[fix] an index error in browse dataset

[LoSealL]

No description provided.

[Copilot]

Pull Request Overview

This PR fixes an index error in the browse dataset functionality for MMSegmentation by correcting the tensor indexing pattern and updates the version number.

• Fixed incorrect tensor indexing that was causing an index error when browsing segmentation datasets
• Bumped version from 0.10.10 to 0.10.11

Reviewed Changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
mmengine/version.py	Version bump to 0.10.11
mmengine/cli/tasks/browse_dataset.py	Fixed tensor indexing from gt_sem_seg[:, idx] to gt_sem_seg[idx]

[Copilot]

The indexing change from gt_sem_seg[:, idx] to gt_sem_seg[idx] appears correct for fixing the index error, but this assumes gt_sem_seg has the batch dimension first. However, without seeing the tensor shape or how gt_sem_seg is constructed, this change could potentially break other cases where the original indexing was correct. Consider adding a comment explaining the expected tensor shape or adding shape validation.

Suggested change

	gt_sem_seg = data_sample.gt_sem_seg.data.copy()
	for idx, img in enumerate(imgs):
	if has_gt:
	gt_sem_seg = data_sample.gt_sem_seg.data.copy()
	# Expect gt_sem_seg to have shape [batch, H, W] or [N, H, W].
	# Validate shape to avoid indexing errors.
	assert gt_sem_seg.ndim == 3, (
	f"Expected gt_sem_seg to have 3 dimensions [batch, H, W], "
	f"but got shape {gt_sem_seg.shape}"
	)
	for idx, img in enumerate(imgs):
	if has_gt:
	# Assign the segmentation mask for the current image.