-
Spotlight Jinja2 Templates for Efficient Pipeline Generation
Large machine learning pipelines can suffer from repetitive edits when stage definitions expand or change.
-
Spotlight Hyperparameter Tuning with Hydra, Optuna, and MLflow
This project integrates Hydra configs, Optuna optimization, and MLflow tracking to streamline hyperparameter tuning.
-
Spotlight Feature Engineering for Reproducibility and Scalability
A strong feature engineering pipeline should maintain clean separation between data ingestion, cleaning, and transformation steps.
-
Overly Commented AI Generated Code
Recently, I've been battling a new adversary: the overly-commented AI-generated code
-
OpenAI News: o3-mini vs o1-mini comparison
Comparing the o3-mini and the o1-mini in terms of response latency, completion tokens, reasoning tokens