A simple Intent classification with Rasa NLU using tensorflow embedding and Of Course, Python
The given dataset(NLP.csv) contains data for 5 different intents.
The implementation goes through some pre-processing, training and testing sample data.
To Run:
Install the required packages with requirements.txt
To verify the installation of rasa_nlu, run these,
pip install rasa_nlu
pip install rasa_nlu[tensorflow]
Use the Jupyter notebook file
(or)
-
Keep all the files in same folder
-
Run the Python file(process.py) to pre-process and generate intermediate files
-
Train a Model using the command
python -m rasa_nlu.train -c nlu_config.yml --data intentClassifiedFile.json -o models --fixed_model_name intentClassifier --project current --verbose
- Run test.py to test sample data
The code is tested and results are stored in TestRun folder for Reference
If you are more curious, Check out here
Cheers!