ansuz / deckard

tooling for evaluating AI

Deckard: A Tool for Evaluating AI

1 - Dependencies

To install this, ensure that you have your favorite library installed. To install deckard along with tensorflow, for example, use

python -m pip install .[tensorflow]

Add the -e flag if you want to edit files:

python -m pip install -e .

Or try the rpi script:

bash rpi.sh

Now, check that deckard works $ python >>> import deckard Then CTRL+D or quit() to quit.

2 - Navigate to your favorite subfolder in examples. One is provided for each framework.

Running dvc repro in that folder will reproduce the experiment outlined in the dvc.yaml. Running python -m deckard will parse the configuration folder, create a params.yaml, and then run dvc repro.

like tears in the rain, this tool is meant for bladerunners. NOT INTENDED FOR USE BY REPLICANTS

Files

.
├── Dockerfile: Constructs a generic Docker image for running experiments
├── LICENSE
├── README.md: this file
├── deckard: Source code
├── examples: Directory containing all the examples
├── rpi.sh: For installation on Raspbian.
├── setup.py : for installation with pip
├── setup.sh : for installation using bash
└── test : test suite

After adding it to your path, you can then run it as a module:

cd examples/power
python -m deckard --config_name mnist.yaml