Ml flow

[ForYourEyesOnlyyy]

Changes Made:

• MLflow Logging Integration: Integrated MLflow logging into the model training process in notebooks/model_playground.ipynb. Every new model trained is automatically logged with versioning enabled.
• Automatic Versioning: For each different model architecture, the best accuracy version is tagged as the champion, allowing easy identification of the best-performing model.
• Champion Version Auto-Selection: Implemented logic that automatically selects and loads the champion version when the corresponding model architecture is loaded, ensuring the most accurate version is always used by default.

Benefits:

• Enhanced Model Tracking: Automatic MLflow logging ensures all model training runs are tracked, making it easier to monitor performance and manage model versions.
• Streamlined Model Selection: The automatic tagging and selection of the champion model simplifies the process of identifying and loading the best-performing models for each architecture.
• Reproducibility: With model versions logged and tagged, experiments are easily reproducible, improving collaboration and future experimentation.