MLP is a fast inference server made in Python via Flask and a few other core Python libraries. Communication is handled via websockets.

Installation

Run the following commands to set up your environment.

git clone https://github.com/h3x4g0ns/mlp
cd mlp
conda create -n mlp_server python=3.11
conda activate mlp_server
pip install -r requirements.txt

Getting Started

For running the server and client together run the following command:

# open up 2 terminals and run the client command in one and the server in another
conda activate mlp_server
make server

conda activate mlp_server
make client

Running unittests

For running unittests run the following command:

make test

# basically runs the following command under the hood
pytest