ml4xcube v1.1.0

[Julia310]

• Introduced create_splits for pre-assigned data splitting and a new evaluation module compatible with trainer classes.
• Expanded the datasets module with train/test splitting, feature scaling, and callback support for PTXrDataset.
• Implemented assign_mask in preprocessing for automatic data cube filtering.
• Renamed multiple modules for clarity and improved MultiProcSampler to automatically include metadata in the data cube.
• Enhanced drop_nan_values, get_statistics, and fill_nan_values methods for better functionality.
• Resolved an issue with GPU transfer of validation data in the trainer class.
• Updated scripts for time series analysis and masked data use cases.