jprdonnelly/MachineLearningNotebooks
1
0
Fork 0
mirror of https://github.com/Azure/MachineLearningNotebooks.git synced 2025-12-19 17:17:04 -05:00
Code Issues Projects Releases Wiki Activity

Revert "Updated notebook folders"

Browse Source 
This reverts commit 06728004b6.
This commit is contained in:
rastala
2018-11-20 10:39:48 -05:00
parent d7127de03c
commit d10b1fa796
114 changed files with 28854 additions and 65033 deletions
Download Patch File Download Diff File Expand all files Collapse all files

26
configuration.ipynb → 00.configuration.ipynb
View File

@@ -140,7 +140,7 @@
"* Your subscription id\n",
"* The resource group name\n",
"\n",
"**Note**: As with other Azure services, there are limits on certain resources (for eg. AmlCompute quota) associated with the Azure Machine Learning service. Please read [this article](https://docs.microsoft.com/en-us/azure/machine-learning/service/how-to-manage-quotas) on the default limits and how to request more quota."
"**Note**: As with other Azure services, there are limits on certain resources (for eg. BatchAI cluster size) associated with the Azure Machine Learning service. Please read [this article](https://docs.microsoft.com/en-us/azure/machine-learning/service/how-to-manage-quotas) on the default limits and how to request more quota."
]
},
{
@@ -210,7 +210,7 @@
"source": [
"## Create compute resources for your training experiments\n",
"\n",
"Many of the subsequent examples use Azure Machine Learning managed compute (AmlCompute) to train models at scale. To create a **CPU** cluster now, run the cell below. The autoscale settings mean that the cluster will scale down to 0 nodes when inactive and up to 4 nodes when busy."
"Many of the subsequent examples use BatchAI clusters to train models at scale. To create a **CPU** cluster now, run the cell below. The autoscale settings mean that the cluster will scale down to 0 nodes when inactive and up to 4 nodes when busy."
]
},
{
@@ -219,7 +219,7 @@
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import ComputeTarget, AmlCompute\n",
"from azureml.core.compute import ComputeTarget, BatchAiCompute\n",
"from azureml.core.compute_target import ComputeTargetException\n",
"\n",
"# Choose a name for your CPU cluster\n",
@@ -230,8 +230,10 @@
" cpu_cluster = ComputeTarget(workspace=ws, name=cpu_cluster_name)\n",
" print('Found existing cluster, use it.')\n",
"except ComputeTargetException:\n",
" compute_config = AmlCompute.provisioning_configuration(vm_size='STANDARD_D2_V2',\n",
" max_nodes=4)\n",
" compute_config = BatchAiCompute.provisioning_configuration(vm_size='STANDARD_D2_V2', \n",
" autoscale_enabled=True,\n",
" cluster_min_nodes=0, \n",
" cluster_max_nodes=4)\n",
" cpu_cluster = ComputeTarget.create(ws, cpu_cluster_name, compute_config)\n",
"\n",
"cpu_cluster.wait_for_completion(show_output=True)"
@@ -250,7 +252,7 @@
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import ComputeTarget, AmlCompute\n",
"from azureml.core.compute import ComputeTarget, BatchAiCompute\n",
"from azureml.core.compute_target import ComputeTargetException\n",
"\n",
"# Choose a name for your GPU cluster\n",
@@ -261,8 +263,10 @@
" gpu_cluster = ComputeTarget(workspace=ws, name=gpu_cluster_name)\n",
" print('Found existing cluster, use it.')\n",
"except ComputeTargetException:\n",
" compute_config = AmlCompute.provisioning_configuration(vm_size='STANDARD_NC6',\n",
" max_nodes=4)\n",
" compute_config = BatchAiCompute.provisioning_configuration(vm_size='STANDARD_NC6', \n",
" autoscale_enabled=True,\n",
" cluster_min_nodes=0, \n",
" cluster_max_nodes=4)\n",
" gpu_cluster = ComputeTarget.create(ws, gpu_cluster_name, compute_config)\n",
"\n",
"gpu_cluster.wait_for_completion(show_output=True)"
@@ -291,9 +295,9 @@
}
],
"kernelspec": {
"display_name": "Python 3",
"display_name": "Python 3.6",
"language": "python",
"name": "python3"
"name": "python36"
},
"language_info": {
"codemirror_mode": {
@@ -305,7 +309,7 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.2"
"version": "3.6.6"
}
},
"nbformat": 4,

808
01.getting-started/01.train-within-notebook/01.train-within-notebook.ipynb Normal file
View File

@@ -0,0 +1,808 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# 01. Train in the Notebook & Deploy Model to ACI\n",
"\n",
"* Load workspace\n",
"* Train a simple regression model directly in the Notebook python kernel\n",
"* Record run history\n",
"* Find the best model in run history and download it.\n",
"* Deploy the model as an Azure Container Instance (ACI)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"1. Make sure you go through the [00. Installation and Configuration](../../00.configuration.ipynb) Notebook first if you haven't. \n",
"\n",
"2. Install following pre-requisite libraries to your conda environment and restart notebook.\n",
"```shell\n",
"(myenv) $ conda install -y matplotlib tqdm scikit-learn\n",
"```\n",
"\n",
"3. Check that ACI is registered for your Azure Subscription. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"!az provider show -n Microsoft.ContainerInstance -o table"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"If ACI is not registered, run following command to register it. Note that you have to be a subscription owner, or this command will fail."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"!az provider register -n Microsoft.ContainerInstance"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Validate Azure ML SDK installation and get version number for debugging purposes"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"install"
]
},
"outputs": [],
"source": [
"from azureml.core import Experiment, Run, Workspace\n",
"import azureml.core\n",
"\n",
"# Check core SDK version number\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize Workspace\n",
"\n",
"Initialize a workspace object from persisted configuration."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"create workspace"
]
},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"print('Workspace name: ' + ws.name, \n",
" 'Azure region: ' + ws.location, \n",
" 'Subscription id: ' + ws.subscription_id, \n",
" 'Resource group: ' + ws.resource_group, sep='\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Set experiment name\n",
"Choose a name for experiment."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"experiment_name = 'train-in-notebook'"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Start a training run in local Notebook"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# load diabetes dataset, a well-known small dataset that comes with scikit-learn\n",
"from sklearn.datasets import load_diabetes\n",
"from sklearn.linear_model import Ridge\n",
"from sklearn.metrics import mean_squared_error\n",
"from sklearn.model_selection import train_test_split\n",
"from sklearn.externals import joblib\n",
"\n",
"X, y = load_diabetes(return_X_y = True)\n",
"columns = ['age', 'gender', 'bmi', 'bp', 's1', 's2', 's3', 's4', 's5', 's6']\n",
"X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=0)\n",
"data = {\n",
" \"train\":{\"X\": X_train, \"y\": y_train}, \n",
" \"test\":{\"X\": X_test, \"y\": y_test}\n",
"}"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Train a simple Ridge model\n",
"Train a very simple Ridge regression model in scikit-learn, and save it as a pickle file."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"reg = Ridge(alpha = 0.03)\n",
"reg.fit(X=data['train']['X'], y=data['train']['y'])\n",
"preds = reg.predict(data['test']['X'])\n",
"print('Mean Squared Error is', mean_squared_error(data['test']['y'], preds))\n",
"joblib.dump(value=reg, filename='model.pkl');"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Add experiment tracking\n",
"Now, let's add Azure ML experiment logging, and upload persisted model into run record as well."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"local run",
"outputs upload"
]
},
"outputs": [],
"source": [
"experiment = Experiment(workspace=ws, name=experiment_name)\n",
"run = experiment.start_logging()\n",
"\n",
"run.tag(\"Description\",\"My first run!\")\n",
"run.log('alpha', 0.03)\n",
"reg = Ridge(alpha=0.03)\n",
"reg.fit(data['train']['X'], data['train']['y'])\n",
"preds = reg.predict(data['test']['X'])\n",
"run.log('mse', mean_squared_error(data['test']['y'], preds))\n",
"joblib.dump(value=reg, filename='model.pkl')\n",
"run.upload_file(name='outputs/model.pkl', path_or_stream='./model.pkl')\n",
"\n",
"run.complete()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We can browse to the recorded run. Please make sure you use Chrome to navigate the run history page."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Simple parameter sweep\n",
"Sweep over alpha values of a sklearn ridge model, and capture metrics and trained model in the Azure ML experiment."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import numpy as np\n",
"import os\n",
"from tqdm import tqdm\n",
"\n",
"model_name = \"model.pkl\"\n",
"\n",
"# list of numbers from 0 to 1.0 with a 0.05 interval\n",
"alphas = np.arange(0.0, 1.0, 0.05)\n",
"\n",
"# try a bunch of alpha values in a Linear Regression (Ridge) model\n",
"for alpha in tqdm(alphas):\n",
" # create a bunch of runs, each train a model with a different alpha value\n",
" with experiment.start_logging() as run:\n",
" # Use Ridge algorithm to build a regression model\n",
" reg = Ridge(alpha=alpha)\n",
" reg.fit(X=data[\"train\"][\"X\"], y=data[\"train\"][\"y\"])\n",
" preds = reg.predict(X=data[\"test\"][\"X\"])\n",
" mse = mean_squared_error(y_true=data[\"test\"][\"y\"], y_pred=preds)\n",
"\n",
" # log alpha, mean_squared_error and feature names in run history\n",
" run.log(name=\"alpha\", value=alpha)\n",
" run.log(name=\"mse\", value=mse)\n",
" run.log_list(name=\"columns\", value=columns)\n",
"\n",
" with open(model_name, \"wb\") as file:\n",
" joblib.dump(value=reg, filename=file)\n",
" \n",
" # upload the serialized model into run history record\n",
" run.upload_file(name=\"outputs/\" + model_name, path_or_stream=model_name)\n",
"\n",
" # now delete the serialized model from local folder since it is already uploaded to run history \n",
" os.remove(path=model_name)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# now let's take a look at the experiment in Azure portal.\n",
"experiment"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Select best model from the experiment\n",
"Load all experiment run metrics recursively from the experiment into a dictionary object."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"runs = {}\n",
"run_metrics = {}\n",
"\n",
"for r in tqdm(experiment.get_runs()):\n",
" metrics = r.get_metrics()\n",
" if 'mse' in metrics.keys():\n",
" runs[r.id] = r\n",
" run_metrics[r.id] = metrics"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now find the run with the lowest Mean Squared Error value"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run_id = min(run_metrics, key = lambda k: run_metrics[k]['mse'])\n",
"best_run = runs[best_run_id]\n",
"print('Best run is:', best_run_id)\n",
"print('Metrics:', run_metrics[best_run_id])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can add tags to your runs to make them easier to catalog"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"query history"
]
},
"outputs": [],
"source": [
"best_run.tag(key=\"Description\", value=\"The best one\")\n",
"best_run.get_tags()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Plot MSE over alpha\n",
"\n",
"Let's observe the best model visually by plotting the MSE values over alpha values:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%matplotlib inline\n",
"import matplotlib\n",
"import matplotlib.pyplot as plt\n",
"\n",
"best_alpha = run_metrics[best_run_id]['alpha']\n",
"min_mse = run_metrics[best_run_id]['mse']\n",
"\n",
"alpha_mse = np.array([(run_metrics[k]['alpha'], run_metrics[k]['mse']) for k in run_metrics.keys()])\n",
"sorted_alpha_mse = alpha_mse[alpha_mse[:,0].argsort()]\n",
"\n",
"plt.plot(sorted_alpha_mse[:,0], sorted_alpha_mse[:,1], 'r--')\n",
"plt.plot(sorted_alpha_mse[:,0], sorted_alpha_mse[:,1], 'bo')\n",
"\n",
"plt.xlabel('alpha', fontsize = 14)\n",
"plt.ylabel('mean squared error', fontsize = 14)\n",
"plt.title('MSE over alpha', fontsize = 16)\n",
"\n",
"# plot arrow\n",
"plt.arrow(x = best_alpha, y = min_mse + 39, dx = 0, dy = -26, ls = '-', lw = 0.4,\n",
" width = 0, head_width = .03, head_length = 8)\n",
"\n",
"# plot \"best run\" text\n",
"plt.text(x = best_alpha - 0.08, y = min_mse + 50, s = 'Best Run', fontsize = 14)\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Register the best model"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Find the model file saved in the run record of best run."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"query history"
]
},
"outputs": [],
"source": [
"for f in best_run.get_file_names():\n",
" print(f)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now we can register this model in the model registry of the workspace"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"register model from history"
]
},
"outputs": [],
"source": [
"model = best_run.register_model(model_name='best_model', model_path='outputs/model.pkl')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Verify that the model has been registered properly. If you have done this several times you'd see the version number auto-increases each time."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"register model from history"
]
},
"outputs": [],
"source": [
"from azureml.core.model import Model\n",
"models = Model.list(workspace=ws, name='best_model')\n",
"for m in models:\n",
" print(m.name, m.version)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can also download the registered model. Afterwards, you should see a `model.pkl` file in the current directory. You can then use it for local testing if you'd like."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"download file"
]
},
"outputs": [],
"source": [
"# remove the model file if it is already on disk\n",
"if os.path.isfile('model.pkl'): \n",
" os.remove('model.pkl')\n",
"# download the model\n",
"model.download(target_dir=\"./\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Scoring script\n",
"\n",
"Now we are ready to build a Docker image and deploy the model in it as a web service. The first step is creating the scoring script. For convenience, we have created the scoring script for you. It is printed below as text, but you can also run `%pfile ./score.py` in a cell to show the file.\n",
"\n",
"Tbe scoring script consists of two functions: `init` that is used to load the model to memory when starting the container, and `run` that makes the prediction when web service is called. Please pay special attention to how the model is loaded in the `init()` function. When Docker image is built for this model, the actual model file is downloaded and placed on disk, and `get_model_path` function returns the local path where the model is placed."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"with open('./score.py', 'r') as scoring_script:\n",
" print(scoring_script.read())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create environment dependency file\n",
"\n",
"We need a environment dependency file `myenv.yml` to specify which libraries are needed by the scoring script when building the Docker image for web service deployment. We can manually create this file, or we can use the `CondaDependencies` API to automatically create this file."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.conda_dependencies import CondaDependencies \n",
"\n",
"myenv = CondaDependencies.create(conda_packages=[\"scikit-learn\"])\n",
"print(myenv.serialize_to_string())\n",
"\n",
"with open(\"myenv.yml\",\"w\") as f:\n",
" f.write(myenv.serialize_to_string())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Deploy web service into an Azure Container Instance\n",
"The deployment process takes the registered model and your scoring scrip, and builds a Docker image. It then deploys the Docker image into Azure Container Instance as a running container with an HTTP endpoint readying for scoring calls. Read more about [Azure Container Instance](https://azure.microsoft.com/en-us/services/container-instances/).\n",
"\n",
"Note ACI is great for quick and cost-effective dev/test deployment scenarios. For production workloads, please use [Azure Kubernentes Service (AKS)](https://azure.microsoft.com/en-us/services/kubernetes-service/) instead. Please follow in struction in [this notebook](11.production-deploy-to-aks.ipynb) to see how that can be done from Azure ML.\n",
" \n",
"** Note: ** The web service creation can take 6-7 minutes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"from azureml.core.webservice import AciWebservice, Webservice\n",
"\n",
"aciconfig = AciWebservice.deploy_configuration(cpu_cores=1, \n",
" memory_gb=1, \n",
" tags={'sample name': 'AML 101'}, \n",
" description='This is a great example.')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Note the below `WebService.deploy_from_model()` function takes a model object registered under the workspace. It then bakes the model file in the Docker image so it can be looked-up using the `Model.get_model_path()` function in `score.py`. \n",
"\n",
"If you have a local model file instead of a registered model object, you can also use the `WebService.deploy()` function which would register the model and then deploy."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"from azureml.core.image import ContainerImage\n",
"image_config = ContainerImage.image_configuration(execution_script=\"score.py\", \n",
" runtime=\"python\", \n",
" conda_file=\"myenv.yml\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"%%time\n",
"# this will take 5-10 minutes to finish\n",
"# you can also use \"az container list\" command to find the ACI being deployed\n",
"service = Webservice.deploy_from_model(name='my-aci-svc',\n",
" deployment_config=aciconfig,\n",
" models=[model],\n",
" image_config=image_config,\n",
" workspace=ws)\n",
"\n",
"service.wait_for_deployment(show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"## Test web service"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"print('web service is hosted in ACI:', service.scoring_uri)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Use the `run` API to call the web service with one row of data to get a prediction."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"import json\n",
"# score the first row from the test set.\n",
"test_samples = json.dumps({\"data\": X_test[0:1, :].tolist()})\n",
"service.run(input_data = test_samples)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Feed the entire test set and calculate the errors (residual values)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"# score the entire test set.\n",
"test_samples = json.dumps({'data': X_test.tolist()})\n",
"\n",
"result = service.run(input_data = test_samples)\n",
"residual = result - y_test"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can also send raw HTTP request to test the web service."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"import requests\n",
"import json\n",
"\n",
"# 2 rows of input data, each with 10 made-up numerical features\n",
"input_data = \"{\\\"data\\\": [[1, 2, 3, 4, 5, 6, 7, 8, 9, 10], [10, 9, 8, 7, 6, 5, 4, 3, 2, 1]]}\"\n",
"\n",
"headers = {'Content-Type':'application/json'}\n",
"\n",
"# for AKS deployment you'd need to the service key in the header as well\n",
"# api_key = service.get_key()\n",
"# headers = {'Content-Type':'application/json', 'Authorization':('Bearer '+ api_key)} \n",
"\n",
"resp = requests.post(service.scoring_uri, input_data, headers = headers)\n",
"print(resp.text)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Residual graph\n",
"Plot a residual value graph to chart the errors on the entire test set. Observe the nice bell curve."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"f, (a0, a1) = plt.subplots(1, 2, gridspec_kw={'width_ratios':[3, 1], 'wspace':0, 'hspace': 0})\n",
"f.suptitle('Residual Values', fontsize = 18)\n",
"\n",
"f.set_figheight(6)\n",
"f.set_figwidth(14)\n",
"\n",
"a0.plot(residual, 'bo', alpha=0.4);\n",
"a0.plot([0,90], [0,0], 'r', lw=2)\n",
"a0.set_ylabel('residue values', fontsize=14)\n",
"a0.set_xlabel('test data set', fontsize=14)\n",
"\n",
"a1.hist(residual, orientation='horizontal', color='blue', bins=10, histtype='step');\n",
"a1.hist(residual, orientation='horizontal', color='blue', alpha=0.2, bins=10);\n",
"a1.set_yticklabels([])\n",
"\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Delete ACI to clean up"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Deleting ACI is super fast!"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"%%time\n",
"service.delete()"
]
}
],
"metadata": {
"authors": [
{
"name": "roastala"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

0
getting-started/train-within-notebook/score.py → 01.getting-started/01.train-within-notebook/score.py
View File

477
01.getting-started/02.train-on-local/02.train-on-local.ipynb Normal file
View File

@@ -0,0 +1,477 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# 02. Train locally\n",
"* Create or load workspace.\n",
"* Create scripts locally.\n",
"* Create `train.py` in a folder, along with a `my.lib` file.\n",
"* Configure & execute a local run in a user-managed Python environment.\n",
"* Configure & execute a local run in a system-managed Python environment.\n",
"* Configure & execute a local run in a Docker environment.\n",
"* Query run metrics to find the best model\n",
"* Register model for operationalization."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"Make sure you go through the [00. Installation and Configuration](00.configuration.ipynb) Notebook first if you haven't."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize Workspace\n",
"\n",
"Initialize a workspace object from persisted configuration."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.workspace import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print(ws.name, ws.resource_group, ws.location, ws.subscription_id, sep='\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create An Experiment\n",
"**Experiment** is a logical container in an Azure ML Workspace. It hosts run records which can include run metrics and output artifacts from your experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Experiment\n",
"experiment_name = 'train-on-local'\n",
"exp = Experiment(workspace=ws, name=experiment_name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## View `train.py`\n",
"\n",
"`train.py` is already created for you."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"with open('./train.py', 'r') as f:\n",
" print(f.read())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Note `train.py` also references a `mylib.py` file."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"with open('./mylib.py', 'r') as f:\n",
" print(f.read())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure & Run\n",
"### User-managed environment\n",
"Below, we use a user-managed run, which means you are responsible to ensure all the necessary packages are available in the Python environment you choose to run the script."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.runconfig import RunConfiguration\n",
"\n",
"# Editing a run configuration property on-fly.\n",
"run_config_user_managed = RunConfiguration()\n",
"\n",
"run_config_user_managed.environment.python.user_managed_dependencies = True\n",
"\n",
"# You can choose a specific Python environment by pointing to a Python path \n",
"#run_config.environment.python.interpreter_path = '/home/johndoe/miniconda3/envs/sdk2/bin/python'"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Submit script to run in the user-managed environment\n",
"Note whole script folder is submitted for execution, including the `mylib.py` file."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import ScriptRunConfig\n",
"\n",
"src = ScriptRunConfig(source_directory='./', script='train.py', run_config=run_config_user_managed)\n",
"run = exp.submit(src)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Get run history details"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Block to wait till run finishes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### System-managed environment\n",
"You can also ask the system to build a new conda environment and execute your scripts in it. The environment is built once and will be reused in subsequent executions as long as the conda dependencies remain unchanged. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.runconfig import RunConfiguration\n",
"from azureml.core.conda_dependencies import CondaDependencies\n",
"\n",
"run_config_system_managed = RunConfiguration()\n",
"\n",
"run_config_system_managed.environment.python.user_managed_dependencies = False\n",
"run_config_system_managed.auto_prepare_environment = True\n",
"\n",
"# Specify conda dependencies with scikit-learn\n",
"cd = CondaDependencies.create(conda_packages=['scikit-learn'])\n",
"run_config_system_managed.environment.python.conda_dependencies = cd"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Submit script to run in the system-managed environment\n",
"A new conda environment is built based on the conda dependencies object. If you are running this for the first time, this might take up to 5 mninutes. But this conda environment is reused so long as you don't change the conda dependencies."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"src = ScriptRunConfig(source_directory=\"./\", script='train.py', run_config=run_config_system_managed)\n",
"run = exp.submit(src)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Get run history details"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Block and wait till run finishes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.wait_for_completion(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Docker-based execution\n",
"**IMPORTANT**: You must have Docker engine installed locally in order to use this execution mode. If your kernel is already running in a Docker container, such as **Azure Notebooks**, this mode will **NOT** work.\n",
"\n",
"You can also ask the system to pull down a Docker image and execute your scripts in it."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run_config_docker = RunConfiguration()\n",
"run_config_docker.environment.python.user_managed_dependencies = False\n",
"run_config_docker.auto_prepare_environment = True\n",
"run_config_docker.environment.docker.enabled = True\n",
"run_config_docker.environment.docker.base_image = azureml.core.runconfig.DEFAULT_CPU_IMAGE\n",
"\n",
"# Specify conda dependencies with scikit-learn\n",
"cd = CondaDependencies.create(conda_packages=['scikit-learn'])\n",
"run_config_docker.environment.python.conda_dependencies = cd\n",
"\n",
"src = ScriptRunConfig(source_directory=\"./\", script='train.py', run_config=run_config_docker)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Submit script to run in the system-managed environment\n",
"A new conda environment is built based on the conda dependencies object. If you are running this for the first time, this might take up to 5 mninutes. But this conda environment is reused so long as you don't change the conda dependencies.\n",
"\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import subprocess\n",
"\n",
"# Check if Docker is installed and Linux containers are enables\n",
"if subprocess.run(\"docker -v\", shell=True) == 0:\n",
" out = subprocess.check_output(\"docker system info\", shell=True, encoding=\"ascii\").split(\"\\n\")\n",
" if not \"OSType: linux\" in out:\n",
" print(\"Switch Docker engine to use Linux containers.\")\n",
" else:\n",
" run = exp.submit(src)\n",
"else:\n",
" print(\"Docker engine not installed.\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#Get run history details\n",
"run"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Query run metrics"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"query history",
"get metrics"
]
},
"outputs": [],
"source": [
"# get all metris logged in the run\n",
"run.get_metrics()\n",
"metrics = run.get_metrics()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Let's find the model that has the lowest MSE value logged."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import numpy as np\n",
"\n",
"best_alpha = metrics['alpha'][np.argmin(metrics['mse'])]\n",
"\n",
"print('When alpha is {1:0.2f}, we have min MSE {0:0.2f}.'.format(\n",
" min(metrics['mse']), \n",
" best_alpha\n",
"))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can also list all the files that are associated with this run record"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.get_file_names()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We know the model `ridge_0.40.pkl` is the best performing model from the eariler queries. So let's register it with the workspace."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# supply a model name, and the full path to the serialized model file.\n",
"model = run.register_model(model_name='best_ridge_model', model_path='./outputs/ridge_0.40.pkl')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(model.name, model.version, model.url)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now you can deploy this model following the example in the 01 notebook."
]
}
],
"metadata": {
"authors": [
{
"name": "roastala"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

0
getting-started/train-on-local/mylib.py → 01.getting-started/02.train-on-local/mylib.py
View File

0
getting-started/train-on-local/train.py → 01.getting-started/02.train-on-local/train.py
View File

289
01.getting-started/03.train-on-aci/03.train-on-aci.ipynb Normal file
View File

@@ -0,0 +1,289 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# 03. Train on Azure Container Instance\n",
"\n",
"* Create Workspace\n",
"* Create `train.py` in the project folder.\n",
"* Configure an ACI (Azure Container Instance) run\n",
"* Execute in ACI"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"Make sure you go through the [00. Installation and Configuration](00.configuration.ipynb) Notebook first if you haven't."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize Workspace\n",
"\n",
"Initialize a workspace object from persisted configuration"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"create workspace"
]
},
"outputs": [],
"source": [
"from azureml.core import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print(ws.name, ws.resource_group, ws.location, ws.subscription_id, sep = '\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create An Experiment\n",
"\n",
"**Experiment** is a logical container in an Azure ML Workspace. It hosts run records which can include run metrics and output artifacts from your experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Experiment\n",
"experiment_name = 'train-on-aci'\n",
"experiment = Experiment(workspace = ws, name = experiment_name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Remote execution on ACI\n",
"\n",
"The training script `train.py` is already created for you. Let's have a look."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"with open('./train.py', 'r') as f:\n",
" print(f.read())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure for using ACI\n",
"Linux-based ACI is available in `West US`, `East US`, `West Europe`, `North Europe`, `West US 2`, `Southeast Asia`, `Australia East`, `East US 2`, and `Central US` regions. See details [here](https://docs.microsoft.com/en-us/azure/container-instances/container-instances-quotas#region-availability)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"configure run"
]
},
"outputs": [],
"source": [
"from azureml.core.runconfig import RunConfiguration\n",
"from azureml.core.conda_dependencies import CondaDependencies\n",
"\n",
"# create a new runconfig object\n",
"run_config = RunConfiguration()\n",
"\n",
"# signal that you want to use ACI to execute script.\n",
"run_config.target = \"containerinstance\"\n",
"\n",
"# ACI container group is only supported in certain regions, which can be different than the region the Workspace is in.\n",
"run_config.container_instance.region = 'eastus2'\n",
"\n",
"# set the ACI CPU and Memory \n",
"run_config.container_instance.cpu_cores = 1\n",
"run_config.container_instance.memory_gb = 2\n",
"\n",
"# enable Docker \n",
"run_config.environment.docker.enabled = True\n",
"\n",
"# set Docker base image to the default CPU-based image\n",
"run_config.environment.docker.base_image = azureml.core.runconfig.DEFAULT_CPU_IMAGE\n",
"\n",
"# use conda_dependencies.yml to create a conda environment in the Docker image for execution\n",
"run_config.environment.python.user_managed_dependencies = False\n",
"\n",
"# auto-prepare the Docker image when used for execution (if it is not already prepared)\n",
"run_config.auto_prepare_environment = True\n",
"\n",
"# specify CondaDependencies obj\n",
"run_config.environment.python.conda_dependencies = CondaDependencies.create(conda_packages=['scikit-learn'])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Submit the Experiment\n",
"Finally, run the training job on the ACI"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"remote run",
"aci"
]
},
"outputs": [],
"source": [
"%%time \n",
"from azureml.core.script_run_config import ScriptRunConfig\n",
"\n",
"script_run_config = ScriptRunConfig(source_directory='./',\n",
" script='train.py',\n",
" run_config=run_config)\n",
"\n",
"run = experiment.submit(script_run_config)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"query history"
]
},
"outputs": [],
"source": [
"# Show run details\n",
"run"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"remote run",
"aci"
]
},
"outputs": [],
"source": [
"%%time\n",
"# Shows output of the run on stdout.\n",
"run.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"get metrics"
]
},
"outputs": [],
"source": [
"# get all metris logged in the run\n",
"run.get_metrics()\n",
"metrics = run.get_metrics()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import numpy as np\n",
"print('When alpha is {1:0.2f}, we have min MSE {0:0.2f}.'.format(\n",
" min(metrics['mse']), \n",
" metrics['alpha'][np.argmin(metrics['mse'])]\n",
"))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# show all the files stored within the run record\n",
"run.get_file_names()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now you can take a model produced here, register it and then deploy as a web service."
]
}
],
"metadata": {
"authors": [
{
"name": "roastala"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

0
getting-started/train-on-aci/train.py → 01.getting-started/03.train-on-aci/train.py
View File

621
01.getting-started/04.train-on-remote-vm/04.train-on-remote-vm.ipynb Normal file
View File

@@ -0,0 +1,621 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# 04. Train in a remote Linux VM\n",
"* Create Workspace\n",
"* Create `train.py` file\n",
"* Create (or attach) DSVM as compute resource.\n",
"* Upoad data files into default datastore\n",
"* Configure & execute a run in a few different ways\n",
" - Use system-built conda\n",
" - Use existing Python environment\n",
" - Use Docker \n",
"* Find the best model in the run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"Make sure you go through the [00. Installation and Configuration](00.configuration.ipynb) Notebook first if you haven't."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize Workspace\n",
"\n",
"Initialize a workspace object from persisted configuration."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print(ws.name, ws.resource_group, ws.location, ws.subscription_id, sep='\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create Experiment\n",
"\n",
"**Experiment** is a logical container in an Azure ML Workspace. It hosts run records which can include run metrics and output artifacts from your experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"experiment_name = 'train-on-remote-vm'\n",
"\n",
"from azureml.core import Experiment\n",
"exp = Experiment(workspace=ws, name=experiment_name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Let's also create a local folder to hold the training script."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"script_folder = './vm-run'\n",
"os.makedirs(script_folder, exist_ok=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Upload data files into datastore\n",
"Every workspace comes with a default datastore (and you can register more) which is backed by the Azure blob storage account associated with the workspace. We can use it to transfer data from local to the cloud, and access it from the compute target."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# get the default datastore\n",
"ds = ws.get_default_datastore()\n",
"print(ds.name, ds.datastore_type, ds.account_name, ds.container_name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Load diabetes data from `scikit-learn` and save it as 2 local files."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.datasets import load_diabetes\n",
"import numpy as np\n",
"\n",
"training_data = load_diabetes()\n",
"np.save(file='./features.npy', arr=training_data['data'])\n",
"np.save(file='./labels.npy', arr=training_data['target'])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now let's upload the 2 files into the default datastore under a path named `diabetes`:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ds.upload_files(['./features.npy', './labels.npy'], target_path='diabetes', overwrite=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## View `train.py`\n",
"\n",
"For convenience, we created a training script for you. It is printed below as a text, but you can also run `%pfile ./train.py` in a cell to show the file. Please pay special attention on how we are loading the features and labels from files in the `data_folder` path, which is passed in as an argument of the training script (shown later)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# copy train.py into the script folder\n",
"import shutil\n",
"shutil.copy('./train.py', os.path.join(script_folder, 'train.py'))\n",
"\n",
"with open(os.path.join(script_folder, './train.py'), 'r') as training_script:\n",
" print(training_script.read())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create Linux DSVM as a compute target\n",
"\n",
"**Note**: If creation fails with a message about Marketplace purchase eligibilty, go to portal.azure.com, start creating DSVM there, and select \"Want to create programmatically\" to enable programmatic creation. Once you've enabled it, you can exit without actually creating VM.\n",
" \n",
"**Note**: By default SSH runs on port 22 and you don't need to specify it. But if for security reasons you switch to a different port (such as 5022), you can specify the port number in the provisioning configuration object."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import DsvmCompute\n",
"from azureml.core.compute_target import ComputeTargetException\n",
"\n",
"compute_target_name = 'mydsvm'\n",
"\n",
"try:\n",
" dsvm_compute = DsvmCompute(workspace=ws, name=compute_target_name)\n",
" print('found existing:', dsvm_compute.name)\n",
"except ComputeTargetException:\n",
" print('creating new.')\n",
" dsvm_config = DsvmCompute.provisioning_configuration(vm_size=\"Standard_D2_v2\")\n",
" dsvm_compute = DsvmCompute.create(ws, name=compute_target_name, provisioning_configuration=dsvm_config)\n",
" dsvm_compute.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Attach an existing Linux DSVM\n",
"You can also attach an existing Linux VM as a compute target. The default port is 22."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"'''\n",
"from azureml.core.compute import RemoteCompute \n",
"# if you want to connect using SSH key instead of username/password you can provide parameters private_key_file and private_key_passphrase \n",
"attached_dsvm_compute = RemoteCompute.attach(workspace=ws,\n",
" name=\"attached_vm\",\n",
" username='<usename>',\n",
" address='<ip_adress_or_fqdn>',\n",
" ssh_port=22,\n",
" password='<password>')\n",
"attached_dsvm_compute.wait_for_completion(show_output=True)\n",
"'''\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure & Run\n",
"First let's create a `DataReferenceConfiguration` object to inform the system what data folder to download to the copmute target."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.runconfig import DataReferenceConfiguration\n",
"dr = DataReferenceConfiguration(datastore_name=ds.name, \n",
" path_on_datastore='diabetes', \n",
" mode='download', # download files from datastore to compute target\n",
" overwrite=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now we can try a few different ways to run the training script in the VM."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Conda run\n",
"You can ask the system to build a conda environment based on your dependency specification, and submit your script to run there. Once the environment is built, and if you don't change your dependencies, it will be reused in subsequent runs."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.runconfig import RunConfiguration\n",
"from azureml.core.conda_dependencies import CondaDependencies\n",
"\n",
"# create a new RunConfig object\n",
"conda_run_config = RunConfiguration(framework=\"python\")\n",
"\n",
"# Set compute target to the Linux DSVM\n",
"conda_run_config.target = dsvm_compute.name\n",
"\n",
"# set the data reference of the run configuration\n",
"conda_run_config.data_references = {ds.name: dr}\n",
"\n",
"# specify CondaDependencies obj\n",
"conda_run_config.environment.python.conda_dependencies = CondaDependencies.create(conda_packages=['scikit-learn'])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Run\n",
"from azureml.core import ScriptRunConfig\n",
"\n",
"src = ScriptRunConfig(source_directory=script_folder, \n",
" script='train.py', \n",
" run_config=conda_run_config, \n",
" # pass the datastore reference as a parameter to the training script\n",
" arguments=['--data-folder', str(ds.as_download())] \n",
" ) \n",
"run = exp.submit(config=src)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Show the run object. You can navigate to the Azure portal to see detailed information about the run."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Native VM run\n",
"You can also configure to use an exiting Python environment in the VM to execute the script without asking the system to create a conda environment for you."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# create a new RunConfig object\n",
"vm_run_config = RunConfiguration(framework=\"python\")\n",
"\n",
"# Set compute target to the Linux DSVM\n",
"vm_run_config.target = dsvm_compute.name\n",
"\n",
"# set the data reference of the run coonfiguration\n",
"conda_run_config.data_references = {ds.name: dr}\n",
"\n",
"# Let system know that you will configure the Python environment yourself.\n",
"vm_run_config.environment.python.user_managed_dependencies = True"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The below run will likely fail because `train.py` needs dependency `azureml`, `scikit-learn` and others, which are not found in that Python environment. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"src = ScriptRunConfig(source_directory=script_folder, \n",
" script='train.py', \n",
" run_config=vm_run_config,\n",
" # pass the datastore reference as a parameter to the training script\n",
" arguments=['--data-folder', str(ds.as_download())])\n",
"run = exp.submit(config=src)\n",
"run.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can choose to SSH into the VM and install Azure ML SDK, and any other missing dependencies, in that Python environment. For demonstration purposes, we simply are going to create another script `train2.py` that doesn't have azureml dependencies, and submit it instead."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile $script_folder/train2.py\n",
"\n",
"print('####################################')\n",
"print('Hello World (without Azure ML SDK)!')\n",
"print('####################################')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now let's try again. And this time it should work fine."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"src = ScriptRunConfig(source_directory=script_folder, \n",
" script='train2.py', \n",
" run_config=vm_run_config)\n",
"run = exp.submit(config=src)\n",
"run.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Note even in this case you get a run record with some basic statistics."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Configure a Docker run with new conda environment on the VM\n",
"You can execute in a Docker container in the VM. If you choose this option, the system will pull down a base Docker image, build a new conda environment in it if you ask for (you can also skip this if you are using a customer Docker image when a preconfigured Python environment), start a container, and run your script in there. This image is also uploaded into your ACR (Azure Container Registry) assoicated with your workspace, an reused if your dependencies don't change in the subsequent runs."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.runconfig import RunConfiguration\n",
"from azureml.core.conda_dependencies import CondaDependencies\n",
"\n",
"\n",
"# Load the \"cpu-dsvm.runconfig\" file (created by the above attach operation) in memory\n",
"docker_run_config = RunConfiguration(framework=\"python\")\n",
"\n",
"# Set compute target to the Linux DSVM\n",
"docker_run_config.target = dsvm_compute.name\n",
"\n",
"# Use Docker in the remote VM\n",
"docker_run_config.environment.docker.enabled = True\n",
"\n",
"# Use CPU base image from DockerHub\n",
"docker_run_config.environment.docker.base_image = azureml.core.runconfig.DEFAULT_CPU_IMAGE\n",
"print('Base Docker image is:', docker_run_config.environment.docker.base_image)\n",
"\n",
"# set the data reference of the run coonfiguration\n",
"docker_run_config.data_references = {ds.name: dr}\n",
"\n",
"# specify CondaDependencies obj\n",
"docker_run_config.environment.python.conda_dependencies = CondaDependencies.create(conda_packages=['scikit-learn'])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Submit the Experiment\n",
"Submit script to run in the Docker image in the remote VM. If you run this for the first time, the system will download the base image, layer in packages specified in the `conda_dependencies.yml` file on top of the base image, create a container and then execute the script in the container."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"src = ScriptRunConfig(source_directory=script_folder, \n",
" script='train.py', \n",
" run_config=docker_run_config,\n",
" # pass the datastore reference as a parameter to the training script\n",
" arguments=['--data-folder', str(ds.as_download())])\n",
"run = exp.submit(config=src)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### View run history details"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Find the best model"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now we have tried various execution modes, we can find the best model from the last run."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# get all metris logged in the run\n",
"run.get_metrics()\n",
"metrics = run.get_metrics()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# find the index where MSE is the smallest\n",
"indices = list(range(0, len(metrics['mse'])))\n",
"min_mse_index = min(indices, key=lambda x: metrics['mse'][x])\n",
"\n",
"print('When alpha is {1:0.2f}, we have min MSE {0:0.2f}.'.format(\n",
" metrics['mse'][min_mse_index], \n",
" metrics['alpha'][min_mse_index]\n",
"))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Clean up compute resource"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"dsvm_compute.delete()"
]
}
],
"metadata": {
"authors": [
{
"name": "haining"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

