v0.4.0

• Fix a bug where filter_noise_subsegments was always enabled, regardless of user parameters;
• Add finetune method.

v0.3.4

• Fix "Assymetric*" -> "asymmetric".

v0.3.3

• Change 'ulysses-fetcher' dependency version from @v1.2.6 -> @master.

v0.3.2

Correct version configuration for Ulysses Fetcher.

v0.3.1

(Rebuilt w/ correct version info in pyproject.toml)

• Set Ulysses Segmenter v3 models as default;
• Implement segmentador.Segmenter.generate_segments_from_ids( method;
• Add segmentador.Segmenter.segment_legal_text(maximum_noise_subsegment_length=...) argument;
• Add post-processing in segmentation pipeline (to remove spurious spaces related to punctuation marks);
• Improve README.

v0.2.3

• Add support for Ulysses Segmenter v2;
• Prevent splitting words composed of multiple subwords;
• Add .to( method to BaseSegmenter;
• Fix missing methods in documentation;
• Change default inference_pooling_operation="sum" to all models.

v0.2.2

• Support for newer versions of Optimum, ONNXRuntime, and ONNX during quantization.

v0.2.1-beta

• Change version from 0.1.0 -> 0.2.1 in pyproject.toml.
• Fix version of Ulysses Fetcher to v0.1.1-beta.
• Improve notebook naming in README.md.

v0.2.0-beta

• Integrate with ulysses-fetcher.
• Activate automated tests in workflows.

Initial pre-release

All library main features.