main_v2.py has thread-safe execution of UI so we will be running that file.

Step 1 - Create virtual env

1. python -m venv ./venv
2. source venv/bin/activate
3. pip install -r requirements.txt

Step 2 - Get omniparser

1. Clone in the same folder from this repo : https://github.com/microsoft/OmniParser
2. pip install -r Omniparser/requirements.txt

Download its weights:

1. cd into the Omniparser directory
2. Run huggingface-cli download microsoft/OmniParser --local-dir weights/ to download latest weights

Test Omniparser is working: (The following code is taken from their README.md)

python gradio_demo.py --icon_detect_model weights/icon_detect_v1_5/model_v1_5.pt --icon_caption_model florence2

Some basic error might show up here as we are dealing with a 3rd party library.

1. Package errors that need to be upgraded (like ultralytics)
2. Module errors in code

Note: Because this is an external repo, some weights may be updated if you're cloning this

Step 3 - Run the server:

1. cd back into parent dir
2. Run uvicorn core_server.server:app
3. Confirm localhost:8000/docs

[Optional] Step 4 - GUI:

1. Run python main_v2.py