This patch improves tokenizer speed by about 10%, and reduces memory usage in the `Vocab` by removing a redundant index. The `vocab._by_orth` and `vocab._by_hash` indexed on different data in v1, but in v2 the orth and the hash are identical.
The patch also fixes an uninitialized variable in the tokenizer, the `has_special` flag. This checks whether a chunk we're tokenizing triggers a special-case rule. If it does, then we avoid caching within the chunk. This check led to incorrectly rejecting some chunks from the cache.
With the `en_core_web_md` model, we now tokenize the IMDB train data at 503,104k words per second. Prior to this patch, we had 465,764k words per second.
Before switching to the regex library and supporting more languages, we had 1.3m words per second for the tokenizer. In order to recover the missing speed, we need to:
* Fix the variable-length lookarounds in the suffix, infix and `token_match` rules
* Improve the performance of the `token_match` regex
* Switch back from the `regex` library to the `re` library.
## 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.
Hi guys,
This is my first spaCy extension. I am excited to able to do this. Please do let me know if there is any suggestions or modifications I need to do. Feel free to use/contribute the repo that I made.
## Description
ExcelCy is a SpaCy toolkit to help improve the data training experiences. It provides easy annotation using Excel file format. It has helper to pre-train entity annotation with phrase and regex matcher pipe.
### Types of change
Update to Universe list in website.
## Checklist
- [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 helper function for reading in JSONL
* Add rule-based NER component
* Fix whitespace
* Add component to factories
* Add tests
* Add option to disable indent on json_dumps compat
Otherwise, reading JSONL back in line by line won't work
* Fix error code
* Fix venv command examples
The documentation refers to `venv`, which is native to Python3.
However, the command examples are as if they were still `virtualenv`,
which is a package independent of `venv`:
- It doesn't need to be installed via `pip`. In fact `pip install venv` would
return an error.
- The correct way to invoke `venv` is `python3 -m venv`, not `venv`, which would
return command not found.
See https://docs.python.org/3/library/venv.html
I suspect the documentation simply replaced all occurrences of `virtualenv` with
`venv`. However they are different modules and are used differently.
* Update comment [ci skip]
List created by taking the 2000 top words from a Wikipedia dump and
removing everything that wasn't hiragana.
Tried going through kanji words and deciding what to keep but there were
too many obvious non-stopwords (東京 was in the top 500) and many other
words where it wasn't clear if they should be included or not.
<!--- Provide a general summary of your changes in the title. -->
## Description
This PR corrects the German lemma form for the word "Rang". Initially, the lemma form was "ringen", which is not correct, because it refers to the verb ("ringen") and not to the noun ("Rang").
### Types of change
The lemma form for "Rang" is corrected to "Rang", see also the [Duden](https://www.duden.de/rechtschreibung/Rang) entry.
## 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.
Matthew Honnibal