/trascritores-colab

gogole colab to make transcriptions, upload to you gogole drive and have fun!

Primary LanguageJupyter NotebookBSD 2-Clause "Simplified" LicenseBSD-2-Clause

YouTube Video Transcription and Media Processing

This repository contains Jupyter Notebooks for transcribing YouTube videos, processing audio and video files, and utilizing various models to transcribe speech into text in different formats.

Notebooks

1. transcritor_en_pt_via_youtube_api.ipynb

This notebook handles transcription of YouTube videos, using the YouTube API and Whisper for English and Portuguese languages.

Key Features:

  • Connects to the YouTube API to retrieve video data.
  • Downloads video/audio and processes the files.
  • Utilizes the Whisper model to generate transcriptions for both English and Portuguese languages.

Required Libraries:

  • whisper
  • youtube_dl
  • torch
  • ffmpeg
  • pydub

2. trascritor_videos_youtube.ipynb

This notebook focuses on downloading and transcribing YouTube videos.

Key Features:

  • Downloads YouTube videos or audio files using the youtube_dl library.
  • Processes audio with Whisper to generate transcriptions.

Required Libraries:

  • whisper
  • youtube_dl
  • ffmpeg
  • tqdm

3. transcritor_formatos_diversos.ipynb

This notebook expands the transcription functionality to support a variety of audio and video formats, beyond YouTube.

Key Features:

  • Accepts various media formats, including .mp3, .mp4, .wav, .opus, and more.
  • Uses pydub and ffmpeg to convert and process files for transcription.
  • Outputs text in multiple formats, including .txt.

Required Libraries:

  • whisper
  • ffmpeg
  • pydub

Setup Instructions

  1. Clone the repository:

    git clone https://github.com/your_username/your_repository.git
cd your_repository

Install the required Python packages:

bash Copiar código pip install -r requirements.txt Run the notebooks using Jupyter or Google Colab.

Dependencies The following libraries are required to run these notebooks:

whisper ffmpeg torch youtube_dl pydub tqdm How to Use Open the notebook in Jupyter or Google Colab. Follow the steps provided in each notebook to transcribe videos or audio files. Transcriptions will be saved in .txt format in the specified directory.