Magic mirror module that takes text as input and responds with machine learning accuracy
This an extension for the MagicMirror to recieve text input and send text answers.
This is a Work In Progress and only has a few commands for now but will get bigger overtime
This module was also made to work with MMM-Hello-Mirror as text inpurt and MMM-TTS as output
- Navigate into your MagicMirror's
modules
folder - clone this git repository
npm install
is needed in cd /MMM-OTIS- (Re)start magic mirror (e.g. with
pm2 restart mm
)
To use this module, add it to the modules array in the config/config.js
file:
modules: [
{
module: 'MMM-OTIS',
position: 'top_right',
}
]
The following properties can be configured:
Option | Description |
---|---|
language |
The language used.
Valid values: en
Default value: en
|
For Module to work you will need to change your python path in the node_helper.js file change the pythonPath variable in options of the runPython function to your python env path
Python dependencies: random, json, sys, os, numpy, torch, torch.nn, torch.utils.data, nltk, nltk.stem.portet
To use MMM-OTIS in your module you have to send a socket notification like this self.sendNotification("REPLY_TO_TEXT", "hello");
.
It will reply with a notification like this: this.sendNotification("MMM-TTS", text);
and then you can run the answer through whichever module you use