Bump to v0.2.0

[kew6688]

Motivation

Refactoring the evaluation workflow makes it correct at scale, easier to change, and faster to iterate.

Modification

Bump version to v0.2.0

BC-breaking (Optional)

Does the modification introduce changes that break the back-compatibility of the downstream repos?
If so, please describe how it breaks the compatibility and how the downstream projects should modify their code to keep compatibility with this PR.

Major breaks in evaluation usage:

• env_type 'vln_multi' => 'internutopia' / 'habitat'
• eval_type vln_multi => vln_distributed
• new eval_settings usa_agent_server
• removed eval_habitat.py
• some modules were refactored to new interface

Use cases (Optional)

If this PR introduces a new feature, it is better to list some use cases here, and update the documentation.

Distributed evaluation now are supported for slurm, torchrun, alicloud
scripts/eval/bash/eval_vln_distributed.sh {habitat|internutopia|internutopia_vec_env} [--config xxx]

Checklist

Pre-commit or other linting tools are used to fix the potential lint issues.
The modification is covered by complete unit tests. If not, please add more unit test to ensure the correctness.
If the modification has potential influence on downstream projects, this PR should be tested with downstream projects.
The documentation has been modified accordingly, like docstring or example tutorials.

Changelog of v0.2.0 (2025/12/4)

Highlights

• Support distributed evaluation for VLN-PE, reducing full benchmark runtime to ~1.6 hours using 16 GPUs (≈13× speedup over single-GPU eval) ([Fix] Habitat Refactor & Support Distributed VLN-PE Evaluation #168)
• Enhance Habitat evaluation flow with DistributedEvaluator and HabitatEnv integrated into the InternNav framework ([Fix] Habitat Refactor & Support Distributed VLN-PE Evaluation #168)
• Support install flags for dependency isolation: [habitat], [isaac], [model] ([FIX] solve import issue, isolate dependency and requirements #135)

New Features

• Support distributed evaluation for VLN-PE ([Fix] Habitat Refactor & Support Distributed VLN-PE Evaluation #168)
• Support a unified evaluation script eval.py, with new Habitat evaluation configs in scripts/eval/configs ([Fix] Habitat Refactor & Support Distributed VLN-PE Evaluation #168)
• Support install flags for dependency isolation ([Fix] Habitat Refactor & Support Distributed VLN-PE Evaluation #168)

Improvements

• Add HabitatEnv with episode pool management ([Fix] Habitat Refactor & Support Distributed VLN-PE Evaluation #168)
• Update InternUtopiaEnv for distributed execution and episode pool management ([Fix] Habitat Refactor & Support Distributed VLN-PE Evaluation #168)
• Enhance episode_loader in VLN-PE with new distributed mode compatibility ([Fix] Habitat Refactor & Support Distributed VLN-PE Evaluation #168)
• Update data_collector to support progress checkpointing and incremental result aggregation in distributed evaluation. ([Fix] Habitat Refactor & Support Distributed VLN-PE Evaluation #168)

Bug Fixes

• Fix logger disabled after Isaac Sim initialization during evaluator bootstrap ([Fix] Habitat Refactor & Support Distributed VLN-PE Evaluation #168)
• Fix dataloader bug where revise_one_data() was incorrectly applied to all datasets ([Fix] Habitat Refactor & Support Distributed VLN-PE Evaluation #168)
• Fix visualization images dimension mismatch during InternVLA-N1 evaluation ([Fix] Habitat Refactor & Support Distributed VLN-PE Evaluation #168)
• Fix distributed evaluation crash in rdp policy ([Fix] Habitat Refactor & Support Distributed VLN-PE Evaluation #168)
• Fix GitHub CI tests ([Fix] Habitat Refactor & Support Distributed VLN-PE Evaluation #168)

Contributors

A total of 3 developers contributed to this release.
@kew6688, @Gariscat, @yuqiang-yang

[Tai-Wang]

"with major codebase refactoring, bug fixes, and stability improvements"
To be specific, if nothing special, just remove it

[Tai-Wang]

distributed evaluation support should be for benchmarks instead of a specific model, InternVLA-N1?

[kew6688]

Changed to InternNav v0.2.0 released — added distributed evaluation support for VLN-PE.