RASA multilingual FAQ bot creator creates all the necessary files to train a RASA chatbot from a Google Docs spreadsheet.
To access the source spreadsheet which this repository uses, click here.
In this particular tutorial, we are working with two languages, English and Hindi in three main topics related to climate resilience with 25 FAQs in total.
pip install -r requirements.txt
If you want to connect it to your own spreadsheet, make sure it's publicly accessible and you specify it's link, the languages involved and the column names in these lines of sheet_to_rasa_v3.py.
data = {
'lang': ['eng', 'hindi'],
'bot_strings_sheet': 'https://docs.google.com/spreadsheets/d/1OpTyjwXZjItPugQpprQ86oaV5xU1fUS0A_QELrQMg5U/edit#gid=1133259128'
}
...
#column_ids
COLUMN_DICT = {
'INTENT' : 'Intent',
'EXAMPLE_HINDI' : 'Sanitized Question',
'EXAMPLE_ENG' : 'Sanitized Question Translation in English',
'RESPONSE_HINDI' : 'Answer Transcription',
'RESPONSE_ENG' : 'Answer Translation in English '
}
python3 sheet_to_rasa_v3.py
This should create following files:
.
├── data_core
│ └── stories.yml
├── data_nlu
│ ├── classifier
│ │ └── nlu_data_classifier.yml
│ ├── eng
│ │ └── nlu_data_eng.yml
│ └── hindi
│ └── nlu_data_hindi.yml
├── domains
│ ├── domain_v3_for_classifier.yml
│ └── domain_v3.yml
├── models
│ ├── core_model
│ └── spacy
│ └── hi
├── README.md
├── requirements.txt
├── responses
│ ├── responses_eng.yml
│ └── responses_hindi.yml
└── sheet_to_rasa_v3.py
11 directories, 11 files