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 bool sort on CUDA #665

Merged
merged 5 commits into from
Dec 8, 2021
Merged

Fix bool sort on CUDA #665

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
f3f88f2
Fix bool sort on CUDA
twsl Dec 7, 2021
8f8542b
Add dedicated gpu test
twsl Dec 7, 2021
025156c
Updated CHANGELOG.md
twsl Dec 7, 2021
8a468ba
Apply suggestions from code review
Borda Dec 8, 2021
7511a1d
Merge branch 'master' into fix/map-gpu
Borda Dec 8, 2021
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
3 changes: 3 additions & 0 deletions CHANGELOG.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -53,6 +53,9 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
### Fixed


- Fixed `torch.sort` currently does not support bool dtype on CUDA ([#665](https://github.com/PyTorchLightning/metrics/pull/665))



## [0.6.1] - 2021-12-06

Expand Down
13 changes: 13 additions & 0 deletions tests/detection/test_map.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -215,6 +215,19 @@ def test_empty_preds():
metric.compute()


_gpu_test_condition = not torch.cuda.is_available()


@pytest.mark.skipif(_pytest_condition, reason="test requires that torchvision=>0.8.0 is installed")
@pytest.mark.skipif(_gpu_test_condition, reason="test requires CUDA availability")
def test_map_gpu():
"""Test predictions on single gpu."""
metric = MAP()
metric = metric.to("cuda")
metric.update(_inputs.preds[0], _inputs.target[0])
metric.compute()


@pytest.mark.skipif(_pytest_condition, reason="test requires that pycocotools and torchvision=>0.8.0 is installed")
def test_empty_metric():
"""Test empty metric."""
Expand Down
6 changes: 4 additions & 2 deletions torchmetrics/detection/map.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -407,7 +407,9 @@ def _evaluate_image(
ignore_area = (areas < area_range[0]) | (areas > area_range[1])

# sort dt highest score first, sort gt ignore last
ignore_area_sorted, gtind = torch.sort(ignore_area)
ignore_area_sorted, gtind = torch.sort(ignore_area.to(torch.uint8))
# Convert to uint8 temporarily and back to bool, because "Sort currently does not support bool dtype on CUDA"
ignore_area_sorted = ignore_area_sorted.to(torch.bool)
gt = gt[gtind]
scores = self.detection_scores[id]
scores_filtered = scores[det_label_mask]
Expand All @@ -429,7 +431,7 @@ def _evaluate_image(
for idx_iou, t in enumerate(self.iou_thresholds):
for idx_det in range(nb_det):
m = MAP._find_best_gt_match(t, nb_gt, gt_matches, idx_iou, gt_ignore, ious, idx_det)
if m is not -1:
if m != -1:
det_ignore[idx_iou, idx_det] = gt_ignore[m]
det_matches[idx_iou, idx_det] = True
gt_matches[idx_iou, m] = True
Expand Down