I am using evidently to detect data drift and for generating classification report comparison and i use streamlit.write method to show report generated by evidently but in dashboard it is showing 'Loading...' . Is there any way to solve this issue?

🎈 Using Streamlit
1

Summary

I want to monitor model and i am using evidently for this that generate data drift report ,classification report etc. it works fine alone but when i try to configure it with streamlit instead of showing report it is just showing ‘Loading…’ .

Steps to reproduce

# ref_data_sample = reference dataset
#prod_data_sample= production dataset(we are taking it locally)
rating_target_drift_dashboard.calculate(ref_data_sample, prod_data_sample, column_mapping=target_column_mapping)
st.write( rating_target_drift_dashboard.show())
2

Hi @sunil_modi,

Here are a few thoughts

  • You seem to use an old Evidently API (with Dashboards) instead of Reports. It is best to use Reports now (unless you fix Evidently to an older version). Here are sample notebooks to show the code: Examples - Evidently Documentation
  • The “loading” issue might refer to the environment you work in. By default, Evidently should work in Jupyter notebook (Jupyter nbextension required) on Mac OS/Linux and in Colab. If you are working in other notebook environments or on Windows, you might need to add an argument to display the report or test suite inline when calling it.
    report.show(mode='inline')
    Here is the relevant docs page: Notebook environments - Evidently Documentation
  • Here is a ready-made tutorial on how to build and host ML monitoring dashboards with Evidently + Streamlit, you might find it useful A tutorial on building ML and data monitoring dashboards with Evidently and Streamlit
1 Like
3

I have successfully implemented it. Appreciate your suggestion!

4

Hi @Mafizul.Islam . Feel free to check out my application which was the combination of python, streamlit, Pycaret and evidently ai.
Advance Model Monitoring