Community-developed components, examples, and tools for TFX
TFX-Addons is available on PyPI for all OS. To install the latest version, run:
pip install tfx-addons
You can then use TFX-Addons like this:
from tfx import v1 as tfx
import tfx_addons as tfxa
# Then you can easily load projects tfxa.{project_name}. For example:
tfxa.feast_examplegen.FeastExampleGen(...)
Developers helping developers. TFX-Addons is a collection of community projects to build new components, examples, libraries, and tools for TFX. The projects are organized under the auspices of the special interest group, SIG TFX-Addons.
Feast ExampleGen Component
An ExampleGen component for ingesting datasets from a Feast Feature Store.
Feature Selection Component
Perform feature selection using various algorithms with this TFX component.
Firebase Publisher Component
A TFX component to publish/update ML models to Firebase ML.
Hugging Face Pusher Component
Pushes a blessed model to the Hugging Face Model Hub. Optionally pushes the application to the Hugging Face Spaces Hub.
Message Exit Handler Component
Handle the completion or failure of a pipeline by notifying users, including any error messages.
MLMD Client Library
Client library to inspect content in ML Metadata populated by TFX pipelines.
Model Card Generator
The ModelCardGenerator takes dataset statistics, model evaluation, and a pushed model to automatically populate parts of a model card.
Pandas Transform Component
Use Pandas dataframes instead of the standard Transform component for your feature engineering. Processing is distributed using Apache Beam for scalability.
Sampling Component
A TFX component to sample data from examples, using probabilistic estimation.
Schema Curation Component
Apply user code to a schema produced by the SchemaGen component, and curate it based on domain knowledge.
XGBoost Evaluator Component
Evaluate XGBoost models by extending the standard Evaluator component.