This is a ClowdFlows 3.0 package which contains widgets for word, sentence and document embeddings. See Widgets to learn more about the included embeddings models.
Install this package by running: pip install cf_text_embeddings
.
To enable it in the ClowdFlows Backend, edit the local_settings.py
file like shown below:
PACKAGE_TREE = [
{
"name": "Text Embeddings",
"packages": ['cf_text_embeddings'],
"order": 1
}
]
Then run ./manage.py import_package cf_text_embeddings
from the ClowdFlows Backend directory to import widgets into the platform.
After package installation, you need to download the models.
The Text Embeddings package supports models for multiple languages. To list available languages:
python -m cf_text_embeddings.downloader -l
To download all the models:
python -m cf_text_embeddings.downloader -d all
You can also download the models for a specific language. For example, to download English models:
python -m cf_text_embeddings.downloader -d en
.