Skip to content

Fixes #1198 - bug with CM in PascalVOC example #1200

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
merged 3 commits into from
Jul 14, 2020
Merged

Fixes #1198 - bug with CM in PascalVOC example #1200

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
bcc5cde
Fixes #1198
vfdev-5 Jul 14, 2020
2ea55cd
Explicit CM compute by all proc and logging by 0 rank proc
vfdev-5 Jul 14, 2020
10f927b
Merge branch 'master' into issue-1198
vfdev-5 Jul 14, 2020
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
19 changes: 10 additions & 9 deletions examples/references/segmentation/pascal_voc2012/code/scripts/training.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -179,7 +179,7 @@ def training(local_rank, config, logger=None):

# Setup evaluators
num_classes = config.num_classes
cm_metric = ConfusionMatrix(num_classes=num_classes)
cm_metric = ConfusionMatrix(num_classes=num_classes, average="recall")

val_metrics = {
"IoU": IoU(cm_metric),
Expand Down Expand Up @@ -241,16 +241,17 @@ def run_validation():
event_name=Events.ITERATION_COMPLETED(once=len(val_loader) // 2),
)

# Log confusion matrix to Trains:
if exp_tracking.has_trains:
from trains import Task
# Log confusion matrix to Trains:
if exp_tracking.has_trains:

@trainer.on(Events.COMPLETED)
def compute_and_log_cm():
cm = cm_metric.compute().cpu().numpy()

trains_logger = Task.current_task().get_logger()
if idist.get_rank() == 0:
from trains import Task

@trainer.on(Events.COMPLETED)
def log_cm():
cm = cm_metric.compute().numpy()
cm = cm / (cm.sum(axis=1)[:, None] + 1e-15)
trains_logger = Task.current_task().get_logger()
trains_logger.report_confusion_matrix(
title="Final Confusion Matrix",
series="cm-preds-gt",
Expand Down