refactor(lidar_centerpoint): add training docs

[kaancolak]

Description

Add training documentation about training lidar CenterPoint model.

Blocks merging of:

• https://github.com/autowarefoundation/autoware-documentation/pull/471

Blocked by:

• https://github.com/autowarefoundation/mmdetection3d/pull/1

Tests performed

Not applicable.

Effects on system behavior

Not applicable.

Pre-review checklist for the PR author

The PR author must check the checkboxes below when creating the PR.

• [X] I've confirmed the contribution guidelines.
• [X] The PR follows the pull request guidelines.

In-review checklist for the PR reviewers

The PR reviewers must check the checkboxes below before approval.

• [ ] The PR follows the pull request guidelines.

Post-review checklist for the PR author

The PR author must check the checkboxes below before merging.

• [ ] There are no open discussions or they are tracked via tickets.

After all checkboxes are checked, anyone who has write access can merge the PR.

[xmfcx]

@kaancolak could you rebase this to latest main?

[github-actions[bot]]

Documentation URL: https://autowarefoundation.github.io/autoware.universe/pr-5570/ Modified URLs:

• https://autowarefoundation.github.io/autoware.universe/pr-5570/perception/lidar_centerpoint/

[xmfcx]

/review

[github-actions[bot]]

PR Review

⏱️ Estimated effort to review [1-5]	4, due to the extensive addition of documentation covering a wide range of steps including environment setup, training, evaluation, and deployment of a model. The complexity and detail of the instructions require careful review to ensure accuracy and clarity.
🧪 Relevant tests	No
🔍 Possible issues	Possible Issue: The documentation does not specify the version of CUDA required for compatibility with PyTorch 1.13.1. This could lead to environment setup issues for users.
	Possible Issue: The documentation assumes the user has prior knowledge of Docker, including building and running containers. This could be a barrier for users unfamiliar with Docker.
🔒 Security concerns	No

Code feedback:

---

relevant file	perception/lidar_centerpoint/README.md
suggestion	Consider specifying the CUDA version compatible with PyTorch 1.13.1 to avoid potential environment setup issues. This clarification will help users ensure they have a compatible setup before proceeding with the installation. [important]
relevant line	conda install pytorch==1.13.1 torchvision==0.14.1 pytorch-cuda=11.6 -c pytorch -c nvidia

---

relevant file	perception/lidar_centerpoint/README.md
suggestion	Add a brief introduction or link to Docker documentation for users who might be unfamiliar with Docker concepts such as building images and running containers. This will make the guide more accessible to a broader audience. [medium]
relevant line	Alternatively, you can use Docker to run the mmdetection3d repository.We provide a Dockerfile to build a Docker image with the mmdetection3d repository and its dependencies.

---

---

✨ Review tool usage guide:

---

Overview: The review tool scans the PR code changes, and generates a PR review which includes several types of feedbacks, such as possible PR issues, security threats and relevant test in the PR. More feedbacks can be added by configuring the tool.

The tool can be triggered automatically every time a new PR is opened, or can be invoked manually by commenting on any PR.

• When commenting, to edit configurations related to the review tool (pr_reviewer section), use the following template:

/review --pr_reviewer.some_config1=... --pr_reviewer.some_config2=...

• With a configuration file, use the following template:

[pr_reviewer]
some_config1=...
some_config2=...

See the review usage page for a comprehensive guide on using this tool.

[xmfcx]

/improve

[github-actions[bot]]

