The new spacy pretrain command implemented BERT/ULMFit/etc-like transfer learning, using our Language Modelling with Approximate Outputs version of BERT's cloze task. Pretraining is convenient, but in some ways it's a bit of a strange solution. All we're doing is initialising the weights. At the same time, we're putting a lot of work into our optimisation so that it's less sensitive to initial conditions, and more likely to find good optima. I discuss this a bit in the pseudo-rehearsal blog post: https://explosion.ai/blog/pseudo-rehearsal-catastrophic-forgetting
Support semi-supervised learning in spacy train
One obvious way to improve these pretraining methods is to do multi-task learning, instead of just transfer learning. This has been shown to work very well: https://arxiv.org/pdf/1809.08370.pdf . This patch makes it easy to do this sort of thing.
Add a new argument to spacy train, --raw-text. This takes a jsonl file with unlabelled data that can be used in arbitrary ways to do semi-supervised learning.
Add a new method to the Language class and to pipeline components, .rehearse(). This is like .update(), but doesn't expect GoldParse objects. It takes a batch of Doc objects, and performs an update on some semi-supervised objective.
Move the BERT-LMAO objective out from spacy/cli/pretrain.py into spacy/_ml.py, so we can create a new pipeline component, ClozeMultitask. This can be specified as a parser or NER multitask in the spacy train command. Example usage:
python -m spacy train en ./tmp ~/data/en-core-web/train/nw.json ~/data/en-core-web/dev/nw.json --pipeline parser --raw-textt ~/data/unlabelled/reddit-100k.jsonl --vectors en_vectors_web_lg --parser-multitasks cloze
Implement rehearsal methods for pipeline components
The new --raw-text argument and nlp.rehearse() method also gives us a good place to implement the the idea in the pseudo-rehearsal blog post in the parser. This works as follows:
Add a new nlp.resume_training() method. This allocates copies of pre-trained models in the pipeline, setting things up for the rehearsal updates. It also returns an optimizer object. This also greatly reduces confusion around the nlp.begin_training() method, which randomises the weights, making it not suitable for adding new labels or otherwise fine-tuning a pre-trained model.
Implement rehearsal updates on the Parser class, making it available for the dependency parser and NER. During rehearsal, the initial model is used to supervise the model being trained. The current model is asked to match the predictions of the initial model on some data. This minimises catastrophic forgetting, by keeping the model's predictions close to the original. See the blog post for details.
Implement rehearsal updates for tagger
Implement rehearsal updates for text categoriz
Currently the TextCategorizer defaults to a fairly complicated model, designed partly around the active learning requirements of Prodigy. The model's a bit slow, and not very GPU-friendly.
This patch implements a straightforward CNN model that still performs pretty well. The replacement model also makes it easy to use the LMAO pretraining, since most of the parameters are in the CNN.
The replacement model has a flag to specify whether labels are mutually exclusive, which defaults to True. This has been a common problem with the text classifier. We'll also now be able to support adding labels to pretrained models again.
Resolves#2934, #2756, #1798, #1748.
Fixes#3027.
* Allow Span.__init__ to take unicode values for the `label` argument.
* Allow `Span.label_` to be writeable.
- [x] I have submitted the spaCy Contributor Agreement.
- [x] I ran the tests, and all new and existing tests passed.
- [ ] My changes don't require a change to the documentation, or if they do, I've added all required information.
* Add todo
* Auto-format
* Update wasabi pin
* Format training results with wasabi
* Remove loading animation from model saving
Currently behaves weirdly
* Inline messages
* Remove unnecessary path2str
Already taken care of by printer
* Inline messages in CLI
* Remove unused function
* Move loading indicator into loading function
* Check for invalid whitespace entities
See #3028. The solution in this patch is pretty debateable.
What we do is give the TokenC struct a .norm field, by repurposing the previously idle .sense attribute. It's nice to repurpose a previous field because it means the TokenC doesn't change size, so even if someone's using the internals very deeply, nothing will break.
The weird thing here is that the TokenC and the LexemeC both have an attribute named NORM. This arguably assists in backwards compatibility. On the other hand, maybe it's really bad! We're changing the semantics of the attribute subtly, so maybe it's better if someone calling lex.norm gets a breakage, and instead is told to write lex.default_norm?
Overall I believe this patch makes the NORM feature work the way we sort of expected it to work. Certainly it's much more like how the docs describe it, and more in line with how we've been directing people to use the norm attribute. We'll also be able to use token.norm to do stuff like spelling correction, which is pretty cool.
Fix a bug in the JSON streaming code that GoldCorpus uses. Escaped
slashes were being handled incorrectly. This bug caused low scores for
French in the early v2.1.0 alphas, because most of the data was not
being read in.
Fittingly, the document that triggered the bug was a Wikipedia article about
Perl. Parsing perl remains difficult!
