* update watson components with output path args to support tekton
* fix store bug and stop batch logs
* update pipeline with explicit helper function
* add missing commit
* testing kevin-wml-iks-pipeline
* switch input/output order
* build v4
* fix the yaml link
* change to v4
* using new v4 api
* update train docker verion to v0.3
* try v4-api
* try v4 author
* test v4 code
* fix compile error
* name upcase
* comment out author name
* fix lowcase runtime
* replace framework with platform
* fix syntax platform
* add version
* add filepath
* add file path
* using wml train code as filepath
* change credentials to s4
* try different version
* try store_library
* wml-v4 code
* update image to v0.4
* change object storage type to s3
* try store with meta no
* add input for store
* missing comma at store yaml
* add missing parameter in kfp
* add argument in store
* add uid
* add meta type
* change deploy
* deployment input
* list deployment
* delete deployment
* adjust input
* play around the input data:
* play around input data
* change input
* deploy works
* fix min
* clean the code
* adjust serve pipeline code
* clean the doc session
* remove hard code
* update yaml
* commit stuff
* test1
* change order
* correct name
* change version
* test
* change1
* change to 0,8
* print runtime version
* change to 0.9
* change2
* change yaml
* change version
* remove command
* change version
* change version
* clean up
* address comments
* fix typo
* SDK - Client - Added a way to set experiment name using environment variables
This is useful for launching notebooks or pipeline files that submit
themselves for execution.
* Switched to subprocess.run which supports env
* Setting the environment variable differently
Looks like `subprocess.run` uses `PATH` to search for the program.
* Convert return code to string
* Changed the way the experiment name is being set
* Changed how the notebook installs the SDK
Notebook is overriding the SDK that's being tested.
* Not installing the KFP SDK package
* Removed the experiment_name from samples and configs.
* Changed the SDK installation lines in samples
Otherwise the sample tests do not correctly test the new SDK code.
Reduced boilerplate code in samples by switching to `kfp.Client().create_run_from_pipeline_package` which:
* compiles the pipeline
* gets or creates the experiment (name is optional)
* generates unique run name
* submits the run
Discovered in #1721
__xrange()__ was removed in Python 3 in favor of an improved version of __range()__. This PR ensures equivalent functionality in both Python 2 and Python 3.
```
./samples/contrib/ibm-samples/watson/source/model-source-code/tf-model/input_data.py💯40: F821 undefined name 'xrange'
fake_image = [1.0 for _ in xrange(784)]
^
./samples/contrib/ibm-samples/watson/source/model-source-code/tf-model/input_data.py:102:41: F821 undefined name 'xrange'
return [fake_image for _ in xrange(batch_size)], [
^
./samples/contrib/ibm-samples/watson/source/model-source-code/tf-model/input_data.py:103:37: F821 undefined name 'xrange'
fake_label for _ in xrange(batch_size)]
^
```
@gaoning777 @Ark-kun Your reviews please.
Tommy Li
Kevin Yu
Alexey Volkov
Christian Clauss
Ning