Lens: comprehensive assessment framework for AI systems
Project description
Lens by Credo AI - Responsible AI Assessment Framework
Lens is a comprehensive assessment framework for AI systems. Lens standardizes model and data assessment, and acts as a central gateway to assessments created in the open source community. In short, Lens connects arbitrary AI models and datasets with Responsible AI tools throughout the ecosystem.
Lens can be run in a notebook, a CI/CD pipeline, or anywhere else you do your ML analytics. It is extensible, and easily customized to your organizations assessments if they are not supported by default.
Though it can be used alone, Lens shows its full value when connected to your organization's Credo AI Platform. Credo AI is an end-to-end AI Governance platform that supports multi-stakeholder alignment, AI assessment (via Lens) and AI risk assesssment.
Dependencies
- Credo AI Lens supports Python 3.7+
- Sphinx (optional for local docs site)
Installation
The latest stable release (and required dependencies) can be installed from PyPI. Note this installation only includes dependencies needed for a small set of modules
pip install credoai-lens
To include additional dependencies needed for some modules and demos, use the following installation command.
pip install credoai-lens[extras]
Modules that require extras include:
- fairness_nlp
Failed pip installation - use conda
Installation sometimes fails on M1 macbooks. Specifically, pandas, scipy, and scikit-learn may fail to build. Installing these packages with the anaconda package manager seems to be the easiest way to circumvent this issue.
For development, the easiest way to interact with Lens is to use anaconda.
conda env create --file environment.yml
Getting Started
To get started, see the quickstart demo: docs/notebooks/quickstart.ipynb
.
For a more detailed example, see docs/notebooks/lens_demos/binaryclassification.ipynb
Documentation
Documentation is hosted by readthedocs.
Configuration
To connect to Credo AI's Governance Platform, enter your connection info in ~/.credoconfig
(in the root directory) using
the below format.
TENANT={tenant name} # Example: credoai
CREDO_URL=<your credo url> # Example: https://api.credo.ai
API_KEY=<your api key> # Example: JSMmd26...
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for credoai_lens-0.0.2-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 7e46640a234931a3ea0db7e4a95dc7b7d4cb6022c00f3b5a41b9beb4ef10bf57 |
|
MD5 | 46a728c8d666c777411a352aeb0978b4 |
|
BLAKE2b-256 | 5629fb09543e81667037e83d63ed8ee7a30edd3470b2f61801274e8f91935948 |