tfserver 0.4.11

Tensor Flow Model Server

Introduce

tfserver is an example for serving gRPC for Tensorflow/Pytorch thing’s models.

It can serve not only through gRPC but also RESTful API with Skitai App Engine and Atila WSGI container.

This project is inspired by issue #176.

I’m so sorry about this soulless manual.

Installation

• from version 0.3, it is now TensorFlow 2+ compatible (deprecated, no longer support)

• from version 0.4, TensorFlow 2+ only

Building Model and Deploy

Please see https://gitlab.com/hansroh/skitai/-/blob/master/tests/level4-2/build_model.py

It is mostly used tensorflow keras and dnn.

Creating Your Own gRPC Server

Please see https://gitlab.com/hansroh/skitai/-/blob/master/tests/examples/tfserve.py

You can know how to serve gRPC service and make yoyr own APIs.

APIs

Please see https://gitlab.com/hansroh/tfserver/-/blob/master/tfserver/export/skitai/__export__.py

APIs to manage models and basic inference.

And for usage see, https://gitlab.com/hansroh/skitai/-/blob/master/tests/level4-2/test_tfserver.py

Release History

• 0.4 (2021. 4)

  • upgrade for tensorflow 2

• 0.3 (2020. 6. 28)

  • add model management APIs

  • reactivate project and compatible with TF2+

• 0.2 (2020. 6. 26): integrated with dnn 0.3

• 0.1b8 (2018. 4. 13): fix grpc trailers, skitai upgrade is required

• 0.1b6 (2018. 3. 19): found works only grpcio 1.4.0

• 0.1b3 (2018. 2. 4): add @app.umounted decorator for clearing resource

• 0.1b2: remove self.tfsess.run (tf.global_variables_initializer())

• 0.1b1 (2018. 1. 28): Beta release

• 0.1a (2018. 1. 4): Alpha release