* Use cosine loss in Cloze multitask
* Fix char_embed for gpu
* Call resume_training for base model in train CLI
* Fix bilstm_depth default in pretrain command
* Implement character-based pretraining objective
* Use chars loss in ClozeMultitask
* Add method to decode predicted characters
* Fix number characters
* Rescale gradients for mlm
* Fix char embed+vectors in ml
* Fix pipes
* Fix pretrain args
* Move get_characters_loss
* Fix import
* Fix import
* Mention characters loss option in pretrain
* Remove broken 'self attention' option in pretrain
* Revert "Remove broken 'self attention' option in pretrain"
This reverts commit 56b820f6af.
* Document 'characters' objective of pretrain
* Update website models for v2.3.0
* Add docs for Chinese word segmentation
* Tighten up Chinese docs section
* Merge branch 'master' into docs/v2.3.0 [ci skip]
* Merge branch 'master' into docs/v2.3.0 [ci skip]
* Auto-format and update version
* Update matcher.md
* Update languages and sorting
* Typo in landing page
* Infobox about token_match behavior
* Add meta and basic docs for Japanese
* POS -> TAG in models table
* Add info about lookups for normalization
* Updates to API docs for v2.3
* Update adding norm exceptions for adding languages
* Add --omit-extra-lookups to CLI API docs
* Add initial draft of "What's New in v2.3"
* Add new in v2.3 tags to Chinese and Japanese sections
* Add tokenizer to migration section
* Add new in v2.3 flags to init-model
* Typo
* More what's new in v2.3
Co-authored-by: Ines Montani <ines@ines.io>
* `debug-data`: determine coverage of provided vectors
* `evaluate`: support `blank:lg` model to make it possible to just evaluate
tokenization
* `init-model`: add option to truncate vectors to N most frequent vectors
from word2vec file
* `train`:
* if training on GPU, only run evaluation/timing on CPU in the first
iteration
* if training is aborted, exit with a non-0 exit status
* Add arch for MishWindowEncoder
* Support mish in tok2vec and conv window >=2
* Pass new tok2vec settings from parser
* Syntax error
* Fix tok2vec setting
* Fix registration of MishWindowEncoder
* Fix receptive field setting
* Fix mish arch
* Pass more options from parser
* Support more tok2vec options in pretrain
* Require thinc 7.3
* Add docs [ci skip]
* Require thinc 7.3.0.dev0 to run CI
* Run black
* Fix typo
* Update Thinc version
Co-authored-by: Ines Montani <ines@ines.io>
* Support train dict format as JSONL
* Add (overly simple) check for dict vs. tuple to read JSONL lines as
either train dicts or train tuples
* Extend JSON/JSONL roundtrip conversion tests using `docs_to_json()`
and `GoldCorpus.train_tuples`
* Revert docs to default JSON output with convert
* Allow vectors name to be specified in init-model
* Document --vectors-name argument to init-model
* Update website/docs/api/cli.md
Co-Authored-By: Ines Montani <ines@ines.io>
* Add doc.cats to spacy.gold at the paragraph level
Support `doc.cats` as `"cats": [{"label": string, "value": number}]` in
the spacy JSON training format at the paragraph level.
* `spacy.gold.docs_to_json()` writes `docs.cats`
* `GoldCorpus` reads in cats in each `GoldParse`
* Update instances of gold_tuples to handle cats
Update iteration over gold_tuples / gold_parses to handle addition of
cats at the paragraph level.
* Add textcat to train CLI
* Add textcat options to train CLI
* Add textcat labels in `TextCategorizer.begin_training()`
* Add textcat evaluation to `Scorer`:
* For binary exclusive classes with provided label: F1 for label
* For 2+ exclusive classes: F1 macro average
* For multilabel (not exclusive): ROC AUC macro average (currently
relying on sklearn)
* Provide user info on textcat evaluation settings, potential
incompatibilities
* Provide pipeline to Scorer in `Language.evaluate` for textcat config
* Customize train CLI output to include only metrics relevant to current
pipeline
* Add textcat evaluation to evaluate CLI
* Fix handling of unset arguments and config params
Fix handling of unset arguments and model confiug parameters in Scorer
initialization.
