CoolFace
Modelpublic

aufklarer/Qwen3-ForcedAligner-0.6B-CoreML-INT4

sourceHugging Faceapache-2.0updated 6mo agoView on Hugging Face
0likes34downloads
Model Card

Qwen3-ForcedAligner-0.6B — CoreML INT4

CoreML conversion of Qwen/Qwen3-ForcedAligner-0.6B with INT4 palettization for Apple Neural Engine.

Predicts word-level timestamps in a single forward pass.

Models

ModelDescriptionQuantization
encoder.mlmodelcAudio encoder (24 layers)INT4 palettized
decoder.mlmodelcText decoder + classify head (28 layers)INT4 palettized

Usage

swift
let aligner = try await CoreMLForcedAligner.fromPretrained(
    modelId: "aufklarer/Qwen3-ForcedAligner-0.6B-CoreML-INT4"
)
let aligned = aligner.align(audio: samples, text: "Hello world", sampleRate: 24000)

Variants

VariantBackendSizeModel ID
CoreML INT4Neural Engine~630 MBaufklarer/Qwen3-ForcedAligner-0.6B-CoreML-INT4
CoreML INT8Neural Engine~1.0 GBaufklarer/Qwen3-ForcedAligner-0.6B-CoreML-INT8
MLX 4-bitGPU~979 MBaufklarer/Qwen3-ForcedAligner-0.6B-4bit
MLX 8-bitGPU~1.4 GBaufklarer/Qwen3-ForcedAligner-0.6B-8bit

Links