CoolFace
Modelpublic

software-mansion/react-native-executorch-whisper-tiny.en

sourceHugging Faceapache-2.0updated 14d agoView on Hugging Face
1likes93kdownloads
README.md60 linesDownload Raw Back to root
1---2license: apache-2.03language:4- en5pipeline_tag: automatic-speech-recognition6library_name: executorch7---8 9# whisper-tiny.en10 11This repository hosts the **whisper-tiny.en** 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 15Upstream model: [whisper-tiny.en](https://huggingface.co/openai/whisper-tiny.en)16 17## Variants18 19| Path | Backend | Precision |20| --- | --- | --- |21| `coreml/whisper_tiny_en_coreml_fp16.pte` | coreml | fp16 |22| `mlx/whisper_tiny_en_mlx_bf16.pte` | mlx | bf16 |23| `mlx/whisper_tiny_en_mlx_int8.pte` | mlx | int8 |24| `vulkan/whisper_tiny_en_vulkan_int8.pte` | vulkan | int8 |25| `vulkan/whisper_tiny_en_vulkan_fp16.pte` | vulkan | fp16 |26| `xnnpack/whisper_tiny_en_xnnpack_int8.pte` | xnnpack | int8 |27| `xnnpack/whisper_tiny_en_xnnpack_fp32.pte` | xnnpack | fp32 |28 29## Repository structure30 31```32config.json                               37 B33coreml/config.json                        1.5 kB34coreml/whisper_tiny_en_coreml_fp16.pte    72.9 MB35mlx/config.json                           2.6 kB36mlx/whisper_tiny_en_mlx_bf16.pte          72.6 MB37mlx/whisper_tiny_en_mlx_int8.pte          57.1 MB38tokenizer.json                            2.3 MB39vulkan/config.json                        2.6 kB40vulkan/whisper_tiny_en_vulkan_fp16.pte    204 MB41vulkan/whisper_tiny_en_vulkan_int8.pte    124 MB42xnnpack/config.json                       2.6 kB43xnnpack/whisper_tiny_en_xnnpack_fp32.pte  222 MB44xnnpack/whisper_tiny_en_xnnpack_int8.pte  168 MB45```46 47## Compatibility48 49These files are published for the **ExecuTorch v1.4.1** runtime. ExecuTorch50gives no forward compatibility guarantee, so an older runtime may fail to load51them.52 53To use them in React Native ExecuTorch, pass the model constant shipped in the54library's model registry to the corresponding task pipeline. See the55[documentation](https://docs.swmansion.com/react-native-executorch/docs/fundamentals/downloading-models).56 57To load these files in your own ExecuTorch runtime, read the58[compatibility note](https://github.com/pytorch/executorch/blob/main/runtime/COMPATIBILITY.md)59first.60