FastDatasetMaker (FDM)

Remove and separate Vocals through transcribed subtitles text from YouTube video (support for local video transcribe and separation vocal through text coming soon)

do not expect much here cuz I'm noob at this, but it works pretty well

How to install

First of all you have to install FFMPEG:

Download here https://github.com/BtbN/FFmpeg-Builds/releases/download/latest/ffmpeg-master-latest-win64-gpl.zip

or check some .zip here https://ffmpeg.org/download.html

unzip in C:\Program Files\ and rename folder as ffmpeg C:\Program Files\ffmpeg

after that create environment variable [important] to exact folder where .exe files stored

win > (type) system > system properties > environment variables > (click) PATH > edit > new > (type)

C:\Program Files\ffmpeg\bin

IF STRUGGLE CHECK VIDEO HOW TO CREATE ENVIRONMENT VARIABLE: https://www.editframe.com/guides/how-to-install-and-start-using-ffmpeg-in-under-10-minutes/Create_Environment_Variable.mp4

---

download FDM_FastDatasetMaker repository as .zip https://github.com/Mika4334/FDM_FastDatasetMaker/archive/refs/heads/main.zip

or simply clone it git clone https://github.com/Mika4334/FDM_FastDatasetMaker.git

Use VSCode Terminal for:

check if repository in folder dir FDM_FastDatasetMaker

open foledr cd .\FDM_FastDatasetMaker\

---

Inside FDM_FastDatasetMaker folder:

create venv python -m venv venv

activate .\venv\Scripts\activate

{optional} upgrade pip python.exe -m pip install --upgrade pip

check if venv is activated

install some reqs pip install -r requirements.txt and wait!!! until all runnning setup.py done and everything Successfully installed

after install torch pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

Done!

Run

open py .\FastDatasetMaker.py

enter any yt link FOR EXAMPLE:

[eng] https://www.youtube.com/watch?v=M7FIvfx5J10 [rus] https://www.youtube.com/watch?v=RL7wMoH6cvA

enjoy :)

audio files will be saved in cutted folder, also .json have time and text of the video

P.S.

Name of the models that will download automatically here model_name_mapper.json or download it from Get models

TODO

[ ] Maybe soon you will be able to use another models

[ ] Also going to release some features soon or you could suggest anythin

Get models

https://github.com/TRvlvr/model_repo/releases/

Real Autors

Anjok07 /ultimatevocalremovergui https://github.com/Anjok07/ultimatevocalremovergui

karaokenerds /python-audio-separator https://github.com/karaokenerds/python-audio-separator

TRvlvr/model_repo https://github.com/TRvlvr/model_repo/releases/