0
getting-started/train-on-remote-vm/train.py → 01.getting-started/04.train-on-remote-vm/train.py
View File

331
01.getting-started/05.train-in-spark/05.train-in-spark.ipynb Normal file
View File

@@ -0,0 +1,331 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# 05. Train in Spark\n",
"* Create Workspace\n",
"* Create Experiment\n",
"* Copy relevant files to the script folder\n",
"* Configure and Run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"Make sure you go through the [00. Installation and Configuration](00.configuration.ipynb) Notebook first if you haven't."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize Workspace\n",
"\n",
"Initialize a workspace object from persisted configuration."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print(ws.name, ws.resource_group, ws.location, ws.subscription_id, sep='\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create Experiment\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"experiment_name = 'train-on-spark'\n",
"\n",
"from azureml.core import Experiment\n",
"exp = Experiment(workspace=ws, name=experiment_name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## View `train-spark.py`\n",
"\n",
"For convenience, we created a training script for you. It is printed below as a text, but you can also run `%pfile ./train-spark.py` in a cell to show the file."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"with open('train-spark.py', 'r') as training_script:\n",
" print(training_script.read())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure & Run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Configure an ACI run\n",
"Before you try running on an actual Spark cluster, you can use a Docker image with Spark already baked in, and run it in ACI(Azure Container Registry)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.runconfig import RunConfiguration\n",
"from azureml.core.conda_dependencies import CondaDependencies\n",
"\n",
"# use pyspark framework\n",
"aci_run_config = RunConfiguration(framework=\"pyspark\")\n",
"\n",
"# use ACI to run the Spark job\n",
"aci_run_config.target = 'containerinstance'\n",
"aci_run_config.container_instance.region = 'eastus2'\n",
"aci_run_config.container_instance.cpu_cores = 1\n",
"aci_run_config.container_instance.memory_gb = 2\n",
"\n",
"# specify base Docker image to use\n",
"aci_run_config.environment.docker.enabled = True\n",
"aci_run_config.environment.docker.base_image = azureml.core.runconfig.DEFAULT_MMLSPARK_CPU_IMAGE\n",
"\n",
"# specify CondaDependencies\n",
"cd = CondaDependencies()\n",
"cd.add_conda_package('numpy')\n",
"aci_run_config.environment.python.conda_dependencies = cd"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Submit script to ACI to run"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import ScriptRunConfig\n",
"\n",
"script_run_config = ScriptRunConfig(source_directory = '.',\n",
" script= 'train-spark.py',\n",
" run_config = aci_run_config)\n",
"run = exp.submit(script_run_config)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Note** you can also create a new VM, or attach an existing VM, and use Docker-based execution to run the Spark job. Please see the `04.train-in-vm` for example on how to configure and run in Docker mode in a VM."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Attach an HDI cluster\n",
"Now we can use a real Spark cluster, HDInsight for Spark, to run this job. To use HDI commpute target:\n",
" 1. Create a Spark for HDI cluster in Azure. Here are some [quick instructions](https://docs.microsoft.com/en-us/azure/hdinsight/spark/apache-spark-jupyter-spark-sql). Make sure you use the Ubuntu flavor, NOT CentOS.\n",
" 2. Enter the IP address, username and password below"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import HDInsightCompute\n",
"from azureml.exceptions import ComputeTargetException\n",
"\n",
"try:\n",
" # if you want to connect using SSH key instead of username/password you can provide parameters private_key_file and private_key_passphrase\n",
" hdi_compute = HDInsightCompute.attach(workspace=ws, \n",
" name=\"myhdi\", \n",
" address=\"<myhdi-ssh>.azurehdinsight.net\", \n",
" ssh_port=22, \n",
" username='<ssh-username>', \n",
" password='<ssh-pwd>')\n",
"\n",
"except ComputeTargetException as e:\n",
" print(\"Caught = {}\".format(e.message))\n",
" \n",
" \n",
"hdi_compute.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Configure HDI run"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.runconfig import RunConfiguration\n",
"from azureml.core.conda_dependencies import CondaDependencies\n",
"\n",
"\n",
"# use pyspark framework\n",
"hdi_run_config = RunConfiguration(framework=\"pyspark\")\n",
"\n",
"# Set compute target to the HDI cluster\n",
"hdi_run_config.target = hdi_compute.name\n",
"\n",
"# specify CondaDependencies object to ask system installing numpy\n",
"cd = CondaDependencies()\n",
"cd.add_conda_package('numpy')\n",
"hdi_run_config.environment.python.conda_dependencies = cd"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Submit the script to HDI"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import ScriptRunConfig\n",
"\n",
"script_run_config = ScriptRunConfig(source_directory = '.',\n",
" script= 'train-spark.py',\n",
" run_config = hdi_run_config)\n",
"run = exp.submit(config=script_run_config)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# get the URL of the run history web page\n",
"run"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# get all metris logged in the run\n",
"metrics = run.get_metrics()\n",
"print(metrics)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"authors": [
{
"name": "aashishb"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

150
01.getting-started/05.train-in-spark/iris.csv Normal file
View File

@@ -0,0 +1,150 @@
5.1,3.5,1.4,0.2,Iris-setosa
4.9,3.0,1.4,0.2,Iris-setosa
4.7,3.2,1.3,0.2,Iris-setosa
4.6,3.1,1.5,0.2,Iris-setosa
5.0,3.6,1.4,0.2,Iris-setosa
5.4,3.9,1.7,0.4,Iris-setosa
4.6,3.4,1.4,0.3,Iris-setosa
5.0,3.4,1.5,0.2,Iris-setosa
4.4,2.9,1.4,0.2,Iris-setosa
4.9,3.1,1.5,0.1,Iris-setosa
5.4,3.7,1.5,0.2,Iris-setosa
4.8,3.4,1.6,0.2,Iris-setosa
4.8,3.0,1.4,0.1,Iris-setosa
4.3,3.0,1.1,0.1,Iris-setosa
5.8,4.0,1.2,0.2,Iris-setosa
5.7,4.4,1.5,0.4,Iris-setosa
5.4,3.9,1.3,0.4,Iris-setosa
5.1,3.5,1.4,0.3,Iris-setosa
5.7,3.8,1.7,0.3,Iris-setosa
5.1,3.8,1.5,0.3,Iris-setosa
5.4,3.4,1.7,0.2,Iris-setosa
5.1,3.7,1.5,0.4,Iris-setosa
4.6,3.6,1.0,0.2,Iris-setosa
5.1,3.3,1.7,0.5,Iris-setosa
4.8,3.4,1.9,0.2,Iris-setosa
5.0,3.0,1.6,0.2,Iris-setosa
5.0,3.4,1.6,0.4,Iris-setosa
5.2,3.5,1.5,0.2,Iris-setosa
5.2,3.4,1.4,0.2,Iris-setosa
4.7,3.2,1.6,0.2,Iris-setosa
4.8,3.1,1.6,0.2,Iris-setosa
5.4,3.4,1.5,0.4,Iris-setosa
5.2,4.1,1.5,0.1,Iris-setosa
5.5,4.2,1.4,0.2,Iris-setosa
4.9,3.1,1.5,0.1,Iris-setosa
5.0,3.2,1.2,0.2,Iris-setosa
5.5,3.5,1.3,0.2,Iris-setosa
4.9,3.1,1.5,0.1,Iris-setosa
4.4,3.0,1.3,0.2,Iris-setosa
5.1,3.4,1.5,0.2,Iris-setosa
5.0,3.5,1.3,0.3,Iris-setosa
4.5,2.3,1.3,0.3,Iris-setosa
4.4,3.2,1.3,0.2,Iris-setosa
5.0,3.5,1.6,0.6,Iris-setosa
5.1,3.8,1.9,0.4,Iris-setosa
4.8,3.0,1.4,0.3,Iris-setosa
5.1,3.8,1.6,0.2,Iris-setosa
4.6,3.2,1.4,0.2,Iris-setosa
5.3,3.7,1.5,0.2,Iris-setosa
5.0,3.3,1.4,0.2,Iris-setosa
7.0,3.2,4.7,1.4,Iris-versicolor
6.4,3.2,4.5,1.5,Iris-versicolor
6.9,3.1,4.9,1.5,Iris-versicolor
5.5,2.3,4.0,1.3,Iris-versicolor
6.5,2.8,4.6,1.5,Iris-versicolor
5.7,2.8,4.5,1.3,Iris-versicolor
6.3,3.3,4.7,1.6,Iris-versicolor
4.9,2.4,3.3,1.0,Iris-versicolor
6.6,2.9,4.6,1.3,Iris-versicolor
5.2,2.7,3.9,1.4,Iris-versicolor
5.0,2.0,3.5,1.0,Iris-versicolor
5.9,3.0,4.2,1.5,Iris-versicolor
6.0,2.2,4.0,1.0,Iris-versicolor
6.1,2.9,4.7,1.4,Iris-versicolor
5.6,2.9,3.6,1.3,Iris-versicolor
6.7,3.1,4.4,1.4,Iris-versicolor
5.6,3.0,4.5,1.5,Iris-versicolor
5.8,2.7,4.1,1.0,Iris-versicolor
6.2,2.2,4.5,1.5,Iris-versicolor
5.6,2.5,3.9,1.1,Iris-versicolor
5.9,3.2,4.8,1.8,Iris-versicolor
6.1,2.8,4.0,1.3,Iris-versicolor
6.3,2.5,4.9,1.5,Iris-versicolor
6.1,2.8,4.7,1.2,Iris-versicolor
6.4,2.9,4.3,1.3,Iris-versicolor
6.6,3.0,4.4,1.4,Iris-versicolor
6.8,2.8,4.8,1.4,Iris-versicolor
6.7,3.0,5.0,1.7,Iris-versicolor
6.0,2.9,4.5,1.5,Iris-versicolor
5.7,2.6,3.5,1.0,Iris-versicolor
5.5,2.4,3.8,1.1,Iris-versicolor
5.5,2.4,3.7,1.0,Iris-versicolor
5.8,2.7,3.9,1.2,Iris-versicolor
6.0,2.7,5.1,1.6,Iris-versicolor
5.4,3.0,4.5,1.5,Iris-versicolor
6.0,3.4,4.5,1.6,Iris-versicolor
6.7,3.1,4.7,1.5,Iris-versicolor
6.3,2.3,4.4,1.3,Iris-versicolor
5.6,3.0,4.1,1.3,Iris-versicolor
5.5,2.5,4.0,1.3,Iris-versicolor
5.5,2.6,4.4,1.2,Iris-versicolor
6.1,3.0,4.6,1.4,Iris-versicolor
5.8,2.6,4.0,1.2,Iris-versicolor
5.0,2.3,3.3,1.0,Iris-versicolor
5.6,2.7,4.2,1.3,Iris-versicolor
5.7,3.0,4.2,1.2,Iris-versicolor
5.7,2.9,4.2,1.3,Iris-versicolor
6.2,2.9,4.3,1.3,Iris-versicolor
5.1,2.5,3.0,1.1,Iris-versicolor
5.7,2.8,4.1,1.3,Iris-versicolor
6.3,3.3,6.0,2.5,Iris-virginica
5.8,2.7,5.1,1.9,Iris-virginica
7.1,3.0,5.9,2.1,Iris-virginica
6.3,2.9,5.6,1.8,Iris-virginica
6.5,3.0,5.8,2.2,Iris-virginica
7.6,3.0,6.6,2.1,Iris-virginica
4.9,2.5,4.5,1.7,Iris-virginica
7.3,2.9,6.3,1.8,Iris-virginica
6.7,2.5,5.8,1.8,Iris-virginica
7.2,3.6,6.1,2.5,Iris-virginica
6.5,3.2,5.1,2.0,Iris-virginica
6.4,2.7,5.3,1.9,Iris-virginica
6.8,3.0,5.5,2.1,Iris-virginica
5.7,2.5,5.0,2.0,Iris-virginica
5.8,2.8,5.1,2.4,Iris-virginica
6.4,3.2,5.3,2.3,Iris-virginica
6.5,3.0,5.5,1.8,Iris-virginica
7.7,3.8,6.7,2.2,Iris-virginica
7.7,2.6,6.9,2.3,Iris-virginica
6.0,2.2,5.0,1.5,Iris-virginica
6.9,3.2,5.7,2.3,Iris-virginica
5.6,2.8,4.9,2.0,Iris-virginica
7.7,2.8,6.7,2.0,Iris-virginica
6.3,2.7,4.9,1.8,Iris-virginica
6.7,3.3,5.7,2.1,Iris-virginica
7.2,3.2,6.0,1.8,Iris-virginica
6.2,2.8,4.8,1.8,Iris-virginica
6.1,3.0,4.9,1.8,Iris-virginica
6.4,2.8,5.6,2.1,Iris-virginica
7.2,3.0,5.8,1.6,Iris-virginica
7.4,2.8,6.1,1.9,Iris-virginica
7.9,3.8,6.4,2.0,Iris-virginica
6.4,2.8,5.6,2.2,Iris-virginica
6.3,2.8,5.1,1.5,Iris-virginica
6.1,2.6,5.6,1.4,Iris-virginica
7.7,3.0,6.1,2.3,Iris-virginica
6.3,3.4,5.6,2.4,Iris-virginica
6.4,3.1,5.5,1.8,Iris-virginica
6.0,3.0,4.8,1.8,Iris-virginica
6.9,3.1,5.4,2.1,Iris-virginica
6.7,3.1,5.6,2.4,Iris-virginica
6.9,3.1,5.1,2.3,Iris-virginica
5.8,2.7,5.1,1.9,Iris-virginica
6.8,3.2,5.9,2.3,Iris-virginica
6.7,3.3,5.7,2.5,Iris-virginica
6.7,3.0,5.2,2.3,Iris-virginica
6.3,2.5,5.0,1.9,Iris-virginica
6.5,3.0,5.2,2.0,Iris-virginica
6.2,3.4,5.4,2.3,Iris-virginica
5.9,3.0,5.1,1.8,Iris-virginica
1 5.1 3.5 1.4 0.2 Iris-setosa
2 4.9 3.0 1.4 0.2 Iris-setosa
3 4.7 3.2 1.3 0.2 Iris-setosa
4 4.6 3.1 1.5 0.2 Iris-setosa
5 5.0 3.6 1.4 0.2 Iris-setosa
6 5.4 3.9 1.7 0.4 Iris-setosa
7 4.6 3.4 1.4 0.3 Iris-setosa
8 5.0 3.4 1.5 0.2 Iris-setosa
9 4.4 2.9 1.4 0.2 Iris-setosa
10 4.9 3.1 1.5 0.1 Iris-setosa
11 5.4 3.7 1.5 0.2 Iris-setosa
12 4.8 3.4 1.6 0.2 Iris-setosa
13 4.8 3.0 1.4 0.1 Iris-setosa
14 4.3 3.0 1.1 0.1 Iris-setosa
15 5.8 4.0 1.2 0.2 Iris-setosa
16 5.7 4.4 1.5 0.4 Iris-setosa
17 5.4 3.9 1.3 0.4 Iris-setosa
18 5.1 3.5 1.4 0.3 Iris-setosa
19 5.7 3.8 1.7 0.3 Iris-setosa
20 5.1 3.8 1.5 0.3 Iris-setosa
21 5.4 3.4 1.7 0.2 Iris-setosa
22 5.1 3.7 1.5 0.4 Iris-setosa
23 4.6 3.6 1.0 0.2 Iris-setosa
24 5.1 3.3 1.7 0.5 Iris-setosa
25 4.8 3.4 1.9 0.2 Iris-setosa
26 5.0 3.0 1.6 0.2 Iris-setosa
27 5.0 3.4 1.6 0.4 Iris-setosa
28 5.2 3.5 1.5 0.2 Iris-setosa
29 5.2 3.4 1.4 0.2 Iris-setosa
30 4.7 3.2 1.6 0.2 Iris-setosa
31 4.8 3.1 1.6 0.2 Iris-setosa
32 5.4 3.4 1.5 0.4 Iris-setosa
33 5.2 4.1 1.5 0.1 Iris-setosa
34 5.5 4.2 1.4 0.2 Iris-setosa
35 4.9 3.1 1.5 0.1 Iris-setosa
36 5.0 3.2 1.2 0.2 Iris-setosa
37 5.5 3.5 1.3 0.2 Iris-setosa
38 4.9 3.1 1.5 0.1 Iris-setosa
39 4.4 3.0 1.3 0.2 Iris-setosa
40 5.1 3.4 1.5 0.2 Iris-setosa
41 5.0 3.5 1.3 0.3 Iris-setosa
42 4.5 2.3 1.3 0.3 Iris-setosa
43 4.4 3.2 1.3 0.2 Iris-setosa
44 5.0 3.5 1.6 0.6 Iris-setosa
45 5.1 3.8 1.9 0.4 Iris-setosa
46 4.8 3.0 1.4 0.3 Iris-setosa
47 5.1 3.8 1.6 0.2 Iris-setosa
48 4.6 3.2 1.4 0.2 Iris-setosa
49 5.3 3.7 1.5 0.2 Iris-setosa
50 5.0 3.3 1.4 0.2 Iris-setosa
51 7.0 3.2 4.7 1.4 Iris-versicolor
52 6.4 3.2 4.5 1.5 Iris-versicolor
53 6.9 3.1 4.9 1.5 Iris-versicolor
54 5.5 2.3 4.0 1.3 Iris-versicolor
55 6.5 2.8 4.6 1.5 Iris-versicolor
56 5.7 2.8 4.5 1.3 Iris-versicolor
57 6.3 3.3 4.7 1.6 Iris-versicolor
58 4.9 2.4 3.3 1.0 Iris-versicolor
59 6.6 2.9 4.6 1.3 Iris-versicolor
60 5.2 2.7 3.9 1.4 Iris-versicolor
61 5.0 2.0 3.5 1.0 Iris-versicolor
62 5.9 3.0 4.2 1.5 Iris-versicolor
63 6.0 2.2 4.0 1.0 Iris-versicolor
64 6.1 2.9 4.7 1.4 Iris-versicolor
65 5.6 2.9 3.6 1.3 Iris-versicolor
66 6.7 3.1 4.4 1.4 Iris-versicolor
67 5.6 3.0 4.5 1.5 Iris-versicolor
68 5.8 2.7 4.1 1.0 Iris-versicolor
69 6.2 2.2 4.5 1.5 Iris-versicolor
70 5.6 2.5 3.9 1.1 Iris-versicolor
71 5.9 3.2 4.8 1.8 Iris-versicolor
72 6.1 2.8 4.0 1.3 Iris-versicolor
73 6.3 2.5 4.9 1.5 Iris-versicolor
74 6.1 2.8 4.7 1.2 Iris-versicolor
75 6.4 2.9 4.3 1.3 Iris-versicolor
76 6.6 3.0 4.4 1.4 Iris-versicolor
77 6.8 2.8 4.8 1.4 Iris-versicolor
78 6.7 3.0 5.0 1.7 Iris-versicolor
79 6.0 2.9 4.5 1.5 Iris-versicolor
80 5.7 2.6 3.5 1.0 Iris-versicolor
81 5.5 2.4 3.8 1.1 Iris-versicolor
82 5.5 2.4 3.7 1.0 Iris-versicolor
83 5.8 2.7 3.9 1.2 Iris-versicolor
84 6.0 2.7 5.1 1.6 Iris-versicolor
85 5.4 3.0 4.5 1.5 Iris-versicolor
86 6.0 3.4 4.5 1.6 Iris-versicolor
87 6.7 3.1 4.7 1.5 Iris-versicolor
88 6.3 2.3 4.4 1.3 Iris-versicolor
89 5.6 3.0 4.1 1.3 Iris-versicolor
90 5.5 2.5 4.0 1.3 Iris-versicolor
91 5.5 2.6 4.4 1.2 Iris-versicolor
92 6.1 3.0 4.6 1.4 Iris-versicolor
93 5.8 2.6 4.0 1.2 Iris-versicolor
94 5.0 2.3 3.3 1.0 Iris-versicolor
95 5.6 2.7 4.2 1.3 Iris-versicolor
96 5.7 3.0 4.2 1.2 Iris-versicolor
97 5.7 2.9 4.2 1.3 Iris-versicolor
98 6.2 2.9 4.3 1.3 Iris-versicolor
99 5.1 2.5 3.0 1.1 Iris-versicolor
100 5.7 2.8 4.1 1.3 Iris-versicolor
101 6.3 3.3 6.0 2.5 Iris-virginica
102 5.8 2.7 5.1 1.9 Iris-virginica
103 7.1 3.0 5.9 2.1 Iris-virginica
104 6.3 2.9 5.6 1.8 Iris-virginica
105 6.5 3.0 5.8 2.2 Iris-virginica
106 7.6 3.0 6.6 2.1 Iris-virginica
107 4.9 2.5 4.5 1.7 Iris-virginica
108 7.3 2.9 6.3 1.8 Iris-virginica
109 6.7 2.5 5.8 1.8 Iris-virginica
110 7.2 3.6 6.1 2.5 Iris-virginica
111 6.5 3.2 5.1 2.0 Iris-virginica
112 6.4 2.7 5.3 1.9 Iris-virginica
113 6.8 3.0 5.5 2.1 Iris-virginica
114 5.7 2.5 5.0 2.0 Iris-virginica
115 5.8 2.8 5.1 2.4 Iris-virginica
116 6.4 3.2 5.3 2.3 Iris-virginica
117 6.5 3.0 5.5 1.8 Iris-virginica
118 7.7 3.8 6.7 2.2 Iris-virginica
119 7.7 2.6 6.9 2.3 Iris-virginica
120 6.0 2.2 5.0 1.5 Iris-virginica
121 6.9 3.2 5.7 2.3 Iris-virginica
122 5.6 2.8 4.9 2.0 Iris-virginica
123 7.7 2.8 6.7 2.0 Iris-virginica
124 6.3 2.7 4.9 1.8 Iris-virginica
125 6.7 3.3 5.7 2.1 Iris-virginica
126 7.2 3.2 6.0 1.8 Iris-virginica
127 6.2 2.8 4.8 1.8 Iris-virginica
128 6.1 3.0 4.9 1.8 Iris-virginica
129 6.4 2.8 5.6 2.1 Iris-virginica
130 7.2 3.0 5.8 1.6 Iris-virginica
131 7.4 2.8 6.1 1.9 Iris-virginica
132 7.9 3.8 6.4 2.0 Iris-virginica
133 6.4 2.8 5.6 2.2 Iris-virginica
134 6.3 2.8 5.1 1.5 Iris-virginica
135 6.1 2.6 5.6 1.4 Iris-virginica
136 7.7 3.0 6.1 2.3 Iris-virginica
137 6.3 3.4 5.6 2.4 Iris-virginica
138 6.4 3.1 5.5 1.8 Iris-virginica
139 6.0 3.0 4.8 1.8 Iris-virginica
140 6.9 3.1 5.4 2.1 Iris-virginica
141 6.7 3.1 5.6 2.4 Iris-virginica
142 6.9 3.1 5.1 2.3 Iris-virginica
143 5.8 2.7 5.1 1.9 Iris-virginica
144 6.8 3.2 5.9 2.3 Iris-virginica
145 6.7 3.3 5.7 2.5 Iris-virginica
146 6.7 3.0 5.2 2.3 Iris-virginica
147 6.3 2.5 5.0 1.9 Iris-virginica
148 6.5 3.0 5.2 2.0 Iris-virginica
149 6.2 3.4 5.4 2.3 Iris-virginica
150 5.9 3.0 5.1 1.8 Iris-virginica

94
01.getting-started/05.train-in-spark/train-spark.py Normal file
View File

@@ -0,0 +1,94 @@
# Copyright (c) Microsoft. All rights reserved.
# Licensed under the MIT license.
import numpy as np
import pyspark
import os
import urllib
import sys
from pyspark.sql.functions import *
from pyspark.ml.classification import *
from pyspark.ml.evaluation import *
from pyspark.ml.feature import *
from pyspark.sql.types import StructType, StructField
from pyspark.sql.types import DoubleType, IntegerType, StringType
from azureml.core.run import Run
# initialize logger
run = Run.get_context()
# start Spark session
spark = pyspark.sql.SparkSession.builder.appName('Iris').getOrCreate()
# print runtime versions
print('****************')
print('Python version: {}'.format(sys.version))
print('Spark version: {}'.format(spark.version))
print('****************')
# load iris.csv into Spark dataframe
schema = StructType([
StructField("sepal-length", DoubleType()),
StructField("sepal-width", DoubleType()),
StructField("petal-length", DoubleType()),
StructField("petal-width", DoubleType()),
StructField("class", StringType())
])
data = spark.read.csv('iris.csv', header=False, schema=schema)
print("First 10 rows of Iris dataset:")
data.show(10)
# vectorize all numerical columns into a single feature column
feature_cols = data.columns[:-1]
assembler = pyspark.ml.feature.VectorAssembler(
inputCols=feature_cols, outputCol='features')
data = assembler.transform(data)
# convert text labels into indices
data = data.select(['features', 'class'])
label_indexer = pyspark.ml.feature.StringIndexer(
inputCol='class', outputCol='label').fit(data)
data = label_indexer.transform(data)
# only select the features and label column
data = data.select(['features', 'label'])
print("Reading for machine learning")
data.show(10)
# change regularization rate and you will likely get a different accuracy.
reg = 0.01
# load regularization rate from argument if present
if len(sys.argv) > 1:
reg = float(sys.argv[1])
# log regularization rate
run.log("Regularization Rate", reg)
# use Logistic Regression to train on the training set
train, test = data.randomSplit([0.70, 0.30])
lr = pyspark.ml.classification.LogisticRegression(regParam=reg)
model = lr.fit(train)
# predict on the test set
prediction = model.transform(test)
print("Prediction")
prediction.show(10)
# evaluate the accuracy of the model using the test set
evaluator = pyspark.ml.evaluation.MulticlassClassificationEvaluator(
metricName='accuracy')
accuracy = evaluator.evaluate(prediction)
print()
print('#####################################')
print('Regularization rate is {}'.format(reg))
print("Accuracy is {}".format(accuracy))
print('#####################################')
print()
# log accuracy
run.log('Accuracy', accuracy)

328
01.getting-started/06.logging-api/06.logging-api.ipynb Normal file
View File

@@ -0,0 +1,328 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# 06. Logging APIs\n",
"This notebook showcase various ways to use the Azure Machine Learning service run logging APIs, and view the results in the Azure portal."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"Make sure you go through the [00. Installation and Configuration](../../00.configuration.ipynb) Notebook first if you haven't. Also make sure you have tqdm and matplotlib installed in the current kernel.\n",
"\n",
"```\n",
"(myenv) $ conda install -y tqdm matplotlib\n",
"```"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Validate Azure ML SDK installation and get version number for debugging purposes"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"install"
]
},
"outputs": [],
"source": [
"from azureml.core import Experiment, Run, Workspace\n",
"import azureml.core\n",
"import numpy as np\n",
"\n",
"# Check core SDK version number\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize Workspace\n",
"\n",
"Initialize a workspace object from persisted configuration."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"create workspace"
]
},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"print('Workspace name: ' + ws.name, \n",
" 'Azure region: ' + ws.location, \n",
" 'Subscription id: ' + ws.subscription_id, \n",
" 'Resource group: ' + ws.resource_group, sep='\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Set experiment\n",
"Create a new experiment (or get the one with such name)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"exp = Experiment(workspace=ws, name='logging-api-test')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Log metrics\n",
"We will start a run, and use the various logging APIs to record different types of metrics during the run."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from tqdm import tqdm\n",
"\n",
"# start logging for the run\n",
"run = exp.start_logging()\n",
"\n",
"# log a string value\n",
"run.log(name='Name', value='Logging API run')\n",
"\n",
"# log a numerical value\n",
"run.log(name='Magic Number', value=42)\n",
"\n",
"# Log a list of values. Note this will generate a single-variable line chart.\n",
"run.log_list(name='Fibonacci', value=[0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89])\n",
"\n",
"# create a dictionary to hold a table of values\n",
"sines = {}\n",
"sines['angle'] = []\n",
"sines['sine'] = []\n",
"\n",
"for i in tqdm(range(-10, 10)):\n",
" # log a metric value repeatedly, this will generate a single-variable line chart.\n",
" run.log(name='Sigmoid', value=1 / (1 + np.exp(-i)))\n",
" angle = i / 2.0\n",
" \n",
" # log a 2 (or more) values as a metric repeatedly. This will generate a 2-variable line chart if you have 2 numerical columns.\n",
" run.log_row(name='Cosine Wave', angle=angle, cos=np.cos(angle))\n",
" \n",
" sines['angle'].append(angle)\n",
" sines['sine'].append(np.sin(angle))\n",
"\n",
"# log a dictionary as a table, this will generate a 2-variable chart if you have 2 numerical columns\n",
"run.log_table(name='Sine Wave', value=sines)\n",
"\n",
"run.complete()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Even after the run is marked completed, you can still log things."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Log an image\n",
"This is how to log a _matplotlib_ pyplot object."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%matplotlib inline\n",
"import matplotlib.pyplot as plt\n",
"angle = np.linspace(-3, 3, 50)\n",
"plt.plot(angle, np.tanh(angle), label='tanh')\n",
"plt.legend(fontsize=12)\n",
"plt.title('Hyperbolic Tangent', fontsize=16)\n",
"plt.grid(True)\n",
"\n",
"run.log_image(name='Hyperbolic Tangent', plot=plt)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Upload a file"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can also upload an abitrary file. First, let's create a dummy file locally."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile myfile.txt\n",
"\n",
"This is a dummy file."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now let's upload this file into the run record as a run artifact, and display the properties after the upload."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"props = run.upload_file(name='myfile_in_the_cloud.txt', path_or_stream='./myfile.txt')\n",
"props.serialize()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Examine the run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now let's take a look at the run detail page in Azure portal. Make sure you checkout the various charts and plots generated/uploaded."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can get all the metrics in that run back."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.get_metrics()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can also see the files uploaded for this run."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.get_file_names()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can also download all the files locally."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"os.makedirs('files', exist_ok=True)\n",
"\n",
"for f in run.get_file_names():\n",
" dest = os.path.join('files', f.split('/')[-1])\n",
" print('Downloading file {} to {}...'.format(f, dest))\n",
" run.download_file(f, dest) "
]
}
],
"metadata": {
"authors": [
{
"name": "haining"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

420
01.getting-started/10.register-model-create-image-deploy-service/10.register-model-create-image-deploy-service.ipynb Normal file
View File

