Node-deepspeech NPM

node-DeepSpeech

What is the node-DeepSpeech?

It started in Jan 23, 2019

this is component using node and morzilla/DeepSpeech.

💡 Motivation

No one made a valid repository for DeepSpeech and we made it.

🏃‍♀️🏃‍♂️QuickStart

1-1. If you want use in your project

$ yarn add node-deepspeech
// OR
$ npm install node-deepspeech

1-2. If you want use in local

// use HTTPS
$ git clone https://github.com/teamthesol/node-DeepSpeech.git
// Or use SSH
$ git clone git@github.com:teamthesol/node-DeepSpeech.git

2-1. if you have training models of deepspeech

Modify your training models directory name to 'models'. follow:

$ mv [your models name] [models name to be changed]

2-2 if you don't have training models of deepspeech, prepare files to be used in advance. follow:

$ npm run pre

edit .env file follw:

Modify file name `.emv.sample's` to `.env`. this is use 'dotenv'.

Edit environment value 'DEEPSPEECH_MODEL_PATH' in `.env`. this is your 'models' path. 'models' is your trainint models. if you run `npm run pre`, edit like 'DEEPSPEECH_MODEL_PATH="./models'(this is only use in node-deepspeech).

but, if you don't want use 'dotenv', run `export DEEPSPEECH_MODEL_PATH="./your/models/path"` in terminal(bash, zsh, fish, etc).

4-1. If you want use in your project

write your project. Now the preparations that we can do are over!

4-2. If you want use in local

// run file-demo // or yarn file-demo
$ npm run file-demo
// run streaming-demo
$ npm run streaing demo // or yarn streaming-demo

🔧 Used Opensources

mozila deepspeech(MPL 2.0) node(MIT)

Script

Title	Description
file-demo	runnung demo using the file path.
streaming-demo	running demo using the microphone.
pre	install pre-information(audio sample, training models).

Development Environments

Title	Description	Default
DEEPSPEECH_MODEL_PATH	your deepspeech 'models' path	./models

Function

Title	Description	Input
dsBuffer	Mozilla/Deepspeech function using the buffer data.	buffer
dsFile	Mozilla/Deepspeech function using the file path.	file path
dsStreaming	Mozilla/Deepspeech function using the microphone.	emmiter(use new EventEmitter()

Usage

The best example of using our service is react-deepspeech.

How can I contribute to this project?

First, read the our contribution guideline. if you have read it, you can contribute to our project by creating issues or contribute to it in pull requests.