spaCy/website/usage/_install/_troubleshooting.jade

200 lines
9.4 KiB
Plaintext
Raw Normal View History

2017-10-03 15:26:20 +03:00
//- 💫 DOCS > USAGE > INSTALL > TROUBLESHOOTING
p
| This section collects some of the most common errors you may come
| across when installing, loading and using spaCy, as well as their solutions.
+aside("Help us improve this guide")
| Did you come across a problem like the ones listed here and want to
| share the solution? You can find the "Suggest edits" button at the
| bottom of this page that points you to the source. We always
| appreciate #[+a(gh("spaCy") + "/pulls") pull requests]!
+h(3, "compatible-model") No compatible model found
+code(false, "text").
No compatible model found for [lang] (spaCy v#{SPACY_VERSION}).
p
| This usually means that the model you're trying to download does not
| exist, or isn't available for your version of spaCy. Check the
| #[+a(gh("spacy-models", "compatibility.json")) compatibility table]
| to see which models are available for your spaCy version. If you're using
| an old version, consider upgrading to the latest release. Note that while
| spaCy supports tokenization for
2017-11-06 15:20:36 +03:00
| #[+a("/usage/models#languages") a variety of languages],
2017-10-03 15:26:20 +03:00
| not all of them come with statistical models. To only use the tokenizer,
| import the language's #[code Language] class instead, for example
| #[code from spacy.fr import French].
+h(3, "symlink-privilege") Symbolic link privilege not held
+code(false, "text").
OSError: symbolic link privilege not held
p
| To create #[+a("/usage/models#usage") shortcut links] that let you
2017-10-03 15:26:20 +03:00
| load models by name, spaCy creates a symbolic link in the
| #[code spacy/data] directory. This means your user needs permission to do
| this. The above error mostly occurs when doing a system-wide installation,
| which will create the symlinks in a system directory. Run the
| #[code download] or #[code link] command as administrator (on Windows,
2018-05-07 22:24:35 +03:00
| you can either right-click on your terminal or shell ans select "Run as
| Administrator"), or use a #[code virtualenv] to install spaCy in a user
| directory, instead of doing a system-wide installation.
2017-10-03 15:26:20 +03:00
+h(3, "no-cache-dir") No such option: --no-cache-dir
+code(false, "text").
no such option: --no-cache-dir
p
| The #[code download] command uses pip to install the models and sets the
| #[code --no-cache-dir] flag to prevent it from requiring too much memory.
| #[+a("https://pip.pypa.io/en/stable/reference/pip_install/#caching") This setting]
| requires pip v6.0 or newer. Run #[code pip install -U pip] to upgrade to
| the latest version of pip. To see which version you have installed,
| run #[code pip --version].
+h(3, "unknown-locale") Unknown locale: UTF-8
+code(false, "text").
ValueError: unknown locale: UTF-8
p
| This error can sometimes occur on OSX and is likely related to a
| still unresolved #[+a("https://bugs.python.org/issue18378") Python bug].
| However, it's easy to fix: just add the following to your
| #[code ~/.bash_profile] or #[code ~/.zshrc] and then run
| #[code source ~/.bash_profile] or #[code source ~/.zshrc].
| Make sure to add #[strong both lines] for #[code LC_ALL] and
| #[code LANG].
+code(false, "bash").
export LC_ALL=en_US.UTF-8
export LANG=en_US.UTF-8
2017-10-03 15:26:20 +03:00
+h(3, "import-error") Import error
+code(false, "text").
Import Error: No module named spacy
p
| This error means that the spaCy module can't be located on your system, or in
| your environment. Make sure you have spaCy installed. If you're using a
| #[code virtualenv], make sure it's activated and check that spaCy is
| installed in that environment otherwise, you're trying to load a system
| installation. You can also run #[code which python] to find out where
| your Python executable is located.
+h(3, "import-error-models") Import error: models
+code(false, "text").
ImportError: No module named 'en_core_web_sm'
p
| As of spaCy v1.7, all models can be installed as Python packages. This means
| that they'll become importable modules of your application. When creating
| #[+a("/usage/models#usage") shortcut links], spaCy will also try
2017-10-03 15:26:20 +03:00
| to import the model to load its meta data. If this fails, it's usually a
| sign that the package is not installed in the current environment.
| Run #[code pip list] or #[code pip freeze] to check which model packages
| you have installed, and install the
| #[+a("/models") correct models] if necessary. If you're
| importing a model manually at the top of a file, make sure to use the name
| of the package, not the shortcut link you've created.
