# CoreML Model: Qwen/Qwen2.5-0.5B-Instruct

## Conversion Details
- **Source Model:** Qwen/Qwen2.5-0.5B-Instruct
- **Sequence Length:** 512
- **Quantization:** float16
- **Compute Units:** CPU_AND_NE
- **Converted At:** 2026-02-12 12:52:51 UTC
- **Runner:** arm64 / macOS 15.7.3
- **CoreMLTools:** 8.0

## Files
- `.mlpackage` — CoreML model package (for development)
- `.mlmodelc` — Compiled CoreML model (for deployment on device)

## Usage in Xcode / Swift
```swift
import CoreML

let config = MLModelConfiguration()
config.computeUnits = .all  // Uses ANE when available
let model = try MLModel(contentsOf: modelURL, configuration: config)
```

## Supported Devices
- iPhone 15 Pro+ (A17 Pro)
- iPhone 16+ (A18/A18 Pro)
- iPad with M1+ chip
- Mac with M1+ chip