v0.2.6

• Bugfix: Pytorch best dev weights selection (#109)
• Bugfix: keep QIDs with no relevant docs returned by the first stage ranker (#110)
• Bugfix: long queries can no longer cause maxseqlen to be exceeded (issue #105)
• Add AutoTokenizer to extractors and a maxqlen option to truncate long queries (#112)
• Add separate testthreshold option to rerank task
• Add Pytorch and Tensorflow CEDRKNRM rerankers
• Trainer improvements, including Tensorflow AMP support and aligning itersize options between TF and Pytorch (#121)

v0.2.5

• Add environment.yml to simplify installation with Conda
• Bugfix: force QIDs in folds.json files to be strings
• Bugfix: add multithreaded option to Pytorch trainer controlling whether a separate worker is used with Pytorch DataLoader. Default to false; true is faster, but conflicting Python packages appear to cause deadlock. This is reproducible on Colab and when using Anaconda's default packages (rather than installing from Miniconda).
• Update docs

v0.2.4.1

Bug fix: allow spaces in paths passed to Anserini (e.g., the collection path)

v0.2.4

• New rerankers: PARADE and BERT-MaxP variants
• Collection iterator: for doc in collection_object: ...
• Experimental support for Birch reranker
• Bug fix: evaluator import should no longer break Python 3.6

v0.2.3

• Report results interpolated with searcher
• Add new, more efficient EmbedText that avoids generating random embeddings for OOV terms. The previous extractor is now SlowEmbedText
• Bug fix: Add missing NF queries file

v0.2.2

• New NFCorpus Collection and Benchmark
• Module cleanup
• Documentation improvements

v0.2.1

• Searcher refactor, with BM25, BM25RM3, BM25PRF, AxiomaticSemanticMatching, DirichletQL, and QLJM supporting lists of parameters
• support setting a relevance level threshold (i.e., trec_eval --level_for_rel)
• miscellaneous cleanup and convenience methods

v0.2.0

Flexible pipeline

• Replace fixed pipeline with a configurable version based on profane
• Add support for Tensorboard, Tensorflow, Tensorflow+TPUs, and TF-Ranking
• Add support for several new datasets, including TREC-COVID and COVID-QA
• Add support for all Anserini searchers
• Unit and integration tests