@@ -0,0 +1,420 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 10. Register Model, Create Image and Deploy Service\n",
"\n",
"This example shows how to deploy a web service in step-by-step fashion:\n",
"\n",
" 1. Register model\n",
" 2. Query versions of models and select one to deploy\n",
" 3. Create Docker image\n",
" 4. Query versions of images\n",
" 5. Deploy the image as web service\n",
" \n",
"**IMPORTANT**:\n",
" * This notebook requires you to first complete \"01.SDK-101-Train-and-Deploy-to-ACI.ipynb\" Notebook\n",
" \n",
"The 101 Notebook taught you how to deploy a web service directly from model in one step. This Notebook shows a more advanced approach that gives you more control over model versions and Docker image versions. "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"Make sure you go through the [00. Installation and Configuration](00.configuration.ipynb) Notebook first if you haven't."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize Workspace\n",
"\n",
"Initialize a workspace object from persisted configuration."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"create workspace"
]
},
"outputs": [],
"source": [
"from azureml.core import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print(ws.name, ws.resource_group, ws.location, ws.subscription_id, sep = '\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Register Model"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can add tags and descriptions to your models. Note you need to have a `sklearn_linreg_model.pkl` file in the current directory. This file is generated by the 01 notebook. The below call registers that file as a model with the same name `sklearn_linreg_model.pkl` in the workspace.\n",
"\n",
"Using tags, you can track useful information such as the name and version of the machine learning library used to train the model. Note that tags must be alphanumeric."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"register model from file"
]
},
"outputs": [],
"source": [
"from azureml.core.model import Model\n",
"import sklearn\n",
"\n",
"library_version = \"sklearn\"+sklearn.__version__.replace(\".\",\"x\")\n",
"\n",
"model = Model.register(model_path = \"sklearn_regression_model.pkl\",\n",
" model_name = \"sklearn_regression_model.pkl\",\n",
" tags = {'area': \"diabetes\", 'type': \"regression\", 'version': library_version},\n",
" description = \"Ridge regression model to predict diabetes\",\n",
" workspace = ws)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can explore the registered models within your workspace and query by tag. Models are versioned. If you call the register_model command many times with same model name, you will get multiple versions of the model with increasing version numbers."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"register model from file"
]
},
"outputs": [],
"source": [
"regression_models = Model.list(workspace=ws, tags=['area'])\n",
"for m in regression_models:\n",
" print(\"Name:\", m.name,\"\\tVersion:\", m.version, \"\\tDescription:\", m.description, m.tags)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can pick a specific model to deploy"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(model.name, model.description, model.version, sep = '\\t')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create Docker Image"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Show `score.py`. Note that the `sklearn_regression_model.pkl` in the `get_model_path` call is referring to a model named `sklearn_linreg_model.pkl` registered under the workspace. It is NOT referenceing the local file."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile score.py\n",
"import pickle\n",
"import json\n",
"import numpy\n",
"from sklearn.externals import joblib\n",
"from sklearn.linear_model import Ridge\n",
"from azureml.core.model import Model\n",
"\n",
"def init():\n",
" global model\n",
" # note here \"sklearn_regression_model.pkl\" is the name of the model registered under\n",
" # this is a different behavior than before when the code is run locally, even though the code is the same.\n",
" model_path = Model.get_model_path('sklearn_regression_model.pkl')\n",
" # deserialize the model file back into a sklearn model\n",
" model = joblib.load(model_path)\n",
"\n",
"# note you can pass in multiple rows for scoring\n",
"def run(raw_data):\n",
" try:\n",
" data = json.loads(raw_data)['data']\n",
" data = numpy.array(data)\n",
" result = model.predict(data)\n",
" # you can return any datatype as long as it is JSON-serializable\n",
" return result.tolist()\n",
" except Exception as e:\n",
" error = str(e)\n",
" return error"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.conda_dependencies import CondaDependencies \n",
"\n",
"myenv = CondaDependencies.create(conda_packages=['numpy','scikit-learn'])\n",
"\n",
"with open(\"myenv.yml\",\"w\") as f:\n",
" f.write(myenv.serialize_to_string())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Note that following command can take few minutes. \n",
"\n",
"You can add tags and descriptions to images. Also, an image can contain multiple models."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"create image"
]
},
"outputs": [],
"source": [
"from azureml.core.image import Image, ContainerImage\n",
"\n",
"image_config = ContainerImage.image_configuration(runtime= \"python\",\n",
" execution_script=\"score.py\",\n",
" conda_file=\"myenv.yml\",\n",
" tags = {'area': \"diabetes\", 'type': \"regression\"},\n",
" description = \"Image with ridge regression model\")\n",
"\n",
"image = Image.create(name = \"myimage1\",\n",
" # this is the model object \n",
" models = [model],\n",
" image_config = image_config, \n",
" workspace = ws)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"create image"
]
},
"outputs": [],
"source": [
"image.wait_for_creation(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"List images by tag and find out the detailed build log for debugging."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"create image"
]
},
"outputs": [],
"source": [
"for i in Image.list(workspace = ws,tags = [\"area\"]):\n",
" print('{}(v.{} [{}]) stored at {} with build log {}'.format(i.name, i.version, i.creation_state, i.image_location, i.image_build_log_uri))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Deploy image as web service on Azure Container Instance\n",
"\n",
"Note that the service creation can take few minutes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"from azureml.core.webservice import AciWebservice\n",
"\n",
"aciconfig = AciWebservice.deploy_configuration(cpu_cores = 1, \n",
" memory_gb = 1, \n",
" tags = {'area': \"diabetes\", 'type': \"regression\"}, \n",
" description = 'Predict diabetes using regression model')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"from azureml.core.webservice import Webservice\n",
"\n",
"aci_service_name = 'my-aci-service-2'\n",
"print(aci_service_name)\n",
"aci_service = Webservice.deploy_from_image(deployment_config = aciconfig,\n",
" image = image,\n",
" name = aci_service_name,\n",
" workspace = ws)\n",
"aci_service.wait_for_deployment(True)\n",
"print(aci_service.state)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test web service"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Call the web service with some dummy input data to get a prediction."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"import json\n",
"\n",
"test_sample = json.dumps({'data': [\n",
" [1,2,3,4,5,6,7,8,9,10], \n",
" [10,9,8,7,6,5,4,3,2,1]\n",
"]})\n",
"test_sample = bytes(test_sample,encoding = 'utf8')\n",
"\n",
"prediction = aci_service.run(input_data=test_sample)\n",
"print(prediction)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Delete ACI to clean up"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"aci_service.delete()"
]
}
],
"metadata": {
"authors": [
{
"name": "raymondl"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

0
getting-started/enable-app-insights-in-production-service/sklearn_regression_model.pkl → 01.getting-started/10.register-model-create-image-deploy-service/sklearn_regression_model.pkl
View File

342
01.getting-started/11.production-deploy-to-aks/11.production-deploy-to-aks.ipynb Normal file
View File

@@ -0,0 +1,342 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Deploying a web service to Azure Kubernetes Service (AKS)\n",
"This notebook shows the steps for deploying a service: registering a model, creating an image, provisioning a cluster (one time action), and deploying a service to it. \n",
"We then test and delete the service, image and model."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Workspace\n",
"from azureml.core.compute import AksCompute, ComputeTarget\n",
"from azureml.core.webservice import Webservice, AksWebservice\n",
"from azureml.core.image import Image\n",
"from azureml.core.model import Model"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import azureml.core\n",
"print(azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Get workspace\n",
"Load existing workspace from the config file info."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.workspace import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print(ws.name, ws.resource_group, ws.location, ws.subscription_id, sep = '\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Register the model\n",
"Register an existing trained model, add descirption and tags."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#Register the model\n",
"from azureml.core.model import Model\n",
"model = Model.register(model_path = \"sklearn_regression_model.pkl\", # this points to a local file\n",
" model_name = \"sklearn_regression_model.pkl\", # this is the name the model is registered as\n",
" tags = {'area': \"diabetes\", 'type': \"regression\"},\n",
" description = \"Ridge regression model to predict diabetes\",\n",
" workspace = ws)\n",
"\n",
"print(model.name, model.description, model.version)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Create an image\n",
"Create an image using the registered model the script that will load and run the model."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile score.py\n",
"import pickle\n",
"import json\n",
"import numpy\n",
"from sklearn.externals import joblib\n",
"from sklearn.linear_model import Ridge\n",
"from azureml.core.model import Model\n",
"\n",
"def init():\n",
" global model\n",
" # note here \"sklearn_regression_model.pkl\" is the name of the model registered under\n",
" # this is a different behavior than before when the code is run locally, even though the code is the same.\n",
" model_path = Model.get_model_path('sklearn_regression_model.pkl')\n",
" # deserialize the model file back into a sklearn model\n",
" model = joblib.load(model_path)\n",
"\n",
"# note you can pass in multiple rows for scoring\n",
"def run(raw_data):\n",
" try:\n",
" data = json.loads(raw_data)['data']\n",
" data = numpy.array(data)\n",
" result = model.predict(data)\n",
" # you can return any data type as long as it is JSON-serializable\n",
" return result.tolist()\n",
" except Exception as e:\n",
" error = str(e)\n",
" return error"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.conda_dependencies import CondaDependencies \n",
"\n",
"myenv = CondaDependencies.create(conda_packages=['numpy','scikit-learn'])\n",
"\n",
"with open(\"myenv.yml\",\"w\") as f:\n",
" f.write(myenv.serialize_to_string())"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.image import ContainerImage\n",
"\n",
"image_config = ContainerImage.image_configuration(execution_script = \"score.py\",\n",
" runtime = \"python\",\n",
" conda_file = \"myenv.yml\",\n",
" description = \"Image with ridge regression model\",\n",
" tags = {'area': \"diabetes\", 'type': \"regression\"}\n",
" )\n",
"\n",
"image = ContainerImage.create(name = \"myimage1\",\n",
" # this is the model object\n",
" models = [model],\n",
" image_config = image_config,\n",
" workspace = ws)\n",
"\n",
"image.wait_for_creation(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Provision the AKS Cluster\n",
"This is a one time setup. You can reuse this cluster for multiple deployments after it has been created. If you delete the cluster or the resource group that contains it, then you would have to recreate it."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Use the default configuration (can also provide parameters to customize)\n",
"prov_config = AksCompute.provisioning_configuration()\n",
"\n",
"aks_name = 'my-aks-9' \n",
"# Create the cluster\n",
"aks_target = ComputeTarget.create(workspace = ws, \n",
" name = aks_name, \n",
" provisioning_configuration = prov_config)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%time\n",
"aks_target.wait_for_completion(show_output = True)\n",
"print(aks_target.provisioning_state)\n",
"print(aks_target.provisioning_errors)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Optional step: Attach existing AKS cluster\n",
"\n",
"If you have existing AKS cluster in your Azure subscription, you can attach it to the Workspace."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"'''\n",
"# Use the default configuration (can also provide parameters to customize)\n",
"resource_id = '/subscriptions/92c76a2f-0e1c-4216-b65e-abf7a3f34c1e/resourcegroups/raymondsdk0604/providers/Microsoft.ContainerService/managedClusters/my-aks-0605d37425356b7d01'\n",
"\n",
"create_name='my-existing-aks' \n",
"# Create the cluster\n",
"aks_target = AksCompute.attach(workspace=ws, name=create_name, resource_id=resource_id)\n",
"# Wait for the operation to complete\n",
"aks_target.wait_for_completion(True)\n",
"'''"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Deploy web service to AKS"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#Set the web service configuration (using default here)\n",
"aks_config = AksWebservice.deploy_configuration()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%time\n",
"aks_service_name ='aks-service-1'\n",
"\n",
"aks_service = Webservice.deploy_from_image(workspace = ws, \n",
" name = aks_service_name,\n",
" image = image,\n",
" deployment_config = aks_config,\n",
" deployment_target = aks_target)\n",
"aks_service.wait_for_deployment(show_output = True)\n",
"print(aks_service.state)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Test the web service\n",
"We test the web sevice by passing data."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%time\n",
"import json\n",
"\n",
"test_sample = json.dumps({'data': [\n",
" [1,2,3,4,5,6,7,8,9,10], \n",
" [10,9,8,7,6,5,4,3,2,1]\n",
"]})\n",
"test_sample = bytes(test_sample,encoding = 'utf8')\n",
"\n",
"prediction = aks_service.run(input_data = test_sample)\n",
"print(prediction)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Clean up\n",
"Delete the service, image and model."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%time\n",
"aks_service.delete()\n",
"image.delete()\n",
"model.delete()"
]
}
],
"metadata": {
"authors": [
{
"name": "raymondl"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

0
getting-started/enable-data-collection-for-models-in-aks/sklearn_regression_model.pkl → 01.getting-started/11.production-deploy-to-aks/sklearn_regression_model.pkl
View File

5
getting-started/enable-data-collection-for-models-in-aks/enable-data-collection-for-models-in-aks.ipynb → 01.getting-started/12.enable-data-collection-for-models-in-aks/12.enable-data-collection-for-models-in-aks.ipynb
View File

@@ -285,10 +285,9 @@
" %%time\n",
" resource_id = '/subscriptions/<subscriptionid>/resourcegroups/<resourcegroupname>/providers/Microsoft.ContainerService/managedClusters/<aksservername>'\n",
" create_name= 'myaks4'\n",
" attach_config = AksCompute.attach_configuration(resource_id=resource_id)\n",
" aks_target = ComputeTarget.attach(workspace = ws, \n",
" aks_target = AksCompute.attach(workspace = ws, \n",
" name = create_name, \n",
" attach_configuration=attach_config)\n",
" resource_id=resource_id)\n",
" ## Wait for the operation to complete\n",
" aks_target.wait_for_provisioning(True)```"
]

0
getting-started/production-deploy-to-aks/sklearn_regression_model.pkl → 01.getting-started/12.enable-data-collection-for-models-in-aks/sklearn_regression_model.pkl
View File

414
01.getting-started/13.enable-app-insights/13.enable-app-insights-in-production-service.ipynb Normal file
View File

@@ -0,0 +1,414 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Enabling App Insights for Services in Production\n",
"With this notebook, you can learn how to enable App Insights for standard service monitoring, plus, we provide examples for doing custom logging within a scoring files in a model. \n",
"\n",
"\n",
"## What does Application Insights monitor?\n",
"It monitors request rates, response times, failure rates, etc. For more information visit [App Insights docs.](https://docs.microsoft.com/en-us/azure/application-insights/app-insights-overview)\n",
"\n",
"\n",
"## What is different compared to standard production deployment process?\n",
"If you want to enable generic App Insights for a service run:\n",
"```python\n",
"aks_service= Webservice(ws, \"aks-w-dc2\")\n",
"aks_service.update(enable_app_insights=True)```\n",
"Where \"aks-w-dc2\" is your service name. You can also do this from the Azure Portal under your Workspace--> deployments--> Select deployment--> Edit--> Advanced Settings--> Select \"Enable AppInsights diagnostics\"\n",
"\n",
"If you want to log custom traces, you will follow the standard deplyment process for AKS and you will:\n",
"1. Update scoring file.\n",
"2. Update aks configuration.\n",
"3. Build new image and deploy it. "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 1. Import your dependencies"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Workspace, Run\n",
"from azureml.core.compute import AksCompute, ComputeTarget\n",
"from azureml.core.webservice import Webservice, AksWebservice\n",
"from azureml.core.image import Image\n",
"from azureml.core.model import Model\n",
"\n",
"import azureml.core\n",
"print(azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 2. Set up your configuration and create a workspace\n",
"Follow Notebook 00 instructions to do this.\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"print(ws.name, ws.resource_group, ws.location, ws.subscription_id, sep = '\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 3. Register Model\n",
"Register an existing trained model, add descirption and tags."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#Register the model\n",
"from azureml.core.model import Model\n",
"model = Model.register(model_path = \"sklearn_regression_model.pkl\", # this points to a local file\n",
" model_name = \"sklearn_regression_model.pkl\", # this is the name the model is registered as\n",
" tags = {'area': \"diabetes\", 'type': \"regression\"},\n",
" description = \"Ridge regression model to predict diabetes\",\n",
" workspace = ws)\n",
"\n",
"print(model.name, model.description, model.version)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 4. *Update your scoring file with custom print statements*\n",
"Here is an example:\n",
"### a. In your init function add:\n",
"```python\n",
"print (\"model initialized\" + time.strftime(\"%H:%M:%S\"))```\n",
"\n",
"### b. In your run function add:\n",
"```python\n",
"print (\"saving input data\" + time.strftime(\"%H:%M:%S\"))\n",
"print (\"saving prediction data\" + time.strftime(\"%H:%M:%S\"))```"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile score.py\n",
"import pickle\n",
"import json\n",
"import numpy \n",
"from sklearn.externals import joblib\n",
"from sklearn.linear_model import Ridge\n",
"from azureml.core.model import Model\n",
"from azureml.monitoring import ModelDataCollector\n",
"import time\n",
"\n",
"def init():\n",
" global model\n",
" #Print statement for appinsights custom traces:\n",
" print (\"model initialized\" + time.strftime(\"%H:%M:%S\"))\n",
" \n",
" # note here \"sklearn_regression_model.pkl\" is the name of the model registered under the workspace\n",
" # this call should return the path to the model.pkl file on the local disk.\n",
" model_path = Model.get_model_path(model_name = 'sklearn_regression_model.pkl')\n",
" \n",
" # deserialize the model file back into a sklearn model\n",
" model = joblib.load(model_path)\n",
" \n",
" global inputs_dc, prediction_dc\n",
" \n",
" # this setup will help us save our inputs under the \"inputs\" path in our Azure Blob\n",
" inputs_dc = ModelDataCollector(model_name=\"sklearn_regression_model\", identifier=\"inputs\", feature_names=[\"feat1\", \"feat2\"]) \n",
" \n",
" # this setup will help us save our ipredictions under the \"predictions\" path in our Azure Blob\n",
" prediction_dc = ModelDataCollector(\"sklearn_regression_model\", identifier=\"predictions\", feature_names=[\"prediction1\", \"prediction2\"]) \n",
" \n",
"# note you can pass in multiple rows for scoring\n",
"def run(raw_data):\n",
" global inputs_dc, prediction_dc\n",
" try:\n",
" data = json.loads(raw_data)['data']\n",
" data = numpy.array(data)\n",
" result = model.predict(data)\n",
" \n",
" #Print statement for appinsights custom traces:\n",
" print (\"saving input data\" + time.strftime(\"%H:%M:%S\"))\n",
" \n",
" #this call is saving our input data into our blob\n",
" inputs_dc.collect(data) \n",
" #this call is saving our prediction data into our blob\n",
" prediction_dc.collect(result)\n",
" \n",
" #Print statement for appinsights custom traces:\n",
" print (\"saving prediction data\" + time.strftime(\"%H:%M:%S\"))\n",
" # you can return any data type as long as it is JSON-serializable\n",
" return result.tolist()\n",
" except Exception as e:\n",
" error = str(e)\n",
" print (error + time.strftime(\"%H:%M:%S\"))\n",
" return error"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 5. *Create myenv.yml file*"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.conda_dependencies import CondaDependencies \n",
"\n",
"myenv = CondaDependencies.create(conda_packages=['numpy','scikit-learn'])\n",
"\n",
"with open(\"myenv.yml\",\"w\") as f:\n",
" f.write(myenv.serialize_to_string())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 6. Create your new Image"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.image import ContainerImage\n",
"\n",
"image_config = ContainerImage.image_configuration(execution_script = \"score.py\",\n",
" runtime = \"python\",\n",
" conda_file = \"myenv.yml\",\n",
" description = \"Image with ridge regression model\",\n",
" tags = {'area': \"diabetes\", 'type': \"regression\"}\n",
" )\n",
"\n",
"image = ContainerImage.create(name = \"myimage1\",\n",
" # this is the model object\n",
" models = [model],\n",
" image_config = image_config,\n",
" workspace = ws)\n",
"\n",
"image.wait_for_creation(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 7. Deploy to AKS service"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create AKS compute if you haven't done so (Notebook 11)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Use the default configuration (can also provide parameters to customize)\n",
"prov_config = AksCompute.provisioning_configuration()\n",
"\n",
"aks_name = 'my-aks-test1' \n",
"# Create the cluster\n",
"aks_target = ComputeTarget.create(workspace = ws, \n",
" name = aks_name, \n",
" provisioning_configuration = prov_config)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%time\n",
"aks_target.wait_for_completion(show_output = True)\n",
"print(aks_target.provisioning_state)\n",
"print(aks_target.provisioning_errors)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"If you already have a cluster you can attach the service to it:"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"```python \n",
"%%time\n",
"resource_id = '/subscriptions/<subscriptionid>/resourcegroups/<resourcegroupname>/providers/Microsoft.ContainerService/managedClusters/<aksservername>'\n",
"create_name= 'myaks4'\n",
"aks_target = AksCompute.attach(workspace = ws, \n",
" name = create_name, \n",
" #esource_id=resource_id)\n",
"## Wait for the operation to complete\n",
"aks_target.wait_for_provisioning(True)```"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### a. *Activate App Insights through updating AKS Webservice configuration*\n",
"In order to enable App Insights in your service you will need to update your AKS configuration file:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#Set the web service configuration\n",
"aks_config = AksWebservice.deploy_configuration(enable_app_insights=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### b. Deploy your service"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%time\n",
"aks_service_name ='aks-w-dc3'\n",
"\n",
"aks_service = Webservice.deploy_from_image(workspace = ws, \n",
" name = aks_service_name,\n",
" image = image,\n",
" deployment_config = aks_config,\n",
" deployment_target = aks_target\n",
" )\n",
"aks_service.wait_for_deployment(show_output = True)\n",
"print(aks_service.state)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 8. Test your service "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%time\n",
"import json\n",
"\n",
"test_sample = json.dumps({'data': [\n",
" [1,28,13,45,54,6,57,8,8,10], \n",
" [101,9,8,37,6,45,4,3,2,41]\n",
"]})\n",
"test_sample = bytes(test_sample,encoding='utf8')\n",
"\n",
"prediction = aks_service.run(input_data=test_sample)\n",
"print(prediction)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 9. See your service telemetry in App Insights\n",
"1. Go to the [Azure Portal](https://portal.azure.com/)\n",
"2. All resources--> Select the subscription/resource group where you created your Workspace--> Select the App Insights type\n",
"3. Click on the AppInsights resource. You'll see a highlevel dashboard with information on Requests, Server response time and availability.\n",
"4. Click on the top banner \"Analytics\"\n",
"5. In the \"Schema\" section select \"traces\" and run your query.\n",
"6. Voila! All your custom traces should be there."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Disable App Insights"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"aks_service.update(enable_app_insights=False)"
]
}
],
"metadata": {
"authors": [
{
"name": "marthalc"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

224
automl/00.configuration.ipynb Normal file
View File

@@ -0,0 +1,224 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 00. Configuration\n",
"\n",
"In this example you will create an Azure Machine Learning `Workspace` object and initialize your notebook directory to easily reload this object from a configuration file. Typically you will only need to run this once per notebook directory, and all other notebooks in this directory or any sub-directories will automatically use the settings you indicate here.\n",
"\n",
"\n",
"## Prerequisites:\n",
"\n",
"Before running this notebook, run the `automl_setup` script described in README.md.\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Register Machine Learning Services Resource Provider\n",
"\n",
"Microsoft.MachineLearningServices only needs to be registed once in the subscription.\n",
"To register it:\n",
"1. Start the Azure portal.\n",
"2. Select your `All services` and then `Subscription`.\n",
"3. Select the subscription that you want to use.\n",
"4. Click on `Resource providers`\n",
"3. Click the `Register` link next to Microsoft.MachineLearningServices"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Check the Azure ML Core SDK Version to Validate Your Installation"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import azureml.core\n",
"\n",
"print(\"SDK Version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize an Azure ML Workspace\n",
"### What is an Azure ML Workspace and Why Do I Need One?\n",
"\n",
"An Azure ML workspace is an Azure resource that organizes and coordinates the actions of many other Azure resources to assist in executing and sharing machine learning workflows. In particular, an Azure ML workspace coordinates storage, databases, and compute resources providing added functionality for machine learning experimentation, operationalization, and the monitoring of operationalized models.\n",
"\n",
"\n",
"### What do I Need?\n",
"\n",
"To create or access an Azure ML workspace, you will need to import the Azure ML library and specify following information:\n",
"* A name for your workspace. You can choose one.\n",
"* Your subscription id. Use the `id` value from the `az account show` command output above.\n",
"* The resource group name. The resource group organizes Azure resources and provides a default region for the resources in the group. The resource group will be created if it doesn't exist. Resource groups can be created and viewed in the [Azure portal](https://portal.azure.com)\n",
"* Supported regions include `eastus2`, `eastus`,`westcentralus`, `southeastasia`, `westeurope`, `australiaeast`, `westus2`, `southcentralus`."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"subscription_id = \"<subscription_id>\"\n",
"resource_group = \"myrg\"\n",
"workspace_name = \"myws\"\n",
"workspace_region = \"eastus2\""
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Creating a Workspace\n",
"If you already have access to an Azure ML workspace you want to use, you can skip this cell. Otherwise, this cell will create an Azure ML workspace for you in the specified subscription, provided you have the correct permissions for the given `subscription_id`.\n",
"\n",
"This will fail when:\n",
"1. The workspace already exists.\n",
"2. You do not have permission to create a workspace in the resource group.\n",
"3. You are not a subscription owner or contributor and no Azure ML workspaces have ever been created in this subscription.\n",
"\n",
"If workspace creation fails for any reason other than already existing, please work with your IT administrator to provide you with the appropriate permissions or to provision the required resources.\n",
"\n",
"**Note:** Creation of a new workspace can take several minutes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Import the Workspace class and check the Azure ML SDK version.\n",
"from azureml.core import Workspace\n",
"\n",
"ws = Workspace.create(name = workspace_name,\n",
" subscription_id = subscription_id,\n",
" resource_group = resource_group, \n",
" location = workspace_region)\n",
"ws.get_details()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configuring Your Local Environment\n",
"You can validate that you have access to the specified workspace and write a configuration file to the default configuration location, `./aml_config/config.json`."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Workspace\n",
"\n",
"ws = Workspace(workspace_name = workspace_name,\n",
" subscription_id = subscription_id,\n",
" resource_group = resource_group)\n",
"\n",
"# Persist the subscription id, resource group name, and workspace name in aml_config/config.json.\n",
"ws.write_config()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can then load the workspace from this config file from any notebook in the current directory."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Load workspace configuration from ./aml_config/config.json file.\n",
"my_workspace = Workspace.from_config()\n",
"my_workspace.get_details()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create a Folder to Host All Sample Projects\n",
"Finally, create a folder where all the sample projects will be hosted."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"\n",
"sample_projects_folder = './sample_projects'\n",
"\n",
"if not os.path.isdir(sample_projects_folder):\n",
" os.mkdir(sample_projects_folder)\n",
" \n",
"print('Sample projects will be created in {}.'.format(sample_projects_folder))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Success!\n",
"Great, you are ready to move on to the rest of the sample notebooks."
]
}
],
"metadata": {
"authors": [
{
"name": "savitam"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

822
automl/01.auto-ml-classification.ipynb
View File

@@ -1,414 +1,414 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 01: Classification with Local Compute\n",
"\n",
"In this example we use the scikit-learn's [digit dataset](http://scikit-learn.org/stable/datasets/index.html#optical-recognition-of-handwritten-digits-dataset) to showcase how you can use AutoML for a simple classification problem.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. Create an `Experiment` in an existing `Workspace`.\n",
"2. Configure AutoML using `AutoMLConfig`.\n",
"3. Train the model using local compute.\n",
"4. Explore the results.\n",
"5. Test the best fitted model.\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# Choose a name for the experiment and specify the project folder.\n",
"experiment_name = 'automl-local-classification'\n",
"project_folder = './sample_projects/automl-local-classification'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace Name'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data = output, index = ['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Load Training Data\n",
"\n",
"This uses scikit-learn's [load_digits](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_digits.html) method."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn import datasets\n",
"\n",
"digits = datasets.load_digits()\n",
"\n",
"# Exclude the first 100 rows from training so that they can be used for test.\n",
"X_train = digits.data[100:,:]\n",
"y_train = digits.target[100:]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML\n",
"\n",
"Instantiate an `AutoMLConfig` object to specify the settings and data used to run the experiment.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**task**|classification or regression|\n",
"|**primary_metric**|This is the metric that you want to optimize. Classification supports the following primary metrics: <br><i>accuracy</i><br><i>AUC_weighted</i><br><i>average_precision_score_weighted</i><br><i>norm_macro_recall</i><br><i>precision_score_weighted</i>|\n",
"|**iteration_timeout_minutes**|Time limit in minutes for each iteration.|\n",
"|**iterations**|Number of iterations. In each iteration AutoML trains a specific pipeline with the data.|\n",
"|**n_cross_validations**|Number of cross validation splits.|\n",
"|**X**|(sparse) array-like, shape = [n_samples, n_features]|\n",
"|**y**|(sparse) array-like, shape = [n_samples, ], [n_samples, n_classes]<br>Multi-class targets. An indicator matrix turns on multilabel classification. This should be an array of integers.|\n",
"|**path**|Relative path to the project folder. AutoML stores configuration files for the experiment under this folder. You can specify a new empty folder.|"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_config = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" primary_metric = 'AUC_weighted',\n",
" iteration_timeout_minutes = 60,\n",
" iterations = 25,\n",
" n_cross_validations = 3,\n",
" verbosity = logging.INFO,\n",
" X = X_train, \n",
" y = y_train,\n",
" path = project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train the Models\n",
"\n",
"Call the `submit` method on the experiment object and pass the run configuration. Execution of local runs is synchronous. Depending on the data and the number of iterations this can run for a while.\n",
"In this example, we specify `show_output = True` to print currently running iterations to the console."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_config, show_output = True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Optionally, you can continue an interrupted local run by calling `continue_experiment` without the `iterations` parameter, or run more iterations for a completed run by specifying the `iterations` parameter:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = local_run.continue_experiment(X = X_train, \n",
" y = y_train, \n",
" show_output = True,\n",
" iterations = 5)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explore the Results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Widget for Monitoring Runs\n",
"\n",
"The widget will first report a \"loading\" status while running the first iteration. After completing the first iteration, an auto-updating graph and table will be shown. The widget will refresh once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"**Note:** The widget displays a link at the bottom. Use this link to open a web interface to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.widgets import RunDetails\n",
"RunDetails(local_run).show() "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"#### Retrieve All Child Runs\n",
"You can also use SDK methods to fetch all the child runs and see individual metrics that we log."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"children = list(local_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
"\n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"rundata"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method returns the best run and the fitted model. The Model includes the pipeline and any pre-processing. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = local_run.get_output()\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Best Model Based on Any Other Metric\n",
"Show the run and the model that has the smallest `log_loss` value:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"lookup_metric = \"log_loss\"\n",
"best_run, fitted_model = local_run.get_output(metric = lookup_metric)\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Model from a Specific Iteration\n",
"Show the run and the model from the third iteration:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"iteration = 3\n",
"third_run, third_model = local_run.get_output(iteration = iteration)\n",
"print(third_run)\n",
"print(third_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test the Best Fitted Model\n",
"\n",
"#### Load Test Data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits = datasets.load_digits()\n",
"X_test = digits.data[:10, :]\n",
"y_test = digits.target[:10]\n",
"images = digits.images[:10]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Testing Our Best Fitted Model\n",
"We will try to predict 2 digits and see how our model works."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Randomly select digits and test.\n",
"for index in np.random.choice(len(y_test), 2, replace = False):\n",
" print(index)\n",
" predicted = fitted_model.predict(X_test[index:index + 1])[0]\n",
" label = y_test[index]\n",
" title = \"Label value = %d Predicted value = %d \" % (label, predicted)\n",
" fig = plt.figure(1, figsize = (3,3))\n",
" ax1 = fig.add_axes((0,0,.8,.8))\n",
" ax1.set_title(title)\n",
" plt.imshow(images[index], cmap = plt.cm.gray_r, interpolation = 'nearest')\n",
" plt.show()"
]
}
],
"metadata": {
"authors": [
{
"name": "savitam"
}
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 01: Classification with Local Compute\n",
"\n",
"In this example we use the scikit-learn's [digit dataset](http://scikit-learn.org/stable/datasets/index.html#optical-recognition-of-handwritten-digits-dataset) to showcase how you can use AutoML for a simple classification problem.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. Create an `Experiment` in an existing `Workspace`.\n",
"2. Configure AutoML using `AutoMLConfig`.\n",
"3. Train the model using local compute.\n",
"4. Explore the results.\n",
"5. Test the best fitted model.\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# Choose a name for the experiment and specify the project folder.\n",
"experiment_name = 'automl-local-classification'\n",
"project_folder = './sample_projects/automl-local-classification'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace Name'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data = output, index = ['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Load Training Data\n",
"\n",
"This uses scikit-learn's [load_digits](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_digits.html) method."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn import datasets\n",
"\n",
"digits = datasets.load_digits()\n",
"\n",
"# Exclude the first 100 rows from training so that they can be used for test.\n",
"X_train = digits.data[100:,:]\n",
"y_train = digits.target[100:]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML\n",
"\n",
"Instantiate an `AutoMLConfig` object to specify the settings and data used to run the experiment.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**task**|classification or regression|\n",
"|**primary_metric**|This is the metric that you want to optimize. Classification supports the following primary metrics: <br><i>accuracy</i><br><i>AUC_weighted</i><br><i>balanced_accuracy</i><br><i>average_precision_score_weighted</i><br><i>precision_score_weighted</i>|\n",
"|**max_time_sec**|Time limit in seconds for each iteration.|\n",
"|**iterations**|Number of iterations. In each iteration AutoML trains a specific pipeline with the data.|\n",
"|**n_cross_validations**|Number of cross validation splits.|\n",
"|**X**|(sparse) array-like, shape = [n_samples, n_features]|\n",
"|**y**|(sparse) array-like, shape = [n_samples, ], [n_samples, n_classes]<br>Multi-class targets. An indicator matrix turns on multilabel classification. This should be an array of integers.|\n",
"|**path**|Relative path to the project folder. AutoML stores configuration files for the experiment under this folder. You can specify a new empty folder.|"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_config = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" primary_metric = 'AUC_weighted',\n",
" max_time_sec = 3600,\n",
" iterations = 50,\n",
" n_cross_validations = 3,\n",
" verbosity = logging.INFO,\n",
" X = X_train, \n",
" y = y_train,\n",
" path = project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train the Models\n",
"\n",
"Call the `submit` method on the experiment object and pass the run configuration. Execution of local runs is synchronous. Depending on the data and the number of iterations this can run for a while.\n",
"In this example, we specify `show_output = True` to print currently running iterations to the console."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_config, show_output = True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Optionally, you can continue an interrupted local run by calling `continue_experiment` without the `iterations` parameter, or run more iterations for a completed run by specifying the `iterations` parameter:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = local_run.continue_experiment(X = X_train, \n",
" y = y_train, \n",
" show_output = True,\n",
" iterations = 5)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explore the Results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Widget for Monitoring Runs\n",
"\n",
"The widget will first report a \"loading\" status while running the first iteration. After completing the first iteration, an auto-updating graph and table will be shown. The widget will refresh once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"**Note:** The widget displays a link at the bottom. Use this link to open a web interface to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.widgets import RunDetails\n",
"RunDetails(local_run).show() "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"#### Retrieve All Child Runs\n",
"You can also use SDK methods to fetch all the child runs and see individual metrics that we log."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"children = list(local_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
"\n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"rundata"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method returns the best run and the fitted model. The Model includes the pipeline and any pre-processing. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = local_run.get_output()\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Best Model Based on Any Other Metric\n",
"Show the run and the model that has the smallest `log_loss` value:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"lookup_metric = \"log_loss\"\n",
"best_run, fitted_model = local_run.get_output(metric = lookup_metric)\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Model from a Specific Iteration\n",
"Show the run and the model from the third iteration:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"iteration = 3\n",
"third_run, third_model = local_run.get_output(iteration = iteration)\n",
"print(third_run)\n",
"print(third_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test the Best Fitted Model\n",
"\n",
"#### Load Test Data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits = datasets.load_digits()\n",
"X_test = digits.data[:10, :]\n",
"y_test = digits.target[:10]\n",
"images = digits.images[:10]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Testing Our Best Fitted Model\n",
"We will try to predict 2 digits and see how our model works."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Randomly select digits and test.\n",
"for index in np.random.choice(len(y_test), 2, replace = False):\n",
" print(index)\n",
" predicted = fitted_model.predict(X_test[index:index + 1])[0]\n",
" label = y_test[index]\n",
" title = \"Label value = %d Predicted value = %d \" % (label, predicted)\n",
" fig = plt.figure(1, figsize = (3,3))\n",
" ax1 = fig.add_axes((0,0,.8,.8))\n",
" ax1.set_title(title)\n",
" plt.imshow(images[index], cmap = plt.cm.gray_r, interpolation = 'nearest')\n",
" plt.show()"
]
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
"metadata": {
"authors": [
{
"name": "savitam"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
"nbformat": 4,
"nbformat_minor": 2
}

