v1.0.3

Hybrid Sample Generator v1.0.3

This release simplifies configuration, improves the MVTec AD 2 workflow, and prepares the project for PyPI distribution.

• Simplified Configuration creation by moving model and anomaly settings into the configuration structure.
• Added config.model.set_model(...) for explicit model selection.
• Automatically synchronizes model input channels with the configured anomaly size.
• Reorganized the MVTec AD 2 example into configuration, presets, studies, pipeline, and review modules.
• Replaced the MVTec MODE switch with explicit workflow functions and CLI commands.
• Added global MVTec defaults with category-specific configuration overrides.
• Preserved saved downstream settings when reopening existing studies.
• Added early validation for invalid categories, duplicate categories, and duplicate study paths.
• Added PyPI metadata, compatible dependency ranges, and optional example dependencies.
• Added Git tag-based versioning through setuptools-scm; all 189 tests pass.