Whishper is a convenient tool that combines speech recognition and translation technologies. The project allows you to upload audio files, automatically receive a text transcription, and, if necessary, a translation into other languages.
Below is a detailed guide on how to install and run Whishper in a development environment.
Required tools
Before you begin, make sure that the following components are installed on your computer:
- Golang version 1.20 or higher — the language in which the server part is implemented.
- NodeJS — used to run the frontend.
- Docker — for running auxiliary services (databases, APIs, etc.).
These tools will ensure the correct operation of the project and an isolated development environment.
Initial project setup
Cloning the repository
First, you need to get the project source code from GitHub. To do this, clone the Whishper repository.
Running Docker containers
After cloning, you need to run the necessary containers. They ensure the operation of the MongoDB database, as well as the services responsible for speech decryption and text translation.
Tip: For convenient viewing of database data, you can install MongoDB Compass, a graphical utility that allows you to view the contents of collections without resorting to the command line.
Launching the server side (Backend)
The next step is to launch the Whishper backend. It is responsible for receiving and processing audio files, interacting with the database, and transferring data to the frontend.
Before launching, specify:
- development mode;
- folder for storing uploaded files;
- addresses of speech recognition and translation services;
- connection to the database.
After that, the server part will be ready to receive requests and process data.
Launching the user interface (Frontend)
To launch the Whishper interface, you need to:
- Go to the folder with the project frontend.
- Create a .env file in it, where the addresses of the API and translation services are specified.
- Install all project dependencies.
- Run the application in development mode.
After a successful launch, the web interface will be available locally — you can open it in a browser and test the system.
Conclusion
Whishper is a powerful tool for automating audio transcription and translation. Thanks to the use of modern technologies and containerization via Docker, the project is easy to deploy and scale.
It is an excellent solution for developers, researchers, and teams working with multilingual audio data.
