Update README.md

README.md

@@ -9,13 +9,19 @@ pipeline_tag: text-to-speech
 
 `wfloat-tts` is a lightweight multi-speaker English VITS text-to-speech model with speaker, emotion, and intensity control.
 
-This repo includes:
+## On-Device packages
 
-- `model.safetensors`: inference weights
-- `config.json`: model config and token mapping
-- `src/wfloat_tts/`: a small Python inference helper
+This Hugging Face repo contains the model files.
 
-The repo is set up for standalone inference from the released model files. You do not need the original training codebase to synthesize speech with it.
+Wfloat also ships packages that distribute and run `wfloat-tts` locally on the user's device.
+
+Available packages:
+
+- [Web](https://github.com/wfloat/wfloat-web) for running in the browser, including mobile browsers
+- [React Native](https://github.com/wfloat/react-native-wfloat) for running locally in iOS and Android apps
+- [Python](https://github.com/wfloat/wfloat-python) for running in Python environments
+
+Missing the platform or framework you need? [Please request it!](https://docs.google.com/forms/d/e/1FAIpQLScLjcb4lkouSQ54ZWDKJ1xlCkUpBFamF1zKRO3fno1vp1Y_IQ/viewform?usp=preview)
 
 ## Sample Outputs
 
@@ -70,6 +76,10 @@ You do not need to pass raw control symbols. The Python helper converts `emotion
 
 ## Install
 
+For running the model from Hugging Face.
+
+Official Python package: [wfloat-python](https://github.com/wfloat/wfloat-python).
+
 ```bash
 pip install -e .
 pip install "piper-phonemize==1.3.0" -f https://k2-fsa.github.io/icefall/piper_phonemize
@@ -163,3 +173,5 @@ Supported emotion labels:
 - `model.safetensors` is the main inference artifact in this repo.
 - `config.json` includes the token mapping needed by the processor.
 - The current release uses a multi-speaker model with 20 speakers.
+- Training code: [https://github.com/wfloat/piper](https://github.com/wfloat/piper)
+- For the checkpoint needed to resume training, message `mitch@wfloat.com`.