Pyannote Diarization — Core ML

Speaker Identification

Speaker diarization: who spoke when. 16 kHz mono, 10s segments.

Core ML conversion of pyannote/pyannote-audio for on-device inference on iPhone, iPad and Mac. Converted with coremltools; the packages are stateless, so all sequencing and buffering lives in your Swift code.


Task	voice activity detection
Upstream	pyannote/pyannote-audio
Packages	1
Download size	5 MB
Minimum iOS	17.0
Peak RAM	~200 MB

Files

File	Size	Compute units	SHA-256
`SpeakerSegmentation.mlpackage.zip`	5 MB	`cpuAndGPU`	`dcfa2b98900f2b99…`
Total	5 MB

compute_units is not a suggestion -- it is the configuration the conversion was verified against. Moving a package to a different compute unit can silently change the numerics (FP16 attention overflow) or crash on the GPU.

Download

hf download mlboydaisuke/coreml-zoo --include "diarization/*" --local-dir ./diarization
unzip './diarization/diarization/*.zip' -d ./diarization

Use in Swift

import CoreML

let config = MLModelConfiguration()
config.computeUnits = .cpuAndGPU   // as converted — see the table above

// Unzip the .mlpackage, drop it into your Xcode target and Xcode compiles it
// at build time:
let model = try SpeakerSegmentation(configuration: config)

// ...or compile a downloaded .mlpackage at runtime:
let compiled = try await MLModel.compileModel(at: mlpackageURL)
let model = try MLModel(contentsOf: compiled, configuration: config)

Demo

Sample app — sample_apps/DiarizationDemo, a standalone SwiftUI project.
Models Zoo — this model is downloadable and runnable inside the Models Zoo app on the App Store, no build required.

Conversion

Script: convert_diarization.py
Pitfalls hit during conversion (FP16 overflow, ANE buffer limits, stride handling): docs/coreml_conversion_notes.md
Model index: CoreML-Models