Add a context manage nlp.memory_zone(), which will begin
memory_zone() blocks on the vocab, string store, and potentially
other components.
Once the memory_zone() block expires, spaCy will free any shared
resources that were allocated for the text-processing that occurred
within the memory_zone. If you create Doc objects within a memory
zone, it's invalid to access them once the memory zone is expired.
The purpose of this is that spaCy creates and stores Lexeme objects
in the Vocab that can be shared between multiple Doc objects. It also
interns strings. Normally, spaCy can't know when all Doc objects using
a Lexeme are out-of-scope, so new Lexemes accumulate in the vocab,
causing memory pressure.
Memory zones solve this problem by telling spaCy "okay none of the
documents allocated within this block will be accessed again". This
lets spaCy free all new Lexeme objects and other data that were
created during the block.
The mechanism is general, so memory_zone() context managers can be
added to other components that could benefit from them, e.g. pipeline
components.
I experimented with adding memory zone support to the tokenizer as well,
for its cache. However, this seems unnecessarily complicated. It makes
more sense to just stick a limit on the cache size. This lets spaCy
benefit from the efficiency advantage of the cache better, because
we can maintain a (bounded) cache even if only small batches of
documents are being processed.
* Update Tokenizer.explain for special cases with whitespace
Update `Tokenizer.explain` to skip special case matches if the exact
text has not been matched due to intervening whitespace.
Enable fuzzy `Tokenizer.explain` tests with additional whitespace
normalization.
* Add unit test for special cases with whitespace, xfail fuzzy tests again
* Use isort with Black profile
* isort all the things
* Fix import cycles as a result of import sorting
* Add DOCBIN_ALL_ATTRS type definition
* Add isort to requirements
* Remove isort from build dependencies check
* Typo
In `Tokenizer.from_bytes`, the exceptions should be loaded last so that
they are only processed once as part of loading the model.
The exceptions are tokenized as phrase matcher patterns in the
background and the internal tokenization needs to be synced with all the
remaining tokenizer settings. If the exceptions are not loaded last,
there are speed regressions for `Tokenizer.from_bytes/disk` vs.
`Tokenizer.add_special_case` as the caches are reloaded more than
necessary during deserialization.
* Don't pass mem pool to new lexeme function
* Remove unused mem from function args
Two methods calling _new_lexeme, get and get_by_orth, took mem arguments
just to call the internal method. That's no longer necessary, so this
cleans it up.
* prettier formatting
* Remove more unused mem args
* Clean up old Matcher call style related stuff
In v2 Matcher.add was called with (key, on_match, *patterns). In v3 this
was changed to (key, patterns, *, on_match=None), but there were various
points where the old call syntax was documented or handled specially.
This removes all those.
The Matcher itself didn't need any code changes, as it just gives a
generic type error. However the PhraseMatcher required some changes
because it would automatically "fix" the old call style.
Surprisingly, the tokenizer was still using the old call style in one
place.
After these changes tests failed in two places:
1. one test for the "new" call style, including the "old" call style. I
removed this test.
2. deserializing the PhraseMatcher fails because the input docs are a
set.
I am not sure why 2 is happening - I guess it's a quirk of the
serialization format? - so for now I just convert the set to a list when
deserializing. The check that the input Docs are a List in the
PhraseMatcher is a new check, but makes it parallel with the other
Matchers, which seemed like the right thing to do.
* Add notes related to input docs / deserialization type
* Remove Typing import
* Remove old note about call style change
* Apply suggestions from code review
Co-authored-by: Adriane Boyd <adrianeboyd@gmail.com>
* Use separate method for setting internal doc representations
In addition to the title change, this changes the internal dict to be a
defaultdict, instead of a dict with frequent use of setdefault.
* Add _add_from_arrays for unpickling
* Cleanup around adding from arrays
This moves adding to internal structures into the private batch method,
and removes the single-add method.
This has one behavioral change for `add`, in that if something is wrong
with the list of input Docs (such as one of the items not being a Doc),
valid items before the invalid one will not be added. Also the callback
will not be updated if anything is invalid. This change should not be
significant.
This also adds a test to check failure when given a non-Doc.
* Update spacy/matcher/phrasematcher.pyx
Co-authored-by: Adriane Boyd <adrianeboyd@gmail.com>
Co-authored-by: Adriane Boyd <adrianeboyd@gmail.com>
* cleanup Tokenizer fields
* remove unused object from vocab
* remove IS_OOV_DEPRECATED
* add back in as FLAG13
* FLAG 18 instead
* import fix
* fix clumpsy fingers
* revert symbol changes in favor of #11352
* bint instead of bool
* Add tokenizer option to allow Matcher handling for all rules
Add tokenizer option `with_faster_rules_heuristics` that determines
whether the special cases applied by the internal `Matcher` are filtered
by whether they contain affixes or space. If `True` (default), the rules
are filtered to prioritize speed over rare edge cases. If `False`, all
rules are included in the final `Matcher`-based pass over the doc.
