Team Ai
Modelpublic

software-mansion/react-native-executorch-phi-4-mini

sourceHugging Facemitupdated 1mo agoView on Hugging Face
1likes550downloads
README.md49 linesDownload Raw Back to root
1---2license: mit3pipeline_tag: text-generation4tags:5- executorch6library_name: executorch7---8 9# phi-4-mini10 11This repository hosts the **phi-4-mini** models exported for the12[React Native ExecuTorch](https://www.npmjs.com/package/react-native-executorch)13library as ExecuTorch `.pte` programs, ready to run on device.14 15## Variants16 17| Path | Backend | Precision |18| --- | --- | --- |19| `mlx/phi_4_mini_mlx_int4.pte` | mlx | 4w |20| `xnnpack/phi_4_mini_xnnpack_bf16.pte` | xnnpack | bf16 |21| `xnnpack/phi_4_mini_xnnpack_8da4w.pte` | xnnpack | 8da4w |22 23## Repository structure24 25```26config.json                           32 B27mlx/config.json                       1.3 kB28mlx/phi_4_mini_mlx_int4.pte           3.0 GB29tokenizer.json                        14.8 MB30tokenizer_config.json                 2.9 kB31xnnpack/config.json                   2.3 kB32xnnpack/phi_4_mini_xnnpack_8da4w.pte  2.6 GB33xnnpack/phi_4_mini_xnnpack_bf16.pte   7.1 GB34```35 36## Compatibility37 38These files are published for the **ExecuTorch v1.4.1** runtime. ExecuTorch39gives no forward compatibility guarantee, so an older runtime may fail to load40them.41 42To use them in React Native ExecuTorch, pass the model constant shipped in the43library's model registry to the corresponding task pipeline. See the44[documentation](https://docs.swmansion.com/react-native-executorch/docs/fundamentals/downloading-models).45 46To load these files in your own ExecuTorch runtime, read the47[compatibility note](https://github.com/pytorch/executorch/blob/main/runtime/COMPATIBILITY.md)48first.49