* Fix batching regression
Some time ago, the spaCy v4 branch switched to the new Thinc v9
schedule. However, this introduced an error in how batching is handed.
In the PR, the batchers were changed to keep track of their step,
so that the step can be passed to the schedule. However, the issue
is that the training loop repeatedly calls the batching functions
(rather than using an infinite generator/iterator). So, the step and
therefore the schedule would be reset each epoch. Before the schedule
switch we didn't have this issue, because the old schedules were
stateful.
This PR fixes this issue by reverting the batching functions to use
a (stateful) generator. Their registry functions do accept a `Schedule`
and we convert `Schedule`s to generators.
* Update batcher docs
* Docstring fixes
* Make minibatch take iterables again as well
* Bump thinc requirement to 9.0.0.dev2
* Use type declaration
* Convert another comment into a proper type declaration
* Add `TrainablePipe.{distill,get_teacher_student_loss}`
This change adds two methods:
- `TrainablePipe::distill` which performs a training step of a
student pipe on a teacher pipe, giving a batch of `Doc`s.
- `TrainablePipe::get_teacher_student_loss` computes the loss
of a student relative to the teacher.
The `distill` or `get_teacher_student_loss` methods are also implemented
in the tagger, edit tree lemmatizer, and parser pipes, to enable
distillation in those pipes and as an example for other pipes.
* Fix stray `Beam` import
* Fix incorrect import
* Apply suggestions from code review
Co-authored-by: Sofie Van Landeghem <svlandeg@users.noreply.github.com>
* Apply suggestions from code review
Co-authored-by: Sofie Van Landeghem <svlandeg@users.noreply.github.com>
* TrainablePipe.distill: use `Iterable[Example]`
* Add Pipe.is_distillable method
* Add `validate_distillation_examples`
This first calls `validate_examples` and then checks that the
student/teacher tokens are the same.
* Update distill documentation
* Add distill documentation for all pipes that support distillation
* Fix incorrect identifier
* Apply suggestions from code review
Co-authored-by: Sofie Van Landeghem <svlandeg@users.noreply.github.com>
* Add comment to explain `is_distillable`
Co-authored-by: Sofie Van Landeghem <svlandeg@users.noreply.github.com>
* Add `training.before_update` callback
This callback can be used to implement training paradigms like gradual (un)freezing of components (e.g: the Transformer) after a certain number of training steps to mitigate catastrophic forgetting during fine-tuning.
* Fix type annotation, default config value
* Generalize arguments passed to the callback
* Update schema
* Pass `epoch` to callback, rename `current_step` to `step`
* Add test
* Simplify test
* Replace config string with `spacy.blank`
* Apply suggestions from code review
Co-authored-by: Adriane Boyd <adrianeboyd@gmail.com>
* Cleanup imports
Co-authored-by: Adriane Boyd <adrianeboyd@gmail.com>
* detect cycle during projectivize
* not complete test to detect cycle in projectivize
* boolean to int type to propagate error
* use unordered_set instead of set
* moved error message to errors
* removed cycle from test case
* use find instead of count
* cycle check: only perform one lookup
* Return bool again from _has_head_as_ancestor
Communicate presence of cycles through an output argument.
* Switch to returning std::pair to encode presence of a cycle
The has_cycle pointer is too easy to misuse. Ideally, we would have a
sum type like Rust's `Result` here, but C++ is not there yet.
* _is_non_proj_arc: clarify what we are returning
* _has_head_as_ancestor: remove count
We are now explicitly checking for cycles, so the algorithm must always
terminate. Either we encounter the head, we find a root, or a cycle.
* _is_nonproj_arc: simplify condition
* Another refactor using C++ exceptions
* Remove unused error code
* Print graph with cycle on exception
* Include .hh files in source package
* Add FIXME comment
* cycle detection test
* find cycle when starting from problematic vertex
Co-authored-by: Daniël de Kok <me@danieldk.eu>
* Alignment: use a simplified ragged type for performance
This introduces the AlignmentArray type, which is a simplified version
of Ragged that performs better on the simple(r) indexing performed for
alignment.
