Document LightGBM quantile regression with FLAML

[Copilot]

Why are these changes needed?

Adds documentation for using FLAML AutoML with LightGBM quantile regression, reflecting the maintainer response in issue #1517. The example shows using task="regression", a custom pinball loss metric, and custom_hp to set LightGBM objective="quantile" and alpha.

Related issue number

Closes #1517

Checks

• I've used pre-commit to lint the changes in this PR (note the same in integrated in our CI checks). (pre-commit is not installed in this environment.)
• I've included any doc changes needed for https://microsoft.github.io/FLAML/. See https://microsoft.github.io/FLAML/docs/Contribute#documentation to build and test documentation locally.
• I've added tests (if relevant) corresponding to the changes introduced in this PR. (Documentation-only change.)
• I've made sure all auto checks have passed. (Automated validation passed locally via available tools; awaiting PR CI.)

[Copilot]

Pull request overview

Adds a new documentation example showing how to use FLAML AutoML with LightGBM quantile regression (per issue #1517), using a custom pinball-loss metric and custom_hp to fix LightGBM’s objective="quantile" and alpha.

Changes:

• Documented a LightGBM quantile regression workflow using task="regression" and a custom metric based on mean_pinball_loss.
• Demonstrated using custom_hp to set fixed LightGBM constructor parameters (objective, alpha) while tuning other hyperparameters.

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.

[thinkall]

@copilot sync with main branch

[Copilot]

@copilot sync with main branch

Fetched the latest origin/main and verified this branch is already synced: a47bce1 is 1 commit ahead and 0 behind, so no merge commit was needed.