This fork is meant to help install DeepLabCut on a local machine. It version-matches with the current version of DeepLabCut installed on TACC resources.
- Prerequisite software:
pip - https://pip.pypa.io/en/stable/installation/
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)- Install the Homebrew package manager.
brew install python@3.7- Install Python version 3.7; one of the other package dependencies, wxPython, is specifically compatible with this version.
/usr/local/opt/python@3.7/bin/python3 -m pip install --upgrade pip wxpython- Install wxPython; it is needed for GUI support.
/usr/local/opt/python@3.7/bin/python3 -m venv --system-site-packages [name]- Create a virtual environment containing wxPython; pass in a
[name]of choice (without square brackets). source [name]/bin/activate- Activate the virtual environment.
pip install tensorflow==1.15 pytest- Install and test TensorFlow version 1.15.
git clone https://github.com/ib-nosirov/dlc.git dlc
cd dlc
pip install -e .- Install DeepLabCut v2.1.10.
- DeepLabCut documentation is the perfect place to start learning how to use DLC.
- https://deeplabcut.github.io/DeepLabCut/docs/standardDeepLabCut_UserGuide.html
I have also included some helpful template files in the Templates folder. The files contain basic files for creating a DLC project and runnings sbatch scripts on a TACC resource. You can even try running python create_project.py from inside a virtual environment to test how a DLC project is made, though make sure the file path dates in the python file are set to your current date.
- One very important command for transporting a created project to and from a TACC resource is the rsync command.
- Move to the directory immediately above your DLC project directory.
Now pass the command
rsync -avz [project name] [username]@maverick2.tacc.utexas.edu:/work2/[your allocation number]/[username]/maverick2/[project name]. (Tip: everything after the ':' is found by passing the pwd command while you are in your TACC workspace.) - More information can be found about TACC resources, and Maverick2 specifically, in the Maverick2 User Guide.
- https://portal.tacc.utexas.edu/user-guides/maverick2