super().__init__(
m,
Expand Down
4 changes: 3 additions & 1 deletion opacus/grad_sample/grad_sample_module_fast_gradient_clipping.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -107,13 +107,15 @@ def __init__(
Raises:
NotImplementedError
If ``strict`` is set to ``True`` and module ``m`` (or any of its
submodules) doesn't have a registered grad sampler function.
submodules) includes a buffer.
"""

super().__init__(
m,
batch_first=batch_first,
loss_reduction=loss_reduction,
strict=strict,
force_functorch=force_functorch,
)
self.trainable_parameters = [p for _, p in trainable_parameters(self._module)]
self.max_grad_norm = max_grad_norm
Expand Down
Loading