feat: Add T1_23dof robot support with WBT training and inference

[yunhanzheng]

Description

This PR adds comprehensive support for the T1_23dof humanoid robot, including configuration redirects, whole-body tracking (WBT) training support, and complete robot model files.

Changes Made

Configuration Updates

• Added src/holosoma/holosoma/config_values/wbt/t1/ with T1-specific configurations
• Updated config values in holosoma, inference, and retargeting modules
• Added T1_23dof redirect support in experiment and robot configurations

New Files Added

• Configurations: T1-specific WBT training configs (command.py, reward.py, termination.py, experiment.py)
• Motion Data: walking.npz for whole-body tracking (564KB)
• Robot Models: Complete T1_23dof robot files (27MB total)
  • XML definition files (t1_23dof.xml, t1_23dof_w_largebox.xml)
  • URDF file (t1_23dof.urdf)
  • STL mesh files (70+ files for 3D model components)

Documentation

• Updated retargeting README.md
• Added configuration examples and references

File Size Note

• Total robot model files: 27MB
• Consistent with existing robots in the project:
  • g1: 64MB
  • t1: 20MB
  • t1_23dof: 27MB (this PR)
• All files are required for T1_23dof robot operation

Testing

• Configurations have been validated locally
• Follows existing project structure and conventions
• Maintains backward compatibility with existing robots

Checklist

• My code follows the project's style guidelines
• I have performed a self-review of my own code
• I have commented my code where necessary
• 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

Related Issues

• No related issues (new feature addition)

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.

[samuelgundry]

Hi @yunhanzheng Thanks for the contribution! This looks like a valuable addition to the ecosystem.

However, we currently do not have the T1 23 DOF internally to validate your changes on sim2real, and we also want to keep our automated nightly/CI coverage across well-supported robots.

To ensure your work is accessible to the community, please consider submitting your changes to the HolosomaExtensions repository instead?   The idea here is to keep community-supported robots (and policies) in the extensions repository while the core repository is kept as the framework and primary robots and policies we are supporting.

A quick note on timing: We have upcoming improvements to the extensions interface that will make it much simpler and more flexible. Please wait until these are made available before porting your changes, as the new system will likely save you significant effort during the integration.

I'll follow-up on this PR here once the extensions repository is updated. Let us know any concerns.

Again, thanks for the effort. This will help the community.

[yunhanzheng]

Thanks for the review and feedback!

I understand the need to keep the core repository focused on well-supported robots. I'm happy to contribute this to the HolosomaExtensions repository instead.

I'll wait for your follow-up on the extensions repository updates. Looking forward to contributing this feature to the community!