/FlockSentry

Open-source affordable system to automatically detect predators and save chickens.

Primary LanguageJupyter NotebookMIT LicenseMIT

FlockSentry

Open-source affordable system to automatically detect predators and save chickens.

Really good documentation goes here. But the exceptionally good documentation goes right here!

Preliminary readme.md

Using anaconda version 4.10.3 (most recent version) - update with following:

conda update -n base -c defaults conda

Create a new environment using chosen python version (3.7?):

conda create -n name_of_env python=3.7

Activate environment:

conda activate name_of_env

Install cudatoolkit:

conda install -c anaconda cudatoolkit

Install Pytorch (1.9.1?):

conda install pytorch torchvision torchaudio -c pytorch

Install numpy:

conda install numpy

Using the YOLOv5 Submodule

To clone the submodule repository:

git submodule update --init --recursive

To push changes to the FlockSentry repo, but not the YOLOv5 repo (like you almost always will), run the following git push command:

git push --recurse-submodules=check

Train the YOLOv5 Model

cd yolov5
python3 train.py --img 640 --batch 16 --epochs 120 --data custom_data.yaml --weights yolov5s.pt