"sklearn/base.py:315: UserWarning" in Docker-Log 0.9.14

[andbez]

Following Warning occurs for every single consumed document:

/usr/local/lib/python3.7/site-packages/sklearn/base.py:315: UserWarning: Trying to unpickle estimator CountVectorizer from version 0.23.2 when using version 0.24.0. This might lead to breaking code or invalid results. Use at your own risk. UserWarning).

Don't know of any problems with paperless. Just to let you know.

[jonaswinkler]

This will automatically go away once paperless updates the classification model, which will happen every hour, but only if your documents have changed, or if the classification_model.pickle file has been deleted.

[andbez]

👍 Thank you.

[rknightion]

This is happening every time I run "document_create_classifier". I thought that that process updated the classification model?

❯ docker-compose run --rm webserver document_create_classifier
Creating paperless_webserver_run ... done
Paperless-ng docker container starting...
Waiting for PostgreSQL to start...
Apply database migrations...
Operations to perform:
  Apply all migrations: admin, auth, authtoken, contenttypes, django_q, documents, paperless_mail, sessions
Running migrations:
  No migrations to apply.
Executing management command document_create_classifier
/usr/local/lib/python3.7/site-packages/sklearn/base.py:315: UserWarning: Trying to unpickle estimator CountVectorizer from version 0.23.2 when using version 0.24.0. This might lead to breaking code or invalid results. Use at your own risk.
  UserWarning)
/usr/local/lib/python3.7/site-packages/sklearn/base.py:315: UserWarning: Trying to unpickle estimator LabelBinarizer from version 0.23.2 when using version 0.24.0. This might lead to breaking code or invalid results. Use at your own risk.
  UserWarning)
/usr/local/lib/python3.7/site-packages/sklearn/base.py:315: UserWarning: Trying to unpickle estimator MLPClassifier from version 0.23.2 when using version 0.24.0. This might lead to breaking code or invalid results. Use at your own risk.
  UserWarning)
❯ docker-compose run --rm webserver document_create_classifier
Creating paperless_webserver_run ... done
Paperless-ng docker container starting...
Waiting for PostgreSQL to start...
Apply database migrations...
Operations to perform:
  Apply all migrations: admin, auth, authtoken, contenttypes, django_q, documents, paperless_mail, sessions
Running migrations:
  No migrations to apply.
Executing management command document_create_classifier
/usr/local/lib/python3.7/site-packages/sklearn/base.py:315: UserWarning: Trying to unpickle estimator CountVectorizer from version 0.23.2 when using version 0.24.0. This might lead to breaking code or invalid results. Use at your own risk.
  UserWarning)
/usr/local/lib/python3.7/site-packages/sklearn/base.py:315: UserWarning: Trying to unpickle estimator LabelBinarizer from version 0.23.2 when using version 0.24.0. This might lead to breaking code or invalid results. Use at your own risk.
  UserWarning)
/usr/local/lib/python3.7/site-packages/sklearn/base.py:315: UserWarning: Trying to unpickle estimator MLPClassifier from version 0.23.2 when using version 0.24.0. This might lead to breaking code or invalid results. Use at your own risk.
  UserWarning)

[jonaswinkler]

It's only updated if there's a change in the data associated with Auto-matching metadata (content changed, documents with Auto-matching metadata added / removed).

There's no abnormal behavior as far as my tests go, so this is fine.

[jonaswinkler]

Added some notes to the trouble shooting section.