Hi 👋, I'm Sheikh Md. Faysal

Machine Learning Engineer

• 📫 How to reach me [email protected]

Connect with me:

Languages and Tools:

Project Title

Model Testing and Monitoring pipeline

Project Overview

Monitoring models after deployed into production is one of the most major task when anyone wants to serve a machine learning model to the world. The way model performed in test data doesnt mean it will perform the same for new user data. Data drift and concept drift may occur and decrease the model performence in production.So it's necessary to monitor continuously and create metrics for error analysis then take action accordingly to model. N.B: All the models and weights here are kept dummy as it's sensitive and not shareable.

Run Locally

Clone the project

  git clone https://github.com/skfaysal/Model-testing-and-monitoring-pipeline.git

Go to the project directory

  cd Model-testing-and-monitoring-pipeline

Create virtual environment using environement.yml

  conda env create -f environment.yml

Activate environment

  conda activate heat_map

For Training Model. We will pass parameters using CLI

 python3 TestModel_cli.py --drmodel models/b5_newpreprocessed_full_fold4.h5
--lfmodel models/model_binary_right_leaft_retina.h5
--imgdata eyepacs_train --savepath output/

For generating confusion matrix and save missclassified images

  cd confusionMatrix
  python3 main.py