Add CUDA graph kernel annotations tutorial (#3915)

This tutorial demonstrates how to use CUDA graph kernel annotations for
semantic profiling traces with custom visualization lanes.

Features:
- End-to-end workflow from graph capture to visualization
- Transformer block example with annotated regions
- Post-processing to merge annotations into profiler traces
- Custom stream assignments for semantic organization

The tutorial includes:
- mark_kernels() context manager usage
- Graph capture with enable_annotations=True
- Profiling and trace post-processing
- Before/after comparison
- Troubleshooting guide

Fixes #ISSUE_NUMBER

## Description
<!--- Describe your changes in detail -->

## Checklist
<!--- Make sure to add `x` to all items in the following checklist: -->
- [ ] The issue that is being fixed is referred in the description (see
above "Fixes #ISSUE_NUMBER")
- [ ] Only one issue is addressed in this pull request
- [ ] Labels from the issue that this PR is fixing are added to this
pull request
- [ ] No unnecessary issues are included into this pull request.

---------

Co-authored-by: yushangdi <yushangdi@users.noreply.github.com>
Co-authored-by: Claude Sonnet 4.5 <noreply@anthropic.com>

Author: yushangdi

.ci/docker/requirements.txt

@@ -26,6 +26,7 @@ fastapi
 matplotlib
 librosa
 torch==2.12
+cuda-bindings>=13.1.0  # Required for CUDA graph annotations tutorial
 torchvision
 torchdata
 networkx