evidently - An open-source framework to evaluate, test and monitor ML and LLM-powered systems.

[ShellLM]

• evidently/README.md at main · evidentlyai/evidently

evidently/README.md at main · evidentlyai/evidently

Evidently

An open-source framework to evaluate, test and monitor ML and LLM-powered systems.

Documentation | Discord Community | Blog | Twitter | Evidently Cloud

🆕 New release

Evidently 0.4.25. LLM evaluation -> Tutorial

📊 What is Evidently?

Evidently is an open-source Python library for ML and LLM evaluation and observability. It helps evaluate, test, and monitor AI-powered systems and data pipelines from experimentation to production.

• 🔡 Works with tabular, text data, and embeddings.
• ✨ Supports predictive and generative systems, from classification to RAG.
• 📚 100+ built-in metrics from data drift detection to LLM judges.
• 🛠️ Python interface for custom metrics and tests.
• 🚦 Both offline evals and live monitoring.
• 💻 Open architecture: easily export data and integrate with existing tools.

Evidently is very modular. You can start with one-off evaluations using Reports or Test Suites in Python or get a real-time monitoring Dashboard service.

1. Reports

Reports compute various data, ML and LLM quality metrics. You can start with Presets or customize.

• Out-of-the-box interactive visuals.
• Best for exploratory analysis and debugging.
• Get results in Python, export as JSON, Python dictionary, HTML, DataFrame, or view in monitoring UI.

Reports

2. Test Suites

Test Suites check for defined conditions on metric values and return a pass or fail result.

• Best for regression testing, CI/CD checks, or data validation pipelines.
• Zero setup option: auto-generate test conditions from the reference dataset.
• Simple syntax to set custom test conditions as gt (greater than), lt (less than), etc.
• Get results in Python, export as JSON, Python dictionary, HTML, DataFrame, or view in monitoring UI.

Test Suite

3. Monitoring Dashboard

Monitoring UI service helps visualize metrics and test results over time.

You can choose:

• Self-host the open-source version. Live demo.
• Sign up for Evidently Cloud (Recommended).

Evidently Cloud offers a generous free tier and extra features like user management, alerting, and no-code evals.

Dashboard

👩‍💻 Install Evidently

Evidently is available as a PyPI package. To install it using pip package manager, run:

pip install evidently

To install Evidently using conda installer, run:

conda install -c conda-forge evidently

▶️ Getting started

Option 1: Test Suites

This is a simple Hello World. Check the Tutorials for more: Tabular data or LLM evaluation.

Import the Test Suite, evaluation Preset and toy tabular dataset.

import pandas as pd

from sklearn import datasets

from evidently.test_suite import TestSuite
from evidently.test_preset import DataStabilityTestPreset

iris_data = datasets.load_iris(as_frame='auto')
iris_frame = iris_data.frame

Split the DataFrame into reference and current. Run the Data Stability Test Suite that will automatically generate checks on column value ranges, missing values, etc. from the reference. Get the output in Jupyter notebook:

data_stability= TestSuite(tests=[
    DataStabilityTestPreset(),
])
data_stability.run(current_data=iris_frame.iloc[:60], reference_data=iris_frame.iloc[60:], column_mapping=None)
data_stability

You can also save an HTML file. You'll need to open it from the destination folder.

data_stability.save_html("file.html")

To get the output as JSON:

data_stability.json()

You can choose other Presets, individual Tests and set conditions.

Option 2: Reports

Import the Report, evaluation Preset and toy tabular dataset.

import pandas as pd

from sklearn import datasets

from evidently.report import Report
from evidently.metric_preset import DataDriftPreset

iris_data = datasets.load_iris(as_frame='auto')
iris_frame = iris_data.frame

Run the Data Drift Report that will compare column distributions between current and reference:

data_drift_report = Report(metrics=[
    DataDriftPreset(),
])

data_drift_report.run(current_data=iris_frame.iloc[:60], reference_data=iris_frame.iloc[60:], column_mapping=None)
data_drift_report

Save the report as HTML. You'll later need to open it from the destination folder.

data_drift_report.save_html("file.html")

To get the output as JSON:

data_drift_report.json()

You can choose other Presets and individual Metrics, including LLM evaluations for text data.

Option 3: ML monitoring dashboard

This launches a demo project in the Evidently UI. Check complete tutorials for Self-hosting or Evidently Cloud.

Recommended step: create a virtual environment and activate it.

pip install virtualenv
virtualenv venv
source venv/bin/activate

After installing Evidently (pip install evidently), run the Evidently UI with the demo projects:

evidently ui --demo-projects all

Access Evidently UI service in your browser. Go to the localhost:8000.

💻 Contributions

We welcome contributions! Read the Guide to learn more.

📚 Documentation

For more information, refer to a complete Documentation. You can start with the tutorials:

• Get Started with Tabular and ML Evaluation
• Get Started with LLM Evaluation
• Self-hosting ML monitoring Dashboard
• Cloud ML monitoring Dashboard

See more examples in the Docs.

How-to guides

Explore the How-to guides to understand specific features in Evidently.

✅ Discord Community

If you want to chat and connect, join our Discord community.

Suggested labels

None

[ShellLM]

Related content

#762 similarity score: 0.9
#830 similarity score: 0.89
#309 similarity score: 0.89
#649 similarity score: 0.89
#499 similarity score: 0.89
#808 similarity score: 0.88