824
automl/02.auto-ml-regression.ipynb
View File

@@ -1,415 +1,415 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 02: Regression with Local Compute\n",
"\n",
"In this example we use the scikit-learn's [diabetes dataset](http://scikit-learn.org/stable/datasets/index.html#diabetes-dataset) to showcase how you can use AutoML for a simple regression problem.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. Create an `Experiment` in an existing `Workspace`.\n",
"2. Configure AutoML using `AutoMLConfig`.\n",
"3. Train the model using local compute.\n",
"4. Explore the results.\n",
"5. Test the best fitted model.\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# Choose a name for the experiment and specify the project folder.\n",
"experiment_name = 'automl-local-regression'\n",
"project_folder = './sample_projects/automl-local-regression'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace Name'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data = output, index = ['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Load Training Data\n",
"This uses scikit-learn's [load_diabetes](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_diabetes.html) method."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Load the diabetes dataset, a well-known built-in small dataset that comes with scikit-learn.\n",
"from sklearn.datasets import load_diabetes\n",
"from sklearn.linear_model import Ridge\n",
"from sklearn.metrics import mean_squared_error\n",
"from sklearn.model_selection import train_test_split\n",
"\n",
"X, y = load_diabetes(return_X_y = True)\n",
"\n",
"columns = ['age', 'gender', 'bmi', 'bp', 's1', 's2', 's3', 's4', 's5', 's6']\n",
"\n",
"X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML\n",
"\n",
"Instantiate an `AutoMLConfig` object to specify the settings and data used to run the experiment.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**task**|classification or regression|\n",
"|**primary_metric**|This is the metric that you want to optimize. Regression supports the following primary metrics: <br><i>spearman_correlation</i><br><i>normalized_root_mean_squared_error</i><br><i>r2_score</i><br><i>normalized_mean_absolute_error</i>|\n",
"|**iteration_timeout_minutes**|Time limit in minutes for each iteration.|\n",
"|**iterations**|Number of iterations. In each iteration AutoML trains a specific pipeline with the data.|\n",
"|**n_cross_validations**|Number of cross validation splits.|\n",
"|**X**|(sparse) array-like, shape = [n_samples, n_features]|\n",
"|**y**|(sparse) array-like, shape = [n_samples, ], [n_samples, n_classes]<br>Multi-class targets. An indicator matrix turns on multilabel classification. This should be an array of integers.|\n",
"|**path**|Relative path to the project folder. AutoML stores configuration files for the experiment under this folder. You can specify a new empty folder.|"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_config = AutoMLConfig(task = 'regression',\n",
" iteration_timeout_minutes = 10,\n",
" iterations = 10,\n",
" primary_metric = 'spearman_correlation',\n",
" n_cross_validations = 5,\n",
" debug_log = 'automl.log',\n",
" verbosity = logging.INFO,\n",
" X = X_train, \n",
" y = y_train,\n",
" path = project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train the Models\n",
"\n",
"Call the `submit` method on the experiment object and pass the run configuration. Execution of local runs is synchronous. Depending on the data and the number of iterations this can run for a while.\n",
"In this example, we specify `show_output = True` to print currently running iterations to the console."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_config, show_output = True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explore the Results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Widget for Monitoring Runs\n",
"\n",
"The widget will first report a \"loading\" status while running the first iteration. After completing the first iteration, an auto-updating graph and table will be shown. The widget will refresh once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"**Note:** The widget displays a link at the bottom. Use this link to open a web interface to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.widgets import RunDetails\n",
"RunDetails(local_run).show() "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"#### Retrieve All Child Runs\n",
"You can also use SDK methods to fetch all the child runs and see individual metrics that we log."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"children = list(local_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
"\n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"rundata"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method returns the best run and the fitted model. The Model includes the pipeline and any pre-processing. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = local_run.get_output()\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Best Model Based on Any Other Metric\n",
"Show the run and the model that has the smallest `root_mean_squared_error` value (which turned out to be the same as the one with largest `spearman_correlation` value):"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"lookup_metric = \"root_mean_squared_error\"\n",
"best_run, fitted_model = local_run.get_output(metric = lookup_metric)\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Model from a Specific Iteration\n",
"Show the run and the model from the third iteration:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"iteration = 3\n",
"third_run, third_model = local_run.get_output(iteration = iteration)\n",
"print(third_run)\n",
"print(third_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test the Best Fitted Model"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Predict on training and test set, and calculate residual values."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"y_pred_train = fitted_model.predict(X_train)\n",
"y_residual_train = y_train - y_pred_train\n",
"\n",
"y_pred_test = fitted_model.predict(X_test)\n",
"y_residual_test = y_test - y_pred_test"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%matplotlib inline\n",
"import matplotlib.pyplot as plt\n",
"import numpy as np\n",
"from sklearn import datasets\n",
"from sklearn.metrics import mean_squared_error, r2_score\n",
"\n",
"# Set up a multi-plot chart.\n",
"f, (a0, a1) = plt.subplots(1, 2, gridspec_kw = {'width_ratios':[1, 1], 'wspace':0, 'hspace': 0})\n",
"f.suptitle('Regression Residual Values', fontsize = 18)\n",
"f.set_figheight(6)\n",
"f.set_figwidth(16)\n",
"\n",
"# Plot residual values of training set.\n",
"a0.axis([0, 360, -200, 200])\n",
"a0.plot(y_residual_train, 'bo', alpha = 0.5)\n",
"a0.plot([-10,360],[0,0], 'r-', lw = 3)\n",
"a0.text(16,170,'RMSE = {0:.2f}'.format(np.sqrt(mean_squared_error(y_train, y_pred_train))), fontsize = 12)\n",
"a0.text(16,140,'R2 score = {0:.2f}'.format(r2_score(y_train, y_pred_train)), fontsize = 12)\n",
"a0.set_xlabel('Training samples', fontsize = 12)\n",
"a0.set_ylabel('Residual Values', fontsize = 12)\n",
"\n",
"# Plot a histogram.\n",
"a0.hist(y_residual_train, orientation = 'horizontal', color = 'b', bins = 10, histtype = 'step');\n",
"a0.hist(y_residual_train, orientation = 'horizontal', color = 'b', alpha = 0.2, bins = 10);\n",
"\n",
"# Plot residual values of test set.\n",
"a1.axis([0, 90, -200, 200])\n",
"a1.plot(y_residual_test, 'bo', alpha = 0.5)\n",
"a1.plot([-10,360],[0,0], 'r-', lw = 3)\n",
"a1.text(5,170,'RMSE = {0:.2f}'.format(np.sqrt(mean_squared_error(y_test, y_pred_test))), fontsize = 12)\n",
"a1.text(5,140,'R2 score = {0:.2f}'.format(r2_score(y_test, y_pred_test)), fontsize = 12)\n",
"a1.set_xlabel('Test samples', fontsize = 12)\n",
"a1.set_yticklabels([])\n",
"\n",
"# Plot a histogram.\n",
"a1.hist(y_residual_test, orientation = 'horizontal', color = 'b', bins = 10, histtype = 'step')\n",
"a1.hist(y_residual_test, orientation = 'horizontal', color = 'b', alpha = 0.2, bins = 10)\n",
"\n",
"plt.show()"
]
}
],
"metadata": {
"authors": [
{
"name": "savitam"
}
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 02: Regression with Local Compute\n",
"\n",
"In this example we use the scikit-learn's [diabetes dataset](http://scikit-learn.org/stable/datasets/index.html#diabetes-dataset) to showcase how you can use AutoML for a simple regression problem.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. Create an `Experiment` in an existing `Workspace`.\n",
"2. Configure AutoML using `AutoMLConfig`.\n",
"3. Train the model using local compute.\n",
"4. Explore the results.\n",
"5. Test the best fitted model.\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# Choose a name for the experiment and specify the project folder.\n",
"experiment_name = 'automl-local-regression'\n",
"project_folder = './sample_projects/automl-local-regression'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace Name'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data = output, index = ['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Load Training Data\n",
"This uses scikit-learn's [load_diabetes](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_diabetes.html) method."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Load the diabetes dataset, a well-known built-in small dataset that comes with scikit-learn.\n",
"from sklearn.datasets import load_diabetes\n",
"from sklearn.linear_model import Ridge\n",
"from sklearn.metrics import mean_squared_error\n",
"from sklearn.model_selection import train_test_split\n",
"\n",
"X, y = load_diabetes(return_X_y = True)\n",
"\n",
"columns = ['age', 'gender', 'bmi', 'bp', 's1', 's2', 's3', 's4', 's5', 's6']\n",
"\n",
"X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML\n",
"\n",
"Instantiate an `AutoMLConfig` object to specify the settings and data used to run the experiment.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**task**|classification or regression|\n",
"|**primary_metric**|This is the metric that you want to optimize. Regression supports the following primary metrics: <br><i>spearman_correlation</i><br><i>normalized_root_mean_squared_error</i><br><i>r2_score</i><br><i>normalized_mean_absolute_error</i>|\n",
"|**max_time_sec**|Time limit in seconds for each iteration.|\n",
"|**iterations**|Number of iterations. In each iteration AutoML trains a specific pipeline with the data.|\n",
"|**n_cross_validations**|Number of cross validation splits.|\n",
"|**X**|(sparse) array-like, shape = [n_samples, n_features]|\n",
"|**y**|(sparse) array-like, shape = [n_samples, ], [n_samples, n_classes]<br>Multi-class targets. An indicator matrix turns on multilabel classification. This should be an array of integers.|\n",
"|**path**|Relative path to the project folder. AutoML stores configuration files for the experiment under this folder. You can specify a new empty folder.|"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_config = AutoMLConfig(task = 'regression',\n",
" max_time_sec = 600,\n",
" iterations = 10,\n",
" primary_metric = 'spearman_correlation',\n",
" n_cross_validations = 5,\n",
" debug_log = 'automl.log',\n",
" verbosity = logging.INFO,\n",
" X = X_train, \n",
" y = y_train,\n",
" path = project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train the Models\n",
"\n",
"Call the `submit` method on the experiment object and pass the run configuration. Execution of local runs is synchronous. Depending on the data and the number of iterations this can run for a while.\n",
"In this example, we specify `show_output = True` to print currently running iterations to the console."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_config, show_output = True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explore the Results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Widget for Monitoring Runs\n",
"\n",
"The widget will first report a \"loading\" status while running the first iteration. After completing the first iteration, an auto-updating graph and table will be shown. The widget will refresh once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"**Note:** The widget displays a link at the bottom. Use this link to open a web interface to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.widgets import RunDetails\n",
"RunDetails(local_run).show() "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"#### Retrieve All Child Runs\n",
"You can also use SDK methods to fetch all the child runs and see individual metrics that we log."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"children = list(local_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
"\n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"rundata"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method returns the best run and the fitted model. The Model includes the pipeline and any pre-processing. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = local_run.get_output()\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Best Model Based on Any Other Metric\n",
"Show the run and the model that has the smallest `root_mean_squared_error` value (which turned out to be the same as the one with largest `spearman_correlation` value):"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"lookup_metric = \"root_mean_squared_error\"\n",
"best_run, fitted_model = local_run.get_output(metric = lookup_metric)\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Model from a Specific Iteration\n",
"Show the run and the model from the third iteration:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"iteration = 3\n",
"third_run, third_model = local_run.get_output(iteration = iteration)\n",
"print(third_run)\n",
"print(third_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test the Best Fitted Model"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Predict on training and test set, and calculate residual values."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"y_pred_train = fitted_model.predict(X_train)\n",
"y_residual_train = y_train - y_pred_train\n",
"\n",
"y_pred_test = fitted_model.predict(X_test)\n",
"y_residual_test = y_test - y_pred_test"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%matplotlib inline\n",
"import matplotlib.pyplot as plt\n",
"import numpy as np\n",
"from sklearn import datasets\n",
"from sklearn.metrics import mean_squared_error, r2_score\n",
"\n",
"# Set up a multi-plot chart.\n",
"f, (a0, a1) = plt.subplots(1, 2, gridspec_kw = {'width_ratios':[1, 1], 'wspace':0, 'hspace': 0})\n",
"f.suptitle('Regression Residual Values', fontsize = 18)\n",
"f.set_figheight(6)\n",
"f.set_figwidth(16)\n",
"\n",
"# Plot residual values of training set.\n",
"a0.axis([0, 360, -200, 200])\n",
"a0.plot(y_residual_train, 'bo', alpha = 0.5)\n",
"a0.plot([-10,360],[0,0], 'r-', lw = 3)\n",
"a0.text(16,170,'RMSE = {0:.2f}'.format(np.sqrt(mean_squared_error(y_train, y_pred_train))), fontsize = 12)\n",
"a0.text(16,140,'R2 score = {0:.2f}'.format(r2_score(y_train, y_pred_train)), fontsize = 12)\n",
"a0.set_xlabel('Training samples', fontsize = 12)\n",
"a0.set_ylabel('Residual Values', fontsize = 12)\n",
"\n",
"# Plot a histogram.\n",
"a0.hist(y_residual_train, orientation = 'horizontal', color = 'b', bins = 10, histtype = 'step');\n",
"a0.hist(y_residual_train, orientation = 'horizontal', color = 'b', alpha = 0.2, bins = 10);\n",
"\n",
"# Plot residual values of test set.\n",
"a1.axis([0, 90, -200, 200])\n",
"a1.plot(y_residual_test, 'bo', alpha = 0.5)\n",
"a1.plot([-10,360],[0,0], 'r-', lw = 3)\n",
"a1.text(5,170,'RMSE = {0:.2f}'.format(np.sqrt(mean_squared_error(y_test, y_pred_test))), fontsize = 12)\n",
"a1.text(5,140,'R2 score = {0:.2f}'.format(r2_score(y_test, y_pred_test)), fontsize = 12)\n",
"a1.set_xlabel('Test samples', fontsize = 12)\n",
"a1.set_yticklabels([])\n",
"\n",
"# Plot a histogram.\n",
"a1.hist(y_residual_test, orientation = 'horizontal', color = 'b', bins = 10, histtype = 'step')\n",
"a1.hist(y_residual_test, orientation = 'horizontal', color = 'b', alpha = 0.2, bins = 10)\n",
"\n",
"plt.show()"
]
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
"metadata": {
"authors": [
{
"name": "savitam"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
"nbformat": 4,
"nbformat_minor": 2
}

485
automl/03.auto-ml-remote-execution.ipynb Normal file
View File

@@ -0,0 +1,485 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 03: Remote Execution using DSVM (Ubuntu)\n",
"\n",
"In this example we use the scikit-learn's [digit dataset](http://scikit-learn.org/stable/datasets/index.html#optical-recognition-of-handwritten-digits-dataset) to showcase how you can use AutoML for a simple classification problem.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you wiil learn how to:\n",
"1. Create an `Experiment` in an existing `Workspace`.\n",
"2. Attach an existing DSVM to a workspace.\n",
"3. Configure AutoML using `AutoMLConfig`.\n",
"4. Train the model using the DSVM.\n",
"5. Explore the results.\n",
"6. Test the best fitted model.\n",
"\n",
"In addition, this notebook showcases the following features:\n",
"- **Parallel** executions for iterations\n",
"- **Asynchronous** tracking of progress\n",
"- **Cancellation** of individual iterations or the entire run\n",
"- Retrieving models for any iteration or logged metric\n",
"- Specifying AutoML settings as `**kwargs`\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# Choose a name for the run history container in the workspace.\n",
"experiment_name = 'automl-remote-dsvm4'\n",
"project_folder = './sample_projects/automl-remote-dsvm4'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace Name'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data = output, index = ['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create a Remote Linux DSVM\n",
"**Note:** If creation fails with a message about Marketplace purchase eligibilty, start creation of a DSVM through the [Azure portal](https://portal.azure.com), and select \"Want to create programmatically\" to enable programmatic creation. Once you've enabled this setting, you can exit the portal without actually creating the DSVM, and creation of the DSVM through the notebook should work.\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import DsvmCompute\n",
"\n",
"dsvm_name = 'mydsvm'\n",
"try:\n",
" dsvm_compute = DsvmCompute(ws, dsvm_name)\n",
" print('Found an existing DSVM.')\n",
"except:\n",
" print('Creating a new DSVM.')\n",
" dsvm_config = DsvmCompute.provisioning_configuration(vm_size = \"Standard_D2_v2\")\n",
" dsvm_compute = DsvmCompute.create(ws, name = dsvm_name, provisioning_configuration = dsvm_config)\n",
" dsvm_compute.wait_for_completion(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create Get Data File\n",
"For remote executions you should author a `get_data.py` file containing a `get_data()` function. This file should be in the root directory of the project. You can encapsulate code to read data either from a blob storage or local disk in this file.\n",
"In this example, the `get_data()` function returns data using scikit-learn's [load_digits](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_digits.html) method."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"if not os.path.exists(project_folder):\n",
" os.makedirs(project_folder)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile $project_folder/get_data.py\n",
"\n",
"from sklearn import datasets\n",
"from scipy import sparse\n",
"import numpy as np\n",
"\n",
"def get_data():\n",
" \n",
" digits = datasets.load_digits()\n",
" X_train = digits.data[100:,:]\n",
" y_train = digits.target[100:]\n",
"\n",
" return { \"X\" : X_train, \"y\" : y_train }"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML <a class=\"anchor\" id=\"Instantiate-AutoML-Remote-DSVM\"></a>\n",
"\n",
"You can specify `automl_settings` as `**kwargs` as well. Also note that you can use a `get_data()` function for local excutions too.\n",
"\n",
"**Note:** When using Remote DSVM, you can't pass Numpy arrays directly to the fit method.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**primary_metric**|This is the metric that you want to optimize. Classification supports the following primary metrics: <br><i>accuracy</i><br><i>AUC_weighted</i><br><i>balanced_accuracy</i><br><i>average_precision_score_weighted</i><br><i>precision_score_weighted</i>|\n",
"|**max_time_sec**|Time limit in seconds for each iteration.|\n",
"|**iterations**|Number of iterations. In each iteration AutoML trains a specific pipeline with the data.|\n",
"|**n_cross_validations**|Number of cross validation splits.|\n",
"|**concurrent_iterations**|Maximum number of iterations to execute in parallel. This should be less than the number of cores on the DSVM.|"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_settings = {\n",
" \"max_time_sec\": 600,\n",
" \"iterations\": 20,\n",
" \"n_cross_validations\": 5,\n",
" \"primary_metric\": 'AUC_weighted',\n",
" \"preprocess\": False,\n",
" \"concurrent_iterations\": 2,\n",
" \"verbosity\": logging.INFO\n",
"}\n",
"\n",
"automl_config = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" path = project_folder, \n",
" compute_target = dsvm_compute,\n",
" data_script = project_folder + \"/get_data.py\",\n",
" **automl_settings\n",
" )\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Note:** The first run on a new DSVM may take several minutes to prepare the environment."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train the Models\n",
"\n",
"Call the `submit` method on the experiment object and pass the run configuration. For remote runs the execution is asynchronous, so you will see the iterations get populated as they complete. You can interact with the widgets and models even when the experiment is running to retrieve the best model up to that point. Once you are satisfied with the model, you can cancel a particular iteration or the whole run.\n",
"\n",
"In this example, we specify `show_output = False` to suppress console output while the run is in progress."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"remote_run = experiment.submit(automl_config, show_output = False)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explore the Results\n",
"\n",
"#### Loading Executed Runs\n",
"In case you need to load a previously executed run, enable the cell below and replace the `run_id` value."
]
},
{
"cell_type": "raw",
"metadata": {},
"source": [
"remote_run = AutoMLRun(experiment=experiment, run_id = 'AutoML_480d3ed6-fc94-44aa-8f4e-0b945db9d3ef')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Widget for Monitoring Runs\n",
"\n",
"The widget will first report a \"loading\" status while running the first iteration. After completing the first iteration, an auto-updating graph and table will be shown. The widget will refresh once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"You can click on a pipeline to see run properties and output logs. Logs are also available on the DSVM under `/tmp/azureml_run/{iterationid}/azureml-logs`\n",
"\n",
"**Note:** The widget displays a link at the bottom. Use this link to open a web interface to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.widgets import RunDetails\n",
"RunDetails(remote_run).show() "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Wait until the run finishes.\n",
"remote_run.wait_for_completion(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"#### Retrieve All Child Runs\n",
"You can also use SDK methods to fetch all the child runs and see individual metrics that we log."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"children = list(remote_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)} \n",
" metricslist[int(properties['iteration'])] = metrics\n",
"\n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"rundata"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Cancelling Runs\n",
"\n",
"You can cancel ongoing remote runs using the `cancel` and `cancel_iteration` functions."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Cancel the ongoing experiment and stop scheduling new iterations.\n",
"# remote_run.cancel()\n",
"\n",
"# Cancel iteration 1 and move onto iteration 2.\n",
"# remote_run.cancel_iteration(1)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method returns the best run and the fitted model. The Model includes the pipeline and any pre-processing. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = remote_run.get_output()\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Best Model Based on Any Other Metric\n",
"Show the run and the model which has the smallest `log_loss` value:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"lookup_metric = \"log_loss\"\n",
"best_run, fitted_model = remote_run.get_output(metric = lookup_metric)\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Model from a Specific Iteration\n",
"Show the run and the model from the third iteration:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"iteration = 3\n",
"third_run, third_model = remote_run.get_output(iteration = iteration)\n",
"print(third_run)\n",
"print(third_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test the Best Fitted Model <a class=\"anchor\" id=\"Testing-the-Fitted-Model-Remote-DSVM\"></a>\n",
"\n",
"#### Load Test Data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits = datasets.load_digits()\n",
"X_test = digits.data[:10, :]\n",
"y_test = digits.target[:10]\n",
"images = digits.images[:10]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Test Our Best Fitted Model"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Randomly select digits and test.\n",
"for index in np.random.choice(len(y_test), 2, replace = False):\n",
" print(index)\n",
" predicted = fitted_model.predict(X_test[index:index + 1])[0]\n",
" label = y_test[index]\n",
" title = \"Label value = %d Predicted value = %d \" % (label, predicted)\n",
" fig = plt.figure(1, figsize=(3,3))\n",
" ax1 = fig.add_axes((0,0,.8,.8))\n",
" ax1.set_title(title)\n",
" plt.imshow(images[index], cmap = plt.cm.gray_r, interpolation = 'nearest')\n",
" plt.show()"
]
}
],
"metadata": {
"authors": [
{
"name": "savitam"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

509
automl/03b.auto-ml-remote-batchai.ipynb Normal file
View File

@@ -0,0 +1,509 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 03: Remote Execution using Batch AI\n",
"\n",
"In this example we use the scikit-learn's [digit dataset](http://scikit-learn.org/stable/datasets/index.html#optical-recognition-of-handwritten-digits-dataset) to showcase how you can use AutoML for a simple classification problem.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you would see\n",
"1. Create an `Experiment` in an existing `Workspace`.\n",
"2. Attach an existing Batch AI compute to a workspace.\n",
"3. Configure AutoML using `AutoMLConfig`.\n",
"4. Train the model using Batch AI.\n",
"5. Explore the results.\n",
"6. Test the best fitted model.\n",
"\n",
"In addition this notebook showcases the following features\n",
"- **Parallel** executions for iterations\n",
"- **Asynchronous** tracking of progress\n",
"- **Cancellation** of individual iterations or the entire run\n",
"- Retrieving models for any iteration or logged metric\n",
"- Specifying AutoML settings as `**kwargs`\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# Choose a name for the run history container in the workspace.\n",
"experiment_name = 'automl-remote-batchai'\n",
"project_folder = './sample_projects/automl-remote-batchai'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace Name'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data = output, index = ['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create Batch AI Cluster\n",
"The cluster is created as Machine Learning Compute and will appear under your workspace.\n",
"\n",
"**Note:** The creation of the Batch AI cluster can take over 10 minutes, please be patient.\n",
"\n",
"As with other Azure services, there are limits on certain resources (e.g. Batch AI cluster size) associated with the Azure Machine Learning service. Please read [this article](https://docs.microsoft.com/en-us/azure/machine-learning/service/how-to-manage-quotas) on the default limits and how to request more quota."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import BatchAiCompute\n",
"from azureml.core.compute import ComputeTarget\n",
"\n",
"# Choose a name for your cluster.\n",
"batchai_cluster_name = \"mybatchai\"\n",
"\n",
"found = False\n",
"# Check if this compute target already exists in the workspace.\n",
"cts = ws.compute_targets\n",
"if batchai_cluster_name in cts and cts[batchai_cluster_name].type == 'BatchAI':\n",
" found = True\n",
" print('Found existing compute target.')\n",
" compute_target = cts[batchai_cluster_name]\n",
" \n",
"if not found:\n",
" print('Creating a new compute target...')\n",
" provisioning_config = BatchAiCompute.provisioning_configuration(vm_size = \"STANDARD_D2_V2\", # for GPU, use \"STANDARD_NC6\"\n",
" #vm_priority = 'lowpriority', # optional\n",
" autoscale_enabled = True,\n",
" cluster_min_nodes = 1, \n",
" cluster_max_nodes = 4)\n",
"\n",
" # Create the cluster.\n",
" compute_target = ComputeTarget.create(ws, batchai_cluster_name, provisioning_config)\n",
" \n",
" # Can poll for a minimum number of nodes and for a specific timeout.\n",
" # If no min_node_count is provided, it will use the scale settings for the cluster.\n",
" compute_target.wait_for_completion(show_output = True, min_node_count = None, timeout_in_minutes = 20)\n",
" \n",
" # For a more detailed view of current Batch AI cluster status, use the 'status' property."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create Get Data File\n",
"For remote executions you should author a `get_data.py` file containing a `get_data()` function. This file should be in the root directory of the project. You can encapsulate code to read data either from a blob storage or local disk in this file.\n",
"In this example, the `get_data()` function returns data using scikit-learn's [load_digits](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_digits.html) method."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"if not os.path.exists(project_folder):\n",
" os.makedirs(project_folder)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile $project_folder/get_data.py\n",
"\n",
"from sklearn import datasets\n",
"from scipy import sparse\n",
"import numpy as np\n",
"\n",
"def get_data():\n",
" \n",
" digits = datasets.load_digits()\n",
" X_train = digits.data\n",
" y_train = digits.target\n",
"\n",
" return { \"X\" : X_train, \"y\" : y_train }"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Instantiate AutoML <a class=\"anchor\" id=\"Instatiate-AutoML-Remote-DSVM\"></a>\n",
"\n",
"You can specify `automl_settings` as `**kwargs` as well. Also note that you can use a `get_data()` function for local excutions too.\n",
"\n",
"**Note:** When using Batch AI, you can't pass Numpy arrays directly to the fit method.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**primary_metric**|This is the metric that you want to optimize. Classification supports the following primary metrics: <br><i>accuracy</i><br><i>AUC_weighted</i><br><i>balanced_accuracy</i><br><i>average_precision_score_weighted</i><br><i>precision_score_weighted</i>|\n",
"|**max_time_sec**|Time limit in seconds for each iteration.|\n",
"|**iterations**|Number of iterations. In each iteration AutoML trains a specific pipeline with the data.|\n",
"|**n_cross_validations**|Number of cross validation splits.|\n",
"|**concurrent_iterations**|Maximum number of iterations that would be executed in parallel. This should be less than the number of cores on the DSVM.|"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_settings = {\n",
" \"max_time_sec\": 120,\n",
" \"iterations\": 20,\n",
" \"n_cross_validations\": 5,\n",
" \"primary_metric\": 'AUC_weighted',\n",
" \"preprocess\": False,\n",
" \"concurrent_iterations\": 5,\n",
" \"verbosity\": logging.INFO\n",
"}\n",
"\n",
"automl_config = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" path = project_folder,\n",
" compute_target = compute_target,\n",
" data_script = project_folder + \"/get_data.py\",\n",
" **automl_settings\n",
" )\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train the Models\n",
"\n",
"Call the `submit` method on the experiment object and pass the run configuration. For remote runs the execution is asynchronous, so you will see the iterations get populated as they complete. You can interact with the widgets and models even when the experiment is running to retrieve the best model up to that point. Once you are satisfied with the model, you can cancel a particular iteration or the whole run.\n",
"In this example, we specify `show_output = False` to suppress console output while the run is in progress."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"remote_run = experiment.submit(automl_config, show_output = False)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explore the Results\n",
"\n",
"#### Loading executed runs\n",
"In case you need to load a previously executed run, enable the cell below and replace the `run_id` value."
]
},
{
"cell_type": "raw",
"metadata": {},
"source": [
"remote_run = AutoMLRun(experiment = experiment, run_id = 'AutoML_5db13491-c92a-4f1d-b622-8ab8d973a058')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Widget for Monitoring Runs\n",
"\n",
"The widget will first report a \"loading\" status while running the first iteration. After completing the first iteration, an auto-updating graph and table will be shown. The widget will refresh once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"You can click on a pipeline to see run properties and output logs. Logs are also available on the DSVM under `/tmp/azureml_run/{iterationid}/azureml-logs`\n",
"\n",
"**Note:** The widget displays a link at the bottom. Use this link to open a web interface to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"remote_run"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.widgets import RunDetails\n",
"RunDetails(remote_run).show() "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Wait until the run finishes.\n",
"remote_run.wait_for_completion(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"#### Retrieve All Child Runs\n",
"You can also use SDK methods to fetch all the child runs and see individual metrics that we log."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"children = list(remote_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
"\n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"rundata"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Cancelling Runs\n",
"\n",
"You can cancel ongoing remote runs using the `cancel` and `cancel_iteration` functions."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Cancel the ongoing experiment and stop scheduling new iterations.\n",
"# remote_run.cancel()\n",
"\n",
"# Cancel iteration 1 and move onto iteration 2.\n",
"# remote_run.cancel_iteration(1)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method returns the best run and the fitted model. The Model includes the pipeline and any pre-processing. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = remote_run.get_output()\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Best Model Based on Any Other Metric\n",
"Show the run and the model which has the smallest `log_loss` value:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"lookup_metric = \"log_loss\"\n",
"best_run, fitted_model = remote_run.get_output(metric = lookup_metric)\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Model from a Specific Iteration\n",
"Show the run and the model from the third iteration:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"iteration = 3\n",
"third_run, third_model = remote_run.get_output(iteration=iteration)\n",
"print(third_run)\n",
"print(third_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Testing the Fitted Model <a class=\"anchor\" id=\"Testing-the-Fitted-Model-Remote-DSVM\"></a>\n",
"\n",
"#### Load Test Data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits = datasets.load_digits()\n",
"X_test = digits.data[:10, :]\n",
"y_test = digits.target[:10]\n",
"images = digits.images[:10]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Testing Our Best Fitted Model"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Randomly select digits and test.\n",
"for index in np.random.choice(len(y_test), 2, replace = False):\n",
" print(index)\n",
" predicted = fitted_model.predict(X_test[index:index + 1])[0]\n",
" label = y_test[index]\n",
" title = \"Label value = %d Predicted value = %d \" % (label, predicted)\n",
" fig = plt.figure(1, figsize=(3,3))\n",
" ax1 = fig.add_axes((0,0,.8,.8))\n",
" ax1.set_title(title)\n",
" plt.imshow(images[index], cmap = plt.cm.gray_r, interpolation = 'nearest')\n",
" plt.show()"
]
}
],
"metadata": {
"authors": [
{
"name": "savitam"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

491
automl/04.auto-ml-remote-execution-text-data-blob-store.ipynb Normal file
View File

