Version 0.2

Model Card Generator

This tool can be used to create a Model Card: a way to share information about a model’s intent, data, architecture, and performance. Use the form below to generate a markdown (*.md) file version of a model card.

The data you enter here is not saved or sent to a server. This is an entirely client-side feature.

Note: A red asterisk (*) indicates a required field.

Accountability

Model name*

Point of contact(s)*

Agencies/Affiliations*

Model acquisition/development method*

Anticipated Use

Division(s) using the model*

Intended application(s)*

Intended stakeholder(s)*

Model Information & Architecture

Current model version*

Version release date*

Changes made since last release (if any)*

License for use*

Type of Model (Classification, Regression, Object Detection, etc.)*

Type of algorithm(s) used*

Dataset & Performance

Source(s) of the training data*

Data collection/generation method*

Number of variables in the dataset*

Number of entries in the dataset*

Percent of data chosen as the training, testing, and validation sets*

Metrics used to rate model performance*

Factors that limit the model's performance (examples: limited dataset, number of nulls/NAs) (if any)?*

Bias Identification & Mitigation

Inclusion of information related to individuals or human populations in the training/testing/validation datasets - Yes/No,
if "yes" include a description of demographic features

Methods used to minimize bias from human judgement

Potential biases found in the training dataset from collection methods, sample size, representation, etc.

Testing/evaluation performed to look for bias in the workflow of the model

Degree of model explainability/transparency

Governance/Compliance

Model/dataset compliance with existing laws and regulations (including privacy protection regulations)