* Temporarily add sklearn requirement
* Remove sklearn version number
* Improve Scorer handling of models without textcats
* Fixing Scorer handling of models without textcats
* Update Scorer output for python 2.7
* Modify inf in Scorer for python 2.7
* Auto-format
Also make small adjustments to make auto-formatting with black easier and produce nicer results
* Move error message to Errors
* Update documentation
* Add cats to annotation JSON format [ci skip]
* Fix tpl flag and docs [ci skip]
* Switch to internal roc_auc_score
Switch to internal `roc_auc_score()` adapted from scikit-learn.
* Add AUCROCScore tests and improve errors/warnings
* Add tests for AUCROCScore and roc_auc_score
* Add missing error for only positive/negative values
* Remove unnecessary warnings and errors
* Make reduced roc_auc_score functions private
Because most of the checks and warnings have been stripped for the
internal functions and access is only intended through `ROCAUCScore`,
make the functions for roc_auc_score adapted from scikit-learn private.
* Check that data corresponds with multilabel flag
Check that the training instances correspond with the multilabel flag,
adding the multilabel flag if required.
* Add textcat score to early stopping check
* Add more checks to debug-data for textcat
* Add example training data for textcat
* Add more checks to textcat train CLI
* Check configuration when extending base model
* Fix typos
* Update textcat example data
* Provide licensing details and licenses for data
* Remove two labels with no positive instances from jigsaw-toxic-comment
data.
Co-authored-by: Ines Montani <ines@ines.io>
* document token ent_kb_id
* document span kb_id
* update pipeline documentation
* prior and context weights as bool's instead
* entitylinker api documentation
* drop for both models
* finish entitylinker documentation
* small fixes
* documentation for KB
* candidate documentation
* links to api pages in code
* small fix
* frequency examples as counts for consistency
* consistent documentation about tensors returned by predict
* add entity linking to usage 101
* add entity linking infobox and KB section to 101
* entity-linking in linguistic features
* small typo corrections
* training example and docs for entity_linker
* predefined nlp and kb
* revert back to similarity encodings for simplicity (for now)
* set prior probabilities to 0 when excluded
* code clean up
* bugfix: deleting kb ID from tokens when entities were removed
* refactor train el example to use either model or vocab
* pretrain_kb example for example kb generation
* add to training docs for KB + EL example scripts
* small fixes
* error numbering
* ensure the language of vocab and nlp stay consistent across serialization
* equality with =
* avoid conflict in errors file
* add error 151
* final adjustements to the train scripts - consistency
* update of goldparse documentation
* small corrections
* push commit
* typo fix
* add candidate API to kb documentation
* update API sidebar with EntityLinker and KnowledgeBase
* remove EL from 101 docs
* remove entity linker from 101 pipelines / rephrase
* custom el model instead of existing model
* set version to 2.2 for EL functionality
* update documentation for 2 CLI scripts
* Updates/bugfixes for NER/IOB converters
* Converter formats `ner` and `iob` use autodetect to choose a converter if
possible
* `iob2json` is reverted to handle sentence-per-line data like
`word1|pos1|ent1 word2|pos2|ent2`
* Fix bug in `merge_sentences()` so the second sentence in each batch isn't
skipped
* `conll_ner2json` is made more general so it can handle more formats with
whitespace-separated columns
* Supports all formats where the first column is the token and the final
column is the IOB tag; if present, the second column is the POS tag
* As in CoNLL 2003 NER, blank lines separate sentences, `-DOCSTART- -X- O O`
separates documents
* Add option for segmenting sentences (new flag `-s`)
* Parser-based sentence segmentation with a provided model, otherwise with
sentencizer (new option `-b` to specify model)
* Can group sentences into documents with `n_sents` as long as sentence
segmentation is available
* Only applies automatic segmentation when there are no existing delimiters
in the data
* Provide info about settings applied during conversion with warnings and
suggestions if settings conflict or might not be not optimal.
* Add tests for common formats
* Add '(default)' back to docs for -c auto
* Add document count back to output
* Revert changes to converter output message
* Use explicit tabs in convert CLI test data
* Adjust/add messages for n_sents=1 default
* Add sample NER data to training examples
* Update README
* Add links in docs to example NER data
* Define msg within converters
* Update pretrain to prevent unintended overwriting of weight files for #3859
* Add '--epoch-start' to pretrain docs
* Add mising pretrain arguments to bash example
* Update doc tag for v2.1.5
* Add error to `get_vectors_loss` for unsupported loss function of `pretrain`
* Add missing "--loss-func" argument to pretrain docs. Update pretrain plac annotations to match docs.
