Add Nemotron-3 diarization

[contentis]

No description provided.

[coderabbitai]

Navigate logical layers of code changes, visualize relationships, and explore their blast radius.

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Repository: NVIDIA/din-deploy/.coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: 7acfe880-70a2-432d-9d79-4faa271eef7f

📥 Commits

Reviewing files that changed from the base of the PR and between 1a9a191 and d7b232c.

📒 Files selected for processing (11)

• CMakeLists.txt
• THIRD_PARTY_NOTICES.md
• asr/diarization/CMakeLists.txt
• asr/diarization/README.md
• asr/diarization/diarization.cpp
• asr/diarization/diarization.h
• asr/diarization/main.cpp
• asr/diarization/model_export/export_diarization.py
• asr/diarization/model_export/mel.py
• asr/diarization/model_export/model.py
• asr/diarization/model_export/validate_diarization.py

Included review availability: Your plan provides up to 12 included reviews per hour; 8 remain after this review.

---

📝 Summary

Summary by CodeRabbit

• New Features
  • Added speaker diarization for 16 kHz audio, with speaker segments and tools to associate speakers with transcript word timestamps.
  • Added command-line support for running diarization with optional transcription and forced alignment.
• Documentation
  • Added guidance for using, configuring, and validating diarization, including model licensing information.

Walkthrough

Adds Nemotron diarization model export and validation tools, a C++ pipeline and CLI, and CMake integration. The CLI can combine diarization with Qwen3 transcription and forced alignment.

Changes

Nemotron diarization

Layer / File(s)	Summary
ONNX export and validation asr/diarization/model_export/*, asr/diarization/README.md, THIRD_PARTY_NOTICES.md	Adds log-mel processing and streaming model export, checkpoint-to-ONNX export, and validation against NeMo and native outputs. Documents export, validation, model behavior, and licensing.
C++ diarization runtime asr/diarization/diarization.h, asr/diarization/diarization.cpp, asr/diarization/CMakeLists.txt, CMakeLists.txt, asr/diarization/README.md	Adds the Config, Segment, Result, and Pipeline APIs. The pipeline runs chunked inference, produces speaker segments, and supports speaker lookup by interval. Adds the library to the CMake build.
CLI and ASR alignment asr/diarization/main.cpp, asr/diarization/CMakeLists.txt, asr/diarization/README.md	Adds a CLI for diarization with optional Qwen3 transcription and forced alignment. Aligned word timestamps include the speaker returned by Result::SpeakerAt.

Priority: ⬇️ Low

Estimated code review effort: 4 (Complex) | ~60 minutes

Merge Risk: ⚪ Minimal · up to d7b23

No concrete issue remains that needs resolution before merge; proceed with normal checks.

🚥 Pre-merge checks | ✅ 3 | ❌ 2

❌ Failed checks (1 warning, 1 inconclusive)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 4.76% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 21 functions across 7 files. (4 skipped: 4…	Write docstrings for the functions missing them to satisfy the coverage threshold.
Description check	❓ Inconclusive	No pull request description was provided, so the relationship between the changes and the intended implementation is not documented.	Add a description that summarizes the Nemotron-3 diarization implementation, including the runtime pipeline, model export and validation tools, CLI integration, and build changes.

✅ Passed checks (3 passed)

Check name	Status	Explanation
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Title check	✅ Passed	The title clearly and concisely describes the main change: adding Nemotron-3 diarization support, including the implementation, CLI, model export tools, validation, and documentation.

Full details: Docstring Coverage

Explanation

Docstring coverage is 4.76% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 21 functions across 7 files. (4 skipped: 4 unsupported.)

• Fix all pre-merge checks with AI

✨ Finishing Touches 💡 1

📝 Generate docstrings 💡

• Commit to this branch
• Create a new PR

🧪 Generate unit tests (beta)

• Commit to this branch
• Create a new PR

---

_{Comment @coderabbitai help to get the list of available commands.}