Updated the story converter code to take the generic utterances and intents CSV and remove punctuation chars from intent names.
Details
The script now requires a separate CSVs for generic utterances and generic intents. The code was modified to parse the two individually and populate the Rasa data Markdown files. Issues also arose when training the Rasa model due to punctuation characters in the Markdown files. Punctuation marks are now removed and story training is now functioning.
Testing
The script was run on the current datasets in the Google Drive. The data was successfully generated and used to train a Rasa Core model. This model was successfully used on the end-to-end Raspberry Pi system.
Summary
Updated the story converter code to take the generic utterances and intents CSV and remove punctuation chars from intent names.
Details
The script now requires a separate CSVs for generic utterances and generic intents. The code was modified to parse the two individually and populate the Rasa data Markdown files. Issues also arose when training the Rasa model due to punctuation characters in the Markdown files. Punctuation marks are now removed and story training is now functioning.
Testing
The script was run on the current datasets in the Google Drive. The data was successfully generated and used to train a Rasa Core model. This model was successfully used on the end-to-end Raspberry Pi system.