@@ -0,0 +1,491 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Auto ML 04: Remote Execution with Text Data from Azure Blob Storage\n",
"\n",
"In this example we use the [Burning Man 2016 dataset](https://innovate.burningman.org/datasets-page/) to showcase how you can use AutoML to handle text data from Azure Blob Storage.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. Create an `Experiment` in an existing `Workspace`.\n",
"2. Attach an existing DSVM to a workspace.\n",
"3. Configure AutoML using `AutoMLConfig`.\n",
"4. Train the model using the DSVM.\n",
"5. Explore the results.\n",
"6. Test the best fitted model.\n",
"\n",
"In addition this notebook showcases the following features\n",
"- **Parallel** executions for iterations\n",
"- **Asynchronous** tracking of progress\n",
"- **Cancellation** of individual iterations or the entire run\n",
"- Retrieving models for any iteration or logged metric\n",
"- Specifying AutoML settings as `**kwargs`\n",
"- Handling **text** data using the `preprocess` flag\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# Choose a name for the run history container in the workspace.\n",
"experiment_name = 'automl-remote-dsvm-blobstore'\n",
"project_folder = './sample_projects/automl-remote-dsvm-blobstore'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data=output, index=['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Attach a Remote Linux DSVM\n",
"To use a remote Docker compute target:\n",
"1. Create a Linux DSVM in Azure, following these [quick instructions](https://docs.microsoft.com/en-us/azure/machine-learning/desktop-workbench/how-to-create-dsvm-hdi). Make sure you use the Ubuntu flavor (not CentOS). Make sure that disk space is available under `/tmp` because AutoML creates files under `/tmp/azureml_run`s. The DSVM should have more cores than the number of parallel runs that you plan to enable. It should also have at least 4GB per core.\n",
"2. Enter the IP address, user name and password below.\n",
"\n",
"**Note:** By default, SSH runs on port 22 and you don't need to change the port number below. If you've configured SSH to use a different port, change `dsvm_ssh_port` accordinglyaddress. [Read more](https://render.githubusercontent.com/documentation/sdk/ssh-issue.md) on changing SSH ports for security reasons."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import RemoteCompute\n",
"import time\n",
"\n",
"# Add your VM information below\n",
"# If a compute with the specified compute_name already exists, it will be used and the dsvm_ip_addr, dsvm_ssh_port, \n",
"# dsvm_username and dsvm_password will be ignored.\n",
"compute_name = 'mydsvm'\n",
"dsvm_ip_addr = '<<ip_addr>>'\n",
"dsvm_ssh_port = 22\n",
"dsvm_username = '<<username>>'\n",
"dsvm_password = '<<password>>'\n",
"\n",
"if compute_name in ws.compute_targets:\n",
" print('Using existing compute.')\n",
" dsvm_compute = ws.compute_targets[compute_name]\n",
"else:\n",
" RemoteCompute.attach(workspace=ws, name=compute_name, address=dsvm_ip_addr, username=dsvm_username, password=dsvm_password, ssh_port=dsvm_ssh_port)\n",
"\n",
" while ws.compute_targets[compute_name].provisioning_state == 'Creating':\n",
" time.sleep(1)\n",
"\n",
" dsvm_compute = ws.compute_targets[compute_name]\n",
" \n",
" if dsvm_compute.provisioning_state == 'Failed':\n",
" print('Attached failed.')\n",
" print(dsvm_compute.provisioning_errors)\n",
" dsvm_compute.delete()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create Get Data File\n",
"For remote executions you should author a `get_data.py` file containing a `get_data()` function. This file should be in the root directory of the project. You can encapsulate code to read data either from a blob storage or local disk in this file.\n",
"In this example, the `get_data()` function returns a [dictionary](README.md#getdata)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"if not os.path.exists(project_folder):\n",
" os.makedirs(project_folder)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile $project_folder/get_data.py\n",
"\n",
"import pandas as pd\n",
"from sklearn.model_selection import train_test_split\n",
"from sklearn.preprocessing import LabelEncoder\n",
"\n",
"def get_data():\n",
" # Load Burning Man 2016 data.\n",
" df = pd.read_csv(\"https://automldemods.blob.core.windows.net/datasets/PlayaEvents2016,_1.6MB,_3.4k-rows.cleaned.2.tsv\",\n",
" delimiter=\"\\t\", quotechar='\"')\n",
" # Get integer labels.\n",
" le = LabelEncoder()\n",
" le.fit(df[\"Label\"].values)\n",
" y = le.transform(df[\"Label\"].values)\n",
" X = df.drop([\"Label\"], axis=1)\n",
"\n",
" X_train, _, y_train, _ = train_test_split(X, y, test_size = 0.1, random_state = 42)\n",
"\n",
" return { \"X\" : X_train, \"y\" : y_train }"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### View data\n",
"\n",
"You can execute the `get_data()` function locally to view the training data."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%run $project_folder/get_data.py\n",
"data_dict = get_data()\n",
"df = data_dict[\"X\"]\n",
"y = data_dict[\"y\"]\n",
"pd.set_option('display.max_colwidth', 15)\n",
"df['Label'] = pd.Series(y, index=df.index)\n",
"df.head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML <a class=\"anchor\" id=\"Instatiate-AutoML-Remote-DSVM\"></a>\n",
"\n",
"You can specify `automl_settings` as `**kwargs` as well. Also note that you can use a `get_data()` function for local excutions too.\n",
"\n",
"**Note:** When using Remote DSVM, you can't pass Numpy arrays directly to the fit method.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**primary_metric**|This is the metric that you want to optimize. Classification supports the following primary metrics: <br><i>accuracy</i><br><i>AUC_weighted</i><br><i>balanced_accuracy</i><br><i>average_precision_score_weighted</i><br><i>precision_score_weighted</i>|\n",
"|**max_time_sec**|Time limit in seconds for each iteration.|\n",
"|**iterations**|Number of iterations. In each iteration AutoML trains a specific pipeline with the data.|\n",
"|**n_cross_validations**|Number of cross validation splits.|\n",
"|**concurrent_iterations**|Maximum number of iterations that would be executed in parallel. This should be less than the number of cores on the DSVM.|\n",
"|**preprocess**|Setting this to *True* enables AutoML to perform preprocessing on the input to handle *missing data*, and to perform some common *feature extraction*.|\n",
"|**max_cores_per_iteration**|Indicates how many cores on the compute target would be used to train a single pipeline.<br>Default is *1*; you can set it to *-1* to use all cores.|"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_settings = {\n",
" \"max_time_sec\": 3600,\n",
" \"iterations\": 10,\n",
" \"n_cross_validations\": 5,\n",
" \"primary_metric\": 'AUC_weighted',\n",
" \"preprocess\": True,\n",
" \"max_cores_per_iteration\": 2\n",
"}\n",
"\n",
"automl_config = AutoMLConfig(task = 'classification',\n",
" path = project_folder,\n",
" compute_target = dsvm_compute,\n",
" data_script = project_folder + \"/get_data.py\",\n",
" **automl_settings\n",
" )\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train the Models <a class=\"anchor\" id=\"Training-the-model-Remote-DSVM\"></a>\n",
"\n",
"Call the `submit` method on the experiment object and pass the run configuration. For remote runs the execution is asynchronous, so you will see the iterations get populated as they complete. You can interact with the widgets and models even when the experiment is running to retrieve the best model up to that point. Once you are satisfied with the model, you can cancel a particular iteration or the whole run."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"remote_run = experiment.submit(automl_config)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Exploring the Results <a class=\"anchor\" id=\"Exploring-the-Results-Remote-DSVM\"></a>\n",
"#### Widget for Monitoring Runs\n",
"\n",
"The widget will first report a \"loading\" status while running the first iteration. After completing the first iteration, an auto-updating graph and table will be shown. The widget will refresh once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"You can click on a pipeline to see run properties and output logs. Logs are also available on the DSVM under `/tmp/azureml_run/{iterationid}/azureml-logs`\n",
"\n",
"**Note:** The widget displays a link at the bottom. Use this link to open a web interface to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.widgets import RunDetails\n",
"RunDetails(remote_run).show() "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"#### Retrieve All Child Runs\n",
"You can also use SDK methods to fetch all the child runs and see individual metrics that we log. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"children = list(remote_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
"\n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"rundata"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Cancelling Runs\n",
"You can cancel ongoing remote runs using the `cancel` and `cancel_iteration` functions."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Cancel the ongoing experiment and stop scheduling new iterations.\n",
"remote_run.cancel()\n",
"\n",
"# Cancel iteration 1 and move onto iteration 2.\n",
"# remote_run.cancel_iteration(1)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method returns the best run and the fitted model. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = remote_run.get_output()\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Best Model Based on Any Other Metric\n",
"Show the run and the model which has the smallest `accuracy` value:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# lookup_metric = \"accuracy\"\n",
"# best_run, fitted_model = remote_run.get_output(metric = lookup_metric)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Model from a Specific Iteration"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"iteration = 0\n",
"zero_run, zero_model = remote_run.get_output(iteration = iteration)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Testing the Fitted Model <a class=\"anchor\" id=\"Testing-the-Fitted-Model-Remote-DSVM\"></a>\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import sklearn\n",
"from sklearn.model_selection import train_test_split\n",
"from sklearn.preprocessing import LabelEncoder\n",
"from pandas_ml import ConfusionMatrix\n",
"\n",
"df = pd.read_csv(\"https://automldemods.blob.core.windows.net/datasets/PlayaEvents2016,_1.6MB,_3.4k-rows.cleaned.2.tsv\",\n",
" delimiter=\"\\t\", quotechar='\"')\n",
"\n",
"# get integer labels\n",
"le = LabelEncoder()\n",
"le.fit(df[\"Label\"].values)\n",
"y = le.transform(df[\"Label\"].values)\n",
"X = df.drop([\"Label\"], axis=1)\n",
"\n",
"_, X_test, _, y_test = train_test_split(X, y, test_size=0.1, random_state=42)\n",
"\n",
"\n",
"ypred = fitted_model.predict(X_test.values)\n",
"\n",
"\n",
"ypred_strings = le.inverse_transform(ypred)\n",
"ytest_strings = le.inverse_transform(y_test)\n",
"\n",
"cm = ConfusionMatrix(ytest_strings, ypred_strings)\n",
"\n",
"print(cm)\n",
"\n",
"cm.plot()"
]
}
],
"metadata": {
"authors": [
{
"name": "savitam"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

756
automl/05.auto-ml-missing-data-Blacklist-Early-Termination.ipynb
View File

@@ -1,381 +1,381 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 05: Blacklisting Models, Early Termination, and Handling Missing Data\n",
"\n",
"In this example we use the scikit-learn's [digit dataset](http://scikit-learn.org/stable/datasets/index.html#optical-recognition-of-handwritten-digits-dataset) to showcase how you can use AutoML for handling missing values in data. We also provide a stopping metric indicating a target for the primary metrics so that AutoML can terminate the run without necessarly going through all the iterations. Finally, if you want to avoid a certain pipeline, we allow you to specify a blacklist of algorithms that AutoML will ignore for this run.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. Create an `Experiment` in an existing `Workspace`.\n",
"2. Configure AutoML using `AutoMLConfig`.\n",
"4. Train the model.\n",
"5. Explore the results.\n",
"6. Test the best fitted model.\n",
"\n",
"In addition this notebook showcases the following features\n",
"- **Blacklisting** certain pipelines\n",
"- Specifying **target metrics** to indicate stopping criteria\n",
"- Handling **missing data** in the input\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# Choose a name for the experiment.\n",
"experiment_name = 'automl-local-missing-data'\n",
"project_folder = './sample_projects/automl-local-missing-data'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data=output, index=['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Creating missing data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from scipy import sparse\n",
"\n",
"digits = datasets.load_digits()\n",
"X_train = digits.data[10:,:]\n",
"y_train = digits.target[10:]\n",
"\n",
"# Add missing values in 75% of the lines.\n",
"missing_rate = 0.75\n",
"n_missing_samples = int(np.floor(X_train.shape[0] * missing_rate))\n",
"missing_samples = np.hstack((np.zeros(X_train.shape[0] - n_missing_samples, dtype=np.bool), np.ones(n_missing_samples, dtype=np.bool)))\n",
"rng = np.random.RandomState(0)\n",
"rng.shuffle(missing_samples)\n",
"missing_features = rng.randint(0, X_train.shape[1], n_missing_samples)\n",
"X_train[np.where(missing_samples)[0], missing_features] = np.nan"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"df = pd.DataFrame(data = X_train)\n",
"df['Label'] = pd.Series(y_train, index=df.index)\n",
"df.head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML\n",
"\n",
"Instantiate an `AutoMLConfig` object to specify the settings and data used to run the experiment. This includes setting `experiment_exit_score`, which should cause the run to complete before the `iterations` count is reached.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**task**|classification or regression|\n",
"|**primary_metric**|This is the metric that you want to optimize. Classification supports the following primary metrics: <br><i>accuracy</i><br><i>AUC_weighted</i><br><i>average_precision_score_weighted</i><br><i>norm_macro_recall</i><br><i>precision_score_weighted</i>|\n",
"|**iteration_timeout_minutes**|Time limit in minutes for each iteration.|\n",
"|**iterations**|Number of iterations. In each iteration AutoML trains a specific pipeline with the data.|\n",
"|**n_cross_validations**|Number of cross validation splits.|\n",
"|**preprocess**|Setting this to *True* enables AutoML to perform preprocessing on the input to handle *missing data*, and to perform some common *feature extraction*.|\n",
"|**experiment_exit_score**|*double* value indicating the target for *primary_metric*. <br>Once the target is surpassed the run terminates.|\n",
"|**blacklist_models**|*List* of *strings* indicating machine learning algorithms for AutoML to avoid in this run.<br><br> Allowed values for **Classification**<br><i>LogisticRegression</i><br><i>SGD</i><br><i>MultinomialNaiveBayes</i><br><i>BernoulliNaiveBayes</i><br><i>SVM</i><br><i>LinearSVM</i><br><i>KNN</i><br><i>DecisionTree</i><br><i>RandomForest</i><br><i>ExtremeRandomTrees</i><br><i>LightGBM</i><br><i>GradientBoosting</i><br><i>TensorFlowDNN</i><br><i>TensorFlowLinearClassifier</i><br><br>Allowed values for **Regression**<br><i>ElasticNet</i><br><i>GradientBoosting</i><br><i>DecisionTree</i><br><i>KNN</i><br><i>LassoLars</i><br><i>SGD</i><br><i>RandomForest</i><br><i>ExtremeRandomTrees</i><br><i>LightGBM</i><br><i>TensorFlowLinearRegressor</i><br><i>TensorFlowDNN</i>|\n",
"|**X**|(sparse) array-like, shape = [n_samples, n_features]|\n",
"|**y**|(sparse) array-like, shape = [n_samples, ], [n_samples, n_classes]<br>Multi-class targets. An indicator matrix turns on multilabel classification. This should be an array of integers.|\n",
"|**path**|Relative path to the project folder. AutoML stores configuration files for the experiment under this folder. You can specify a new empty folder.|"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_config = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" primary_metric = 'AUC_weighted',\n",
" iteration_timeout_minutes = 60,\n",
" iterations = 20,\n",
" n_cross_validations = 5,\n",
" preprocess = True,\n",
" experiment_exit_score = 0.9984,\n",
" blacklist_models = ['KNN','LinearSVM'],\n",
" verbosity = logging.INFO,\n",
" X = X_train, \n",
" y = y_train,\n",
" path = project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train the Models\n",
"\n",
"Call the `submit` method on the experiment object and pass the run configuration. Execution of local runs is synchronous. Depending on the data and the number of iterations this can run for a while.\n",
"In this example, we specify `show_output = True` to print currently running iterations to the console."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_config, show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explore the Results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Widget for Monitoring Runs\n",
"\n",
"The widget will first report a \"loading\" status while running the first iteration. After completing the first iteration, an auto-updating graph and table will be shown. The widget will refresh once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"**Note:** The widget displays a link at the bottom. Use this link to open a web interface to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.widgets import RunDetails\n",
"RunDetails(local_run).show() "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"#### Retrieve All Child Runs\n",
"You can also use SDK methods to fetch all the child runs and see individual metrics that we log."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"children = list(local_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
"\n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"rundata"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method returns the best run and the fitted model. The Model includes the pipeline and any pre-processing. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = local_run.get_output()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Best Model Based on Any Other Metric\n",
"Show the run and the model which has the smallest `accuracy` value:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# lookup_metric = \"accuracy\"\n",
"# best_run, fitted_model = local_run.get_output(metric = lookup_metric)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Model from a Specific Iteration\n",
"Show the run and the model from the third iteration:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# iteration = 3\n",
"# best_run, fitted_model = local_run.get_output(iteration = iteration)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Testing the best Fitted Model"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits = datasets.load_digits()\n",
"X_test = digits.data[:10, :]\n",
"y_test = digits.target[:10]\n",
"images = digits.images[:10]\n",
"\n",
"# Randomly select digits and test.\n",
"for index in np.random.choice(len(y_test), 2, replace = False):\n",
" print(index)\n",
" predicted = fitted_model.predict(X_test[index:index + 1])[0]\n",
" label = y_test[index]\n",
" title = \"Label value = %d Predicted value = %d \" % (label, predicted)\n",
" fig = plt.figure(1, figsize=(3,3))\n",
" ax1 = fig.add_axes((0,0,.8,.8))\n",
" ax1.set_title(title)\n",
" plt.imshow(images[index], cmap = plt.cm.gray_r, interpolation = 'nearest')\n",
" plt.show()\n"
]
}
],
"metadata": {
"authors": [
{
"name": "savitam"
}
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 05: Blacklisting Models, Early Termination, and Handling Missing Data\n",
"\n",
"In this example we use the scikit-learn's [digit dataset](http://scikit-learn.org/stable/datasets/index.html#optical-recognition-of-handwritten-digits-dataset) to showcase how you can use AutoML for handling missing values in data. We also provide a stopping metric indicating a target for the primary metrics so that AutoML can terminate the run without necessarly going through all the iterations. Finally, if you want to avoid a certain pipeline, we allow you to specify a blacklist of algorithms that AutoML will ignore for this run.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. Create an `Experiment` in an existing `Workspace`.\n",
"2. Configure AutoML using `AutoMLConfig`.\n",
"4. Train the model.\n",
"5. Explore the results.\n",
"6. Test the best fitted model.\n",
"\n",
"In addition this notebook showcases the following features\n",
"- **Blacklisting** certain pipelines\n",
"- Specifying **target metrics** to indicate stopping criteria\n",
"- Handling **missing data** in the input\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# Choose a name for the experiment.\n",
"experiment_name = 'automl-local-missing-data'\n",
"project_folder = './sample_projects/automl-local-missing-data'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data=output, index=['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Creating missing data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from scipy import sparse\n",
"\n",
"digits = datasets.load_digits()\n",
"X_train = digits.data[10:,:]\n",
"y_train = digits.target[10:]\n",
"\n",
"# Add missing values in 75% of the lines.\n",
"missing_rate = 0.75\n",
"n_missing_samples = int(np.floor(X_train.shape[0] * missing_rate))\n",
"missing_samples = np.hstack((np.zeros(X_train.shape[0] - n_missing_samples, dtype=np.bool), np.ones(n_missing_samples, dtype=np.bool)))\n",
"rng = np.random.RandomState(0)\n",
"rng.shuffle(missing_samples)\n",
"missing_features = rng.randint(0, X_train.shape[1], n_missing_samples)\n",
"X_train[np.where(missing_samples)[0], missing_features] = np.nan"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"df = pd.DataFrame(data = X_train)\n",
"df['Label'] = pd.Series(y_train, index=df.index)\n",
"df.head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML\n",
"\n",
"Instantiate an `AutoMLConfig` object to specify the settings and data used to run the experiment. This includes setting `exit_score`, which should cause the run to complete before the `iterations` count is reached.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**task**|classification or regression|\n",
"|**primary_metric**|This is the metric that you want to optimize. Classification supports the following primary metrics: <br><i>accuracy</i><br><i>AUC_weighted</i><br><i>balanced_accuracy</i><br><i>average_precision_score_weighted</i><br><i>precision_score_weighted</i>|\n",
"|**max_time_sec**|Time limit in seconds for each iteration.|\n",
"|**iterations**|Number of iterations. In each iteration AutoML trains a specific pipeline with the data.|\n",
"|**n_cross_validations**|Number of cross validation splits.|\n",
"|**preprocess**|Setting this to *True* enables AutoML to perform preprocessing on the input to handle *missing data*, and to perform some common *feature extraction*.|\n",
"|**exit_score**|*double* value indicating the target for *primary_metric*. <br>Once the target is surpassed the run terminates.|\n",
"|**blacklist_algos**|*List* of *strings* indicating machine learning algorithms for AutoML to avoid in this run.<br><br> Allowed values for **Classification**<br><i>LogisticRegression</i><br><i>SGDClassifierWrapper</i><br><i>NBWrapper</i><br><i>BernoulliNB</i><br><i>SVCWrapper</i><br><i>LinearSVMWrapper</i><br><i>KNeighborsClassifier</i><br><i>DecisionTreeClassifier</i><br><i>RandomForestClassifier</i><br><i>ExtraTreesClassifier</i><br><i>LightGBMClassifier</i><br><br>Allowed values for **Regression**<br><i>ElasticNet<i><br><i>GradientBoostingRegressor<i><br><i>DecisionTreeRegressor<i><br><i>KNeighborsRegressor<i><br><i>LassoLars<i><br><i>SGDRegressor<i><br><i>RandomForestRegressor<i><br><i>ExtraTreesRegressor<i>|\n",
"|**X**|(sparse) array-like, shape = [n_samples, n_features]|\n",
"|**y**|(sparse) array-like, shape = [n_samples, ], [n_samples, n_classes]<br>Multi-class targets. An indicator matrix turns on multilabel classification. This should be an array of integers.|\n",
"|**path**|Relative path to the project folder. AutoML stores configuration files for the experiment under this folder. You can specify a new empty folder.|"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_config = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" primary_metric = 'AUC_weighted',\n",
" max_time_sec = 3600,\n",
" iterations = 20,\n",
" n_cross_validations = 5,\n",
" preprocess = True,\n",
" exit_score = 0.9984,\n",
" blacklist_algos = ['KNeighborsClassifier','LinearSVMWrapper'],\n",
" verbosity = logging.INFO,\n",
" X = X_train, \n",
" y = y_train,\n",
" path = project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train the Models\n",
"\n",
"Call the `submit` method on the experiment object and pass the run configuration. Execution of local runs is synchronous. Depending on the data and the number of iterations this can run for a while.\n",
"In this example, we specify `show_output = True` to print currently running iterations to the console."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_config, show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explore the Results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Widget for Monitoring Runs\n",
"\n",
"The widget will first report a \"loading\" status while running the first iteration. After completing the first iteration, an auto-updating graph and table will be shown. The widget will refresh once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"**Note:** The widget displays a link at the bottom. Use this link to open a web interface to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.widgets import RunDetails\n",
"RunDetails(local_run).show() "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"#### Retrieve All Child Runs\n",
"You can also use SDK methods to fetch all the child runs and see individual metrics that we log."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"children = list(local_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
"\n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"rundata"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method returns the best run and the fitted model. The Model includes the pipeline and any pre-processing. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = local_run.get_output()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Best Model Based on Any Other Metric\n",
"Show the run and the model which has the smallest `accuracy` value:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# lookup_metric = \"accuracy\"\n",
"# best_run, fitted_model = local_run.get_output(metric = lookup_metric)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Model from a Specific Iteration\n",
"Show the run and the model from the third iteration:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# iteration = 3\n",
"# best_run, fitted_model = local_run.get_output(iteration = iteration)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Testing the best Fitted Model"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits = datasets.load_digits()\n",
"X_test = digits.data[:10, :]\n",
"y_test = digits.target[:10]\n",
"images = digits.images[:10]\n",
"\n",
"# Randomly select digits and test.\n",
"for index in np.random.choice(len(y_test), 2, replace = False):\n",
" print(index)\n",
" predicted = fitted_model.predict(X_test[index:index + 1])[0]\n",
" label = y_test[index]\n",
" title = \"Label value = %d Predicted value = %d \" % (label, predicted)\n",
" fig = plt.figure(1, figsize=(3,3))\n",
" ax1 = fig.add_axes((0,0,.8,.8))\n",
" ax1.set_title(title)\n",
" plt.imshow(images[index], cmap = plt.cm.gray_r, interpolation = 'nearest')\n",
" plt.show()\n"
]
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
"metadata": {
"authors": [
{
"name": "savitam"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
"nbformat": 4,
"nbformat_minor": 2
}

762
automl/06.auto-ml-sparse-data-train-test-split.ipynb
View File

@@ -1,384 +1,384 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 06: Train Test Split and Handling Sparse Data\n",
"\n",
"In this example we use the scikit-learn's [20newsgroup](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.fetch_20newsgroups.html) to showcase how you can use AutoML for handling sparse data and how to specify custom cross validations splits.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. Create an `Experiment` in an existing `Workspace`.\n",
"2. Configure AutoML using `AutoMLConfig`.\n",
"4. Train the model.\n",
"5. Explore the results.\n",
"6. Test the best fitted model.\n",
"\n",
"In addition this notebook showcases the following features\n",
"- Explicit train test splits \n",
"- Handling **sparse data** in the input"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# choose a name for the experiment\n",
"experiment_name = 'automl-local-missing-data'\n",
"# project folder\n",
"project_folder = './sample_projects/automl-local-missing-data'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data=output, index=['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Creating Sparse Data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.datasets import fetch_20newsgroups\n",
"from sklearn.feature_extraction.text import HashingVectorizer\n",
"from sklearn.model_selection import train_test_split\n",
"\n",
"remove = ('headers', 'footers', 'quotes')\n",
"categories = [\n",
" 'alt.atheism',\n",
" 'talk.religion.misc',\n",
" 'comp.graphics',\n",
" 'sci.space',\n",
"]\n",
"data_train = fetch_20newsgroups(subset = 'train', categories = categories,\n",
" shuffle = True, random_state = 42,\n",
" remove = remove)\n",
"\n",
"X_train, X_valid, y_train, y_valid = train_test_split(data_train.data, data_train.target, test_size = 0.33, random_state = 42)\n",
"\n",
"\n",
"vectorizer = HashingVectorizer(stop_words = 'english', alternate_sign = False,\n",
" n_features = 2**16)\n",
"X_train = vectorizer.transform(X_train)\n",
"X_valid = vectorizer.transform(X_valid)\n",
"\n",
"summary_df = pd.DataFrame(index = ['No of Samples', 'No of Features'])\n",
"summary_df['Train Set'] = [X_train.shape[0], X_train.shape[1]]\n",
"summary_df['Validation Set'] = [X_valid.shape[0], X_valid.shape[1]]\n",
"summary_df"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML\n",
"\n",
"Instantiate an `AutoMLConfig` object to specify the settings and data used to run the experiment.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**task**|classification or regression|\n",
"|**primary_metric**|This is the metric that you want to optimize. Classification supports the following primary metrics: <br><i>accuracy</i><br><i>AUC_weighted</i><br><i>average_precision_score_weighted</i><br><i>norm_macro_recall</i><br><i>precision_score_weighted</i>|\n",
"|**iteration_timeout_minutes**|Time limit in minutes for each iteration.|\n",
"|**iterations**|Number of iterations. In each iteration AutoML trains a specific pipeline with the data.|\n",
"|**preprocess**|Setting this to *True* enables AutoML to perform preprocessing on the input to handle *missing data*, and to perform some common *feature extraction*.<br>**Note:** If input data is sparse, you cannot use *True*.|\n",
"|**X**|(sparse) array-like, shape = [n_samples, n_features]|\n",
"|**y**|(sparse) array-like, shape = [n_samples, ], [n_samples, n_classes]<br>Multi-class targets. An indicator matrix turns on multilabel classification. This should be an array of integers.|\n",
"|**X_valid**|(sparse) array-like, shape = [n_samples, n_features] for the custom validation set.|\n",
"|**y_valid**|(sparse) array-like, shape = [n_samples, ], [n_samples, n_classes]<br>Multi-class targets. An indicator matrix turns on multilabel classification for the custom validation set.|\n",
"|**path**|Relative path to the project folder. AutoML stores configuration files for the experiment under this folder. You can specify a new empty folder.|"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_config = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" primary_metric = 'AUC_weighted',\n",
" iteration_timeout_minutes = 60,\n",
" iterations = 5,\n",
" preprocess = False,\n",
" verbosity = logging.INFO,\n",
" X = X_train, \n",
" y = y_train,\n",
" X_valid = X_valid, \n",
" y_valid = y_valid, \n",
" path = project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train the Models\n",
"\n",
"Call the `submit` method on the experiment object and pass the run configuration. Execution of local runs is synchronous. Depending on the data and the number of iterations this can run for a while.\n",
"In this example, we specify `show_output = True` to print currently running iterations to the console."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_config, show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explore the Results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Widget for Monitoring Runs\n",
"\n",
"The widget will first report a \"loading\" status while running the first iteration. After completing the first iteration, an auto-updating graph and table will be shown. The widget will refresh once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"**Note:** The widget displays a link at the bottom. Use this link to open a web interface to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.widgets import RunDetails\n",
"RunDetails(local_run).show() "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"#### Retrieve All Child Runs\n",
"You can also use SDK methods to fetch all the child runs and see individual metrics that we log."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"children = list(local_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
" \n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"rundata"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method returns the best run and the fitted model. The Model includes the pipeline and any pre-processing. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = local_run.get_output()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Best Model Based on Any Other Metric\n",
"Show the run and the model which has the smallest `accuracy` value:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# lookup_metric = \"accuracy\"\n",
"# best_run, fitted_model = local_run.get_output(metric = lookup_metric)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Model from a Specific Iteration\n",
"Show the run and the model from the third iteration:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# iteration = 3\n",
"# best_run, fitted_model = local_run.get_output(iteration = iteration)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Testing the Best Fitted Model"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Load test data.\n",
"from pandas_ml import ConfusionMatrix\n",
"\n",
"data_test = fetch_20newsgroups(subset = 'test', categories = categories,\n",
" shuffle = True, random_state = 42,\n",
" remove = remove)\n",
"\n",
"X_test = vectorizer.transform(data_test.data)\n",
"y_test = data_test.target\n",
"\n",
"# Test our best pipeline.\n",
"\n",
"y_pred = fitted_model.predict(X_test)\n",
"y_pred_strings = [data_test.target_names[i] for i in y_pred]\n",
"y_test_strings = [data_test.target_names[i] for i in y_test]\n",
"\n",
"cm = ConfusionMatrix(y_test_strings, y_pred_strings)\n",
"print(cm)\n",
"cm.plot()"
]
}
],
"metadata": {
"authors": [
{
"name": "savitam"
}
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 06: Train Test Split and Handling Sparse Data\n",
"\n",
"In this example we use the scikit-learn's [20newsgroup](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.fetch_20newsgroups.html) to showcase how you can use AutoML for handling sparse data and how to specify custom cross validations splits.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. Create an `Experiment` in an existing `Workspace`.\n",
"2. Configure AutoML using `AutoMLConfig`.\n",
"4. Train the model.\n",
"5. Explore the results.\n",
"6. Test the best fitted model.\n",
"\n",
"In addition this notebook showcases the following features\n",
"- Explicit train test splits \n",
"- Handling **sparse data** in the input"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# choose a name for the experiment\n",
"experiment_name = 'automl-local-missing-data'\n",
"# project folder\n",
"project_folder = './sample_projects/automl-local-missing-data'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data=output, index=['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Creating Sparse Data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.datasets import fetch_20newsgroups\n",
"from sklearn.feature_extraction.text import HashingVectorizer\n",
"from sklearn.model_selection import train_test_split\n",
"\n",
"remove = ('headers', 'footers', 'quotes')\n",
"categories = [\n",
" 'alt.atheism',\n",
" 'talk.religion.misc',\n",
" 'comp.graphics',\n",
" 'sci.space',\n",
"]\n",
"data_train = fetch_20newsgroups(subset = 'train', categories = categories,\n",
" shuffle = True, random_state = 42,\n",
" remove = remove)\n",
"\n",
"X_train, X_valid, y_train, y_valid = train_test_split(data_train.data, data_train.target, test_size = 0.33, random_state = 42)\n",
"\n",
"\n",
"vectorizer = HashingVectorizer(stop_words = 'english', alternate_sign = False,\n",
" n_features = 2**16)\n",
"X_train = vectorizer.transform(X_train)\n",
"X_valid = vectorizer.transform(X_valid)\n",
"\n",
"summary_df = pd.DataFrame(index = ['No of Samples', 'No of Features'])\n",
"summary_df['Train Set'] = [X_train.shape[0], X_train.shape[1]]\n",
"summary_df['Validation Set'] = [X_valid.shape[0], X_valid.shape[1]]\n",
"summary_df"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML\n",
"\n",
"Instantiate an `AutoMLConfig` object to specify the settings and data used to run the experiment.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**task**|classification or regression|\n",
"|**primary_metric**|This is the metric that you want to optimize. Classification supports the following primary metrics: <br><i>accuracy</i><br><i>AUC_weighted</i><br><i>balanced_accuracy</i><br><i>average_precision_score_weighted</i><br><i>precision_score_weighted</i>|\n",
"|**max_time_sec**|Time limit in seconds for each iteration.|\n",
"|**iterations**|Number of iterations. In each iteration AutoML trains a specific pipeline with the data.|\n",
"|**preprocess**|Setting this to *True* enables AutoML to perform preprocessing on the input to handle *missing data*, and to perform some common *feature extraction*.<br>**Note:** If input data is sparse, you cannot use *True*.|\n",
"|**X**|(sparse) array-like, shape = [n_samples, n_features]|\n",
"|**y**|(sparse) array-like, shape = [n_samples, ], [n_samples, n_classes]<br>Multi-class targets. An indicator matrix turns on multilabel classification. This should be an array of integers.|\n",
"|**X_valid**|(sparse) array-like, shape = [n_samples, n_features] for the custom validation set.|\n",
"|**y_valid**|(sparse) array-like, shape = [n_samples, ], [n_samples, n_classes]<br>Multi-class targets. An indicator matrix turns on multilabel classification for the custom validation set.|\n",
"|**path**|Relative path to the project folder. AutoML stores configuration files for the experiment under this folder. You can specify a new empty folder.|"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_config = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" primary_metric = 'AUC_weighted',\n",
" max_time_sec = 3600,\n",
" iterations = 5,\n",
" preprocess = False,\n",
" verbosity = logging.INFO,\n",
" X = X_train, \n",
" y = y_train,\n",
" X_valid = X_valid, \n",
" y_valid = y_valid, \n",
" path = project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train the Models\n",
"\n",
"Call the `submit` method on the experiment object and pass the run configuration. Execution of local runs is synchronous. Depending on the data and the number of iterations this can run for a while.\n",
"In this example, we specify `show_output = True` to print currently running iterations to the console."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_config, show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explore the Results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Widget for Monitoring Runs\n",
"\n",
"The widget will first report a \"loading\" status while running the first iteration. After completing the first iteration, an auto-updating graph and table will be shown. The widget will refresh once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"**Note:** The widget displays a link at the bottom. Use this link to open a web interface to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.widgets import RunDetails\n",
"RunDetails(local_run).show() "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"#### Retrieve All Child Runs\n",
"You can also use SDK methods to fetch all the child runs and see individual metrics that we log."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"children = list(local_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
" \n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"rundata"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method returns the best run and the fitted model. The Model includes the pipeline and any pre-processing. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = local_run.get_output()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Best Model Based on Any Other Metric\n",
"Show the run and the model which has the smallest `accuracy` value:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# lookup_metric = \"accuracy\"\n",
"# best_run, fitted_model = local_run.get_output(metric = lookup_metric)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Model from a Specific Iteration\n",
"Show the run and the model from the third iteration:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# iteration = 3\n",
"# best_run, fitted_model = local_run.get_output(iteration = iteration)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Testing the Best Fitted Model"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Load test data.\n",
"from pandas_ml import ConfusionMatrix\n",
"\n",
"data_test = fetch_20newsgroups(subset = 'test', categories = categories,\n",
" shuffle = True, random_state = 42,\n",
" remove = remove)\n",
"\n",
"X_test = vectorizer.transform(data_test.data)\n",
"y_test = data_test.target\n",
"\n",
"# Test our best pipeline.\n",
"\n",
"y_pred = fitted_model.predict(X_test)\n",
"y_pred_strings = [data_test.target_names[i] for i in y_pred]\n",
"y_test_strings = [data_test.target_names[i] for i in y_test]\n",
"\n",
"cm = ConfusionMatrix(y_test_strings, y_pred_strings)\n",
"print(cm)\n",
"cm.plot()"
]
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
"metadata": {
"authors": [
{
"name": "savitam"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
"nbformat": 4,
"nbformat_minor": 2
}

