mirror of
https://github.com/home-assistant/home-assistant.io.git
synced 2025-05-16 16:08:58 +00:00
52 lines
3.7 KiB
Markdown
52 lines
3.7 KiB
Markdown
---
|
|
title: "Installing a local Assist pipeline"
|
|
---
|
|
|
|
In Home Assistant, the Assist pipelines are made up of various components that together form a voice assistant.
|
|
|
|
For each component you can choose from different options. We have prepared a speech-to-text and text-to-speech option that runs fully local.
|
|
|
|
The speech-to-text option is [Whisper](https://github.com/openai/whisper). It's an open source AI model that supports [various languages](https://github.com/openai/whisper#available-models-and-languages). We use a forked version called [faster-whisper](https://github.com/guillaumekln/faster-whisper). On a Raspberry Pi 4, it takes around 8 seconds to process incoming voice commands. On an Intel NUC it is done in under a second.
|
|
|
|
For text-to-speech we have developed [Piper](https://github.com/rhasspy/piper). Piper is a fast, local neural text to speech system that sounds great and is optimized for the Raspberry Pi 4. It supports [many languages](https://rhasspy.github.io/piper-samples/). On a Raspberry Pi, using medium quality models, it can generate 1.6s of voice in a second.
|
|
|
|
## Installing a local Assist pipeline
|
|
|
|
For the quickest way to get your local Assist pipeline started, follow these steps:
|
|
|
|
1. Install the add-ons to convert text into speech and vice versa.
|
|
* Install the {% my supervisor_addon addon="core_whisper" title="**Whisper**" %} and the {% my supervisor_addon addon="core_piper" title="**Piper**" %} add-ons.
|
|
|
|
* Start both add-ons.
|
|
* Once the add-ons are started, head over to the integrations under {% my integrations title="**Settings** > **Devices & Services**" %}.
|
|
* You should now see Piper and Whisper being discovered by the [Wyoming integration](/integrations/wyoming/).
|
|
|
|
* For both integrations, select **Configure**.
|
|
* Once the setup is complete, you should see both Piper and Whisper in one integration.
|
|
|
|
* **Whisper** converts speech into text.
|
|
* **Piper** converts text into speech.
|
|
* **Wyoming** is the protocol they are both using to communicate.
|
|
1. Setup your assistant.
|
|
* Go to **Settings** > **Voice assistants** and select **Add assistant**.
|
|
|
|
* Enter a name. You can pick any name that is meaningful to you.
|
|
* Select the language that you want to speak.
|
|
* Under **Conversation agent**, select **Home Assistant**.
|
|
* Under **Speech-to-text**, select **faster-whisper**.
|
|
* Under **Text-to-speech**, select **piper**.
|
|
* Depending on your language, you may be able to select different language variants.
|
|
1. That's it. You ensured your voice commands can be processed locally on your device.
|
|
1. If you haven't done so yet, [expose your devices to Assist](/docs/assist/voice_remote_expose_devices/#exposing-your-devices).
|
|
* Otherwise you won't be able to control them by voice.
|
|
|
|
|
|
## Fine-tuning Whisper and Piper for your setup
|
|
|
|
You would like to tweak the configuration for better performance or accuracy?
|
|
View some of the options in the video below. Explained by Mike Hansen, creator of Rhasspy, Piper, and Wyoming.
|
|
|
|
<lite-youtube videoid="Tk-pnm7FY7c" videoStartAt="1589" videotitle="Configure your local Assist pipeline for your setup"></lite-youtube>
|
|
|
|
The options are also documented in the add-on itself. Go to the {% my supervisor_addon addon="core_whisper" title="**Whisper**" %} or the {% my supervisor_addon addon="core_piper" title="**Piper**" %} add-on and open the **Documentation** page.
|