Performance

Performance

NLP Performance tests

Under the performance tab, different sections can be found to test NLP performance. It is possible to test the NLP performance by enabling multiple NLP engines. This is useful to test the performance of your chatbot in relation to actual input from your audience.

In addition, it is possible to upload utterances in batch to check the NLP performance on the expected utterances from your audience. Of these utterances, it is advised to roughly import 70% as the trained dataset and the remaining 30% afterwards to test your model.


    • Related Articles

    • v2.1.7

      Release date: 16-03-2021 The list below shows the features that were introduced, changed or removed with the release of v2.1.7. New Features Blits Automate This enabled automatic selection of the best NLU engine for your bot. The blits algoritme, ...

    • v2.1.5

      Release date: 25-12-2020 The list below shows the features that were introduced, changed or removed with the release of v2.1.5. New Features Performance increase on all channels 50+ bugs Various styling and bug fixes are part of this deploy

    • v2.1.8

      Release date: 02-04-2021 Performance improvements Horizontal autoscaling is even faster Shared queue between the multiple app instances introduced More efficient cron job process Faster app caching Rasa Entities are now performant Validation blocks ...

    • About bot analytics

      Intro Every bot comes with its own monthly analytics dashboard. This is used to track usage and the performance of your bot. It also gives suggestions to improve your bot. The standard bot metrics include: Amount of interactions (amount) Intents ...

    • v2.0.7

      The list below shows the features that were introduced, changed or removed with the release of v2.0.7. In this version, several bugs worth mentioning were resolved, while also UX improvements were made. Features You can test the NLP performance ...