Downloads · 30 days
5
13% of all-time downloads
Parakeet-TDT-v3-CoreML-INT4
aitytech/Parakeet-TDT-v3-CoreML-INT4
Parakeet-TDT-v3-CoreML-INT4 is a machine learning model from aitytech. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as cc-by-4.0.
CoreML conversion of NVIDIA Parakeet-TDT 0.6B v2 with INT4-quantized encoder for Apple Neural Engine acceleration.
All-time downloads
40
Public
Repo size
675 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin346 MB · 100%
At a glance
- License
- cc-by-4.0
- Access
- Public
- Created
- Mar 1, 2026
- Updated
- Mar 1, 2026
- SHA
- e548cac2
Base models
Model card
From the Hugging Face model README
Parakeet TDT v3 — CoreML INT4
CoreML conversion of NVIDIA Parakeet-TDT 0.6B v2 with INT4-quantized encoder for Apple Neural Engine acceleration.
Models
| Model | Description | Compute | Quantization |
|---|---|---|---|
encoder.mlmodelc | FastConformer encoder (24L, 1024 hidden) | CPU + Neural Engine | INT4 palettized |
decoder.mlmodelc | LSTM prediction network (2L, 640 hidden) | CPU + Neural Engine | FP16 |
joint.mlmodelc | TDT dual-head joint (token + duration logits) | CPU + Neural Engine | FP16 |
Additional Files
| File | Description |
|---|---|
vocab.json | SentencePiece vocabulary (1024 tokens) |
config.json | Model configuration |
Notes
- Mel preprocessing is done in Swift using Accelerate/vDSP (not CoreML) because
torch.stfttracing bakes audio length as a constant, breaking per-feature normalization for variable-length inputs. - Encoder uses
EnumeratedShapes(100–3000 mel frames, covering 1–30s audio) to avoid BNNS crashes with dynamic shapes. - Performance: ~110x RTF on M4 Pro via Neural Engine.
Usage
Used by qwen3-asr-swift ParakeetASR module:
let model = try await ParakeetASRModel.fromPretrained()
let text = try model.transcribeAudio(samples, sampleRate: 16000)
- License
- cc-by-4.0
- Languages
- en
- Created
- Mar 1, 2026
- Updated
- Mar 1, 2026