666
automl/07.auto-ml-exploring-previous-runs.ipynb
View File

@@ -1,336 +1,336 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 07: Exploring Previous Runs\n",
"\n",
"In this example we present some examples on navigating previously executed runs. We also show how you can download a fitted model for any previous run.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. List all experiments in a workspace.\n",
"2. List all AutoML runs in an experiment.\n",
"3. Get details for an AutoML run, including settings, run widget, and all metrics.\n",
"4. Download a fitted pipeline for any iteration.\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# List all AutoML Experiments in a Workspace"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"import re\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.run import Run\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"experiment_list = Experiment.list(workspace=ws)\n",
"\n",
"summary_df = pd.DataFrame(index = ['No of Runs'])\n",
"pattern = re.compile('^AutoML_[^_]*$')\n",
"for experiment in experiment_list:\n",
" all_runs = list(experiment.get_runs())\n",
" automl_runs = []\n",
" for run in all_runs:\n",
" if(pattern.match(run.id)):\n",
" automl_runs.append(run) \n",
" summary_df[experiment.name] = [len(automl_runs)]\n",
" \n",
"pd.set_option('display.max_colwidth', -1)\n",
"summary_df.T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# List AutoML runs for an experiment\n",
"Set `experiment_name` to any experiment name from the result of the Experiment.list cell to load the AutoML runs."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"experiment_name = 'automl-local-classification' # Replace this with any project name from previous cell.\n",
"\n",
"proj = ws.experiments[experiment_name]\n",
"summary_df = pd.DataFrame(index = ['Type', 'Status', 'Primary Metric', 'Iterations', 'Compute', 'Name'])\n",
"pattern = re.compile('^AutoML_[^_]*$')\n",
"all_runs = list(proj.get_runs(properties={'azureml.runsource': 'automl'}))\n",
"automl_runs_project = []\n",
"for run in all_runs:\n",
" if(pattern.match(run.id)):\n",
" properties = run.get_properties()\n",
" tags = run.get_tags()\n",
" amlsettings = eval(properties['RawAMLSettingsString'])\n",
" if 'iterations' in tags:\n",
" iterations = tags['iterations']\n",
" else:\n",
" iterations = properties['num_iterations']\n",
" summary_df[run.id] = [amlsettings['task_type'], run.get_details()['status'], properties['primary_metric'], iterations, properties['target'], amlsettings['name']]\n",
" if run.get_details()['status'] == 'Completed':\n",
" automl_runs_project.append(run.id)\n",
" \n",
"from IPython.display import HTML\n",
"projname_html = HTML(\"<h3>{}</h3>\".format(proj.name))\n",
"\n",
"from IPython.display import display\n",
"display(projname_html)\n",
"display(summary_df.T)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Get details for an AutoML run\n",
"\n",
"Copy the project name and run id from the previous cell output to find more details on a particular run."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run_id = automl_runs_project[0] # Replace with your own run_id from above run ids\n",
"assert (run_id in summary_df.keys()), \"Run id not found! Please set run id to a value from above run ids\"\n",
"\n",
"from azureml.widgets import RunDetails\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"ml_run = AutoMLRun(experiment = experiment, run_id = run_id)\n",
"\n",
"summary_df = pd.DataFrame(index = ['Type', 'Status', 'Primary Metric', 'Iterations', 'Compute', 'Name', 'Start Time', 'End Time'])\n",
"properties = ml_run.get_properties()\n",
"tags = ml_run.get_tags()\n",
"status = ml_run.get_details()\n",
"amlsettings = eval(properties['RawAMLSettingsString'])\n",
"if 'iterations' in tags:\n",
" iterations = tags['iterations']\n",
"else:\n",
" iterations = properties['num_iterations']\n",
"start_time = None\n",
"if 'startTimeUtc' in status:\n",
" start_time = status['startTimeUtc']\n",
"end_time = None\n",
"if 'endTimeUtc' in status:\n",
" end_time = status['endTimeUtc']\n",
"summary_df[ml_run.id] = [amlsettings['task_type'], status['status'], properties['primary_metric'], iterations, properties['target'], amlsettings['name'], start_time, end_time]\n",
"display(HTML('<h3>Runtime Details</h3>'))\n",
"display(summary_df)\n",
"\n",
"#settings_df = pd.DataFrame(data = amlsettings, index = [''])\n",
"display(HTML('<h3>AutoML Settings</h3>'))\n",
"display(amlsettings)\n",
"\n",
"display(HTML('<h3>Iterations</h3>'))\n",
"RunDetails(ml_run).show() \n",
"\n",
"children = list(ml_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
"\n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"display(HTML('<h3>Metrics</h3>'))\n",
"display(rundata)\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Download fitted models"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Download the Best Model for Any Given Metric"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"metric = 'AUC_weighted' # Replace with a metric name.\n",
"best_run, fitted_model = ml_run.get_output(metric = metric)\n",
"fitted_model"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Download the Model for Any Given Iteration"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"iteration = 1 # Replace with an iteration number.\n",
"best_run, fitted_model = ml_run.get_output(iteration = iteration)\n",
"fitted_model"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Register fitted model for deployment\n",
"If neither `metric` nor `iteration` are specified in the `register_model` call, the iteration with the best primary metric is registered."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"description = 'AutoML Model'\n",
"tags = None\n",
"ml_run.register_model(description = description, tags = tags)\n",
"ml_run.model_id # Use this id to deploy the model as a web service in Azure."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Register the Best Model for Any Given Metric"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"metric = 'AUC_weighted' # Replace with a metric name.\n",
"description = 'AutoML Model'\n",
"tags = None\n",
"ml_run.register_model(description = description, tags = tags, metric = metric)\n",
"print(ml_run.model_id) # Use this id to deploy the model as a web service in Azure."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Register the Model for Any Given Iteration"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"iteration = 1 # Replace with an iteration number.\n",
"description = 'AutoML Model'\n",
"tags = None\n",
"ml_run.register_model(description = description, tags = tags, iteration = iteration)\n",
"print(ml_run.model_id) # Use this id to deploy the model as a web service in Azure."
]
}
],
"metadata": {
"authors": [
{
"name": "savitam"
}
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 07: Exploring Previous Runs\n",
"\n",
"In this example we present some examples on navigating previously executed runs. We also show how you can download a fitted model for any previous run.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. List all experiments in a workspace.\n",
"2. List all AutoML runs in an experiment.\n",
"3. Get details for an AutoML run, including settings, run widget, and all metrics.\n",
"4. Download a fitted pipeline for any iteration.\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# List all AutoML Experiments in a Workspace"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"import re\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.run import Run\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"experiment_list = Experiment.list(workspace=ws)\n",
"\n",
"summary_df = pd.DataFrame(index = ['No of Runs'])\n",
"pattern = re.compile('^AutoML_[^_]*$')\n",
"for experiment in experiment_list:\n",
" all_runs = list(experiment.get_runs())\n",
" automl_runs = []\n",
" for run in all_runs:\n",
" if(pattern.match(run.id)):\n",
" automl_runs.append(run) \n",
" summary_df[experiment.name] = [len(automl_runs)]\n",
" \n",
"pd.set_option('display.max_colwidth', -1)\n",
"summary_df.T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# List AutoML runs for an experiment\n",
"Set `experiment_name` to any experiment name from the result of the Experiment.list cell to load the AutoML runs."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"experiment_name = 'automl-local-classification' # Replace this with any project name from previous cell.\n",
"\n",
"proj = ws.experiments[experiment_name]\n",
"summary_df = pd.DataFrame(index = ['Type', 'Status', 'Primary Metric', 'Iterations', 'Compute', 'Name'])\n",
"pattern = re.compile('^AutoML_[^_]*$')\n",
"all_runs = list(proj.get_runs(properties={'azureml.runsource': 'automl'}))\n",
"automl_runs_project = []\n",
"for run in all_runs:\n",
" if(pattern.match(run.id)):\n",
" properties = run.get_properties()\n",
" tags = run.get_tags()\n",
" amlsettings = eval(properties['RawAMLSettingsString'])\n",
" if 'iterations' in tags:\n",
" iterations = tags['iterations']\n",
" else:\n",
" iterations = properties['num_iterations']\n",
" summary_df[run.id] = [amlsettings['task_type'], run.get_details()['status'], properties['primary_metric'], iterations, properties['target'], amlsettings['name']]\n",
" if run.get_details()['status'] == 'Completed':\n",
" automl_runs_project.append(run.id)\n",
" \n",
"from IPython.display import HTML\n",
"projname_html = HTML(\"<h3>{}</h3>\".format(proj.name))\n",
"\n",
"from IPython.display import display\n",
"display(projname_html)\n",
"display(summary_df.T)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Get details for an AutoML run\n",
"\n",
"Copy the project name and run id from the previous cell output to find more details on a particular run."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run_id = automl_runs_project[0] # Replace with your own run_id from above run ids\n",
"assert (run_id in summary_df.keys()), \"Run id not found! Please set run id to a value from above run ids\"\n",
"\n",
"from azureml.train.widgets import RunDetails\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"ml_run = AutoMLRun(experiment = experiment, run_id = run_id)\n",
"\n",
"summary_df = pd.DataFrame(index = ['Type', 'Status', 'Primary Metric', 'Iterations', 'Compute', 'Name', 'Start Time', 'End Time'])\n",
"properties = ml_run.get_properties()\n",
"tags = ml_run.get_tags()\n",
"status = ml_run.get_details()\n",
"amlsettings = eval(properties['RawAMLSettingsString'])\n",
"if 'iterations' in tags:\n",
" iterations = tags['iterations']\n",
"else:\n",
" iterations = properties['num_iterations']\n",
"start_time = None\n",
"if 'startTimeUtc' in status:\n",
" start_time = status['startTimeUtc']\n",
"end_time = None\n",
"if 'endTimeUtc' in status:\n",
" end_time = status['endTimeUtc']\n",
"summary_df[ml_run.id] = [amlsettings['task_type'], status['status'], properties['primary_metric'], iterations, properties['target'], amlsettings['name'], start_time, end_time]\n",
"display(HTML('<h3>Runtime Details</h3>'))\n",
"display(summary_df)\n",
"\n",
"#settings_df = pd.DataFrame(data = amlsettings, index = [''])\n",
"display(HTML('<h3>AutoML Settings</h3>'))\n",
"display(amlsettings)\n",
"\n",
"display(HTML('<h3>Iterations</h3>'))\n",
"RunDetails(ml_run).show() \n",
"\n",
"children = list(ml_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
"\n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"display(HTML('<h3>Metrics</h3>'))\n",
"display(rundata)\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Download fitted models"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Download the Best Model for Any Given Metric"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"metric = 'AUC_weighted' # Replace with a metric name.\n",
"best_run, fitted_model = ml_run.get_output(metric = metric)\n",
"fitted_model"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Download the Model for Any Given Iteration"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"iteration = 4 # Replace with an iteration number.\n",
"best_run, fitted_model = ml_run.get_output(iteration = iteration)\n",
"fitted_model"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Register fitted model for deployment\n",
"If neither `metric` nor `iteration` are specified in the `register_model` call, the iteration with the best primary metric is registered."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"description = 'AutoML Model'\n",
"tags = None\n",
"ml_run.register_model(description = description, tags = tags)\n",
"ml_run.model_id # Use this id to deploy the model as a web service in Azure."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Register the Best Model for Any Given Metric"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"metric = 'AUC_weighted' # Replace with a metric name.\n",
"description = 'AutoML Model'\n",
"tags = None\n",
"ml_run.register_model(description = description, tags = tags, metric = metric)\n",
"print(ml_run.model_id) # Use this id to deploy the model as a web service in Azure."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Register the Model for Any Given Iteration"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"iteration = 4 # Replace with an iteration number.\n",
"description = 'AutoML Model'\n",
"tags = None\n",
"ml_run.register_model(description = description, tags = tags, iteration = iteration)\n",
"print(ml_run.model_id) # Use this id to deploy the model as a web service in Azure."
]
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
"metadata": {
"authors": [
{
"name": "savitam"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
"nbformat": 4,
"nbformat_minor": 2
}

1132
automl/08.auto-ml-remote-execution-with-DataStore.ipynb
View File

File diff suppressed because it is too large Load Diff

995
automl/09.auto-ml-classification-with-deployment.ipynb
View File

@@ -1,501 +1,500 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 09: Classification with Deployment\n",
"\n",
"In this example we use the scikit learn's [digit dataset](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_digits.html) to showcase how you can use AutoML for a simple classification problem and deploy it to an Azure Container Instance (ACI).\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. Create an experiment using an existing workspace.\n",
"2. Configure AutoML using `AutoMLConfig`.\n",
"3. Train the model using local compute.\n",
"4. Explore the results.\n",
"5. Register the model.\n",
"6. Create a container image.\n",
"7. Create an Azure Container Instance (ACI) service.\n",
"8. Test the ACI service.\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import json\n",
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# choose a name for experiment\n",
"experiment_name = 'automl-local-classification'\n",
"# project folder\n",
"project_folder = './sample_projects/automl-local-classification'\n",
"\n",
"experiment=Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data=output, index=['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML\n",
"\n",
"Instantiate a AutoMLConfig object. This defines the settings and data used to run the experiment.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**task**|classification or regression|\n",
"|**primary_metric**|This is the metric that you want to optimize. Classification supports the following primary metrics: <br><i>accuracy</i><br><i>AUC_weighted</i><br><i>average_precision_score_weighted</i><br><i>norm_macro_recall</i><br><i>precision_score_weighted</i>|\n",
"|**iteration_timeout_minutes**|Time limit in minutes for each iteration.|\n",
"|**iterations**|Number of iterations. In each iteration AutoML trains a specific pipeline with the data.|\n",
"|**n_cross_validations**|Number of cross validation splits.|\n",
"|**X**|(sparse) array-like, shape = [n_samples, n_features]|\n",
"|**y**|(sparse) array-like, shape = [n_samples, ], [n_samples, n_classes]<br>Multi-class targets. An indicator matrix turns on multilabel classification. This should be an array of integers.|\n",
"|**path**|Relative path to the project folder. AutoML stores configuration files for the experiment under this folder. You can specify a new empty folder.|"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits = datasets.load_digits()\n",
"X_train = digits.data[10:,:]\n",
"y_train = digits.target[10:]\n",
"\n",
"automl_config = AutoMLConfig(task = 'classification',\n",
" name = experiment_name,\n",
" debug_log = 'automl_errors.log',\n",
" primary_metric = 'AUC_weighted',\n",
" iteration_timeout_minutes = 20,\n",
" iterations = 10,\n",
" n_cross_validations = 2,\n",
" verbosity = logging.INFO,\n",
" X = X_train, \n",
" y = y_train,\n",
" path = project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train the Models\n",
"\n",
"Call the `submit` method on the experiment object and pass the run configuration. Execution of local runs is synchronous. Depending on the data and the number of iterations this can run for a while.\n",
"In this example, we specify `show_output = True` to print currently running iterations to the console."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_config, show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method on `automl_classifier` returns the best run and the fitted model for the last invocation. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = local_run.get_output()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Register the Fitted Model for Deployment\n",
"If neither `metric` nor `iteration` are specified in the `register_model` call, the iteration with the best primary metric is registered."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"description = 'AutoML Model'\n",
"tags = None\n",
"model = local_run.register_model(description = description, tags = tags)\n",
"local_run.model_id # This will be written to the script file later in the notebook."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create Scoring Script"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile score.py\n",
"import pickle\n",
"import json\n",
"import numpy\n",
"import azureml.train.automl\n",
"from sklearn.externals import joblib\n",
"from azureml.core.model import Model\n",
"\n",
"\n",
"def init():\n",
" global model\n",
" model_path = Model.get_model_path(model_name = '<<modelid>>') # this name is model.id of model that we want to deploy\n",
" # deserialize the model file back into a sklearn model\n",
" model = joblib.load(model_path)\n",
"\n",
"def run(rawdata):\n",
" try:\n",
" data = json.loads(rawdata)['data']\n",
" data = numpy.array(data)\n",
" result = model.predict(data)\n",
" except Exception as e:\n",
" result = str(e)\n",
" return json.dumps({\"error\": result})\n",
" return json.dumps({\"result\":result.tolist()})"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create a YAML File for the Environment"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"To ensure the fit results are consistent with the training results, the SDK dependency versions need to be the same as the environment that trains the model. Details about retrieving the versions can be found in notebook [12.auto-ml-retrieve-the-training-sdk-versions](12.auto-ml-retrieve-the-training-sdk-versions.ipynb)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"experiment_name = 'automl-local-classification'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"ml_run = AutoMLRun(experiment = experiment, run_id = local_run.id)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"dependencies = ml_run.get_run_sdk_dependencies(iteration = 7)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"for p in ['azureml-train-automl', 'azureml-sdk', 'azureml-core']:\n",
" print('{}\\t{}'.format(p, dependencies[p]))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.conda_dependencies import CondaDependencies\n",
"\n",
"myenv = CondaDependencies.create(conda_packages=['numpy','scikit-learn'], pip_packages=['azureml-sdk[automl]'])\n",
"\n",
"conda_env_file_name = 'myenv.yml'\n",
"myenv.save_to_file('.', conda_env_file_name)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Substitute the actual version number in the environment file.\n",
"# This is not strictly needed in this notebook because the model should have been generated using the current SDK version.\n",
"# However, we include this in case this code is used on an experiment from a previous SDK version.\n",
"\n",
"with open(conda_env_file_name, 'r') as cefr:\n",
" content = cefr.read()\n",
"\n",
"with open(conda_env_file_name, 'w') as cefw:\n",
" cefw.write(content.replace(azureml.core.VERSION, dependencies['azureml-sdk']))\n",
"\n",
"# Substitute the actual model id in the script file.\n",
"\n",
"script_file_name = 'score.py'\n",
"\n",
"with open(script_file_name, 'r') as cefr:\n",
" content = cefr.read()\n",
"\n",
"with open(script_file_name, 'w') as cefw:\n",
" cefw.write(content.replace('<<modelid>>', local_run.model_id))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create a Container Image"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.image import Image, ContainerImage\n",
"\n",
"image_config = ContainerImage.image_configuration(runtime= \"python\",\n",
" execution_script = script_file_name,\n",
" conda_file = conda_env_file_name,\n",
" tags = {'area': \"digits\", 'type': \"automl_classification\"},\n",
" description = \"Image for automl classification sample\")\n",
"\n",
"image = Image.create(name = \"automlsampleimage\",\n",
" # this is the model object \n",
" models = [model],\n",
" image_config = image_config, \n",
" workspace = ws)\n",
"\n",
"image.wait_for_creation(show_output = True)\n",
"\n",
"if image.creation_state == 'Failed':\n",
" print(\"Image build log at: \" + image.image_build_log_uri)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Deploy the Image as a Web Service on Azure Container Instance"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.webservice import AciWebservice\n",
"\n",
"aciconfig = AciWebservice.deploy_configuration(cpu_cores = 1, \n",
" memory_gb = 1, \n",
" tags = {'area': \"digits\", 'type': \"automl_classification\"}, \n",
" description = 'sample service for Automl Classification')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.webservice import Webservice\n",
"\n",
"aci_service_name = 'automl-sample-01'\n",
"print(aci_service_name)\n",
"aci_service = Webservice.deploy_from_image(deployment_config = aciconfig,\n",
" image = image,\n",
" name = aci_service_name,\n",
" workspace = ws)\n",
"aci_service.wait_for_deployment(True)\n",
"print(aci_service.state)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Delete a Web Service"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#aci_service.delete()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Get Logs from a Deployed Web Service"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#aci_service.get_logs()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test a Web Service"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#Randomly select digits and test\n",
"digits = datasets.load_digits()\n",
"X_test = digits.data[:10, :]\n",
"y_test = digits.target[:10]\n",
"images = digits.images[:10]\n",
"\n",
"for index in np.random.choice(len(y_test), 3, replace = False):\n",
" print(index)\n",
" test_sample = json.dumps({'data':X_test[index:index + 1].tolist()})\n",
" predicted = aci_service.run(input_data = test_sample)\n",
" label = y_test[index]\n",
" predictedDict = json.loads(predicted)\n",
" title = \"Label value = %d Predicted value = %s \" % ( label,predictedDict['result'][0])\n",
" fig = plt.figure(1, figsize = (3,3))\n",
" ax1 = fig.add_axes((0,0,.8,.8))\n",
" ax1.set_title(title)\n",
" plt.imshow(images[index], cmap = plt.cm.gray_r, interpolation = 'nearest')\n",
" plt.show()"
]
}
],
"metadata": {
"authors": [
{
"name": "savitam"
}
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 09: Classification with Deployment\n",
"\n",
"In this example we use the scikit learn's [digit dataset](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_digits.html) to showcase how you can use AutoML for a simple classification problem and deploy it to an Azure Container Instance (ACI).\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. Create an experiment using an existing workspace.\n",
"2. Configure AutoML using `AutoMLConfig`.\n",
"3. Train the model using local compute.\n",
"4. Explore the results.\n",
"5. Register the model.\n",
"6. Create a container image.\n",
"7. Create an Azure Container Instance (ACI) service.\n",
"8. Test the ACI service.\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import json\n",
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# choose a name for experiment\n",
"experiment_name = 'automl-local-classification'\n",
"# project folder\n",
"project_folder = './sample_projects/automl-local-classification'\n",
"\n",
"experiment=Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data=output, index=['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML\n",
"\n",
"Instantiate a AutoMLConfig object. This defines the settings and data used to run the experiment.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**task**|classification or regression|\n",
"|**primary_metric**|This is the metric that you want to optimize. Classification supports the following primary metrics: <br><i>accuracy</i><br><i>AUC_weighted</i><br><i>balanced_accuracy</i><br><i>average_precision_score_weighted</i><br><i>precision_score_weighted</i>|\n",
"|**max_time_sec**|Time limit in seconds for each iteration.|\n",
"|**iterations**|Number of iterations. In each iteration AutoML trains a specific pipeline with the data.|\n",
"|**n_cross_validations**|Number of cross validation splits.|\n",
"|**X**|(sparse) array-like, shape = [n_samples, n_features]|\n",
"|**y**|(sparse) array-like, shape = [n_samples, ], [n_samples, n_classes]<br>Multi-class targets. An indicator matrix turns on multilabel classification. This should be an array of integers.|\n",
"|**path**|Relative path to the project folder. AutoML stores configuration files for the experiment under this folder. You can specify a new empty folder.|"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits = datasets.load_digits()\n",
"X_train = digits.data[10:,:]\n",
"y_train = digits.target[10:]\n",
"\n",
"automl_config = AutoMLConfig(task = 'classification',\n",
" name = experiment_name,\n",
" debug_log = 'automl_errors.log',\n",
" primary_metric = 'AUC_weighted',\n",
" max_time_sec = 1200,\n",
" iterations = 10,\n",
" n_cross_validations = 2,\n",
" verbosity = logging.INFO,\n",
" X = X_train, \n",
" y = y_train,\n",
" path = project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train the Models\n",
"\n",
"Call the `submit` method on the experiment object and pass the run configuration. Execution of local runs is synchronous. Depending on the data and the number of iterations this can run for a while.\n",
"In this example, we specify `show_output = True` to print currently running iterations to the console."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_config, show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method on `automl_classifier` returns the best run and the fitted model for the last invocation. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = local_run.get_output()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Register the Fitted Model for Deployment\n",
"If neither `metric` nor `iteration` are specified in the `register_model` call, the iteration with the best primary metric is registered."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"description = 'AutoML Model'\n",
"tags = None\n",
"model = local_run.register_model(description = description, tags = tags)\n",
"local_run.model_id # This will be written to the script file later in the notebook."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create Scoring Script"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile score.py\n",
"import pickle\n",
"import json\n",
"import numpy\n",
"from sklearn.externals import joblib\n",
"from azureml.core.model import Model\n",
"\n",
"\n",
"def init():\n",
" global model\n",
" model_path = Model.get_model_path(model_name = '<<modelid>>') # this name is model.id of model that we want to deploy\n",
" # deserialize the model file back into a sklearn model\n",
" model = joblib.load(model_path)\n",
"\n",
"def run(rawdata):\n",
" try:\n",
" data = json.loads(rawdata)['data']\n",
" data = numpy.array(data)\n",
" result = model.predict(data)\n",
" except Exception as e:\n",
" result = str(e)\n",
" return json.dumps({\"error\": result})\n",
" return json.dumps({\"result\":result.tolist()})"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create a YAML File for the Environment"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"To ensure the fit results are consistent with the training results, the SDK dependency versions need to be the same as the environment that trains the model. Details about retrieving the versions can be found in notebook [12.auto-ml-retrieve-the-training-sdk-versions](12.auto-ml-retrieve-the-training-sdk-versions.ipynb)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"experiment_name = 'automl-local-classification'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"ml_run = AutoMLRun(experiment = experiment, run_id = local_run.id)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"dependencies = ml_run.get_run_sdk_dependencies(iteration = 7)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"for p in ['azureml-train-automl', 'azureml-sdk', 'azureml-core']:\n",
" print('{}\\t{}'.format(p, dependencies[p]))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile myenv.yml\n",
"name: myenv\n",
"channels:\n",
" - defaults\n",
"dependencies:\n",
" - pip:\n",
" - numpy==1.14.2\n",
" - scikit-learn==0.19.2\n",
" - azureml-sdk[notebooks,automl]==<<azureml-version>>"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Substitute the actual version number in the environment file.\n",
"\n",
"conda_env_file_name = 'myenv.yml'\n",
"\n",
"with open(conda_env_file_name, 'r') as cefr:\n",
" content = cefr.read()\n",
"\n",
"with open(conda_env_file_name, 'w') as cefw:\n",
" cefw.write(content.replace('<<azureml-version>>', dependencies['azureml-sdk']))\n",
"\n",
"# Substitute the actual model id in the script file.\n",
"\n",
"script_file_name = 'score.py'\n",
"\n",
"with open(script_file_name, 'r') as cefr:\n",
" content = cefr.read()\n",
"\n",
"with open(script_file_name, 'w') as cefw:\n",
" cefw.write(content.replace('<<modelid>>', local_run.model_id))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create a Container Image"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.image import Image, ContainerImage\n",
"\n",
"image_config = ContainerImage.image_configuration(runtime= \"python\",\n",
" execution_script = script_file_name,\n",
" conda_file = conda_env_file_name,\n",
" tags = {'area': \"digits\", 'type': \"automl_classification\"},\n",
" description = \"Image for automl classification sample\")\n",
"\n",
"image = Image.create(name = \"automlsampleimage\",\n",
" # this is the model object \n",
" models = [model],\n",
" image_config = image_config, \n",
" workspace = ws)\n",
"\n",
"image.wait_for_creation(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Deploy the Image as a Web Service on Azure Container Instance"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.webservice import AciWebservice\n",
"\n",
"aciconfig = AciWebservice.deploy_configuration(cpu_cores = 1, \n",
" memory_gb = 1, \n",
" tags = {'area': \"digits\", 'type': \"automl_classification\"}, \n",
" description = 'sample service for Automl Classification')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.webservice import Webservice\n",
"\n",
"aci_service_name = 'automl-sample-01'\n",
"print(aci_service_name)\n",
"aci_service = Webservice.deploy_from_image(deployment_config = aciconfig,\n",
" image = image,\n",
" name = aci_service_name,\n",
" workspace = ws)\n",
"aci_service.wait_for_deployment(True)\n",
"print(aci_service.state)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Delete a Web Service"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#aci_service.delete()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Get Logs from a Deployed Web Service"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#aci_service.get_logs()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test a Web Service"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#Randomly select digits and test\n",
"digits = datasets.load_digits()\n",
"X_test = digits.data[:10, :]\n",
"y_test = digits.target[:10]\n",
"images = digits.images[:10]\n",
"\n",
"for index in np.random.choice(len(y_test), 3, replace = False):\n",
" print(index)\n",
" test_sample = json.dumps({'data':X_test[index:index + 1].tolist()})\n",
" predicted = aci_service.run(input_data = test_sample)\n",
" label = y_test[index]\n",
" predictedDict = json.loads(predicted)\n",
" title = \"Label value = %d Predicted value = %s \" % ( label,predictedDict['result'][0])\n",
" fig = plt.figure(1, figsize = (3,3))\n",
" ax1 = fig.add_axes((0,0,.8,.8))\n",
" ax1.set_title(title)\n",
" plt.imshow(images[index], cmap = plt.cm.gray_r, interpolation = 'nearest')\n",
" plt.show()"
]
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
"metadata": {
"authors": [
{
"name": "savitam"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
"nbformat": 4,
"nbformat_minor": 2
}

582
automl/10.auto-ml-multi-output-example.ipynb
View File

@@ -1,294 +1,294 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 10: Multi-output\n",
"\n",
"This notebook shows how to use AutoML to train multi-output problems by leveraging the correlation between the outputs using indicator vectors.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Transformer Functions\n",
"The transformations of inputs `X` and `y` are happening as follows, e.g. `y = {y_1, y_2}`, then `X` becomes\n",
" \n",
"`X 1 0`\n",
" \n",
"`X 0 1`\n",
"\n",
"and `y` becomes,\n",
"\n",
"`y_1`\n",
"\n",
"`y_2`"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from scipy import sparse\n",
"from scipy import linalg\n",
"\n",
"#Transformer functions\n",
"def multi_output_transform_x_y(X, y):\n",
" X_new = multi_output_transformer_x(X, y.shape[1])\n",
" y_new = multi_output_transform_y(y)\n",
" return X_new, y_new\n",
"\n",
"def multi_output_transformer_x(X, number_of_columns_y):\n",
" indicator_vecs = linalg.block_diag(*([np.ones((X.shape[0], 1))] * number_of_columns_y))\n",
" if sparse.issparse(X):\n",
" X_new = sparse.vstack(np.tile(X, number_of_columns_y))\n",
" indicator_vecs = sparse.coo_matrix(indicator_vecs)\n",
" X_new = sparse.hstack((X_new, indicator_vecs))\n",
" else:\n",
" X_new = np.tile(X, (number_of_columns_y, 1))\n",
" X_new = np.hstack((X_new, indicator_vecs))\n",
" return X_new\n",
"\n",
"def multi_output_transform_y(y):\n",
" return y.reshape(-1, order=\"F\")\n",
"\n",
"def multi_output_inverse_transform_y(y, number_of_columns_y):\n",
" return y.reshape((-1, number_of_columns_y), order = \"F\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## AutoML Experiment Setup"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# Choose a name for the experiment and specify the project folder.\n",
"experiment_name = 'automl-local-multi-output'\n",
"project_folder = './sample_projects/automl-local-multi-output'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data = output, index = ['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create a Random Dataset for Test Purposes"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"rng = np.random.RandomState(1)\n",
"X_train = np.sort(200 * rng.rand(600, 1) - 100, axis = 0)\n",
"y_train = np.array([np.pi * np.sin(X_train).ravel(), np.pi * np.cos(X_train).ravel()]).T\n",
"y_train += (0.5 - rng.rand(*y_train.shape))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Perform X and y transformation using the transformer function."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"X_train_transformed, y_train_transformed = multi_output_transform_x_y(X_train, y_train)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Configure AutoML using the transformed results."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_config = AutoMLConfig(task = 'regression',\n",
" debug_log = 'automl_errors_multi.log',\n",
" primary_metric = 'r2_score',\n",
" iterations = 10,\n",
" n_cross_validations = 2,\n",
" verbosity = logging.INFO,\n",
" X = X_train_transformed,\n",
" y = y_train_transformed,\n",
" path = project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Fit the Transformed Data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_config, show_output = True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Get the best fit model.\n",
"best_run, fitted_model = local_run.get_output()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Generate random data set for predicting.\n",
"X_test = np.sort(200 * rng.rand(200, 1) - 100, axis = 0)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Transform predict data.\n",
"X_test_transformed = multi_output_transformer_x(X_test, y_train.shape[1])\n",
"\n",
"# Predict and inverse transform the prediction.\n",
"y_predict = fitted_model.predict(X_test_transformed)\n",
"y_predict = multi_output_inverse_transform_y(y_predict, y_train.shape[1])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(y_predict)"
]
}
],
"metadata": {
"authors": [
{
"name": "savitam"
}
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 10: Multi-output\n",
"\n",
"This notebook shows how to use AutoML to train multi-output problems by leveraging the correlation between the outputs using indicator vectors.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Transformer Functions\n",
"The transformations of inputs `X` and `y` are happening as follows, e.g. `y = {y_1, y_2}`, then `X` becomes\n",
" \n",
"`X 1 0`\n",
" \n",
"`X 0 1`\n",
"\n",
"and `y` becomes,\n",
"\n",
"`y_1`\n",
"\n",
"`y_2`"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from scipy import sparse\n",
"from scipy import linalg\n",
"\n",
"#Transformer functions\n",
"def multi_output_transform_x_y(X, y):\n",
" X_new = multi_output_transformer_x(X, y.shape[1])\n",
" y_new = multi_output_transform_y(y)\n",
" return X_new, y_new\n",
"\n",
"def multi_output_transformer_x(X, number_of_columns_y):\n",
" indicator_vecs = linalg.block_diag(*([np.ones((X.shape[0], 1))] * number_of_columns_y))\n",
" if sparse.issparse(X):\n",
" X_new = sparse.vstack(np.tile(X, number_of_columns_y))\n",
" indicator_vecs = sparse.coo_matrix(indicator_vecs)\n",
" X_new = sparse.hstack((X_new, indicator_vecs))\n",
" else:\n",
" X_new = np.tile(X, (number_of_columns_y, 1))\n",
" X_new = np.hstack((X_new, indicator_vecs))\n",
" return X_new\n",
"\n",
"def multi_output_transform_y(y):\n",
" return y.reshape(-1, order=\"F\")\n",
"\n",
"def multi_output_inverse_transform_y(y, number_of_columns_y):\n",
" return y.reshape((-1, number_of_columns_y), order = \"F\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## AutoML Experiment Setup"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# Choose a name for the experiment and specify the project folder.\n",
"experiment_name = 'automl-local-multi-output'\n",
"project_folder = './sample_projects/automl-local-multi-output'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data = output, index = ['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create a Random Dataset for Test Purposes"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"rng = np.random.RandomState(1)\n",
"X_train = np.sort(200 * rng.rand(600, 1) - 100, axis = 0)\n",
"y_train = np.array([np.pi * np.sin(X_train).ravel(), np.pi * np.cos(X_train).ravel()]).T\n",
"y_train += (0.5 - rng.rand(*y_train.shape))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Perform X and y transformation using the transformer function."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"X_train_transformed, y_train_transformed = multi_output_transform_x_y(X_train, y_train)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Configure AutoML using the transformed results."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_config = AutoMLConfig(task = 'regression',\n",
" debug_log = 'automl_errors_multi.log',\n",
" primary_metric = 'r2_score',\n",
" iterations = 10,\n",
" n_cross_validations = 2,\n",
" verbosity = logging.INFO,\n",
" X = X_train_transformed,\n",
" y = y_train_transformed,\n",
" path = project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Fit the Transformed Data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_config, show_output = True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Get the best fit model.\n",
"best_run, fitted_model = local_run.get_output()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Generate random data set for predicting.\n",
"X_test = np.sort(200 * rng.rand(200, 1) - 100, axis = 0)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Transform predict data.\n",
"X_test_transformed = multi_output_transformer_x(X_test, y_train.shape[1])\n",
"\n",
"# Predict and inverse transform the prediction.\n",
"y_predict = fitted_model.predict(X_test_transformed)\n",
"y_predict = multi_output_inverse_transform_y(y_predict, y_train.shape[1])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(y_predict)"
]
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
"metadata": {
"authors": [
{
"name": "savitam"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
"nbformat": 4,
"nbformat_minor": 2
}

