haipipe-nn-2-tuner

What problem does it solve?

This Skill provides a standardized interface for wrapping diverse machine learning algorithms, ensuring consistent data handling, training, inference, and model serialization across different libraries.

Core Features & Use Cases

• Algorithm Agnosticism: Wraps any algorithm with a uniform contract, regardless of its underlying library.
• Data Transformation: Handles conversion of input data into algorithm-specific formats.
• Model Lifecycle Management: Manages training, inference, saving, and loading of models.
• Use Case: Integrate a new time-series forecasting algorithm (e.g., from a new research paper) into an existing pipeline by creating a Tuner that adheres to this contract, without modifying the pipeline's orchestration logic.

Quick Start

Implement the get_tfm_data, fit, infer, save_model, and load_model methods within a new class inheriting from ModelTuner.