* Add missing quotation marks
* Add check for empty input file to CLI pretrain
* Raise error if JSONL is not a dict or contains neither `tokens` nor `text` key
* Skip empty values for correct pretrain keys and log a counter as warning
* Add tests for CLI pretrain core function make_docs.
* Add a short hint for the `tokens` key to the CLI pretrain docs
* Add success message to CLI pretrain
* Update model loading to fix the tests
* Skip empty values and do not create docs out of it
<!--- Provide a general summary of your changes in the title. -->
When using `spacy pretrain`, the model is saved only after every epoch. But each epoch can be very big since `pretrain` is used for language modeling tasks. So I added a `--save-every` option in the CLI to save after every `--save-every` batches.
## Description
<!--- Use this section to describe your changes. If your changes required
testing, include information about the testing environment and the tests you
ran. If your test fixes a bug reported in an issue, don't forget to include the
issue number. If your PR is still a work in progress, that's totally fine – just
include a note to let us know. -->
To test...
Save this file to `sample_sents.jsonl`
```
{"text": "hello there."}
{"text": "hello there."}
{"text": "hello there."}
{"text": "hello there."}
{"text": "hello there."}
{"text": "hello there."}
{"text": "hello there."}
{"text": "hello there."}
{"text": "hello there."}
{"text": "hello there."}
{"text": "hello there."}
{"text": "hello there."}
{"text": "hello there."}
{"text": "hello there."}
{"text": "hello there."}
{"text": "hello there."}
```
Then run `--save-every 2` when pretraining.
```bash
spacy pretrain sample_sents.jsonl en_core_web_md here -nw 1 -bs 1 -i 10 --save-every 2
```
And it should save the model to the `here/` folder after every 2 batches. The models that are saved during an epoch will have a `.temp` appended to the save name.
At the end the training, you should see these files (`ls here/`):
```bash
config.json model2.bin model5.bin model8.bin
log.jsonl model2.temp.bin model5.temp.bin model8.temp.bin
model0.bin model3.bin model6.bin model9.bin
model0.temp.bin model3.temp.bin model6.temp.bin model9.temp.bin
model1.bin model4.bin model7.bin
model1.temp.bin model4.temp.bin model7.temp.bin
```
### Types of change
<!-- What type of change does your PR cover? Is it a bug fix, an enhancement
or new feature, or a change to the documentation? -->
This is a new feature to `spacy pretrain`.
🌵 **Unfortunately, I haven't been able to test this because compiling from source is not working (cythonize error).**
```
Processing matcher.pyx
[Errno 2] No such file or directory: '/Users/mwu/github/spaCy/spacy/matcher.pyx'
Traceback (most recent call last):
File "/Users/mwu/github/spaCy/bin/cythonize.py", line 169, in <module>
run(args.root)
File "/Users/mwu/github/spaCy/bin/cythonize.py", line 158, in run
process(base, filename, db)
File "/Users/mwu/github/spaCy/bin/cythonize.py", line 124, in process
preserve_cwd(base, process_pyx, root + ".pyx", root + ".cpp")
File "/Users/mwu/github/spaCy/bin/cythonize.py", line 87, in preserve_cwd
func(*args)
File "/Users/mwu/github/spaCy/bin/cythonize.py", line 63, in process_pyx
raise Exception("Cython failed")
Exception: Cython failed
Traceback (most recent call last):
File "setup.py", line 276, in <module>
setup_package()
File "setup.py", line 209, in setup_package
generate_cython(root, "spacy")
File "setup.py", line 132, in generate_cython
raise RuntimeError("Running cythonize failed")
RuntimeError: Running cythonize failed
```
Edit: Fixed! after deleting all `.cpp` files: `find spacy -name "*.cpp" | xargs rm`
## Checklist
<!--- Before you submit the PR, go over this checklist and make sure you can
tick off all the boxes. [] -> [x] -->
- [x] I have submitted the spaCy Contributor Agreement.
- [x] I ran the tests, and all new and existing tests passed.
- [x] My changes don't require a change to the documentation, or if they do, I've added all required information.
Add and document CLI options for batch size, max doc length, min doc length for `spacy pretrain`.
Also improve CLI output.
Closes#3216
## Checklist
<!--- Before you submit the PR, go over this checklist and make sure you can
tick off all the boxes. [] -> [x] -->
- [x] I have submitted the spaCy Contributor Agreement.
- [x] I ran the tests, and all new and existing tests passed.
- [x] My changes don't require a change to the documentation, or if they do, I've added all required information.
