Skip to content

Commit

Permalink
Adding a sample for serving component (kubeflow#1830)
Browse files Browse the repository at this point in the history 
* Adding a sample for serving component

* removed typo / updated based on PR feedback

* fixing the jupyter rendering issue

* adding pip3 for tensorflow

* Fixed spelling error in VERSION

* fix indentation based on review feedback
  • Loading branch information
@SinaChavoshi @k8s-ci-robot
SinaChavoshi authored and k8s-ci-robot committed Aug 16, 2019
1 parent 54ff3e6 commit a6f3f40
Showing 1 changed file with 297 additions and 0 deletions.
297 changes: 297 additions & 0 deletions samples/core/model_serving_component/model_serving_component.ipynb
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,297 @@
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"# Copyright 2019 Google Inc. All Rights Reserved.\n",
"#\n",
"# Licensed under the Apache License, Version 2.0 (the \"License\");\n",
"# you may not use this file except in compliance with the License.\n",
"# You may obtain a copy of the License at\n",
"#\n",
"# http://www.apache.org/licenses/LICENSE-2.0\n",
"#\n",
"# Unless required by applicable law or agreed to in writing, software\n",
"# distributed under the License is distributed on an \"AS IS\" BASIS,\n",
"# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n",
"# See the License for the specific language governing permissions and\n",
"# limitations under the License."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Install Pipeline SDK - This only needs to be ran once in the enviroment. \n",
"!pip3 install kfp --upgrade\n",
"!pip3 install tensorflow==1.14"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## KubeFlow Pipelines Serving Component\n",
"In this notebook, we will demo:\n",
"\n",
"* Saving a Keras model in a format compatible with TF Serving\n",
"* Creating a pipeline to serve a trained model within a KubeFlow cluster\n",
"\n",
"Reference documentation:\n",
"* https://www.tensorflow.org/tfx/serving/architecture\n",
"* https://www.tensorflow.org/beta/guide/keras/saving_and_serializing\n",
"* https://www.kubeflow.org/docs/components/serving/tfserving_new/"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Setup\n"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"tags": [
"parameters"
]
},
"outputs": [],
"source": [
"# Set your output and project. !!!Must Do before you can proceed!!!\n",
"PROJECT_NAME = 'Your-Gcp-Project-ID' #'Your-GCP-Project-ID'\n",
"MODEL_NAME = 'Model-Name' # Model name matching TF_serve naming requirements \n",
"EXPERIMENT_NAME = 'serving_component'\n",
"MODEL_VERSION = '1' # A number representing the version model \n",
"OUTPUT_BUCKET = 'gs://%s-serving-component' % PROJECT_NAME # A GCS bucket for asset outputs\n",
"KUBEFLOW_DEPLOYER_IMAGE = 'gcr.io/ml-pipeline/ml-pipeline-kubeflow-deployer:fe639f41661d8e17fcda64ff8242127620b80ba0'\n",
"MODEL_PATH = '%s/%s' % (OUTPUT_BUCKET,MODEL_NAME) \n",
"MODEL_VERSION_PATH = '%s/%s/%s' % (OUTPUT_BUCKET,MODEL_NAME,MODEL_VERSION)"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Creating gs://chavoshi-dev-2-serving-component/...\n"]
}
],
"source": [
"!gsutil mb {OUTPUT_BUCKET}"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [],
"source": [
"#Get or create an experiment and submit a pipeline run\n",
"import kfp\n",
"client = kfp.Client()\n",
"\n",
"try:\n",
" experiment = client.get_experiment(experiment_name=EXPERIMENT_NAME)\n",
"except:\n",
" experiment = client.create_experiment(EXPERIMENT_NAME)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Load a Keras Model \n",
"Loading a pretrained Keras model to use as an example. "
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [],
"source": [
"import tensorflow as tf"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"WARNING: Logging before flag parsing goes to stderr.\n",
"W0813 22:25:44.771417 139829630277440 deprecation.py:506] From /opt/conda/lib/python3.6/site-packages/tensorflow/python/ops/init_ops.py:1251: calling VarianceScaling.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version.\n",
"Instructions for updating:\n",