* Reset all caches when reloading special cases
* Revert "Reset all caches when reloading special cases"
This reverts commit 4ef6bd171d.
* Initialize max_length properly
* Add new tag to API docs
* Rename to faster heuristics
* Fix infix as prefix in Tokenizer.explain
Update `Tokenizer.explain` to align with the `Tokenizer` algorithm:
* skip infix matches that are prefixes in the current substring
* Update tokenizer pseudocode in docs
* Ignore prefix in suffix matches
Ignore the currently matched prefix when looking for suffix matches in
the tokenizer. Otherwise a lookbehind in the suffix pattern may match
incorrectly due the presence of the prefix in the token string.
* Move °[cfkCFK]. to a tokenizer exception
* Adjust exceptions for same tokenization as v3.1
* Also update test accordingly
* Continue to split . after °CFK if ° is not a prefix
* Exclude new ° exceptions for pl
* Switch back to default tokenization of "° C ."
* Revert "Exclude new ° exceptions for pl"
This reverts commit 952013a5b4.
* Add exceptions for °C for hu
* Replace all basestring references with unicode
`basestring` was a compatability type introduced by Cython to make
dealing with utf-8 strings in Python2 easier. In Python3 it is
equivalent to the unicode (or str) type.
I replaced all references to basestring with unicode, since that was
used elsewhere, but we could also just replace them with str, which
shoudl also be equivalent.
All tests pass locally.
* Replace all references to unicode type with str
Since we only support python3 this is simpler.
* Remove all references to unicode type
This removes all references to the unicode type across the codebase and
replaces them with `str`, which makes it more drastic than the prior
commits. In order to make this work importing `unicode_literals` had to
be removed, and one explicit unicode literal also had to be removed (it
is unclear why this is necessary in Cython with language level 3, but
without doing it there were errors about implicit conversion).
When `unicode` is used as a type in comments it was also edited to be
`str`.
Additionally `coding: utf8` headers were removed from a few files.
* Pass excludes when serializing vocab
Additional minor bug fix:
* Deserialize vocab in `EntityLinker.from_disk`
* Add test for excluding strings on load
* Fix formatting
* Fix tokenizer cache flushing
Fix/simplify tokenizer init detection in order to fix cache flushing
when properties are modified.
* Remove init reloading logic
* Remove logic disabling `_reload_special_cases` on init
* Setting `rules` last in `__init__` (as before) means that setting
other properties doesn't reload any special cases
* Reset `rules` first in `from_bytes` so that setting other properties
during deserialization doesn't reload any special cases
unnecessarily
* Reset all properties in `Tokenizer.from_bytes` to allow any settings
to be `None`
* Also reset special matcher when special cache is flushed
* Remove duplicate special case validation
* Add test for special cases flushing
* Extend test for tokenizer deserialization of None values
* Update Tokenizer.explain with special matches
Update `Tokenizer.explain` and the pseudo-code in the docs to include
the processing of special cases that contain affixes or whitespace.
* Handle optional settings in explain
* Add test for special matches in explain
Add test for `Tokenizer.explain` for special cases containing affixes.
* Add regression test
* Run PhraseMatcher on Spans
* Add test for PhraseMatcher on Spans and Docs
* Add SCA
* Add test with 3 matches in Doc, 1 match in Span
* Update docs
* Use doc.length for find_matches in tokenizer
Co-authored-by: Adriane Boyd <adrianeboyd@gmail.com>
* Override language defaults for null token and URL match
When the serialized `token_match` or `url_match` is `None`, override the
language defaults to preserve `None` on deserialization.
* Fix fixtures in tests
Preserve `token.spacy` corresponding to the span end token in the
original doc rather than adjusting for the current offset.
* If not modifying in place, this checks in the original document
(`doc.c` rather than `tokens`).
* If modifying in place, the document has not been modified past the
current span start position so the value at the current span end
position is valid.
* Refactor the Scorer to improve flexibility
Refactor the `Scorer` to improve flexibility for arbitrary pipeline
components.
