Nama-Diarization-Khmer · Nāma នាម Khmer speaker diarization

Work in progress: no model weights here yet. This repo will hold the first model when it is trained and tested.

Nāma ("who") will tell who is speaking when, in Khmer meetings, calls and podcasts, so Sota can write "who said what". It belongs to the same family as Sota (ASR, listens), Chakkhu (OCR, reads), Tossana (captioning, sees), Akkhara (keyboard, types) and VoxPocket (TTS, speaks).

Planned design (≈6M parameters, estimate)

• Speaker embeddings + clustering, tested on Khmer speech
• Pairs with Sota for speaker-labelled transcripts
• Export: int8 ONNX

Status

Step	Result
Test existing open models on Khmer	planned
First model	soon

License

To be decided with the first release.

Citation

@misc{ing2026nama,
  author    = {Ing, Muyleang},
  title     = {Nāma: Khmer speaker diarization},
  year      = {2026},
  publisher = {Hugging Face},
  url       = {https://huggingface.co/muyleanging/Nama-Diarization-Khmer}
}