Addition of multiconformer prediction model and functionality

[rodvrees]

This pull request introduces several changes to the im2deep project, focusing on improving command-line interface (CLI) argument consistency, adding support for a multi-conformer model, and enhancing the calibration and prediction functionalities. Below are the most significant changes:

CLI Argument Consistency:

• Updated CLI argument names to use hyphens instead of underscores for consistency (README.md, im2deep/__main__.py). [1] [2]

Multi-Conformer Model Support:

• Added a new --multi option to enable the use of a multi-conformer model in addition to the classical model (im2deep/__main__.py). [1] [2]
• Integrated the predict_multi function to handle multi-conformer predictions and included necessary imports and configurations (im2deep/im2deep.py, im2deep/predict_multi.py, im2deep/utils.py). [1] [2] [3]

Calibration and Prediction Enhancements:

• Enhanced the predict_ccs function to support multi-conformer predictions and output the results accordingly (im2deep/im2deep.py). [1] [2] [3]
• Improved the linear_calibration function to correctly handle the use_charge_state parameter (im2deep/calibrate.py).

These changes collectively enhance the functionality, consistency, and flexibility of the im2deep tool, particularly with the addition of multi-conformer model support.