496
automl/11.auto-ml-sample-weight.ipynb
View File

@@ -1,251 +1,251 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 11: Sample Weight\n",
"\n",
"In this example we use the scikit-learn's [digit dataset](http://scikit-learn.org/stable/datasets/index.html#optical-recognition-of-handwritten-digits-dataset) to showcase how you can use sample weight with AutoML. Sample weight is used where some sample values are more important than others.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to configure AutoML to use `sample_weight` and you will see the difference sample weight makes to the test results.\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# Choose names for the regular and the sample weight experiments.\n",
"experiment_name = 'non_sample_weight_experiment'\n",
"sample_weight_experiment_name = 'sample_weight_experiment'\n",
"\n",
"project_folder = './sample_projects/automl-local-classification'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"sample_weight_experiment=Experiment(ws, sample_weight_experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace Name'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data = output, index = ['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML\n",
"\n",
"Instantiate two `AutoMLConfig` objects. One will be used with `sample_weight` and one without."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits = datasets.load_digits()\n",
"X_train = digits.data[100:,:]\n",
"y_train = digits.target[100:]\n",
"\n",
"# The example makes the sample weight 0.9 for the digit 4 and 0.1 for all other digits.\n",
"# This makes the model more likely to classify as 4 if the image it not clear.\n",
"sample_weight = np.array([(0.9 if x == 4 else 0.01) for x in y_train])\n",
"\n",
"automl_classifier = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" primary_metric = 'AUC_weighted',\n",
" iteration_timeout_minutes = 60,\n",
" iterations = 10,\n",
" n_cross_validations = 2,\n",
" verbosity = logging.INFO,\n",
" X = X_train, \n",
" y = y_train,\n",
" path = project_folder)\n",
"\n",
"automl_sample_weight = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" primary_metric = 'AUC_weighted',\n",
" iteration_timeout_minutes = 60,\n",
" iterations = 10,\n",
" n_cross_validations = 2,\n",
" verbosity = logging.INFO,\n",
" X = X_train, \n",
" y = y_train,\n",
" sample_weight = sample_weight,\n",
" path = project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train the Models\n",
"\n",
"Call the `submit` method on the experiment objects and pass the run configuration. Execution of local runs is synchronous. Depending on the data and the number of iterations this can run for a while.\n",
"In this example, we specify `show_output = True` to print currently running iterations to the console."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_classifier, show_output = True)\n",
"sample_weight_run = sample_weight_experiment.submit(automl_sample_weight, show_output = True)\n",
"\n",
"best_run, fitted_model = local_run.get_output()\n",
"best_run_sample_weight, fitted_model_sample_weight = sample_weight_run.get_output()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test the Best Fitted Model\n",
"\n",
"#### Load Test Data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits = datasets.load_digits()\n",
"X_test = digits.data[:100, :]\n",
"y_test = digits.target[:100]\n",
"images = digits.images[:100]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Compare the Models\n",
"The prediction from the sample weight model is more likely to correctly predict 4's. However, it is also more likely to predict 4 for some images that are not labelled as 4."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Randomly select digits and test.\n",
"for index in range(0,len(y_test)):\n",
" predicted = fitted_model.predict(X_test[index:index + 1])[0]\n",
" predicted_sample_weight = fitted_model_sample_weight.predict(X_test[index:index + 1])[0]\n",
" label = y_test[index]\n",
" if predicted == 4 or predicted_sample_weight == 4 or label == 4:\n",
" title = \"Label value = %d Predicted value = %d Prediced with sample weight = %d\" % (label, predicted, predicted_sample_weight)\n",
" fig = plt.figure(1, figsize=(3,3))\n",
" ax1 = fig.add_axes((0,0,.8,.8))\n",
" ax1.set_title(title)\n",
" plt.imshow(images[index], cmap = plt.cm.gray_r, interpolation = 'nearest')\n",
" plt.show()"
]
}
],
"metadata": {
"authors": [
{
"name": "savitam"
}
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 11: Sample Weight\n",
"\n",
"In this example we use the scikit-learn's [digit dataset](http://scikit-learn.org/stable/datasets/index.html#optical-recognition-of-handwritten-digits-dataset) to showcase how you can use sample weight with AutoML. Sample weight is used where some sample values are more important than others.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to configure AutoML to use `sample_weight` and you will see the difference sample weight makes to the test results.\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# Choose names for the regular and the sample weight experiments.\n",
"experiment_name = 'non_sample_weight_experiment'\n",
"sample_weight_experiment_name = 'sample_weight_experiment'\n",
"\n",
"project_folder = './sample_projects/automl-local-classification'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"sample_weight_experiment=Experiment(ws, sample_weight_experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace Name'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data = output, index = ['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML\n",
"\n",
"Instantiate two `AutoMLConfig` objects. One will be used with `sample_weight` and one without."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits = datasets.load_digits()\n",
"X_train = digits.data[100:,:]\n",
"y_train = digits.target[100:]\n",
"\n",
"# The example makes the sample weight 0.9 for the digit 4 and 0.1 for all other digits.\n",
"# This makes the model more likely to classify as 4 if the image it not clear.\n",
"sample_weight = np.array([(0.9 if x == 4 else 0.01) for x in y_train])\n",
"\n",
"automl_classifier = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" primary_metric = 'AUC_weighted',\n",
" max_time_sec = 3600,\n",
" iterations = 10,\n",
" n_cross_validations = 2,\n",
" verbosity = logging.INFO,\n",
" X = X_train, \n",
" y = y_train,\n",
" path = project_folder)\n",
"\n",
"automl_sample_weight = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" primary_metric = 'AUC_weighted',\n",
" max_time_sec = 3600,\n",
" iterations = 10,\n",
" n_cross_validations = 2,\n",
" verbosity = logging.INFO,\n",
" X = X_train, \n",
" y = y_train,\n",
" sample_weight = sample_weight,\n",
" path = project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train the Models\n",
"\n",
"Call the `submit` method on the experiment objects and pass the run configuration. Execution of local runs is synchronous. Depending on the data and the number of iterations this can run for a while.\n",
"In this example, we specify `show_output = True` to print currently running iterations to the console."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_classifier, show_output = True)\n",
"sample_weight_run = sample_weight_experiment.submit(automl_sample_weight, show_output = True)\n",
"\n",
"best_run, fitted_model = local_run.get_output()\n",
"best_run_sample_weight, fitted_model_sample_weight = sample_weight_run.get_output()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test the Best Fitted Model\n",
"\n",
"#### Load Test Data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits = datasets.load_digits()\n",
"X_test = digits.data[:100, :]\n",
"y_test = digits.target[:100]\n",
"images = digits.images[:100]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Compare the Models\n",
"The prediction from the sample weight model is more likely to correctly predict 4's. However, it is also more likely to predict 4 for some images that are not labelled as 4."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Randomly select digits and test.\n",
"for index in range(0,len(y_test)):\n",
" predicted = fitted_model.predict(X_test[index:index + 1])[0]\n",
" predicted_sample_weight = fitted_model_sample_weight.predict(X_test[index:index + 1])[0]\n",
" label = y_test[index]\n",
" if predicted == 4 or predicted_sample_weight == 4 or label == 4:\n",
" title = \"Label value = %d Predicted value = %d Prediced with sample weight = %d\" % (label, predicted, predicted_sample_weight)\n",
" fig = plt.figure(1, figsize=(3,3))\n",
" ax1 = fig.add_axes((0,0,.8,.8))\n",
" ax1.set_title(title)\n",
" plt.imshow(images[index], cmap = plt.cm.gray_r, interpolation = 'nearest')\n",
" plt.show()"
]
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
"metadata": {
"authors": [
{
"name": "savitam"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.5"
}
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.5"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
"nbformat": 4,
"nbformat_minor": 2
}

448
automl/12.auto-ml-retrieve-the-training-sdk-versions.ipynb
View File

@@ -1,227 +1,227 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 12: Retrieving Training SDK Versions\n",
"\n",
"This example shows how to find the SDK versions used for an experiment.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Train models using AutoML"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# Choose a name for the experiment and specify the project folder.\n",
"experiment_name = 'automl-local-classification'\n",
"project_folder = './sample_projects/automl-local-classification'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data=output, index=['']).T"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits = datasets.load_digits()\n",
"X_train = digits.data[10:,:]\n",
"y_train = digits.target[10:]\n",
"\n",
"automl_config = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" primary_metric = 'AUC_weighted',\n",
" iterations = 3,\n",
" n_cross_validations = 2,\n",
" verbosity = logging.INFO,\n",
" X = X_train, \n",
" y = y_train,\n",
" path = project_folder)\n",
"\n",
"local_run = experiment.submit(automl_config, show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Retrieve the SDK versions from RunHistory"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"To get the SDK versions from RunHistory, first the run id needs to be recorded. This can either be done by copying it from the output message or by retrieving it after each run."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Use a run id copied from an output message.\n",
"#run_id = 'AutoML_c0585b1f-a0e6-490b-84c7-3a099468b28e'\n",
"\n",
"# Retrieve the run id from a run.\n",
"run_id = local_run.id\n",
"print(run_id)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Initialize a new `AutoMLRun` object."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"experiment_name = 'automl-local-classification'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"ml_run = AutoMLRun(experiment = experiment, run_id = run_id)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Get parent training SDK versions."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ml_run.get_run_sdk_dependencies()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Get the traning SDK versions of a specific run."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ml_run.get_run_sdk_dependencies(iteration = 2)"
]
}
],
"metadata": {
"authors": [
{
"name": "savitam"
}
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 12: Retrieving Training SDK Versions\n",
"\n",
"This example shows how to find the SDK versions used for an experiment.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Train models using AutoML"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# Choose a name for the experiment and specify the project folder.\n",
"experiment_name = 'automl-local-classification'\n",
"project_folder = './sample_projects/automl-local-classification'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data=output, index=['']).T"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits = datasets.load_digits()\n",
"X_train = digits.data[10:,:]\n",
"y_train = digits.target[10:]\n",
"\n",
"automl_config = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" primary_metric = 'AUC_weighted',\n",
" iterations = 3,\n",
" n_cross_validations = 2,\n",
" verbosity = logging.INFO,\n",
" X = X_train, \n",
" y = y_train,\n",
" path = project_folder)\n",
"\n",
"local_run = experiment.submit(automl_config, show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Retrieve the SDK versions from RunHistory"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"To get the SDK versions from RunHistory, first the run id needs to be recorded. This can either be done by copying it from the output message or by retrieving it after each run."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Use a run id copied from an output message.\n",
"#run_id = 'AutoML_c0585b1f-a0e6-490b-84c7-3a099468b28e'\n",
"\n",
"# Retrieve the run id from a run.\n",
"run_id = local_run.id\n",
"print(run_id)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Initialize a new `AutoMLRun` object."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"experiment_name = 'automl-local-classification'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"ml_run = AutoMLRun(experiment = experiment, run_id = run_id)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Get parent training SDK versions."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ml_run.get_run_sdk_dependencies()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Get the traning SDK versions of a specific run."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ml_run.get_run_sdk_dependencies(iteration = 2)"
]
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
"metadata": {
"authors": [
{
"name": "savitam"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
"nbformat": 4,
"nbformat_minor": 2
}

886
automl/13a.auto-ml-dataprep.ipynb
View File

@@ -1,446 +1,446 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 13: Prepare Data using `azureml.dataprep` for Local Execution\n",
"In this example we showcase how you can use the `azureml.dataprep` SDK to load and prepare data for AutoML. `azureml.dataprep` can also be used standalone; full documentation can be found [here](https://github.com/Microsoft/PendletonDocs).\n",
"\n",
"Make sure you have executed the [setup](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. Define data loading and preparation steps in a `Dataflow` using `azureml.dataprep`.\n",
"2. Pass the `Dataflow` to AutoML for a local run.\n",
"3. Pass the `Dataflow` to AutoML for a remote run."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"\n",
"import pandas as pd\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"import azureml.dataprep as dprep\n",
"from azureml.train.automl import AutoMLConfig"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
" \n",
"# choose a name for experiment\n",
"experiment_name = 'automl-dataprep-local'\n",
"# project folder\n",
"project_folder = './sample_projects/automl-dataprep-local'\n",
" \n",
"experiment = Experiment(ws, experiment_name)\n",
" \n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace Name'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data = output, index = ['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Loading Data using DataPrep"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# You can use `smart_read_file` which intelligently figures out delimiters and datatypes of a file.\n",
"# The data referenced here was pulled from `sklearn.datasets.load_digits()`.\n",
"simple_example_data_root = 'https://dprepdata.blob.core.windows.net/automl-notebook-data/'\n",
"X = dprep.smart_read_file(simple_example_data_root + 'X.csv').skip(1) # Remove the header row.\n",
"\n",
"# You can also use `read_csv` and `to_*` transformations to read (with overridable delimiter)\n",
"# and convert column types manually.\n",
"# Here we read a comma delimited file and convert all columns to integers.\n",
"y = dprep.read_csv(simple_example_data_root + 'y.csv').to_long(dprep.ColumnSelector(term='.*', use_regex = True))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Review the Data Preparation Result\n",
"\n",
"You can peek the result of a Dataflow at any range using `skip(i)` and `head(j)`. Doing so evaluates only `j` records for all the steps in the Dataflow, which makes it fast even against large datasets."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"X.skip(1).head(5)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML\n",
"\n",
"This creates a general AutoML settings object applicable for both local and remote runs."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_settings = {\n",
" \"iteration_timeout_minutes\" : 10,\n",
" \"iterations\" : 2,\n",
" \"primary_metric\" : 'AUC_weighted',\n",
" \"preprocess\" : False,\n",
" \"verbosity\" : logging.INFO,\n",
" \"n_cross_validations\": 3\n",
"}"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Local Run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Pass Data with `Dataflow` Objects\n",
"\n",
"The `Dataflow` objects captured above can be passed to the `submit` method for a local run. AutoML will retrieve the results from the `Dataflow` for model training."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_config = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" X = X,\n",
" y = y,\n",
" **automl_settings)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_config, show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explore the Results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Widget for Monitoring Runs\n",
"\n",
"The widget will first report a \"loading\" status while running the first iteration. After completing the first iteration, an auto-updating graph and table will be shown. The widget will refresh once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"**Note:** The widget displays a link at the bottom. Use this link to open a web interface to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.widgets import RunDetails\n",
"RunDetails(local_run).show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Retrieve All Child Runs\n",
"You can also use SDK methods to fetch all the child runs and see individual metrics that we log."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"children = list(local_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
" \n",
"import pandas as pd\n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"rundata"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method returns the best run and the fitted model. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = local_run.get_output()\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Best Model Based on Any Other Metric\n",
"Show the run and the model that has the smallest `log_loss` value:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"lookup_metric = \"log_loss\"\n",
"best_run, fitted_model = local_run.get_output(metric = lookup_metric)\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Model from a Specific Iteration\n",
"Show the run and the model from the first iteration:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"iteration = 0\n",
"best_run, fitted_model = local_run.get_output(iteration = iteration)\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test the Best Fitted Model\n",
"\n",
"#### Load Test Data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn import datasets\n",
"\n",
"digits = datasets.load_digits()\n",
"X_test = digits.data[:10, :]\n",
"y_test = digits.target[:10]\n",
"images = digits.images[:10]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Testing Our Best Fitted Model\n",
"We will try to predict 2 digits and see how our model works."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#Randomly select digits and test\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import random\n",
"import numpy as np\n",
"\n",
"for index in np.random.choice(len(y_test), 2, replace = False):\n",
" print(index)\n",
" predicted = fitted_model.predict(X_test[index:index + 1])[0]\n",
" label = y_test[index]\n",
" title = \"Label value = %d Predicted value = %d \" % (label, predicted)\n",
" fig = plt.figure(1, figsize=(3,3))\n",
" ax1 = fig.add_axes((0,0,.8,.8))\n",
" ax1.set_title(title)\n",
" plt.imshow(images[index], cmap = plt.cm.gray_r, interpolation = 'nearest')\n",
" plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Appendix"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Capture the `Dataflow` Objects for Later Use in AutoML\n",
"\n",
"`Dataflow` objects are immutable and are composed of a list of data preparation steps. A `Dataflow` object can be branched at any point for further usage."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# sklearn.digits.data + target\n",
"digits_complete = dprep.smart_read_file('https://dprepdata.blob.core.windows.net/automl-notebook-data/digits-complete.csv')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"`digits_complete` (sourced from `sklearn.datasets.load_digits()`) is forked into `dflow_X` to capture all the feature columns and `dflow_y` to capture the label column."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits_complete.to_pandas_dataframe().shape\n",
"labels_column = 'Column64'\n",
"dflow_X = digits_complete.drop_columns(columns = [labels_column])\n",
"dflow_y = digits_complete.keep_columns(columns = [labels_column])"
]
}
],
"metadata": {
"authors": [
{
"name": "savitam"
}
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 13: Prepare Data using `azureml.dataprep` for Local Execution\n",
"In this example we showcase how you can use the `azureml.dataprep` SDK to load and prepare data for AutoML. `azureml.dataprep` can also be used standalone; full documentation can be found [here](https://github.com/Microsoft/PendletonDocs).\n",
"\n",
"Make sure you have executed the [setup](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. Define data loading and preparation steps in a `Dataflow` using `azureml.dataprep`.\n",
"2. Pass the `Dataflow` to AutoML for a local run.\n",
"3. Pass the `Dataflow` to AutoML for a remote run."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"\n",
"import pandas as pd\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"import azureml.dataprep as dprep\n",
"from azureml.train.automl import AutoMLConfig"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
" \n",
"# choose a name for experiment\n",
"experiment_name = 'automl-dataprep-local'\n",
"# project folder\n",
"project_folder = './sample_projects/automl-dataprep-local'\n",
" \n",
"experiment = Experiment(ws, experiment_name)\n",
" \n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace Name'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data = output, index = ['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Loading Data using DataPrep"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# You can use `smart_read_file` which intelligently figures out delimiters and datatypes of a file.\n",
"# The data referenced here was pulled from `sklearn.datasets.load_digits()`.\n",
"simple_example_data_root = 'https://dprepdata.blob.core.windows.net/automl-notebook-data/'\n",
"X = dprep.smart_read_file(simple_example_data_root + 'X.csv').skip(1) # Remove the header row.\n",
"\n",
"# You can also use `read_csv` and `to_*` transformations to read (with overridable delimiter)\n",
"# and convert column types manually.\n",
"# Here we read a comma delimited file and convert all columns to integers.\n",
"y = dprep.read_csv(simple_example_data_root + 'y.csv').to_long(dprep.ColumnSelector(term='.*', use_regex = True))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Review the Data Preparation Result\n",
"\n",
"You can peek the result of a Dataflow at any range using `skip(i)` and `head(j)`. Doing so evaluates only `j` records for all the steps in the Dataflow, which makes it fast even against large datasets."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"X.skip(1).head(5)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML\n",
"\n",
"This creates a general AutoML settings object applicable for both local and remote runs."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_settings = {\n",
" \"max_time_sec\" : 600,\n",
" \"iterations\" : 2,\n",
" \"primary_metric\" : 'AUC_weighted',\n",
" \"preprocess\" : False,\n",
" \"verbosity\" : logging.INFO,\n",
" \"n_cross_validations\": 3\n",
"}"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Local Run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Pass Data with `Dataflow` Objects\n",
"\n",
"The `Dataflow` objects captured above can be passed to the `submit` method for a local run. AutoML will retrieve the results from the `Dataflow` for model training."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_config = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" X = X,\n",
" y = y,\n",
" **automl_settings)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_config, show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explore the Results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Widget for Monitoring Runs\n",
"\n",
"The widget will first report a \"loading\" status while running the first iteration. After completing the first iteration, an auto-updating graph and table will be shown. The widget will refresh once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"**Note:** The widget displays a link at the bottom. Use this link to open a web interface to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.widgets import RunDetails\n",
"RunDetails(local_run).show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Retrieve All Child Runs\n",
"You can also use SDK methods to fetch all the child runs and see individual metrics that we log."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"children = list(local_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
" \n",
"import pandas as pd\n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"rundata"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method returns the best run and the fitted model. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = local_run.get_output()\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Best Model Based on Any Other Metric\n",
"Show the run and the model that has the smallest `log_loss` value:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"lookup_metric = \"log_loss\"\n",
"best_run, fitted_model = local_run.get_output(metric = lookup_metric)\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Model from a Specific Iteration\n",
"Show the run and the model from the first iteration:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"iteration = 0\n",
"best_run, fitted_model = local_run.get_output(iteration = iteration)\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test the Best Fitted Model\n",
"\n",
"#### Load Test Data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn import datasets\n",
"\n",
"digits = datasets.load_digits()\n",
"X_test = digits.data[:10, :]\n",
"y_test = digits.target[:10]\n",
"images = digits.images[:10]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Testing Our Best Fitted Model\n",
"We will try to predict 2 digits and see how our model works."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#Randomly select digits and test\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import random\n",
"import numpy as np\n",
"\n",
"for index in np.random.choice(len(y_test), 2, replace = False):\n",
" print(index)\n",
" predicted = fitted_model.predict(X_test[index:index + 1])[0]\n",
" label = y_test[index]\n",
" title = \"Label value = %d Predicted value = %d \" % (label, predicted)\n",
" fig = plt.figure(1, figsize=(3,3))\n",
" ax1 = fig.add_axes((0,0,.8,.8))\n",
" ax1.set_title(title)\n",
" plt.imshow(images[index], cmap = plt.cm.gray_r, interpolation = 'nearest')\n",
" plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Appendix"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Capture the `Dataflow` Objects for Later Use in AutoML\n",
"\n",
"`Dataflow` objects are immutable and are composed of a list of data preparation steps. A `Dataflow` object can be branched at any point for further usage."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# sklearn.digits.data + target\n",
"digits_complete = dprep.smart_read_file('https://dprepdata.blob.core.windows.net/automl-notebook-data/digits-complete.csv')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"`digits_complete` (sourced from `sklearn.datasets.load_digits()`) is forked into `dflow_X` to capture all the feature columns and `dflow_y` to capture the label column."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits_complete.to_pandas_dataframe().shape\n",
"labels_column = 'Column64'\n",
"dflow_X = digits_complete.drop_columns(columns = [labels_column])\n",
"dflow_y = digits_complete.keep_columns(columns = [labels_column])"
]
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
"metadata": {
"authors": [
{
"name": "savitam"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.5"
}
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.5"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
"nbformat": 4,
"nbformat_minor": 2
}

962
automl/13b.auto-ml-dataprep-remote-execution.ipynb
View File

@@ -1,495 +1,473 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 13: Prepare Data using `azureml.dataprep` for Remote Execution (DSVM)\n",
"In this example we showcase how you can use the `azureml.dataprep` SDK to load and prepare data for AutoML. `azureml.dataprep` can also be used standalone; full documentation can be found [here](https://github.com/Microsoft/PendletonDocs).\n",
"\n",
"Make sure you have executed the [setup](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. Define data loading and preparation steps in a `Dataflow` using `azureml.dataprep`.\n",
"2. Pass the `Dataflow` to AutoML for a local run.\n",
"3. Pass the `Dataflow` to AutoML for a remote run."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import time\n",
"\n",
"import pandas as pd\n",
"\n",
"import azureml.core\n",
"from azureml.core.compute import DsvmCompute\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"import azureml.dataprep as dprep\n",
"from azureml.train.automl import AutoMLConfig"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
" \n",
"# choose a name for experiment\n",
"experiment_name = 'automl-dataprep-remote-dsvm'\n",
"# project folder\n",
"project_folder = './sample_projects/automl-dataprep-remote-dsvm'\n",
" \n",
"experiment = Experiment(ws, experiment_name)\n",
" \n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace Name'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data = output, index = ['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Loading Data using DataPrep"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# You can use `smart_read_file` which intelligently figures out delimiters and datatypes of a file.\n",
"# The data referenced here was pulled from `sklearn.datasets.load_digits()`.\n",
"simple_example_data_root = 'https://dprepdata.blob.core.windows.net/automl-notebook-data/'\n",
"X = dprep.smart_read_file(simple_example_data_root + 'X.csv').skip(1) # Remove the header row.\n",
"\n",
"# You can also use `read_csv` and `to_*` transformations to read (with overridable delimiter)\n",
"# and convert column types manually.\n",
"# Here we read a comma delimited file and convert all columns to integers.\n",
"y = dprep.read_csv(simple_example_data_root + 'y.csv').to_long(dprep.ColumnSelector(term='.*', use_regex = True))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Review the Data Preparation Result\n",
"\n",
"You can peek the result of a Dataflow at any range using `skip(i)` and `head(j)`. Doing so evaluates only `j` records for all the steps in the Dataflow, which makes it fast even against large datasets."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"X.skip(1).head(5)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML\n",
"\n",
"This creates a general AutoML settings object applicable for both local and remote runs."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_settings = {\n",
" \"iteration_timeout_minutes\" : 10,\n",
" \"iterations\" : 2,\n",
" \"primary_metric\" : 'AUC_weighted',\n",
" \"preprocess\" : False,\n",
" \"verbosity\" : logging.INFO,\n",
" \"n_cross_validations\": 3\n",
"}"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Remote Run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create or Attach a Remote Linux DSVM"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"dsvm_name = 'mydsvmd'\n",
"\n",
"try:\n",
" while ws.compute_targets[dsvm_name].provisioning_state == 'Creating':\n",
" time.sleep(1)\n",
" \n",
" dsvm_compute = DsvmCompute(ws, dsvm_name)\n",
" print('Found existing DVSM.')\n",
"except:\n",
" print('Creating a new DSVM.')\n",
" dsvm_config = DsvmCompute.provisioning_configuration(vm_size = \"Standard_D2_v2\")\n",
" dsvm_compute = DsvmCompute.create(ws, name = dsvm_name, provisioning_configuration = dsvm_config)\n",
" dsvm_compute.wait_for_completion(show_output = True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.runconfig import RunConfiguration\n",
"from azureml.core.conda_dependencies import CondaDependencies\n",
"\n",
"conda_run_config = RunConfiguration(framework=\"python\")\n",
"\n",
"conda_run_config.target = dsvm_compute\n",
"\n",
"cd = CondaDependencies.create(pip_packages=['azureml-sdk[automl]'], conda_packages=['numpy'])\n",
"conda_run_config.environment.python.conda_dependencies = cd"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Pass Data with `Dataflow` Objects\n",
"\n",
"The `Dataflow` objects captured above can also be passed to the `submit` method for a remote run. AutoML will serialize the `Dataflow` object and send it to the remote compute target. The `Dataflow` will not be evaluated locally."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_config = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" path = project_folder,\n",
" run_configuration=conda_run_config,\n",
" X = X,\n",
" y = y,\n",
" **automl_settings)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"remote_run = experiment.submit(automl_config, show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explore the Results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Widget for Monitoring Runs\n",
"\n",
"The widget will first report a \"loading\" status while running the first iteration. After completing the first iteration, an auto-updating graph and table will be shown. The widget will refresh once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"**Note:** The widget displays a link at the bottom. Use this link to open a web interface to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.widgets import RunDetails\n",
"RunDetails(remote_run).show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Retrieve All Child Runs\n",
"You can also use SDK methods to fetch all the child runs and see individual metrics that we log."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"children = list(remote_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
" \n",
"import pandas as pd\n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"rundata"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method returns the best run and the fitted model. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = remote_run.get_output()\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Best Model Based on Any Other Metric\n",
"Show the run and the model that has the smallest `log_loss` value:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"lookup_metric = \"log_loss\"\n",
"best_run, fitted_model = remote_run.get_output(metric = lookup_metric)\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Model from a Specific Iteration\n",
"Show the run and the model from the first iteration:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"iteration = 0\n",
"best_run, fitted_model = remote_run.get_output(iteration = iteration)\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test the Best Fitted Model\n",
"\n",
"#### Load Test Data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn import datasets\n",
"\n",
"digits = datasets.load_digits()\n",
"X_test = digits.data[:10, :]\n",
"y_test = digits.target[:10]\n",
"images = digits.images[:10]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Testing Our Best Fitted Model\n",
"We will try to predict 2 digits and see how our model works."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#Randomly select digits and test\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import random\n",
"import numpy as np\n",
"\n",
"for index in np.random.choice(len(y_test), 2, replace = False):\n",
" print(index)\n",
" predicted = fitted_model.predict(X_test[index:index + 1])[0]\n",
" label = y_test[index]\n",
" title = \"Label value = %d Predicted value = %d \" % (label, predicted)\n",
" fig = plt.figure(1, figsize=(3,3))\n",
" ax1 = fig.add_axes((0,0,.8,.8))\n",
" ax1.set_title(title)\n",
" plt.imshow(images[index], cmap = plt.cm.gray_r, interpolation = 'nearest')\n",
" plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Appendix"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Capture the `Dataflow` Objects for Later Use in AutoML\n",
"\n",
"`Dataflow` objects are immutable and are composed of a list of data preparation steps. A `Dataflow` object can be branched at any point for further usage."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# sklearn.digits.data + target\n",
"digits_complete = dprep.smart_read_file('https://dprepdata.blob.core.windows.net/automl-notebook-data/digits-complete.csv')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"`digits_complete` (sourced from `sklearn.datasets.load_digits()`) is forked into `dflow_X` to capture all the feature columns and `dflow_y` to capture the label column."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits_complete.to_pandas_dataframe().shape\n",
"labels_column = 'Column64'\n",
"dflow_X = digits_complete.drop_columns(columns = [labels_column])\n",
"dflow_y = digits_complete.keep_columns(columns = [labels_column])"
]
}
],
"metadata": {
"authors": [
{
"name": "savitam"
}
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 13: Prepare Data using `azureml.dataprep` for Remote Execution (DSVM)\n",
"In this example we showcase how you can use the `azureml.dataprep` SDK to load and prepare data for AutoML. `azureml.dataprep` can also be used standalone; full documentation can be found [here](https://github.com/Microsoft/PendletonDocs).\n",
"\n",
"Make sure you have executed the [setup](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. Define data loading and preparation steps in a `Dataflow` using `azureml.dataprep`.\n",
"2. Pass the `Dataflow` to AutoML for a local run.\n",
"3. Pass the `Dataflow` to AutoML for a remote run."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"\n",
"import pandas as pd\n",
"\n",
"import azureml.core\n",
"from azureml.core.compute import DsvmCompute\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"import azureml.dataprep as dprep\n",
"from azureml.train.automl import AutoMLConfig"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
" \n",
"# choose a name for experiment\n",
"experiment_name = 'automl-dataprep-remote-dsvm'\n",
"# project folder\n",
"project_folder = './sample_projects/automl-dataprep-remote-dsvm'\n",
" \n",
"experiment = Experiment(ws, experiment_name)\n",
" \n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace Name'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data = output, index = ['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Loading Data using DataPrep"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# You can use `smart_read_file` which intelligently figures out delimiters and datatypes of a file.\n",
"# The data referenced here was pulled from `sklearn.datasets.load_digits()`.\n",
"simple_example_data_root = 'https://dprepdata.blob.core.windows.net/automl-notebook-data/'\n",
"X = dprep.smart_read_file(simple_example_data_root + 'X.csv').skip(1) # Remove the header row.\n",
"\n",
"# You can also use `read_csv` and `to_*` transformations to read (with overridable delimiter)\n",
"# and convert column types manually.\n",
"# Here we read a comma delimited file and convert all columns to integers.\n",
"y = dprep.read_csv(simple_example_data_root + 'y.csv').to_long(dprep.ColumnSelector(term='.*', use_regex = True))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Review the Data Preparation Result\n",
"\n",
"You can peek the result of a Dataflow at any range using `skip(i)` and `head(j)`. Doing so evaluates only `j` records for all the steps in the Dataflow, which makes it fast even against large datasets."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"X.skip(1).head(5)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML\n",
"\n",
"This creates a general AutoML settings object applicable for both local and remote runs."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_settings = {\n",
" \"max_time_sec\" : 600,\n",
" \"iterations\" : 2,\n",
" \"primary_metric\" : 'AUC_weighted',\n",
" \"preprocess\" : False,\n",
" \"verbosity\" : logging.INFO,\n",
" \"n_cross_validations\": 3\n",
"}"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Remote Run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create or Attach a Remote Linux DSVM"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"dsvm_name = 'mydsvm'\n",
"try:\n",
" dsvm_compute = DsvmCompute(ws, dsvm_name)\n",
" print('Found existing DVSM.')\n",
"except:\n",
" print('Creating a new DSVM.')\n",
" dsvm_config = DsvmCompute.provisioning_configuration(vm_size = \"Standard_D2_v2\")\n",
" dsvm_compute = DsvmCompute.create(ws, name = dsvm_name, provisioning_configuration = dsvm_config)\n",
" dsvm_compute.wait_for_completion(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Pass Data with `Dataflow` Objects\n",
"\n",
"The `Dataflow` objects captured above can also be passed to the `submit` method for a remote run. AutoML will serialize the `Dataflow` object and send it to the remote compute target. The `Dataflow` will not be evaluated locally."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_config = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" path = project_folder,\n",
" compute_target = dsvm_compute,\n",
" X = X,\n",
" y = y,\n",
" **automl_settings)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"remote_run = experiment.submit(automl_config, show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explore the Results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Widget for Monitoring Runs\n",
"\n",
"The widget will first report a \"loading\" status while running the first iteration. After completing the first iteration, an auto-updating graph and table will be shown. The widget will refresh once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"**Note:** The widget displays a link at the bottom. Use this link to open a web interface to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.widgets import RunDetails\n",
"RunDetails(remote_run).show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Retrieve All Child Runs\n",
"You can also use SDK methods to fetch all the child runs and see individual metrics that we log."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"children = list(remote_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
" \n",
"import pandas as pd\n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"rundata"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method returns the best run and the fitted model. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = remote_run.get_output()\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Best Model Based on Any Other Metric\n",
"Show the run and the model that has the smallest `log_loss` value:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"lookup_metric = \"log_loss\"\n",
"best_run, fitted_model = remote_run.get_output(metric = lookup_metric)\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Model from a Specific Iteration\n",
"Show the run and the model from the first iteration:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"iteration = 0\n",
"best_run, fitted_model = remote_run.get_output(iteration = iteration)\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test the Best Fitted Model\n",
"\n",
"#### Load Test Data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn import datasets\n",
"\n",
"digits = datasets.load_digits()\n",
"X_test = digits.data[:10, :]\n",
"y_test = digits.target[:10]\n",
"images = digits.images[:10]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Testing Our Best Fitted Model\n",
"We will try to predict 2 digits and see how our model works."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#Randomly select digits and test\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import random\n",
"import numpy as np\n",
"\n",
"for index in np.random.choice(len(y_test), 2, replace = False):\n",
" print(index)\n",
" predicted = fitted_model.predict(X_test[index:index + 1])[0]\n",
" label = y_test[index]\n",
" title = \"Label value = %d Predicted value = %d \" % (label, predicted)\n",
" fig = plt.figure(1, figsize=(3,3))\n",
" ax1 = fig.add_axes((0,0,.8,.8))\n",
" ax1.set_title(title)\n",
" plt.imshow(images[index], cmap = plt.cm.gray_r, interpolation = 'nearest')\n",
" plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Appendix"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Capture the `Dataflow` Objects for Later Use in AutoML\n",
"\n",
"`Dataflow` objects are immutable and are composed of a list of data preparation steps. A `Dataflow` object can be branched at any point for further usage."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# sklearn.digits.data + target\n",
"digits_complete = dprep.smart_read_file('https://dprepdata.blob.core.windows.net/automl-notebook-data/digits-complete.csv')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"`digits_complete` (sourced from `sklearn.datasets.load_digits()`) is forked into `dflow_X` to capture all the feature columns and `dflow_y` to capture the label column."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits_complete.to_pandas_dataframe().shape\n",
"labels_column = 'Column64'\n",
"dflow_X = digits_complete.drop_columns(columns = [labels_column])\n",
"dflow_y = digits_complete.keep_columns(columns = [labels_column])"
]
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
"metadata": {
"authors": [
{
"name": "savitam"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.5"
}
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.5"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
"nbformat": 4,
"nbformat_minor": 2
}

