To make it easier to create rasanlu.md
and domain.yml
file.
This package helps you convert a preformatted .csv
file to Rasa acceptable file formats.
The standard format to be followed while creating a .csv
is shown in nlu_sample_format_for_conversion.csv
file. This CSV file is converted to rasa format nlu.md
and domain.yml
file.
Run command download_csv()
to download sample file that you can fill.
from easy-rasa import download_csv
download_csv()
This will download a preloaded file called nlu_sample_format_for_conversion.csv
. Search the directory where you ran the download_csv()
command to find the file.
Fill the .csv
file by follwing these instructions:
- First rows consist of intent names. Add any number of intents in the first row.
- The succeeding rows. i.e. after the intent names row, add all your training phrases/sentences.
- save the file
- Call the function/method
create_rasa_nlu(csv_path, create_files_path)
. Don't forget to specify path to the CSV file using the path parameter, and path where the output files need to be created via thecreate_files_path
parameter.
from easy-rasa import create_rasa_files
create_rasa_files(csv_path, create_files_path)
#csv_path and create_files_path can both be ./
Check the directory (create_files_path) to file respective nlu.md
and domain.yml
files.