Automating Machine Learning Testing using GitHub Actions and DeepChecks
-
Updated
Jul 16, 2024 - Python
Automating Machine Learning Testing using GitHub Actions and DeepChecks
Modelling and prediction of default + deployment via AWS Sagemaker
Repository contains the detail about ML model deployment and building end-to-end ML pipeline for production
Comparison between several Python data profile libraries.
Credit risk analytics project for loan default ranking with Python, XGBoost, SHAP, threshold strategy, monitoring, and governance.
End-to-end MLOps pipeline for Telco Customer Churn. Hydra + DVC + MLflow + XGBoost (hyperopt-tuned with scale_pos_weight) + BentoML 1.2 service + GitHub Actions deploying to GHCR. Test-set-leakage-free training, 38 tests (Pandera + pytest-steps + Deepchecks + HTTP integration), all 8 phases adversarially reviewed by Opus.
Deep learning-driven cache eviction policy using LSTM forecasting and uncertainty estimation, deployed as a production microservice with a full MLOps pipeline.
End-to-end MLOps CI/CD pipeline for a FastAPI bean-leaf disease classifier (ViT). GitHub Actions lints, validates data with Deepchecks, builds a Docker image, pushes to Amazon ECR and deploys to ECS Fargate on every push to main, going from commit to live service in about 14 minutes.
Stock-risk MLOps pipeline: fallback data ingestion, DeepChecks validation, Prefect training, FastAPI + Docker serving, CI.
Independent SR 11-7-style model validation of an ML trading strategy - deepchecks + giskard, formal report, reproducible. Verdict: not approved for deployment.
"End-to-end MLOps pipeline for image classification using ZenML, MLflow, and TensorFlow. Features automated training, continuous deployment, drift detection with Deepchecks, and a Streamlit frontend."
To associate your repository with the deepchecks topic, visit your repo's landing page and select "manage topics."