Add documentation how to train traffic light fine detector model

[lexavtanke]

Checklist

• I've read the contribution guidelines.
• I've searched other issues and no duplicate issues were found.
• I've agreed with the maintainers that I can plan this task.

Description

Write instruction how to train traffic light detection models and prepare them for the inference in the autoware.

• docs(traffic_light_fine_detector): update add training documentation autoware_universe#5704
• create a repo for training script and configs
• provide sample dataset
• [update autoware.documentation page ](docs(how-to-guides): add training docs for traffic_light_fine_detector #489)

Purpose

Provide autoware users ability to easily train traffic light detector models with their own data specific to their case or area.

Possible approaches

Create separate lightweight helping repo with the training scripts and configuration files based of mmdetection v2.28.2 (as used in TierIV internal usage).
Write instructions how to properly prepare dataset for the task, how to prepare env for training, how to run training, how to convert model to onnx.

Definition of done

Training documentation, conversion script and sample dataset are provided.

[stale]

This pull request has been automatically marked as stale because it has not had recent activity.

[vividf]

@lexavtanke Kindly ask what is the status of this one