About

The app takes a link to a VKontakte post and visualizes its comments using a model from Hugging Face.

Run tests

python -m tests.utilsTests

How to launch:

1. Clone reopsiroty

git clone https://github.com/i-pankrat/comment-visualizer-vk

2. Create venv

cd comment-visualizer-vk
python -m venv venv
. venv/bin/activate

3. Install requirements

pip install -r requirements.txt

4. Install torch
5. Log to hugging faces

huggingface-cli login

6. Add your vk access token to your environment variable. Create an application in vk to get the access token. You can create app here. Then go to setttings of the created app and find the 'Service token' there. That's what you need to put insted of YOUR_ACESS_TOKEN in the command below.

export VK_ACCESS_TOKEN=YOUR_ACESS_TOKEN

7. Start the application!

cd src/
streamlit run steamlit_app.py