Skip to content

Latest commit

 

History

History
30 lines (16 loc) · 691 Bytes

File metadata and controls

30 lines (16 loc) · 691 Bytes

Deploying keras model with periodic trainer

This code presents an approach to train keras models periodically which make your predictions always trained by latest dataset.

Requirements

docker

docker-compose

Installation:

git clone https://github.com/patron-labs/periodic_keras_model_training_example

cd periodic_keras_model_training_example

mkdir data

mkdir ml_models

docker-compose up --build -d

Documentation:

Please check this medium post.

You can import keras_model_deploy_example.json to insomnia-rest and postman to test endpoints in api service.