v0.1.5

What's Changed

• Refactor cacheable metrics and speed up BERTScore by @de9uch1 in #31
• Implement an option for deduplication in mbrs-generate by @de9uch1 in #32

Full Changelog: v0.1.4...v0.1.5

v0.1.4

What's Changed

• Implement xCOMET-lite (Larionov et al., 2024) by @de9uch1 in #25
• Implement MetricX by @de9uch1 in #26
• Migrate the package manager to uv by @de9uch1 in #27
• Support fastchrf by @de9uch1 in #28
• Implement BERTScore by @de9uch1 in #29
• Implement plug-in module loader by @de9uch1 in #30

Full Changelog: v0.1.3...v0.1.4

v0.1.3

What's Changed

• Implement Diverse MBR (Jinnai et al., 2024) by @de9uch1 in #24

Full Changelog: v0.1.2...v0.1.3

v0.1.2

What's Changed

• Implement reference aggregation (DeNero et al., 2009; Vamvas and Sennrich, 2024) by @de9uch1 in #8
• Rename modules and classes by @de9uch1 in #9
• Install CI/CD by @de9uch1 in #10
• Add documents by @de9uch1 in #11
• Calculate log-probabilities in mbrs-generate by @de9uch1 in #12
• Speed up pairwise_scores of BLEURT by @de9uch1 in #13
• Use multiprocessing for pairwise_scores of BLEU and chrF by @de9uch1 in #14
• Fix a bug of multiprocessing in mecab-python3 tokenizer used in BLEU by @de9uch1 in #15
• Treat empty inputs in MetricBLEURT.pairwise_scores() by @de9uch1 in #16
• Fix the model-based estimation by @de9uch1 in #17
• Add documents of implementation lists by @de9uch1 in #18
• Ignore unnecessary arguments in corpus_score() of BLEURT by @de9uch1 in #19
• Add icon by @de9uch1 in #20
• Use multiprocessing for all scores() methods in BLEU, chrF, and TER by @de9uch1 in #21
• Implement YAML-based configuration by @de9uch1 in #22
• Add documents of YAML configuration by @de9uch1 in #23

Full Changelog: v0.1.1...v0.1.2

v0.1.1

What's Changed

• Add supports for XCOMET metrics by @de9uch1 in #1
• Add the citation of this software by @de9uch1 in #2
• Implement Probabilistic MBR (Trabelsi et al., 2024) by @de9uch1 in #3
• Support multiple scoring methods of UnifiedMetric by @de9uch1 in #4
• Implement MetricBLEURT by @de9uch1 in #5
• Add a scoring script by @de9uch1 in #6
• Implement model-based estimation (Jinnai et al., 2024) by @de9uch1 in #7

Full Changelog: v0.1.0...v0.1.1

v0.1.0

First release