EvalScope integration: one-command evaluation support for PerceptionBench

Hi MoonshotAI/PerceptionBench maintainers,

EvalScope now supports PerceptionBench. This gives users another way to run the benchmark when they want a standardized evaluation workflow around OpenAI-compatible model endpoints, saved predictions, scoring, and report generation.

This may be useful for users who find this repository while setting up PerceptionBench and want a quick path to evaluate a hosted or locally served model.

Example:

evalscope eval \
  --model YOUR_MODEL \
  --api-url OPENAI_API_COMPAT_URL \
  --api-key EMPTY_TOKEN \
  --datasets perception_bench \
  --limit 10

Docs: https://evalscope.readthedocs.io/en/latest/benchmarks/perception_bench.html
EvalScope repo: https://github.com/modelscope/evalscope

Some benchmarks may require additional assets, media files, or judge-model configuration for a full run; the EvalScope guide above includes the benchmark-specific setup notes.

If this is useful for your community, we can follow up with a small README/Docs PR linking to the EvalScope usage guide.

Start by checking the README and existing documentation to find the appropriate location for an evaluation workflow link. Add a concise EvalScope section with the supplied command and guide URL, then verify that the example and benchmark-specific setup notes are accurate.