Skip to content

Convolution default fixed, extensive tests #321

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

Merged
blackencino merged 16 commits into from
Nov 5, 2025
Merged

Convolution default fixed, extensive tests #321

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
16 commits
Select commit Hold shift + click to select a range
9b32fcf
WIP on convolution ground truth tests
blackencino Oct 20, 2025
50579b6
WIP on convolution ground truth testing
blackencino Oct 20, 2025
98aee19
WIP Convolution Ground Truth - single and multiple impulses work
blackencino Oct 21, 2025
8bc65a6
Merge branch 'main' into feature/conv_test_minimal_ground_truth
blackencino Oct 21, 2025
57934b2
WIP on (still quite broken) convolution ground truth
blackencino Oct 22, 2025
d4ab569
WIP on ground truth kernel
blackencino Oct 23, 2025
63cddc6
WIP on convolution fixes
blackencino Oct 24, 2025
11b9611
Merge branch 'main' into feature/conv_test_minimal_ground_truth
blackencino Oct 24, 2025
3fc3b53
Tinkering with convolultion tests
blackencino Oct 29, 2025
864998c
Fixed upstream merge
blackencino Nov 4, 2025
df645ea
Default convolution tests are robust, and it passes
blackencino Nov 5, 2025
2202784
Added multiple impulse testing for conv default
blackencino Nov 5, 2025
6a3bb15
Merge remote-tracking branch 'upstream/main' into feature/conv_test_m…
blackencino Nov 5, 2025
9118d8d
Removed debug prints, spelling errors
blackencino Nov 5, 2025
c61fcee
Removed unused test files, fixed cmake
blackencino Nov 5, 2025
5f71351
Fixed the not-implemented checks
blackencino Nov 5, 2025
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
32 changes: 24 additions & 8 deletions fvdb/convolution_plan.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -189,12 +189,16 @@ def from_grid_batch(
kernel_size = to_Vec3i(kernel_size, value_constraint=ValueConstraint.POSITIVE)
stride = to_Vec3i(stride, value_constraint=ValueConstraint.POSITIVE)

if not _vec_is_all(kernel_size, kernel_size[0].item()):
raise NotImplementedError("Non-uniform kernel sizes are not currently supported")
if not _vec_is_all(stride, 1):
raise NotImplementedError("Strides not equal to 1 are not currently supported")

if not _vec_is_all(stride, stride[0].item()):
raise NotImplementedError("Non-uniform strides are not currently supported")

backend = expert_config.get("backend", "default")
if backend != "default":
raise NotImplementedError("Non-default backends are not currently supported")

if backend in ["dense", "halo", "lggs"]:
if target_grid is not None:
raise ValueError("Target grid must be None for dense, halo, and lggs backends.")
Expand Down Expand Up @@ -267,12 +271,16 @@ def from_grid_batch_transposed(
kernel_size = to_Vec3i(kernel_size, value_constraint=ValueConstraint.POSITIVE)
stride = to_Vec3i(stride, value_constraint=ValueConstraint.POSITIVE)

if not _vec_is_all(kernel_size, kernel_size[0].item()):
raise NotImplementedError("Non-uniform kernel sizes are not currently supported")
if not _vec_is_all(stride, 1):
raise NotImplementedError("Strides not equal to 1 are not currently supported")

if not _vec_is_all(stride, stride[0].item()):
raise NotImplementedError("Non-uniform strides are not currently supported")

backend = expert_config.get("backend", "default")
if backend != "default":
raise NotImplementedError("Non-default backends are not currently supported")

if backend == "dense":
if target_grid is not None:
raise ValueError("Target grid must be None for dense backend, transposed.")
Expand Down Expand Up @@ -348,12 +356,16 @@ def from_grid(
kernel_size = to_Vec3i(kernel_size, value_constraint=ValueConstraint.POSITIVE)
stride = to_Vec3i(stride, value_constraint=ValueConstraint.POSITIVE)

if not _vec_is_all(kernel_size, kernel_size[0].item()):
raise NotImplementedError("Non-uniform kernel sizes are not currently supported")
if not _vec_is_all(stride, 1):
raise NotImplementedError("Strides not equal to 1 are not currently supported")

if not _vec_is_all(stride, stride[0].item()):
raise NotImplementedError("Non-uniform strides are not currently supported")

backend = expert_config.get("backend", "default")
if backend != "default":
raise NotImplementedError("Non-default backends are not currently supported")

if backend in ["dense", "halo", "lggs"]:
if target_grid is not None:
raise ValueError("Target grid must be None for dense, halo, and lggs backends.")
Expand Down Expand Up @@ -424,12 +436,16 @@ def from_grid_transposed(
kernel_size = to_Vec3i(kernel_size, value_constraint=ValueConstraint.POSITIVE)
stride = to_Vec3i(stride, value_constraint=ValueConstraint.POSITIVE)

if not _vec_is_all(kernel_size, kernel_size[0].item()):
raise NotImplementedError("Non-uniform kernel sizes are not currently supported")
if not _vec_is_all(stride, 1):
raise NotImplementedError("Strides not equal to 1 are not currently supported")

if not _vec_is_all(stride, stride[0].item()):
raise NotImplementedError("Non-uniform strides are not currently supported")

backend = expert_config.get("backend", "default")
if backend != "default":
raise NotImplementedError("Non-default backends are not currently supported")

if backend == "dense":
if target_grid is not None:
raise ValueError("Target grid must be None for dense backend, transposed.")
Expand Down
Loading
Loading