mltraq 0.0.36

A lightweight experimentation toolkit for data scientists.

---

Open source experiment tracking API with ML performance analysis to build better models faster, facilitating collaboration and transparency within the team and with stakeholders.

---

• Documentation: https://www.mltraq.com
• Source code: https://github.com/elehcimd/mltraq

---

Key features

• Fast and efficient: start tracking experiments with a few lines of code.
• Distributed: work on experiments independently and upstream them for sharing.
• Accessible: Simple SQL tables queriable with SQL, Pandas and Python API.
• Structured types: track Python types, Numpy arrays, Pandas dataframes and series.
• Parallel execution: define experiments as steps with parameter grids and execute them.
• Light checkpointing: save time by reloading and continuing your experiments anywhere.
• Steps library: enjoy pre-built steps for tracking, testing, analysis and reporting.

Requirements

• Python 3.7+
• SQLAlchemy, Pandas, and Joblib (installed as dependencies)

Installation

pip install mltraq

License

This project is licensed under the terms of the BSD 3-Clause License.