OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Evaluation & Observability

Model Analysis

Model Analysis is an open-source Python library, TensorFlow Model Analysis (TFMA), for evaluating TensorFlow models on large datasets in a distributed manner. It computes trainer-defined metrics across data slices and supports visualizing those metrics in Jupyter notebooks.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Model Analysis — OpenViST project profile in AI Evaluation & Observability.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use TFMA when a TensorFlow evaluation workflow needs distributed evaluation over large datasets and metrics broken down by data slice.

Who it is for

  • Teams evaluating TensorFlow models against large datasets.
  • Practitioners who need to compare model metrics across data slices.
  • TensorFlow developers building distributed model-evaluation workflows.

Documented uses

  • Evaluate a TensorFlow model over a large dataset using metrics defined in its trainer.
  • Compute metrics for different slices of evaluation data.
  • Inspect slice-level metrics through Jupyter notebook visualizations.
Before you choose

Requirements & limits

The project is open source under the Apache-2.0 license. The repository documents installation from PyPI and source builds.

  • TensorFlow must be installed separately; TFMA does not declare an explicit dependency on the TensorFlow PyPI package.
  • The repository publishes compatible-version pairings for TFMA and related dependencies, so package versions should be selected together.
  • Building from source requires correctly installed Protocol Buffers tooling, including protoc.
  • The README cautions that backward-incompatible changes may occur before version 1.0.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Build a TFMA wheel from source

Create and install a wheel from a local clone of the TFMA repository.

Before you start

  • Python 3
  • protoc and its required includes
  • setuptools and wheel

Steps

  1. Create and activate a Python virtual environment.

  2. Install setuptools and wheel.

  3. Clone the repository and change into its directory.

  4. Run the command below to build a wheel in the dist directory.

    Documented command
    python3 setup.py bdist_wheel

  5. Change into dist and install the generated wheel with pip.

Expected outcome

A TFMA wheel is built in `dist` and can be installed from that directory.

Guide limitations

  • The build requires Protocol Buffers to be installed correctly, including protoc and its referenced includes.
  • TensorFlow must be installed for TFMA, though the package does not declare an explicit TensorFlow PyPI dependency.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
12 October 2026
Original website title
tensorflow/model-analysis · Official homepage title · retained in its source language
How we review sourcesSuggest a correction