This example demonstrates how to use Flair NER model with Label Studio.
- Build and start the Machine Learning backend on
http://localhost:9090
docker-compose up
- Validate that the backend is running
$ curl http://localhost:9090/health
{"status":"UP"}
- Create a project in Label Studio. Then from the Model page in the project settings, connect the model. The default URL is
http://localhost:9090
.
<View>
<Labels name="label" toName="text">
<Label value="PER" background="red"/>
<Label value="ORG" background="darkorange"/>
<Label value="LOC" background="orange"/>
<Label value="MISC" background="green"/>
</Labels>
<Text name="text" value="$text"/>
</View>
FLAIR_MODEL_NAME
: The name of the Flair model to use. Default isner
. See all options here