0.6.0

🧰 Maintenance

Update Tensorflow version to 2.8 @elaubsch (#46)

This PR updates the base image for the Dockerfile of this repo to Tensorflow 2.8.

0.5.0

🧰 Maintenance

Drop support for Python 2.7 and 3.5. @willgraf (#45)

Python 2.7 and Python 3.5 have reached EOL.

Update year of all copyrights. @willgraf (#44)

Bump serving version to 2.5.3-gpu. @willgraf (#43)

Add new release-drafter GitHub Workflow. @willgraf (#42)

0.4.0

0.4.0

Features

• Replace TravisCI with GitHub Actions (#38)

Breaking Changes

• Upgrade TensorFlow Serving base image to 2.3.0-gpu (#40)
• Deprecate CLOUD_PROVIDER and require STORAGE_BUCKET to contain the bucket protocol (#41)

0.3.0

0.3.0

Features

• Split the original Docker container into 2, (a configuration file writer container and a model server container) to reduce overall image size. (#35)

Bugfixes

• Updated the server container to only include configuration files if the option is turned on, preventing any conflicts. (#35)

Breaking Changes

• Splitting the containers is not backward-compatible, and any downstream uses will be affected. To run the model server, the configuration files must be pre-created and mounted.

0.2.1 Upgrade tf-serving base to 1.15.0

1.15.0 is the final 1.x release for tensorflow-serving. This version supports all of the default models for deepcell.org.

0.2.0 Upgrade tf-serving base to 1.14.0

upgrade to tensorflow-serving 1.14.0 (#28)

TFS with model and batching configs

Add MonitoringConfigWriter and BatchingConfigWriter (#25)

* create ConfigWriter base class, refactor TFServingConfigWriter to ModelConfigWriter

* add BatchConfigWriter and MonitoringConfigWriter to the writers module

* pass environment variables to argparser, call a function for each config file type