/quick-serve

Deploy ML models quickly in <3 lines of code!

Primary LanguageJupyter NotebookBSD 3-Clause "New" or "Revised" LicenseBSD-3-Clause

quickserve

Quickly build inference engines from jupyter notebooks and deploy on web and edge in 3 lines of code.

[Under development]

Notes regarding getting started

  • Please check requirements.txt for list of dependencies. Install scripts will be added later.

  • The pre-trained models used in the demo notebooks/submission videos are from a CV library built by one of the members. For evaluation purposes on local machines, it can be downloaded here.

  • ngrok is also a dependency. Future commits will focus on API based authentication but for now, ngrok should be authenticated from CLI.

  • All features pass build on linux machine (Ubuntu 18.04).

Get started

  • Clone the repo!
$ git clone https://github.com/pranjaldatta/quick-serve.git
  • Install with pip
$ pip install .

Authors