git clone https://github.com/diopthe20/nanobot
git submodule update --init --recursive

• Develop a deployment strategy

We will use some vision model to get the information about what we will see in the screen

I used EasyOCR for recognize some text in the screen during gathering. You can go to https://huggingface.co/spaces/tomofi/EasyOCR to test with EasyOCR

We take the environment state as an event and send it to event handler

Label with Label Studio, Export to YOLO Format and then Upload to ROBOFLOW to export to the right format for YOLO

Train with YOLO

=> Predict from the screen stream

Currently this project in development. The current phase is try out new probilities