feat: sync mlflow

[Zeyi-Lin]

Description

This pull request introduces the synchronization of MLflow with SwanLab, including the necessary imports and functionality to support this integration. The most important changes include adding the sync_mlflow function, updating the __all__ variable to include sync_mlflow, and modifying the version in package.json.

MLflow synchronization:

• swanlab/sync/mlflow.py: Added a new module for MLflow synchronization, which includes functions to synchronize MLflow operations (such as logging parameters, metrics, and models) with SwanLab.

Updates to imports:

• swanlab/__init__.py: Added sync_mlflow to the list of imported functions from swanlab.sync.
• swanlab/sync/__init__.py: Updated the __all__ variable to include sync_mlflow.

Version update:

• swanlab/package.json: Changed the version from "development" to "0.5.1".