docs: say examples support offline evaluation only

[fuergaosi233]

Motivation

examples/README.md on main contradicts itself about online evaluation.

Same file, same page:

1. Supported-looking claim (now fixed):

   Optional config sections provide online/offline evaluation with <run_id>-best selection

2. Contradictory claim (left as-is; matches code):

   Online evaluation is also unsupported.

Code on main (specforge/config/schema.py) rejects online eval:

if self.data.eval_data_path:
    raise ValueError(
        "data.eval_data_path is unsupported; online evaluation is not "
        "supported by the server-only capture path"
    )
...
if self.data.eval_hidden_states_path and mode != "offline":
    raise ValueError(
        "data.eval_hidden_states_path requires an offline training data source"
    )

Tests: test_online_evaluation_is_explicitly_unsupported and
test_offline_eval_source_requires_offline_training in
tests/test_config/test_schema.py.

Modifications

One wording change in examples/README.md: online/offline evaluation → offline evaluation.

The earlier "Online evaluation is also unsupported." sentence is unchanged.

Related Issues

N/A

Accuracy Test

N/A (docs only)

Benchmark & Profiling

N/A (docs only)

Checklist

• Format your code according to the Code Formatting with Pre-Commit.
• Add unit tests as outlined in the Running Unit Tests.
• Update documentation / docstrings / example tutorials as needed.
• Provide throughput / latency benchmark results and accuracy evaluation results as needed.
• For reviewers: If you haven't made any contributions to this PR and are only assisting with merging the main branch, please remove yourself as a co-author when merging the PR.
• Please feel free to join our Slack channel to discuss your PR.