[PyTorch Debug] Custom feature tutorial.

[pggPL]

Description

This PR adds short custom feature tutorial to precision debug tools docs.

Type of change

• Documentation change (change only to the documentation, either a fix or a new content)

• Bug fix (non-breaking change which fixes an issue)

• New feature (non-breaking change which adds functionality)

• Breaking change (fix or feature that would cause existing functionality to not work as expected)

• Infra/Build change

• Code refactoring

• I have read and followed the contributing guidelines

• The functionality is complete

• I have commented my code, particularly in hard-to-understand areas

• I have made corresponding changes to the documentation

• My changes generate no new warnings

• I have added tests that prove my fix is effective or that my feature works

• New and existing unit tests pass locally with my changes

[pggPL]

It's better to debug this PR with docs generated, instead of looking into changes.

[Copilot]

Pull Request Overview

This PR adds a tutorial for creating custom features in PyTorch debug tools, demonstrating how to implement and use a custom PercentageGreaterThanThreshold feature.

• Adds a custom feature implementation that calculates the percentage of tensor values exceeding a threshold
• Provides utility functions for plotting statistics from debug logs
• Includes YAML configuration example showing how to enable custom features alongside built-in ones

Reviewed Changes

Copilot reviewed 4 out of 5 changed files in this pull request and generated 2 comments.

File	Description
docs/debug/custom_feature_dir/percentage_greater_than_threshold.py	Implements the custom PercentageGreaterThanThreshold feature class with tensor inspection methods
docs/debug/custom_feature_dir/utils.py	Provides plotting utilities to visualize statistics from debug logs
docs/debug/custom_feature_dir/custom_feature_example_config.yaml	Configuration file demonstrating how to enable both custom and built-in features
docs/debug.rst	Adds reference to the new custom feature tutorial notebook

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[Copilot]

The union type syntax torch.Tensor | QuantizedTensor uses Python 3.10+ syntax. Consider using Union[torch.Tensor, QuantizedTensor] from the typing module for better backward compatibility, especially since Optional from typing is already imported.

[Copilot]

The import re statement should be moved to the top of the file with other imports (after line 9) following Python's PEP 8 style guidelines.