Skip to main content

Explains DIETClassifier model predictions in Rasa chatbot framework.

Project description

DIME (Dual Interpretable Model-agnostic Explanations) is mainly aimed at Explaining DIET Classifiers in RASA 2.8.X. Models.

Features 🦄

  • Explains RASA DIET Classifiers using feature importance
  • Generates dual feature importance scores (Global FI + Local FI)
  • Efficient
  • Total confidence drop as the feature importance score
  • Able to explain both local and REST Rasa models
  • Easy to use DIME CLI
  • GUI with a dedicated server on-demand
  • Generate, Store, Download, Upload, Peak DIME explanations. Read more on docs
  • Supports Sinhalese unicode / fully Sinhala-English code-switchable

What's Cooking? 🍪

  • DIME for Jupyter Notebooks
  • Stopwords List Generation
  • DIME Example Notebooks
  • DIME for non-DIET text classification models

Limitations and Known Issues 🤏🏽

  • Global Importance is disabled for REST models due to performance bottlenecks
  • Explaining RASA models locally on Notebooks such as CoLab is not supported yet due to dependency issues
  • Benchmark tests are in progress

📒 Docs:
📦 PyPi:
🪵 Full Changelog:

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

dime_xai-1.2.1.tar.gz (20.3 MB view hashes)

Uploaded source

Built Distribution

dime_xai-1.2.1-py3-none-any.whl (20.3 MB view hashes)

Uploaded py3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page