Now we follow the official documents to build and test tensorflow serving with latest dev docker image. The examples can not be run successfully because the tensorflow version is 2.1.0 in tensorflow/serving:latest-devel while the examples code like mnist_saved_model.py still uses TensorFlow 1.x APIs.

We should upgrade the examples to TensorFlow 2 so that users can build and test with latest

Hello Vespa Team,
Can you please consider support a properties-file which is available during run-time along with the model ? So some meta-data e.g. threshold/label, etc can be associated with the model.
This is for the stateless evaluation of the models like (XGBoost, TensorFlow, Onnx, etc) which is supported in vespa.

Thank you,
Pinank

It would be good to update our current Kubeflow example to use 1.0, as well as add a page on the Seldon Docs on "Kubeflow Integration". If possible it would also be good to update it to leverage prepackaged model servers instead of the reusable containers as that is now the way that we're looking to integrate with model training frameworks. Example: https://docs.seldon.io/projects/seldon-core/en

Error

File "./gen_client/gen_client.py", line 76, in gen_tensorflow_client
    if dtype == int(tf.int8) or dtype == int(tf.uint8) or dtype == int(

TypeError: int() argument must be a string, a bytes-like object or a number, not 'DType'

change

dtype = input_opname_dtype_map[opname]
if dtype in [6, 4, 5, 17, 3, 22]:
    default_value = 1
elif dtype in [9, 2

Issue

Wrong description in CLI

$ hs model  --help
Usage: hs model [OPTIONS] COMMAND [ARGS]...
  Working with data profiles

$ hs --help
...
Commands:
  app      Working with data profiles <-----
  apply    Apply resources from files or directory.
  cluster  Utilities to manage hs clusters
  model    Working with data profiles <-----
  profile  Working with data profile