* Individual pipeline components provide their own `evaluate` methods
that score a list of `Example`s and return a dictionary of scores
* `Scorer` is initialized either:
* with a provided pipeline containing components to be scored
* with a default pipeline containing the built-in statistical
components (senter, tagger, morphologizer, parser, ner)
* `Scorer.score` evaluates a list of `Example`s and returns a dictionary
of scores referring to the scores provided by the components in the
pipeline
Significant differences:
* `tags_acc` is renamed to `tag_acc` to be consistent with `token_acc`
and the new `morph_acc`, `pos_acc`, and `lemma_acc`
* Scoring is no longer cumulative: `Scorer.score` scores a list of
examples rather than a single example and does not retain any state
about previously scored examples
* PRF values in the returned scores are no longer multiplied by 100
* Add kwargs to Morphologizer.evaluate
* Create generalized scoring methods in Scorer
* Generalized static scoring methods are added to `Scorer`
* Methods require an attribute (either on Token or Doc) that is
used to key the returned scores
Naming differences:
* `uas`, `las`, and `las_per_type` in the scores dict are renamed to
`dep_uas`, `dep_las`, and `dep_las_per_type`
Scoring differences:
* `Doc.sents` is now scored as spans rather than on sentence-initial
token positions so that `Doc.sents` and `Doc.ents` can be scored with
the same method (this lowers scores since a single incorrect sentence
start results in two incorrect spans)
* Simplify / extend hasattr check for eval method
* Add hasattr check to tokenizer scoring
* Simplify to hasattr check for component scoring
* Reset Example alignment if docs are set
Reset the Example alignment if either doc is set in case the
tokenization has changed.
* Add PRF tokenization scoring for tokens as spans
Add PRF scores for tokens as character spans. The scores are:
* token_acc: # correct tokens / # gold tokens
* token_p/r/f: PRF for (token.idx, token.idx + len(token))
* Add docstring to Scorer.score_tokenization
* Rename component.evaluate() to component.score()
* Update Scorer API docs
* Update scoring for positive_label in textcat
* Fix TextCategorizer.score kwargs
* Update Language.evaluate docs
* Update score names in default config
* Update with WIP
* Update with WIP
* Update with pipeline serialization
* Update types and pipe factories
* Add deep merge, tidy up and add tests
* Fix pipe creation from config
* Don't validate default configs on load
* Update spacy/language.py
Co-authored-by: Ines Montani <ines@ines.io>
* Adjust factory/component meta error
* Clean up factory args and remove defaults
* Add test for failing empty dict defaults
* Update pipeline handling and methods
* provide KB as registry function instead of as object
* small change in test to make functionality more clear
* update example script for EL configuration
* Fix typo
* Simplify test
* Simplify test
* splitting pipes.pyx into separate files
* moving default configs to each component file
* fix batch_size type
* removing default values from component constructors where possible (TODO: test 4725)
* skip instead of xfail
* Add test for config -> nlp with multiple instances
* pipeline.pipes -> pipeline.pipe
* Tidy up, document, remove kwargs
* small cleanup/generalization for Tok2VecListener
* use DEFAULT_UPSTREAM field
* revert to avoid circular imports
* Fix tests
* Replace deprecated arg
* Make model dirs require config
* fix pickling of keyword-only arguments in constructor
* WIP: clean up and integrate full config
* Add helper to handle function args more reliably
Now also includes keyword-only args
* Fix config composition and serialization
* Improve config debugging and add visual diff
* Remove unused defaults and fix type
* Remove pipeline and factories from meta
* Update spacy/default_config.cfg
Co-authored-by: Sofie Van Landeghem <svlandeg@users.noreply.github.com>
* Update spacy/default_config.cfg
* small UX edits
* avoid printing stack trace for debug CLI commands
* Add support for language-specific factories
* specify the section of the config which holds the model to debug
* WIP: add Language.from_config
* Update with language data refactor WIP
* Auto-format
* Add backwards-compat handling for Language.factories
* Update morphologizer.pyx
* Fix morphologizer
* Update and simplify lemmatizers
* Fix Japanese tests
* Port over tagger changes
* Fix Chinese and tests
* Update to latest Thinc
* WIP: xfail first Russian lemmatizer test
* Fix component-specific overrides
* fix nO for output layers in debug_model
* Fix default value
* Fix tests and don't pass objects in config
* Fix deep merging
* Fix lemma lookup data registry
Only load the lookups if an entry is available in the registry (and if spacy-lookups-data is installed)
* Add types
* Add Vocab.from_config
* Fix typo
* Fix tests
* Make config copying more elegant
* Fix pipe analysis
* Fix lemmatizers and is_base_form
* WIP: move language defaults to config
* Fix morphology type
* Fix vocab
* Remove comment
* Update to latest Thinc
* Add morph rules to config
* Tidy up
* Remove set_morphology option from tagger factory
* Hack use_gpu
* Move [pipeline] to top-level block and make [nlp.pipeline] list
Allows separating component blocks from component order – otherwise, ordering the config would mean a changed component order, which is bad. Also allows initial config to define more components and not use all of them
* Fix use_gpu and resume in CLI
* Auto-format
* Remove resume from config
* Fix formatting and error
* [pipeline] -> [components]
* Fix types
* Fix tagger test: requires set_morphology?
Co-authored-by: Sofie Van Landeghem <svlandeg@users.noreply.github.com>
Co-authored-by: svlandeg <sofie.vanlandeghem@gmail.com>
Co-authored-by: Matthew Honnibal <honnibal+gh@gmail.com>
