Skip to content
/ blog_demo Public

Data, configs, and code for the walkthrough from the 3DB blogpost: https://gradientscience.org/3db

Notifications You must be signed in to change notification settings

3db/blog_demo

Repository files navigation

Demo for 3DB Blog Post

This is the code accompanying the 3DB blog post/walkthrough. You can find an abbreviated guide on running the code below; see the post for more details and context.

Running this demo

  1. Clone https://github.com/3db/3db
  2. Run bash setup.sh (assumes unzip is installed) to download the studio environment
  3. Run curl https://raw.githubusercontent.com/3db/installers/main/linux_x86_64.sh | bash /dev/stdin threedb to install 3DB
  4. Run conda activate threedb
  5. Run your desired experiment, replacing EXPERIMENT with one of {backgrounds, texture_swap, part_of_object}:
export BLENDER_DATA=data/EXPERIMENT RESULTS_FOLDER=results/
threedb_master $BLENDER_DATA EXPERIMENT.yaml $RESULTS_FOLDER \$SLURM_STEP_RESV_PORTS
threedb_workers 1 $BLENDER_DATA 5555
  1. Analyze the results using the included notebooks (see the analysis/ folder), or by launching the dashboard:
python -m threedb.dashboard $RESULTS_FOLDER