v1.2.9

• Add stance detection model.

v1.2.8

• Add new sentence models: legal_sroberta_v0 and legal_sroberta_v1;
• Add datasets for Ulysses SentEval.

v1.2.7

• Add 3 new SBERTs for Brazilian legal semantic mapping (task = sentence_similarity):
  • sbert_1mil_amana;
  • sbert_650k_nheeng;
  • sbert_map2doc_v1.

v.1.2.6

(Rebuilt w/ correct pyproject.toml version)

New Resources:

• Ulysses Segmenter models v3.

Minor fixes

• Improve Exception messages when resource name is unknown.

v1.2.4

• Add Ulysses Segmenter V2 models.

v1.2.3

• Add quantization task with the following resource:
  • ulysses_tesemo_v2_subset_static_quantization.

v1.2.2

• Add resouce: ulysses_LaBSE_30000.

v1.2.1-beta

• Update probing dataset hashes.

v1.2.0-beta

• Add new legal probing tasks.

v1.1.0-beta

• Add dataset_sp_court_cases* resources; and
• Update dataset_wikipedia_ptbr* hashes.