spaCy/spacy/ml
Raphael Mitsch 1f23c615d7
Refactor KB for easier customization (#11268)
* Add implementation of batching + backwards compatibility fixes. Tests indicate issue with batch disambiguation for custom singular entity lookups.

* Fix tests. Add distinction w.r.t. batch size.

* Remove redundant and add new comments.

* Adjust comments. Fix variable naming in EL prediction.

* Fix mypy errors.

* Remove KB entity type config option. Change return types of candidate retrieval functions to Iterable from Iterator. Fix various other issues.

* Update spacy/pipeline/entity_linker.py

Co-authored-by: Paul O'Leary McCann <polm@dampfkraft.com>

* Update spacy/pipeline/entity_linker.py

Co-authored-by: Paul O'Leary McCann <polm@dampfkraft.com>

* Update spacy/kb_base.pyx

Co-authored-by: Paul O'Leary McCann <polm@dampfkraft.com>

* Update spacy/kb_base.pyx

Co-authored-by: Paul O'Leary McCann <polm@dampfkraft.com>

* Update spacy/pipeline/entity_linker.py

Co-authored-by: Paul O'Leary McCann <polm@dampfkraft.com>

* Add error messages to NotImplementedErrors. Remove redundant comment.

* Fix imports.

* Remove redundant comments.

* Rename KnowledgeBase to InMemoryLookupKB and BaseKnowledgeBase to KnowledgeBase.

* Fix tests.

* Update spacy/errors.py

Co-authored-by: Sofie Van Landeghem <svlandeg@users.noreply.github.com>

* Move KB into subdirectory.

* Adjust imports after KB move to dedicated subdirectory.

* Fix config imports.

* Move Candidate + retrieval functions to separate module. Fix other, small issues.

* Fix docstrings and error message w.r.t. class names. Fix typing for candidate retrieval functions.

* Update spacy/kb/kb_in_memory.pyx

Co-authored-by: Sofie Van Landeghem <svlandeg@users.noreply.github.com>

* Update spacy/ml/models/entity_linker.py

Co-authored-by: Sofie Van Landeghem <svlandeg@users.noreply.github.com>

* Fix typing.

* Change typing of mentions to be Span instead of Union[Span, str].

* Update docs.

* Update EntityLinker and _architecture docs.

* Update website/docs/api/entitylinker.md

Co-authored-by: Paul O'Leary McCann <polm@dampfkraft.com>

* Adjust message for E1046.

* Re-add section for Candidate in kb.md, add reference to dedicated page.

* Update docs and docstrings.

* Re-add section + reference for KnowledgeBase.get_alias_candidates() in docs.

* Update spacy/kb/candidate.pyx

* Update spacy/kb/kb_in_memory.pyx

* Update spacy/pipeline/legacy/entity_linker.py

* Remove canididate.md. Remove mistakenly added config snippet in entity_linker.py.

Co-authored-by: Paul O'Leary McCann <polm@dampfkraft.com>
Co-authored-by: Sofie Van Landeghem <svlandeg@users.noreply.github.com>
2022-09-08 10:38:07 +02:00
..
models Refactor KB for easier customization (#11268) 2022-09-08 10:38:07 +02:00
__init__.py Auto-format code with black (#9530) 2021-10-22 13:03:10 +02:00
_character_embed.py 🏷 Add Mypy check to CI and ignore all existing Mypy errors (#9167) 2021-10-14 15:21:40 +02:00
_precomputable_affine.py Fix compatibility with CuPy 9.x (#11194) 2022-07-26 10:52:01 +02:00
callbacks.py Add NVTX ranges to TrainablePipe components (#10965) 2022-06-30 11:28:12 +02:00
extract_ngrams.py 🏷 Add Mypy check to CI and ignore all existing Mypy errors (#9167) 2021-10-14 15:21:40 +02:00
extract_spans.py Fix entity linker batching (#9669) 2022-03-04 09:17:36 +01:00
featureextractor.py Fix import 2020-10-02 01:12:34 +02:00
parser_model.pxd Parser: use C saxpy/sgemm provided by the Ops implementation (#10773) 2022-05-27 11:20:52 +02:00
parser_model.pyx precompute_hiddens/Parser: do not look up CPU ops (3.4) (#11069) 2022-07-05 10:53:42 +02:00
staticvectors.py Fix issues for Mypy 0.950 and Pydantic 1.9.0 (#10786) 2022-05-25 09:33:54 +02:00
tb_framework.py TransitionBasedParser.v1 to legacy (#8586) 2021-07-06 15:26:45 +02:00