Add Onnx inference info

README.md

@@ -163,22 +163,26 @@ python3 matcha_vocos_inference.py --output_path=/output/path --text_input="Bon d
 
 #### ONNX
 
-We also release ONNXs version of the models, you can use it via the OVOS plugins. 
-Just install the plugin from pypi. 
+We also release ONNXs version of the models, you can deploy the models throug a simple api:
+
+Steps:
+
+    Clone the repo. git clone https://github.com/langtech-bsc/minimal-tts-api
+
+```bash
+#Build and run the container.
+cd minimal-tts-api
+docker build -t minimal-tts-api .
+docker run -p 8000:8000 -t minimal-tts-api
+```
 
 ```bash
-pip install ovos-tts-plugin-matxa-multispeaker-cat
+#Test with a simple request.
+curl -X POST   http://0.0.0.0:8000/api/tts   -H "Content-Type: application/json"   -d '{"text":"Bon dia","voice":"quim","accent":"balear","type":"text"}'   | aplay -t wav -
 ```
 
-and synthesize 
 
-```python
-from ovos_tts_plugin_matxa_multispeaker_cat import MatxaCatalanTTSPlugin
 
-sent = "Això és una prova de síntesi de veu."
-tts = MatxaCatalanTTSPlugin()
-tts.get_tts(sent, "test.wav", voice="valencia/gina")
-```
 
 ### For Training