📝 Integration of yt-dlp, Faster Whisper(from Whisper), ChatGPT, and Notion with GUI!
- Quickly take notes into Notion by simply pasting the URL/video/audio/subtitle/text file path!
- Support thousands of sites including YouTube and BiliBili, check supported sites.
- Utilize the GPU computing power of your local machine. Take notes with Whisper model for free!
Note
Spported Languages: English, 简体中文, 繁體中文, Español, Français, Deutsch, Português, Русский, 日本語, العربية, हिन्दी, 한국어, Italiano
Note
The corresponding subtitles will be downloaded if available. -> No time/GPU comsuming for transcription.
Spported Subtitles: srt, vtt, ass
Note
Update for v1.3.0: Now support for GPT-4o mini. "Significantly smarter, cheaper, and just as fast as GPT-3.5 Turbo," as OpenAI advertised.
- Your own OpenAI API key (You need to have a balance)
- Your own Notion API key
- Your own Notion Database ID (How to create a database)
- Permissions of your database for Notion API
- FFmpeg: if you don't have FFmpeg, get it from Download FFmpeg
- CUDA Toolkit 12
- cuBLAS for CUDA 12
- cuDNN 8 for CUDA 12
You can use CPU and Whisper API if you don't want to use a GPU
Tip
An easier way for me to use cuBLAS and cuDNN 8, in Faster Whisper GitHub:
Download the libraries from Purfview's repository (Windows & Linux)
Purfview's whisper-standalone-win provides the required NVIDIA libraries for Windows & Linux in a single archive. Decompress the archive and place the libraries in a directory included in the
PATH
.
You can also place the archives in the same folder as the executable or Python code below
Note
Before using ChatGPT API and Whisper API(not necessary), check Pricing - OpenAI
Usage: Executable by PyInstaller (Windows)
- Download Executable
- Run it and setup the API tokens
- Paste the link/file path and then setup link/file language and your preference
- Clone the repo
- Install Python packages
pip install -r requirements.txt
- Run Python script and setup the API keys by GUI (like in Usage: Executable)
python main.py
- Paste the link/file path and then setup link/file language and your preference (like in Usage: Executable)
- App crashes and exits suddenly after transcribing: The app will save transcription into a
.txt
file in the same folder. Use it to take notes. Or choose a smaller Whisper model.