* AlignmentArray: raise an error when using unsupported index
* AlignmentArray: move error messages to Errors
* AlignmentArray: remove simlified ... with simplifications
* AlignmentArray: fix typo that broke a[n:n] indexing
* Tagger: use unnormalized probabilities for inference
Using unnormalized softmax avoids use of the relatively expensive exp function,
which can significantly speed up non-transformer models (e.g. I got a speedup
of 27% on a German tagging + parsing pipeline).
* Add spacy.Tagger.v2 with configurable normalization
Normalization of probabilities is disabled by default to improve
performance.
* Update documentation, models, and tests to spacy.Tagger.v2
* Move Tagger.v1 to spacy-legacy
* docs/architectures: run prettier
* Unnormalized softmax is now a Softmax_v2 option
* Require thinc 8.0.14 and spacy-legacy 3.0.9
* Migrate regressions 1-1000
* Move serialize test to correct file
* Remove tests that won't work in v3
* Migrate regressions 1000-1500
Removed regression test 1250 because v3 doesn't support the old LEX
scheme anymore.
* Add missing imports in serializer tests
* Migrate tests 1500-2000
* Migrate regressions from 2000-2500
* Migrate regressions from 2501-3000
* Migrate regressions from 3000-3501
* Migrate regressions from 3501-4000
* Migrate regressions from 4001-4500
* Migrate regressions from 4501-5000
* Migrate regressions from 5001-5501
* Migrate regressions from 5501 to 7000
* Migrate regressions from 7001 to 8000
* Migrate remaining regression tests
* Fixing missing imports
* Update docs with new system [ci skip]
* Update CONTRIBUTING.md
- Fix formatting
- Update wording
* Remove lemmatizer tests in el lang
* Move a few tests into the general tokenizer
* Separate Doc and DocBin tests
* Add all symbols in Unicode Currency Symbols block
In #8102 it came up that the rupee symbol was treated different from
dollar / euro / yen symbols. This adds many symbols not already
included.
* Fix test
* Fix training test
* extend span scorer with consider_label and allow_overlap
* unit test for spans y2x overlap
* add score_spans unit test
* docs for new fields in scorer.score_spans
* rename to include_label
* spell out if-else for clarity
* rename to 'labeled'
Co-authored-by: Adriane Boyd <adrianeboyd@gmail.com>
* Switch converters to generator functions
To reduce the memory usage when converting large corpora, refactor the
convert methods to be generator functions.
* Update tests
* When checking for token alignments, check not only that the tokens are
identical but that the character positions are both at the start of a
token.
It's possible for the tokens to be identical even though the two
tokens aren't aligned one-to-one in a case like `["a'", "''"]` vs.
`["a", "''", "'"]`, where the middle tokens are identical but should not
be aligned on the token level at character position 2 since it's the
start of one token but the middle of another.
* Use the lowercased version of the token texts to create the
character-to-token alignment because lowercasing can change the string
length (e.g., for `İ`, see the not-a-bug bug report:
https://bugs.python.org/issue34723)
* Replace pytokenizations with internal alignment
Replace pytokenizations with internal alignment algorithm that is
restricted to only allow differences in whitespace and capitalization.
* Rename `spacy.training.align` to `spacy.training.alignment` to contain
the `Alignment` dataclass
* Implement `get_alignments` in `spacy.training.align`
* Refactor trailing whitespace handling
* Remove unnecessary exception for empty docs
Allow a non-empty whitespace-only doc to be aligned with an empty doc
* Remove empty docs exceptions completely
* Refactor Token morph setting
* Remove `Token.morph_`
* Add `Token.set_morph()`
* `0` resets `token.c.morph` to unset
* Any other values are passed to `Morphology.add`
* Add token.morph setter to set from MorphAnalysis
* Support data augmentation in Corpus
* Note initial docs for data augmentation
* Add augmenter to quickstart
* Fix flake8
* Format
* Fix test
* Update spacy/tests/training/test_training.py
* Improve data augmentation arguments
* Update templates
* Move randomization out into caller
* Refactor
* Update spacy/training/augment.py
* Update spacy/tests/training/test_training.py
* Fix augment
* Fix test
