Dear, contributor!
Please, find the list of simple steps how to wrap your Neural Network and be able to sell it on our AI Marketplace
Quickstart
Prepare working environment (if you have’nt do this yet):
python3 -m venv env source env/bin/activate
Install the latest Cookiecutter if you haven’t installed it yet (this requires Cookiecutter 1.4.0 or higher):
pip install -U cookiecutter
Clone a cookiecutter-neuromarket-model project:
git clone https://github.com/neuromarket-ai/cookiecutter-neuromarket-model
Generate a Python package project:
cookiecutter <path_to_project>
Then:
- Install the dev requirements into a virtualenv. (
pip install -r requirements_dev.txt
) - Add a
requirements.txt
file that specifies the packages you will need for your project and their versions. For more info see the `pip docs for requirements files`_. - Use
requirements-cpu.txt
andrequirements-cuda.txt
files to specify CPU and CUDA specific the packages you will need for your project and their versions (i.e.onnxruntime
andonnxruntime-gpu
). - Wrap your AI assets into Model following guideline in the code
- Create CLI and Demo API for your project
- Build distributable
whl
files with the command:make dist - Zip and share with us your *.whl and *.zip file and we will publish it in neuromarket.ai
For more details, see the cookiecutter-pypackage tutorial.