Skip to content

Latest commit

 

History

History
79 lines (59 loc) · 2.81 KB

README.en.md

File metadata and controls

79 lines (59 loc) · 2.81 KB

🧰 Requirements - 🛠️ Configuration - 🚀 Deploy - ⚖️ License - 🌍 Español

A WhatApp bot to transcribe and summarize audio messages.

Note You can also join the waiting list to test the alpha version.

Requirements

You will need:

  • NodeJS >= 16
  • WhatsApp API token.
  • OpenAI API token.
  • (optional) If you want to use "Google Cloud Speech to Text":
    • A Google Cloud Storage bucket, to store temporally files.
    • ffmpeg installed to transform audios to a format accepted by the service.
  • (optional) If you want to use "whisper", you only need to have it install.
  • (optional) If you want to use a task queue, the project is compatible with Google Cloud Tasks.

Configuration

Copy the .env.example file as .env or add the following env vars in your environment.

OPEN_AI_GPT3_TOKEN="" // OpenAI token.

WHATSAPP_TOKEN="" // WhatsApp sender token.
WHATSAPP_API_VERSION="v15.0"
WHATSAPP_SENDER="" // the phone that will send the messages.
WHATSAPP_VERIFY_TOKEN="" // a random token to verify your webhook with WhatsApp.

FILE_DOWNLOAD_PATH="/tmp" // The folder in which the temporal files are going to be stored.

GCLOUD_BUCKET="" // Your Google Cloud Storage Bucket.
GCLOUD_TASK_QUEUE="" // You queue from Google Cloud Task.
GCLOUD_LOCATION="" // Your Google Cloud Task region.
GCLOUD_TASK_MODE="" // A value between "appengine" (if you use AppEngine in Google Cloud Tasks) or "http" if you use other service.
GCLOUD_PROJECT="" // Your Google Cloud project name.

TASK_URL="" // The URL of your service. Example: https://tu-servicio.com/tasks
TASK_SERVICE="" // A value between "google" (if you want to use Google Cloud Tasks) or "http" (if you don't want a task service).

SPEECH_TO_TEXT_SERVICE="" // A value between "google" (if you want to use Google Cloud Speech to Text) or "whisper" (if you want to use whisper).

WHISPER_MODEL="base" // Whisper model you want to use
WHISPER_LANGUAGE="Spanish" // The language whisper is going to use to transcribe (if empty whisper will auto detect the language)

Deploy

yarn install
yarn build // to build ts into js
yarn start // to start the server

Contributing

Thank you for considering contributing to the DevOps for Stream Deck. Feel free to send in any pull requests.

Support the project

If you like the project, you can subscribe to my Twitch channel, where I do live coding of this and other projects.

License

This project is under the CC-BY-NC-4.0 license.