Feature/progress and threads

[RalfG]

Added

• Add plain calibration function to core to avoid code duplication and add user friendly entry point for calibration of PSMs
• In piece-wise linear calibration: Introduce min_samples_per_segment to avoid setting badly fit anchor points in sparse RT regions

Changed

• Simplify and improve training and inference progress bars
• Update function order in _features module; add todo for fixed modifications
• Change to SplineCalibration by default after finetuning
• Saved model files with 'promote buffers' onnx fix, avoiding requirement to apply fix on every model load
• In piece-wise linear calibration: Set defaults to 20 splits and use_median false to avoid jagged calibration lines; introduce min_samples_per_segment to avoid setting badly fit anchor points in sparse RT regions

Fixed

• Fix missing encoding for terminal modifications

Removed

• Remove calibration parameter from finetune_and_predict: This option is never useful as you can't have a fitted calibration for a model that doesn't exist yet.
• Remove partial model freeze support for finetuning, as this mostly reduces finetuning performance
• Remove 'simplified' option in spline calibration.