[serge] Fix 12 integration tests for model dac failing with output_mismatch (tensor values differ (6), other (6))

[sergereview]

Original CI failure

• Failure group: 1/1: 12 integration tests for model dacfailing withoutput_mismatch (tensor values differ (6), other (6))
• tests/models/dac/test_modeling_dac.py::DacIntegrationTest::test_integration_0_dac_16khz [multi-gpu] (output_mismatch, seen 7/7)

Relates to #46940

The DAC integration tests fail because DacVectorQuantize.decode_latents computes the codebook distance with misplaced parentheses. The current expression adds the codebook norm instead of subtracting it, so the nearest-neighbor lookup does not minimize Euclidean distance. This causes the quantized codes and decoded audio to drift from the expected values.

Fix the distance formula to negate the full squared-distance expression, matching the equivalent EnCodec/XCodec codebooks and the mathematical intent. This is a one-character parenthesis fix in src/transformers/models/dac/modeling_dac.py.

---

This change was produced automatically by serge from a CI failure report. The patch was generated by an LLM and applied by serge; review before merging.

serge v0.1.0 · model: moonshotai/Kimi-K2.7-Code · 27 LLM turns · 30 tool calls · 58.3s · 555794 in / 10331 out tokens

[github-actions]

[For maintainers] Suggested jobs to run (before merge)

run-slow: dac

[github-actions]

CI Dashboard: View test results in Grafana

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.