fix bug saving / loading exported training data

[gbeane]

Recently I changed the ClassifierType from an IntEnum to a an Enum class with string values (inherits from both str and Enum)

This introduced a bug loading the training data in the jabs-classify train command.

[Copilot]

Pull request overview

This PR fixes a bug introduced when ClassifierType was changed from an IntEnum to a string-based Enum. The fix ensures backward compatibility by supporting both the old integer format (1 = Random Forest, 2 = XGBoost) and the new string format when loading training data.

Key Changes:

• Modified load_training_data to handle both integer and string classifier type values
• Updated export_training_data to save the enum's string value instead of converting to string

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
src/jabs/project/read_training.py	Added backward-compatible loading logic for classifier_type attribute that handles both old integer and new string formats
src/jabs/project/export_training.py	Changed to save classifier_type.value instead of str(classifier_type) for consistent string storage

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[bergsalex]

Is this change related to the rest of the PR?

[gbeane]

no, this shouldn't have been in this PR but it's a non-functional change (self._label_closest is assigned another value a few lines later)

[keithshep]

looks good