Add MLX model & restructurize the repository

.gitattributes

@@ -33,4 +33,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
-xnnpack/** filter=lfs diff=lfs merge=lfs -text
+xnnpack/** filter=lfs diff=lfs merge=lfs -text
+mlx/** filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,3 +1,47 @@
 ---
 license: openrail
 ---
+
+# Supertonic 3 TTS
+
+This repository hosts the **Supertonic 3** model for the [React Native ExecuTorch](https://github.com/software-mansion/react-native-executorch) library. It performs text-to-speech synthesis supporting **30+ languages** (🇬🇧 English, 🇪🇸 Spanish, 🇫🇷 French, 🇩🇪 German, 🇵🇱 Polish, 🇮🇳 Hindi, 🇯🇵 Japanese, 🇨🇳 Chinese, and many more), with a single voice style per language.
+
+The model is composed of four sub-models that run sequentially:
+
+1. **Duration predictor** — estimates speech duration from text
+2. **Text encoder** — encodes text into a style-conditioned representation
+3. **Vector estimator** — flow-matching denoiser that generates the audio latent
+4. **Vocoder** — decodes the latent into a 44.1 kHz waveform
+
+## Compatibility
+
+These models were exported using **v1.3.1** of ExecuTorch and no forward compatibility is guaranteed. Older versions of the runtime may not work with these files.
+
+The models are intended to be used within the React Native ExecuTorch package. If you want to use them outside the package, make sure your runtime is compatible with the ExecuTorch version used to export the `.pte` files and follow the example scripts to run the models.
+
+## Backends
+
+| Backend | Description | RTF (Apple Silicon) |
+|---------|-------------|---------------------|
+| `xnnpack` | CPU-optimized via XNNPACK delegate | ~0.07 (14× faster than real-time) |
+| `mlx` | Apple Silicon GPU via MLX delegate | ~0.026 (38× faster than real-time) |
+
+## Repository Structure
+
+```
+.
+├── config.json              # Backend-agnostic model manifest
+├── unicode_indexer.json     # Character-to-id mapping for text preprocessing
+├── voices/                  # Pre-computed speaker embeddings
+│   ├── M1.json
+│   ├── M2.json
+│   ├── F1.json
+│   └── ...
+├── xnnpack/                 # XNNPACK-exported .pte files
+└── mlx/                     # MLX-exported .pte files
+```
+
+Each `.pte` file exposes two methods:
+
+- `forward` — the sub-model inference
+- `get_dynamic_dims_forward` — returns per-input `[rank, 3]` shape constraints (`[min, max, step]`) for runtime input validation

config.json

@@ -0,0 +1,3 @@
+{
+  "modelName": "supertonic"
+}

xnnpack/tts.json → mlx/config.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:42078d3aef1cd43ab43021f3c54f47d2d75ceb4e75f627f118890128b06a0d09
-size 8253
+oid sha256:c779e14e47a87185026c5adccfc46263fd46d31367c82cd7ff6e9b62c7a5faa5
+size 6460

xnnpack/duration_predictor.pte → mlx/duration_predictor_mlx_fp32.pte

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:11688dd5b20661ccf0c419d5744241cfb45a194754d60b4dd7b86e9b0d888264
-size 3578372
+oid sha256:fd6eb5abeedb9def30b04a225f0c3816aa8c21383b5c552c6f64c91845170b50
+size 3518884

xnnpack/text_encoder.pte → mlx/text_encoder_mlx_fp32.pte

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:94ef59b924e13ee9d01ec2135c09a223de81fc81dabaf81ae86b6dd75a3c7e1f
-size 36208640
+oid sha256:f5455739ff38100edc04cbe0fcca42a60792f3ba61e1edcd2426d8db47222db2
+size 36093604

mlx/vector_estimator_mlx_fp32.pte

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:80e1c81f86b927e389fd0d6506e7bfd837c984a40ad4e0344342676bdbb414f6
+size 256235916

mlx/vocoder_mlx_fp32.pte

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b478056c9143ea531954d822cc15892f6d01e9a75ddd4c4788cafadd70c5b65a
+size 101400868

xnnpack/{unicode_indexer.json → config.json}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9bf7346e43883a81f8645c81224f786d43c5b57f3641f6e7671a7d6c493cb24f
-size 277676
+oid sha256:78dc916dc0035e57ffc4fb227b2937c48f1010834343032816b7b021a799ec02
+size 6484

xnnpack/duration_predictor_xnnpack_fp32.pte

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77638ffde9137f060b4ed2fd024894e1ff812ec1098a22363bdf2aa72202e046
+size 3572484

xnnpack/text_encoder_xnnpack_fp32.pte

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6647a2ff007022577111482c5735c5b756b3e09df3a649a1427295ef9a210f2c
+size 36198528

xnnpack/vector_estimator.pte

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:5ecbc66791174a484f4eddb73eefd3aa308605dc305887a2e7f89e63350c9ddc
-size 256463744

xnnpack/vector_estimator_xnnpack_fp32.pte

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48fcb112df733302c1cc305c4cb8de64e18cfda04373397d17f77e30dc2d745b
+size 256437888

xnnpack/vocoder.pte

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:d47129247148fd01870188de5e7f3e18381a8088f73c4715457836a7fd5cf68e
-size 101435136

xnnpack/vocoder_xnnpack_fp32.pte

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7933c5a2b6cccedb5851be5de0d549dfda5cc228f4bfd9096726182a9b04d08a
+size 101429632