PR Code Suggestions

Category	Suggestions
Enhancement	Add a step to verify the installation of Miniconda. Consider adding a step to verify the successful installation of Miniconda or any other prerequisites. This can be done by running a simple command that checks the version or lists the installed packages. This step will help users ensure that the installation was successful before proceeding to the next steps. perception/lidar_centerpoint/README.md [82] **Step 1.** Download and install Miniconda from the [official website](https://mmpretrain.readthedocs.io/en/latest/get_started.html). +**Step 1.1.** Verify the installation by running `conda --version`.
	Add explanations for Docker run command flags to aid user understanding. For the Docker run command, it's beneficial to include an explanation for each flag used, especially for users who might not be familiar with Docker. This helps in understanding the purpose of each option and how it affects the container's behavior. perception/lidar_centerpoint/README.md [142] docker run --gpus all --shm-size=8g -it -v {DATA_DIR}:/mmdetection3d/data mmdetection3d +# Explanation: +# --gpus all: Allows the container to use all available GPUs. +# --shm-size=8g: Sets the shared memory size to 8GB, which is useful for data loading. +# -it: Runs the container in interactive mode with a tty. +# -v {DATA_DIR}:/mmdetection3d/data: Mounts the host directory {DATA_DIR} to the container's /mmdetection3d/data directory.
	Add an explanation of the evaluation outcome and how to interpret it. For the evaluation section, it would be beneficial to include a brief explanation of the expected outcome or what constitutes a successful evaluation. This could include information on the metrics used for evaluation, expected ranges for these metrics, or how to interpret the results. This guidance can help users better understand the evaluation process and what to look for in the results. perception/lidar_centerpoint/README.md [210] Run evaluation +# After running the evaluation, you will receive a report detailing the performance of the model on the dataset. Look for metrics such as precision, recall, and mAP (mean Average Precision) to gauge the model's accuracy. A higher mAP value indicates better performance.
Best practice	Specify the CUDA version compatible with PyTorch to avoid compatibility issues. It's recommended to specify the version of CUDA that is compatible with the installed version of PyTorch. This ensures that users install a version of CUDA that is tested and guaranteed to work with the specified PyTorch version, avoiding potential compatibility issues. perception/lidar_centerpoint/README.md [96] -conda install pytorch==1.13.1 torchvision==0.14.1 pytorch-cuda=11.6 -c pytorch -c nvidia +conda install pytorch==1.13.1 torchvision==0.14.1 cudatoolkit=11.6 -c pytorch -c nvidia
	Include a note about storage requirements for the NuScenes dataset. When providing instructions for dataset preparation, it's helpful to include a note or warning about the storage requirements. Large datasets like NuScenes can consume significant disk space, and users should be made aware of this to ensure they have adequate storage available before downloading and extracting the dataset. perception/lidar_centerpoint/README.md [147] **Step 1.** Download the NuScenes dataset from the [official website](https://www.nuscenes.org/download) and extract the dataset to a folder of your choice. +> **Note:** The NuScenes dataset is large and requires significant disk space. Ensure you have enough storage available before proceeding.

---

✨ Improve tool usage guide:

---

Overview: The improve tool scans the PR code changes, and automatically generates suggestions for improving the PR code. The tool can be triggered automatically every time a new PR is opened, or can be invoked manually by commenting on a PR.

• When commenting, to edit configurations related to the improve tool (pr_code_suggestions section), use the following template:

/improve --pr_code_suggestions.some_config1=... --pr_code_suggestions.some_config2=...

• With a configuration file, use the following template:

[pr_code_suggestions]
some_config1=...
some_config2=...

See the improve usage page for a comprehensive guide on using this tool.

[codecov[bot]]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 14.99%. Comparing base (e24f48d) to head (ef15632). Report is 150 commits behind head on main.

:exclamation: Current head ef15632 differs from pull request most recent head bb2a69e

Please upload reports for the commit bb2a69e to get more accurate results.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #5570      +/-   ##
==========================================
- Coverage   15.10%   14.99%   -0.12%     
==========================================
  Files        1922     1933      +11     
  Lines      134796   133074    -1722     
  Branches    43473    39720    -3753     
==========================================
- Hits        20364    19951     -413     
+ Misses      91780    90888     -892     
+ Partials    22652    22235     -417     

Flag	Coverage Δ		*Carryforward flag
differential	0.00% <ø> (?)
total	14.99% <ø> (-0.12%)	:arrow_down:	Carriedforward from 3345fd8

*This pull request uses carry forward flags. Click here to find out more.

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[Shin-kyoto]

@kaancolak

It seems that there is an issue with spell check. Could you please resolve it?