"Call initializer instance with the dtype argument instead of passing it to the constructor\n"
]
}
],
"source": [
"model = tf.keras.applications.NASNetMobile(input_shape=None,\n",
" include_top=True,\n",
" weights='imagenet',\n",
" input_tensor=None,\n",
" pooling=None,\n",
" classes=1000)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Saved the Model for TF-Serve\n",
"Save the model using keras export_saved_model function. Note that specifically for TF-Serve the output directory should be structure as model_name/model_version/saved_model."
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"W0813 22:26:30.974738 139829630277440 deprecation.py:506] From /opt/conda/lib/python3.6/site-packages/tensorflow/python/ops/init_ops.py:97: calling Zeros.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version.\n",
"Instructions for updating:\n",
"Call initializer instance with the dtype argument instead of passing it to the constructor\n",
"W0813 22:26:30.992686 139829630277440 deprecation.py:506] From /opt/conda/lib/python3.6/site-packages/tensorflow/python/ops/init_ops.py:97: calling Ones.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version.\n",
"Instructions for updating:\n",
"Call initializer instance with the dtype argument instead of passing it to the constructor\n",
"W0813 22:26:31.067510 139829630277440 deprecation.py:506] From /opt/conda/lib/python3.6/site-packages/tensorflow/python/ops/init_ops.py:97: calling GlorotUniform.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version.\n",
"Instructions for updating:\n",
"Call initializer instance with the dtype argument instead of passing it to the constructor\n",
"W0813 22:27:34.353115 139829630277440 deprecation.py:323] From /opt/conda/lib/python3.6/site-packages/tensorflow/python/saved_model/signature_def_utils_impl.py:201: build_tensor_info (from tensorflow.python.saved_model.utils_impl) is deprecated and will be removed in a future version.\n",
"Instructions for updating:\n",
"This function will only be available through the v1 compatibility library as tf.compat.v1.saved_model.utils.build_tensor_info or tf.compat.v1.saved_model.build_tensor_info.\n"
]
}
],
"source": [
"tf.keras.experimental.export_saved_model(model,MODEL_VERSION_PATH)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create a pipeline using KFP TF-Serve component\n"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [],
"source": [
"def kubeflow_deploy_op():\n",
" return dsl.ContainerOp(\n",
" name = 'deploy',\n",
" image = KUBEFLOW_DEPLOYER_IMAGE,\n",
" arguments = [\n",
" '--model-export-path', MODEL_PATH,\n",
" '--server-name', MODEL_NAME,\n",
" ]\n",
" )"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [],
"source": [
"import kfp\n",
"import kfp.dsl as dsl\n",
"from kfp.gcp import use_gcp_secret\n",
"\n",
"# The pipeline definition\n",
"@dsl.pipeline(\n",
" name='Sample model deployer',\n",
" description='Sample for deploying models using KFP model serving component'\n",
")\n",
"def model_server():\n",
" deploy = kubeflow_deploy_op().apply(use_gcp_secret('user-gcp-sa'))"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [],
"source": [
"pipeline_func = model_server\n",
"pipeline_filename = pipeline_func.__name__ + '.pipeline.zip'\n",
"\n",
"import kfp.compiler as compiler\n",
"compiler.Compiler().compile(pipeline_func, pipeline_filename)"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"Run link <a href=\"/pipeline/#/runs/details/8fb3c380-be19-11e9-b782-42010a8000a3\" target=\"_blank\" >here</a>"
],
"text/plain": [
"<IPython.core.display.HTML object>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"#Specify pipeline argument values\n",
"arguments = {}\n",
"\n",
"#Submit a pipeline run\n",
"run_name = pipeline_func.__name__ + ' run'\n",
"run_result = client.run_pipeline(experiment.id, run_name, pipeline_filename, arguments)\n",
"\n",
"#This link leads to the run information page. \n",
"#Note: There is a bug in JupyterLab that modifies the URL and makes the link stop working"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.7"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

0 comments on commit a6f3f40

Please sign in to comment.