Skip to content

Transcribe video/audio file and process transcription using OpenAI

Notifications You must be signed in to change notification settings

radoslavdodek/TranscribePlusAI

Repository files navigation

Transcribe Video/Audio File and Process Transcription Using OpenAI

Prerequisites

The tools listed below need to be installed:

Installation

  • Set OPENAI_API_KEY environment variable (You can create an API KEY here: https://platform.openai.com/api-keys).

    You can do so in ~/.zshrc or ~/.bashrc file:

    export OPENAI_API_KEY="<YOUR_KEY>"
  • Clone this repository to a local directory:

    git clone [email protected]:radoslavdodek/TranscribePlusAI.git
    cd TranscribePlusAI
  • Install python virtual environment and required python packages:

    ./setup.sh

Usage

Here's how to use the command on mp3 or mp4 files:

./transcribe_plus_ai.sh "<YOUR_FILE_PATH>"

If you're working with mp4 file, the command will first convert it to mp3 file. Then, it will use the AWS Transcribe API to create a transcription of the audio. This transcription will be stored in a file named <YOUR_FILE_PATH>.transcription (the file extension can be configured in CONFIG.yaml file).

Once the transcription has been created, the script will process it using the OpenAI API, using the prompt specified in the PROMPT_TEMPLATE.txt file. The results of this process will be saved in a file named <YOUR_FILE_PATH>.output (the file extension can be configured in CONFIG.yaml file).

Watch for new mp4 files in the directory

If you would like to automatically call this script when a new mp4 file is added to a specific directory, you can configure it as follows (valid for Ubuntu):

  • Create new file called auto_transcribe_new_mp4s.sh:
#!/bin/bash

DIRECTORY_TO_WATCH=<CHANGE_TO_YOUR_DIRECTORY>
TRANSCRIBE_SCRIPT=<CHANGE_TO_TRANSCRIBE_SCRIPT_PATH>

inotifywait --include 'mp4' -m "${DIRECTORY_TO_WATCH}" -e moved_to --format '%w%f' |
    while IFS=' ' read -r fname
    do
        [ -f "${fname}" ] && "${TRANSCRIBE_SCRIPT}" "${fname}"
    done

Note: don't forget to specify DIRECTORY_TO_WATCH and TRANSCRIBE_SCRIPT in the above script.

We need to start this file at the startup: Create new file: ~/.config/autostart/auto_transcribe_new_mp4s.sh.desktop

[Desktop Entry]
Type=Application
Exec=<YOUR_SCRIPT_PATH>/auto_transcribe_new_mp4s.sh
Hidden=false
NoDisplay=false
X-GNOME-Autostart-enabled=true
Name=Transcribe videos automatically

Note: Don't forget to change the YOUR_SCRIPT_PATH in the above file.

About

Transcribe video/audio file and process transcription using OpenAI

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published