+h(3, "vocab-strings") File not found: vocab/strings.json
+code(false, "text").
FileNotFoundError: No such file or directory: [...]/vocab/strings.json
p
| This error may occur when using #[code spacy.load()] to load
| a language model either because you haven't set up a
| #[+a("/usage/models#usage") shortcut link] for it, or because it
| doesn't actually exist. Set up a link for the model
2017-10-03 15:26:20 +03:00
| you want to load. This can either be an installed model package, or a
| local directory containing the model data. If you want to use one of the
| #[+a("/usage/models#languages") alpha tokenizers] for
| languages that don't yet have a statistical model, you should import its
| #[code Language] class instead, for example
2017-11-06 15:17:09 +03:00
| #[code from spacy.lang.bn import Bengali]. You can also use
| #[+api("top-level#spacy.blank") #[code spacy.blank]] to create a blank
| instance, e.g. #[code nlp = spacy.blank('bn')].
2017-10-03 15:26:20 +03:00
+h(3, "command-not-found") Command not found
+code(false, "text").
command not found: spacy
p
| This error may occur when running the #[code spacy] command from the
| command line. spaCy does not currently add an entry to our #[code PATH]
| environment variable, as this can lead to unexpected results, especially
| when using #[code virtualenv]. Instead, spaCy adds an auto-alias that
| maps #[code spacy] to #[code python -m spacy]. If this is not working as
| expected, run the command with #[code python -m], yourself
| for example #[code python -m spacy download en]. For more info on this,
2017-11-06 15:17:09 +03:00
| see the #[+api("cli#download") #[code download]] command.
2017-10-03 15:26:20 +03:00
+h(3, "module-load") 'module' object has no attribute 'load'
+code(false, "text").
AttributeError: 'module' object has no attribute 'load'
p
| While this could technically have many causes, including spaCy being
| broken, the most likely one is that your script's file or directory name
| is "shadowing" the module e.g. your file is called #[code spacy.py],
| or a directory you're importing from is called #[code spacy]. So, when
| using spaCy, never call anything else #[code spacy].
+h(3, "pron-lemma") Pronoun lemma is returned as #[code -PRON-]
+code.
doc = nlp(u'They are')
print(doc[0].lemma_)
# -PRON-
p
| This is in fact expected behaviour and not a bug.
| Unlike verbs and common nouns, there's no clear base form of a personal
| pronoun. Should the lemma of "me" be "I", or should we normalize person
| as well, giving "it" — or maybe "he"? spaCy's solution is to introduce a
| novel symbol, #[code -PRON-], which is used as the lemma for
| all personal pronouns. For more info on this, see the
2017-11-06 15:17:09 +03:00
| #[+a("/api/annotation#lemmatization") lemmatization specs].
+h(3, "catastrophic-forgetting") NER model doesn't recognise other entities anymore after training
p
| If your training data only contained new entities and you didn't mix in
| any examples the model previously recognised, it can cause the model to
| "forget" what it had previously learned. This is also referred to as the
| #[+a("https://explosion.ai/blog/pseudo-rehearsal-catastrophic-forgetting", true) "catastrophic forgetting problem"].
| A solution is to pre-label some text, and mix it with the new text in
| your updates. You can also do this by running spaCy over some text,
| extracting a bunch of entities the model previously recognised correctly,
| and adding them to your training examples.
+h(3, "unhashable-list") Unhashable type: 'list'
+code(false, "text").
TypeError: unhashable type: 'list'
p
| If you're training models, writing them to disk, and versioning them with
| git, you might encounter this error when trying to load them in a Windows
| environment. This happens because a default install of Git for Windows is
| configured to automatically convert Unix-style end-of-line characters
| (LF) to Windows-style ones (CRLF) during file checkout (and the reverse
| when commiting). While that's mostly fine for text files, a trained model
| written to disk has some binary files that should not go through this
| conversion. When they do, you get the error above. You can fix it by
2018-07-09 19:35:46 +03:00
| either changing your #[+a("https://git-scm.com/book/en/v2/Customizing-Git-Git-Configuration") #[code core.autocrlf]]
| setting to #[code "false"], or by commiting a #[+a("https://git-scm.com/docs/gitattributes") #[code .gitattributes] file]
| to your repository to tell git on which files or folders it shouldn't do
2018-07-09 19:35:46 +03:00
| LF-to-CRLF conversion, with an entry like #[code path/to/spacy/model/** -text].
| After you've done either of these, clone your repository again.