Speech to Text with OpenAI API

This script transcribes an audio file using OpenAI's Whisper model and optionally post-processes the transcription with GPT-4o for corrections. The transcription and corrected text are saved to text files.

Features

• Transcribes audio files to text using OpenAI's Whisper model.
• Optionally post-processes the transcription with GPT-4o to correct spelling and punctuation.
• Progress bars for uploading the file.
• Saves the transcription and corrected text to text files.

Supported Audio Formats

• MP3 (.mp3)
• MP4 (.mp4)
• MPEG (.mpeg)
• MPGA (.mpga)
• M4A (.m4a)
• WAV (.wav)
• WEBM (.webm)

Prerequisites

• Node.js (v14 or later)
• npm (Node package manager)

Installation

1. Clone the repository:

   git clone https://github.com/o-Oby/speech-to-text.git
   cd speech-to-text

2. Install dependencies:

   npm install fs path form-data axios readline-sync openai progress chalk

Setup

1. Update API Key:

   Open the transcribe_and_postprocess.js file and replace the placeholder API key with your actual OpenAI API key.

   const configuration = new Configuration({
     apiKey: 'your-api-key-here', // Replace with your actual API key
   });

2. Update File Path:

   Ensure the file path to your audio file is correct in the transcribeFile function.

   const filePath = path.resolve('path/to/your/audio/file.m4a'); // Replace with your actual file path

Usage

1. Run the script:

   node transcribe_and_postprocess.js

2. Follow the prompts:

   • The script will ask if you want to post-process the transcription with GPT-4o.
   • Respond with yes or no.

Files

• transcription.txt: Contains the initial transcription of the audio file.
• corrected_transcription.txt: Contains the corrected transcription (if post-processed with GPT-4o).

Notes

• File uploads are currently limited to 25 MB. Ensure your audio file size does not exceed this limit.

License

This project is licensed under the MIT License