374
automl/14.auto-ml-model-explanation.ipynb
View File

@@ -1,374 +0,0 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 14: Explain classification model and visualize the explanation\n",
"\n",
"In this example we use the sklearn's [iris dataset](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_iris.html) to showcase how you can use the AutoML Classifier for a simple classification problem.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you would see\n",
"1. Creating an Experiment in an existing Workspace\n",
"2. Instantiating AutoMLConfig\n",
"3. Training the Model using local compute and explain the model\n",
"4. Visualization model's feature importance in widget\n",
"5. Explore best model's explanation\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Install AzureML Explainer SDK "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"!pip install azureml_sdk[explain]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create Experiment\n",
"\n",
"As part of the setup you have already created a <b>Workspace</b>. For AutoML you would need to create an <b>Experiment</b>. An <b>Experiment</b> is a named object in a <b>Workspace</b>, which is used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"import pandas as pd\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# choose a name for experiment\n",
"experiment_name = 'automl-local-classification'\n",
"# project folder\n",
"project_folder = './sample_projects/automl-local-classification-model-explanation'\n",
"\n",
"experiment=Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace Name'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data = output, index = ['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Load Iris Data Set"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn import datasets\n",
"\n",
"iris = datasets.load_iris()\n",
"y = iris.target\n",
"X = iris.data\n",
"\n",
"features = iris.feature_names\n",
"\n",
"from sklearn.model_selection import train_test_split\n",
"X_train, X_test, y_train, y_test = train_test_split(X,\n",
" y,\n",
" test_size=0.1,\n",
" random_state=100,\n",
" stratify=y)\n",
"\n",
"X_train = pd.DataFrame(X_train, columns=features)\n",
"X_test = pd.DataFrame(X_test, columns=features)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Instantiate Auto ML Config\n",
"\n",
"Instantiate a AutoMLConfig object. This defines the settings and data used to run the experiment.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**task**|classification or regression|\n",
"|**primary_metric**|This is the metric that you want to optimize. Classification supports the following primary metrics: <br><i>accuracy</i><br><i>AUC_weighted</i><br><i>average_precision_score_weighted</i><br><i>norm_macro_recall</i><br><i>precision_score_weighted</i>|\n",
"|**max_time_sec**|Time limit in minutes for each iterations|\n",
"|**iterations**|Number of iterations. In each iteration Auto ML trains the data with a specific pipeline|\n",
"|**X**|(sparse) array-like, shape = [n_samples, n_features]|\n",
"|**y**|(sparse) array-like, shape = [n_samples, ], [n_samples, n_classes]<br>Multi-class targets. An indicator matrix turns on multilabel classification. This should be an array of integers. |\n",
"|**X_valid**|(sparse) array-like, shape = [n_samples, n_features]|\n",
"|**y_valid**|(sparse) array-like, shape = [n_samples, ], [n_samples, n_classes]|\n",
"|**model_explainability**|Indicate to explain each trained pipeline or not |\n",
"|**path**|Relative path to the project folder. AutoML stores configuration files for the experiment under this folder. You can specify a new empty folder. |"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_config = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" primary_metric = 'AUC_weighted',\n",
" iteration_timeout_minutes = 200,\n",
" iterations = 10,\n",
" verbosity = logging.INFO,\n",
" X = X_train, \n",
" y = y_train,\n",
" X_valid = X_test,\n",
" y_valid = y_test,\n",
" model_explainability=True,\n",
" path=project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Training the Model\n",
"\n",
"You can call the submit method on the experiment object and pass the run configuration. For Local runs the execution is synchronous. Depending on the data and number of iterations this can run for while.\n",
"You will see the currently running iterations printing to the console."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_config, show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Exploring the results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Widget for monitoring runs\n",
"\n",
"The widget will sit on \"loading\" until the first iteration completed, then you will see an auto-updating graph and table show up. It refreshed once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"NOTE: The widget displays a link at the bottom. This links to a web-ui to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.widgets import RunDetails\n",
"RunDetails(local_run).show() "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"child_run = next(local_run.get_children())\n",
"RunDetails(child_run).show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The *get_output* method on automl_classifier returns the best run and the fitted model for the last *fit* invocation. There are overloads on *get_output* that allow you to retrieve the best run and fitted model for *any* logged metric or a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = local_run.get_output()\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Best Model 's explanation\n",
"\n",
"Retrieve the explanation from the best_run. And explanation information includes:\n",
"\n",
"1.\tshap_values: The explanation information generated by shap lib\n",
"2.\texpected_values: The expected value of the model applied to set of X_train data.\n",
"3.\toverall_summary: The model level feature importance values sorted in descending order\n",
"4.\toverall_imp: The feature names sorted in the same order as in overall_summary\n",
"5.\tper_class_summary: The class level feature importance values sorted in descending order. Only available for the classification case\n",
"6.\tper_class_imp: The feature names sorted in the same order as in per_class_summary. Only available for the classification case"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.automl.automlexplainer import retrieve_model_explanation\n",
"\n",
"shap_values, expected_values, overall_summary, overall_imp, per_class_summary, per_class_imp = \\\n",
" retrieve_model_explanation(best_run)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(overall_summary)\n",
"print(overall_imp)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(per_class_summary)\n",
"print(per_class_imp)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Beside retrieve the existed model explanation information, explain the model with different train/test data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.automl.automlexplainer import explain_model\n",
"\n",
"shap_values, expected_values, overall_summary, overall_imp, per_class_summary, per_class_imp = \\\n",
" explain_model(fitted_model, X_train, X_test)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(overall_summary)\n",
"print(overall_imp)"
]
}
],
"metadata": {
"authors": [
{
"name": "xif"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

426
automl/15a.auto-ml-classification-ensemble.ipynb Normal file
View File

@@ -0,0 +1,426 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 15a: Classification with ensembling on local compute\n",
"\n",
"In this example we use the scikit-learn's [digit dataset](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_digits.html) to showcase how you can use AutoML for a simple classification problem.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. Create an `Experiment` in an existing `Workspace`.\n",
"2. Configure AutoML using `AutoMLConfig` which enables an extra ensembling iteration.\n",
"3. Train the model using local compute.\n",
"4. Explore the results.\n",
"5. Test the best fitted model.\n",
"\n",
"<b>Disclaimers / Limitations </b>\n",
"- Currently only Train/Validation split is supported; support for cross-validation will be coming soon.\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# Choose a name for the experiment and specify the project folder.\n",
"experiment_name = 'automl-local-classification'\n",
"project_folder = './sample_projects/automl-local-classification'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace Name'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data = output, index = ['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Load Training Data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn import datasets\n",
"\n",
"digits = datasets.load_digits()\n",
"\n",
"# Exclude the first 50 rows from training so that they can be used for test.\n",
"X_train = digits.data[150:,:]\n",
"y_train = digits.target[150:]\n",
"X_valid = digits.data[50:150]\n",
"y_valid = digits.target[50:150]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML\n",
"\n",
"Instantiate an `AutoMLConfig` object to specify the settings and data used to run the experiment.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**task**|classification or regression|\n",
"|**primary_metric**|This is the metric that you want to optimize. Classification supports the following primary metrics: <br><i>accuracy</i><br><i>AUC_weighted</i><br><i>balanced_accuracy</i><br><i>average_precision_score_weighted</i><br><i>precision_score_weighted</i>|\n",
"|**max_time_sec**|Time limit in seconds for each iteration.|\n",
"|**iterations**|Number of iterations. In each iteration AutoML trains a specific pipeline with the data.|\n",
"|**n_cross_validations**|Number of cross validation splits.|\n",
"|**X**|(sparse) array-like, shape = [n_samples, n_features]|\n",
"|**y**|(sparse) array-like, shape = [n_samples, ], [n_samples, n_classes]<br>Multi-class targets. An indicator matrix turns on multilabel classification. This should be an array of integers.|\n",
"|**X_valid**|(sparse) array-like, shape = [n_samples, n_features]|\n",
"|**y_valid**|(sparse) array-like, shape = [n_samples, ], [n_samples, n_classes]<br>Multi-class targets. An indicator matrix turns on multilabel classification. This should be an array of integers.|\n",
"|**enable_ensembling**|Flag to enable an ensembling iteration after all the other iterations complete.|\n",
"|**ensemble_iterations**|Number of iterations during which we choose a fitted pipeline to be part of the final ensemble.|\n",
"|**path**|Relative path to the project folder. AutoML stores configuration files for the experiment under this folder. You can specify a new empty folder.|"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_config = AutoMLConfig(task = 'classification',\n",
" debug_log = 'classification.log',\n",
" primary_metric = 'AUC_weighted',\n",
" max_time_sec = 3600,\n",
" iterations = 10,\n",
" verbosity = logging.INFO,\n",
" X = X_train, \n",
" y = y_train,\n",
" X_valid = X_valid,\n",
" y_valid = y_valid,\n",
" enable_ensembling = True,\n",
" ensemble_iterations = 5,\n",
" path = project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train the Model\n",
"\n",
"Call the `submit` method on the experiment object and pass the run configuration. Execution of local runs is synchronous. Depending on the data and the number of iterations this can run for a while.\n",
"In this example, we specify `show_output = True` to print currently running iterations to the console."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_config, show_output = True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Optionally, you can continue an interrupted local run by calling `continue_experiment` without the `iterations` parameter, or run more iterations for a completed run by specifying the `iterations` parameter:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = local_run.continue_experiment(X = X_train, \n",
" y = y_train,\n",
" X_valid = X_valid,\n",
" y_valid = y_valid,\n",
" show_output = True,\n",
" iterations = 5)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explore the Results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Widget for Monitoring Runs\n",
"\n",
"The widget will first report a \"loading\" status while running the first iteration. After completing the first iteration, an auto-updating graph and table will be shown. The widget will refresh once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"**Note:** The widget displays a link at the bottom. Use this link to open a web interface to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.widgets import RunDetails\n",
"RunDetails(local_run).show() "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"#### Retrieve All Child Runs\n",
"You can also use SDK methods to fetch all the child runs and see individual metrics that we log."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"children = list(local_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
"\n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"rundata"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method on `automl_classifier` returns the best run and the fitted model for the last invocation. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = local_run.get_output()\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Best Model Based on Any Other Metric\n",
"Show the run and the model that has the smallest `log_loss` value:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"lookup_metric = \"log_loss\"\n",
"best_run, fitted_model = local_run.get_output(metric = lookup_metric)\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Model from a Specific Iteration\n",
"Show the run and the model from the third iteration:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"iteration = 3\n",
"third_run, third_model = local_run.get_output(iteration = iteration)\n",
"print(third_run)\n",
"print(third_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test the Best Fitted Model\n",
"\n",
"#### Load Test Data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits = datasets.load_digits()\n",
"X_test = digits.data[:10, :]\n",
"y_test = digits.target[:10]\n",
"images = digits.images[:10]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Testing Our Best Pipeline\n",
"We will try to predict 2 digits and see how our model works."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Randomly select digits and test.\n",
"for index in np.random.choice(len(y_test), 2, replace = False):\n",
" print(index)\n",
" predicted = fitted_model.predict(X_test[index:index + 1])[0]\n",
" label = y_test[index]\n",
" title = \"Label value = %d Predicted value = %d \" % (label, predicted)\n",
" fig = plt.figure(1, figsize = (3,3))\n",
" ax1 = fig.add_axes((0,0,.8,.8))\n",
" ax1.set_title(title)\n",
" plt.imshow(images[index], cmap = plt.cm.gray_r, interpolation = 'nearest')\n",
" plt.show()"
]
}
],
"metadata": {
"authors": [
{
"name": "ratanase"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

442
automl/15b.auto-ml-regression-ensemble.ipynb Normal file
View File

@@ -0,0 +1,442 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 15b: Regression with ensembling on remote compute\n",
"\n",
"In this example we use the scikit-learn's [diabetes dataset](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_diabetes.html) to showcase how you can use AutoML for a simple regression problem.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. Create an `Experiment` in an existing `Workspace`.\n",
"2. Configure AutoML using `AutoMLConfig`which enables an extra ensembling iteration.\n",
"3. Train the model using remote compute.\n",
"4. Explore the results.\n",
"5. Test the best fitted model.\n",
"\n",
"<b>Disclaimers / Limitations </b>\n",
"- Currently only Train/Validation split is supported; support for cross-validation will be coming soon.\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# Choose a name for the experiment and specify the project folder.\n",
"experiment_name = 'automl-local-regression'\n",
"project_folder = './sample_projects/automl-local-regression'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace Name'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data = output, index = ['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create a Remote Linux DSVM\n",
"**Note:** If creation fails with a message about Marketplace purchase eligibilty, start creation of a DSVM through the [Azure portal](https://portal.azure.com), and select \"Want to create programmatically\" to enable programmatic creation. Once you've enabled this setting, you can exit the portal without actually creating the DSVM, and creation of the DSVM through the notebook should work."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import DsvmCompute\n",
"\n",
"dsvm_name = 'mydsvm'\n",
"try:\n",
" dsvm_compute = DsvmCompute(ws, dsvm_name)\n",
" print('Found an existing DSVM.')\n",
"except:\n",
" print('Creating a new DSVM.')\n",
" dsvm_config = DsvmCompute.provisioning_configuration(vm_size = \"Standard_D2_v2\")\n",
" dsvm_compute = DsvmCompute.create(ws, name = dsvm_name, provisioning_configuration = dsvm_config)\n",
" dsvm_compute.wait_for_completion(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create Get Data File\n",
"For remote executions you should author a `get_data.py` file containing a `get_data()` function. This file should be in the root directory of the project. You can encapsulate code to read data either from a blob storage or local disk in this file.\n",
"In this example, the `get_data()` function returns data using scikit-learn's `diabetes` dataset."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile $project_folder/get_data.py\n",
"\n",
"# Load the diabetes dataset, a well-known built-in small dataset that comes with scikit-learn.\n",
"from sklearn.datasets import load_diabetes\n",
"from sklearn.linear_model import Ridge\n",
"from sklearn.metrics import mean_squared_error\n",
"from sklearn.model_selection import train_test_split\n",
"\n",
"def get_data():\n",
" X, y = load_diabetes(return_X_y = True)\n",
"\n",
" columns = ['age', 'gender', 'bmi', 'bp', 's1', 's2', 's3', 's4', 's5', 's6']\n",
"\n",
" X_train, X_temp, y_train, y_temp = train_test_split(X, y, test_size = 0.2, random_state = 0)\n",
" X_valid, X_test, y_valid, y_test = train_test_split(X_temp, y_temp, test_size = 0.5, random_state = 0)\n",
" return { \"X\" : X_train, \"y\" : y_train, \"X_valid\": X_valid, \"y_valid\": y_valid }"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML\n",
"\n",
"Instantiate an `AutoMLConfig` object to specify the settings and data used to run the experiment.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**task**|classification or regression|\n",
"|**primary_metric**|This is the metric that you want to optimize. Regression supports the following primary metrics: <br><i>spearman_correlation</i><br><i>normalized_root_mean_squared_error</i><br><i>r2_score</i><br><i>normalized_mean_absolute_error</i>|\n",
"|**max_time_sec**|Time limit in seconds for each iteration.|\n",
"|**iterations**|Number of iterations. In each iteration AutoML trains a specific pipeline with the data.|\n",
"|**enable_ensembling**|Flag to enable an ensembling iteration after all the other iterations complete.|\n",
"|**ensemble_iterations**|Number of iterations during which we choose a fitted pipeline to be part of the final ensemble.|\n",
"|**path**|Relative path to the project folder. AutoML stores configuration files for the experiment under this folder. You can specify a new empty folder.|"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_config = AutoMLConfig(task = 'regression',\n",
" max_time_sec = 600,\n",
" iterations = 20,\n",
" primary_metric = 'spearman_correlation',\n",
" debug_log = 'regression.log',\n",
" verbosity = logging.INFO,\n",
" compute_target = dsvm_compute,\n",
" data_script = project_folder + \"/get_data.py\",\n",
" enable_ensembling = True,\n",
" ensemble_iterations = 5,\n",
" path = project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train the Model\n",
"\n",
"Call the `submit` method on the experiment object and pass the run configuration. Execution of local runs is synchronous. Depending on the data and the number of iterations this can run for a while.\n",
"In this example, we specify `show_output = True` to print currently running iterations to the console."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_config, show_output = True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explore the Results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Widget for Monitoring Runs\n",
"\n",
"The widget will first report a \"loading\" status while running the first iteration. After completing the first iteration, an auto-updating graph and table will be shown. The widget will refresh once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"**Note:** The widget displays a link at the bottom. Use this link to open a web interface to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.widgets import RunDetails\n",
"RunDetails(local_run).show() "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"#### Retrieve All Child Runs\n",
"You can also use SDK methods to fetch all the child runs and see individual metrics that we log."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"children = list(local_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
"\n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"rundata"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method on `automl_classifier` returns the best run and the fitted model for the last invocation. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = local_run.get_output()\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Best Model Based on Any Other Metric\n",
"Show the run and the model that has the smallest `root_mean_squared_error` value."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"lookup_metric = \"root_mean_squared_error\"\n",
"best_run, fitted_model = local_run.get_output(metric = lookup_metric)\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test the Best Model (Ensemble)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Predict on training and test set, and calculate residual values."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.datasets import load_diabetes\n",
"from sklearn.linear_model import Ridge\n",
"from sklearn.metrics import mean_squared_error\n",
"from sklearn.model_selection import train_test_split\n",
"\n",
"X, y = load_diabetes(return_X_y = True)\n",
"\n",
"X_train, X_temp, y_train, y_temp = train_test_split(X, y, test_size = 0.2, random_state = 0)\n",
"X_valid, X_test, y_valid, y_test = train_test_split(X_temp, y_temp, test_size = 0.5, random_state = 0)\n",
"\n",
"\n",
"y_pred_train = fitted_model.predict(X_train)\n",
"y_residual_train = y_train - y_pred_train\n",
"\n",
"y_pred_test = fitted_model.predict(X_test)\n",
"y_residual_test = y_test - y_pred_test"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%matplotlib inline\n",
"import matplotlib.pyplot as plt\n",
"import numpy as np\n",
"from sklearn import datasets\n",
"from sklearn.metrics import mean_squared_error, r2_score\n",
"\n",
"# Set up a multi-plot chart.\n",
"f, (a0, a1) = plt.subplots(1, 2, gridspec_kw = {'width_ratios':[1, 1], 'wspace':0, 'hspace': 0})\n",
"f.suptitle('Regression Residual Values', fontsize = 18)\n",
"f.set_figheight(6)\n",
"f.set_figwidth(16)\n",
"\n",
"# Plot residual values of training set.\n",
"a0.axis([0, 360, -200, 200])\n",
"a0.plot(y_residual_train, 'bo', alpha = 0.5)\n",
"a0.plot([-10,360],[0,0], 'r-', lw = 3)\n",
"a0.text(16,170,'RMSE = {0:.2f}'.format(np.sqrt(mean_squared_error(y_train, y_pred_train))), fontsize = 12)\n",
"a0.text(16,140,'R2 score = {0:.2f}'.format(r2_score(y_train, y_pred_train)), fontsize = 12)\n",
"a0.set_xlabel('Training samples', fontsize = 12)\n",
"a0.set_ylabel('Residual Values', fontsize = 12)\n",
"\n",
"# Plot a histogram.\n",
"a0.hist(y_residual_train, orientation = 'horizontal', color = 'b', bins = 10, histtype = 'step');\n",
"a0.hist(y_residual_train, orientation = 'horizontal', color = 'b', alpha = 0.2, bins = 10);\n",
"\n",
"# Plot residual values of test set.\n",
"a1.axis([0, 90, -200, 200])\n",
"a1.plot(y_residual_test, 'bo', alpha = 0.5)\n",
"a1.plot([-10,360],[0,0], 'r-', lw = 3)\n",
"a1.text(5,170,'RMSE = {0:.2f}'.format(np.sqrt(mean_squared_error(y_test, y_pred_test))), fontsize = 12)\n",
"a1.text(5,140,'R2 score = {0:.2f}'.format(r2_score(y_test, y_pred_test)), fontsize = 12)\n",
"a1.set_xlabel('Test samples', fontsize = 12)\n",
"a1.set_yticklabels([])\n",
"\n",
"# Plot a histogram.\n",
"a1.hist(y_residual_test, orientation = 'horizontal', color = 'b', bins = 10, histtype = 'step')\n",
"a1.hist(y_residual_test, orientation = 'horizontal', color = 'b', alpha = 0.2, bins = 10)\n",
"\n",
"plt.show()"
]
}
],
"metadata": {
"authors": [
{
"name": "ratanase"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

390
automl/17.auto-ml-classification_with_tensorflow.ipynb
View File

@@ -1,390 +0,0 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 17: Classification with Local Compute with Tesnorflow DNNClassifier and LinearClassifier using whitelist models feature.\n",
"\n",
"In this example we use the scikit-learn's [digit dataset](http://scikit-learn.org/stable/datasets/index.html#optical-recognition-of-handwritten-digits-dataset) to showcase how you can use AutoML for a simple classification problem.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"This notebooks shows how can automl can be trained on a a selected list of models,see the readme.md for the models.\n",
"This trains the model exclusively on tensorflow based models.\n",
"\n",
"In this notebook you will learn how to:\n",
"1. Create an `Experiment` in an existing `Workspace`.\n",
"2. Configure AutoML using `AutoMLConfig`.\n",
"3. Train the model on a whilelisted models using local compute. \n",
"4. Explore the results.\n",
"5. Test the best fitted model.\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create an Experiment\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import random\n",
"\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# Choose a name for the experiment and specify the project folder.\n",
"experiment_name = 'automl-local-classification'\n",
"project_folder = './sample_projects/automl-local-classification'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace Name'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data = output, index = ['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Load Training Data\n",
"\n",
"This uses scikit-learn's [load_digits](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_digits.html) method."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn import datasets\n",
"\n",
"digits = datasets.load_digits()\n",
"\n",
"# Exclude the first 100 rows from training so that they can be used for test.\n",
"X_train = digits.data[100:,:]\n",
"y_train = digits.target[100:]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure AutoML\n",
"\n",
"Instantiate an `AutoMLConfig` object to specify the settings and data used to run the experiment.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**task**|classification or regression|\n",
"|**primary_metric**|This is the metric that you want to optimize. Classification supports the following primary metrics: <br><i>accuracy</i><br><i>AUC_weighted</i><br><i>balanced_accuracy</i><br><i>average_precision_score_weighted</i><br><i>precision_score_weighted</i>|\n",
"|**iteration_timeout_minutes**|Time limit in minutes for each iteration.|\n",
"|**iterations**|Number of iterations. In each iteration AutoML trains a specific pipeline with the data.|\n",
"|**n_cross_validations**|Number of cross validation splits.|\n",
"|**X**|(sparse) array-like, shape = [n_samples, n_features]|\n",
"|**y**|(sparse) array-like, shape = [n_samples, ], [n_samples, n_classes]<br>Multi-class targets. An indicator matrix turns on multilabel classification. This should be an array of integers.|\n",
"|**path**|Relative path to the project folder. AutoML stores configuration files for the experiment under this folder. You can specify a new empty folder.|"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_config = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" primary_metric = 'AUC_weighted',\n",
" iteration_timeout_minutes = 60,\n",
" iterations = 10,\n",
" n_cross_validations = 3,\n",
" verbosity = logging.INFO,\n",
" X = X_train, \n",
" y = y_train,\n",
" enable_tf=True,\n",
" whitelist_models=[\"TensorFlowLinearClassifier\", \"TensorFlowDNN\"],\n",
" path = project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train the Models\n",
"\n",
"Call the `submit` method on the experiment object and pass the run configuration. Execution of local runs is synchronous. Depending on the data and the number of iterations this can run for a while.\n",
"In this example, we specify `show_output = True` to print currently running iterations to the console."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_config, show_output = True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explore the Results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Widget for Monitoring Runs\n",
"\n",
"The widget will first report a \"loading\" status while running the first iteration. After completing the first iteration, an auto-updating graph and table will be shown. The widget will refresh once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"**Note:** The widget displays a link at the bottom. Use this link to open a web interface to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.widgets import RunDetails\n",
"RunDetails(local_run).show() "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"#### Retrieve All Child Runs\n",
"You can also use SDK methods to fetch all the child runs and see individual metrics that we log."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"children = list(local_run.get_children())\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
"\n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"rundata"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method returns the best run and the fitted model. The Model includes the pipeline and any pre-processing. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = local_run.get_output()\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Best Model Based on Any Other Metric\n",
"Show the run and the model that has the smallest `log_loss` value:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"lookup_metric = \"log_loss\"\n",
"best_run, fitted_model = local_run.get_output(metric = lookup_metric)\n",
"print(best_run)\n",
"print(fitted_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Model from a Specific Iteration\n",
"Show the run and the model from the third iteration:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"iteration = 3\n",
"third_run, third_model = local_run.get_output(iteration = iteration)\n",
"print(third_run)\n",
"print(third_model)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test the Best Fitted Model\n",
"\n",
"#### Load Test Data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"digits = datasets.load_digits()\n",
"X_test = digits.data[:10, :]\n",
"y_test = digits.target[:10]\n",
"images = digits.images[:10]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Testing Our Best Fitted Model\n",
"We will try to predict 2 digits and see how our model works."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Randomly select digits and test.\n",
"for index in np.random.choice(len(y_test), 2, replace = False):\n",
" print(index)\n",
" predicted = fitted_model.predict(X_test[index:index + 1])[0]\n",
" label = y_test[index]\n",
" title = \"Label value = %d Predicted value = %d \" % (label, predicted)\n",
" fig = plt.figure(1, figsize = (3,3))\n",
" ax1 = fig.add_axes((0,0,.8,.8))\n",
" ax1.set_title(title)\n",
" plt.imshow(images[index], cmap = plt.cm.gray_r, interpolation = 'nearest')\n",
" plt.show()"
]
}
],
"metadata": {
"authors": [
{
"name": "savitam"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

394
automl/18.auto-ml-timeseries.ipynb
View File

@@ -1,394 +0,0 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# AutoML 18: Energy Demand Forecasting\n",
"\n",
"In this example, we show how AutoML can be used for energy demand forecasting.\n",
"\n",
"Make sure you have executed the [00.configuration](00.configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you would see\n",
"1. Creating an Experiment in an existing Workspace\n",
"2. Instantiating AutoMLConfig with new task type \"forecasting\" for timeseries data training, and other timeseries related settings: for this dataset we use the basic one: \"time_column_name\" \n",
"3. Training the Model using local compute\n",
"4. Exploring the results\n",
"5. Testing the fitted model"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create Experiment\n",
"\n",
"As part of the setup you have already created a <b>Workspace</b>. For AutoML you would need to create an <b>Experiment</b>. An <b>Experiment</b> is a named object in a <b>Workspace</b>, which is used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import azureml.core\n",
"import pandas as pd\n",
"import numpy as np\n",
"import os\n",
"import logging\n",
"\n",
"\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.train.automl import AutoMLConfig\n",
"from azureml.train.automl.run import AutoMLRun\n",
"from matplotlib import pyplot as plt\n",
"from matplotlib.pyplot import imshow\n",
"from sklearn.metrics import mean_absolute_error, mean_squared_error, r2_score"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# choose a name for the run history container in the workspace\n",
"experiment_name = 'automl-energydemandforecasting'\n",
"# project folder\n",
"project_folder = './sample_projects/automl-local-energydemandforecasting'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Run History Name'] = experiment_name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"pd.DataFrame(data=output, index=['']).T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Read Data\n",
"Read energy demanding data from file, and preview data."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"data = pd.read_csv(\"nyc_energy.csv\", parse_dates=['timeStamp'])\n",
"data.head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Split the data to train and test\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"train = data[data['timeStamp'] < '2017-02-01']\n",
"test = data[data['timeStamp'] >= '2017-02-01']\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Prepare the test data, we will feed X_test to the fitted model and get prediction"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"y_test = test.pop('demand').values\n",
"X_test = test"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Split the train data to train and valid\n",
"\n",
"Use one month's data as valid data\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"X_train = train[train['timeStamp'] < '2017-01-01']\n",
"X_valid = train[train['timeStamp'] >= '2017-01-01']\n",
"y_train = X_train.pop('demand').values\n",
"y_valid = X_valid.pop('demand').values\n",
"print(X_train.shape)\n",
"print(y_train.shape)\n",
"print(X_valid.shape)\n",
"print(y_valid.shape)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Instantiate Auto ML Config\n",
"\n",
"Instantiate a AutoMLConfig object. This defines the settings and data used to run the experiment.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**task**|forecasting|\n",
"|**primary_metric**|This is the metric that you want to optimize.<br> Forecasting supports the following primary metrics <br><i>spearman_correlation</i><br><i>normalized_root_mean_squared_error</i><br><i>r2_score</i><br><i>normalized_mean_absolute_error</i>\n",
"|**iterations**|Number of iterations. In each iteration, Auto ML trains a specific pipeline on the given data|\n",
"|**iteration_timeout_minutes**|Time limit in minutes for each iteration.|\n",
"|**X**|(sparse) array-like, shape = [n_samples, n_features]|\n",
"|**y**|(sparse) array-like, shape = [n_samples, ], [n_samples, n_classes]<br>Multi-class targets. An indicator matrix turns on multilabel classification. This should be an array of integers. |\n",
"|**X_valid**|Data used to evaluate a model in a iteration. (sparse) array-like, shape = [n_samples, n_features]|\n",
"|**y_valid**|Data used to evaluate a model in a iteration. (sparse) array-like, shape = [n_samples, ], [n_samples, n_classes]<br>Multi-class targets. An indicator matrix turns on multilabel classification. This should be an array of integers. |\n",
"|**path**|Relative path to the project folder. AutoML stores configuration files for the experiment under this folder. You can specify a new empty folder. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"time_column_name = 'timeStamp'\n",
"automl_settings = {\n",
" \"time_column_name\": time_column_name,\n",
"}\n",
"\n",
"\n",
"automl_config = AutoMLConfig(task = 'forecasting',\n",
" debug_log = 'automl_nyc_energy_errors.log',\n",
" primary_metric='normalized_root_mean_squared_error',\n",
" iterations = 10,\n",
" iteration_timeout_minutes = 5,\n",
" X = X_train,\n",
" y = y_train,\n",
" X_valid = X_valid,\n",
" y_valid = y_valid,\n",
" path=project_folder,\n",
" verbosity = logging.INFO,\n",
" **automl_settings)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Training the Model\n",
"\n",
"You can call the submit method on the experiment object and pass the run configuration. For Local runs the execution is synchronous. Depending on the data and number of iterations this can run for while.\n",
"You will see the currently running iterations printing to the console."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"local_run = experiment.submit(automl_config, show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best Model\n",
"Below we select the best pipeline from our iterations. The get_output method on automl_classifier returns the best run and the fitted model for the last fit invocation. There are overloads on get_output that allow you to retrieve the best run and fitted model for any logged metric or a particular iteration."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, fitted_model = local_run.get_output()\n",
"fitted_model.steps"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test the Best Fitted Model\n",
"\n",
"Predict on training and test set, and calculate residual values."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"y_pred = fitted_model.predict(X_test)\n",
"y_pred"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Define a Check Data Function\n",
"\n",
"Remove the nan values from y_test to avoid error when calculate metrics "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"def _check_calc_input(y_true, y_pred, rm_na=True):\n",
" \"\"\"\n",
" Check that 'y_true' and 'y_pred' are non-empty and\n",
" have equal length.\n",
"\n",
" :param y_true: Vector of actual values\n",
" :type y_true: array-like\n",
"\n",
" :param y_pred: Vector of predicted values\n",
" :type y_pred: array-like\n",
"\n",
" :param rm_na:\n",
" If rm_na=True, remove entries where y_true=NA and y_pred=NA.\n",
" :type rm_na: boolean\n",
"\n",
" :return:\n",
" Tuple (y_true, y_pred). if rm_na=True,\n",
" the returned vectors may differ from their input values.\n",
" :rtype: Tuple with 2 entries\n",
" \"\"\"\n",
" if len(y_true) != len(y_pred):\n",
" raise ValueError(\n",
" 'the true values and prediction values do not have equal length.')\n",
" elif len(y_true) == 0:\n",
" raise ValueError(\n",
" 'y_true and y_pred are empty.')\n",
" # if there is any non-numeric element in the y_true or y_pred,\n",
" # the ValueError exception will be thrown.\n",
" y_true = np.array(y_true).astype(float)\n",
" y_pred = np.array(y_pred).astype(float)\n",
" if rm_na:\n",
" # remove entries both in y_true and y_pred where at least\n",
" # one element in y_true or y_pred is missing\n",
" y_true_rm_na = y_true[~(np.isnan(y_true) | np.isnan(y_pred))]\n",
" y_pred_rm_na = y_pred[~(np.isnan(y_true) | np.isnan(y_pred))]\n",
" return (y_true_rm_na, y_pred_rm_na)\n",
" else:\n",
" return y_true, y_pred"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Use the Check Data Function to remove the nan values from y_test to avoid error when calculate metrics "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"y_test,y_pred = _check_calc_input(y_test,y_pred)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Calculate metrics for the prediction\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(\"[Test Data] \\nRoot Mean squared error: %.2f\" % np.sqrt(mean_squared_error(y_test, y_pred)))\n",
"# Explained variance score: 1 is perfect prediction\n",
"print('mean_absolute_error score: %.2f' % mean_absolute_error(y_test, y_pred))\n",
"print('R2 score: %.2f' % r2_score(y_test, y_pred))\n",
"\n",
"\n",
"\n",
"# Plot outputs\n",
"test_pred = plt.scatter(y_test, y_pred, color='b')\n",
"test_test = plt.scatter(y_test, y_test, color='g')\n",
"plt.legend((test_pred, test_test), ('prediction', 'truth'), loc='upper left', fontsize=8)\n",
"plt.show()"
]
}
],
"metadata": {
"authors": [
{
"name": "xiaga"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

126
automl/README.md
View File

@@ -1,7 +1,6 @@
# Table of Contents
1. [Automated ML Introduction](#introduction)
1. [Running samples in Azure Notebooks](#jupyter)
1. [Running samples in Azure Databricks](#databricks)
1. [Running samples in a Local Conda environment](#localconda)
1. [Automated ML SDK Sample Notebooks](#samples)
1. [Documentation](#documentation)
@@ -9,7 +8,7 @@
1. [Troubleshooting](#troubleshooting)
<a name="introduction"></a>
# Automated ML introduction
# Automated ML introduction
Automated machine learning (automated ML) builds high quality machine learning models for you by automating model and hyperparameter selection. Bring a labelled dataset that you want to build a model for, automated ML will give you a high quality machine learning model that you can use for predictions.
@@ -17,32 +16,20 @@ If you are new to Data Science, AutoML will help you get jumpstarted by simplify
If you are an experienced data scientist, AutoML will help increase your productivity by intelligently performing the model and hyperparameter selection for your training and generates high quality models much quicker than manually specifying several combinations of the parameters and running training jobs. AutoML provides visibility and access to all the training jobs and the performance characteristics of the models to help you further tune the pipeline if you desire.
Below are the three execution environments supported by AutoML.
<a name="jupyter"></a>
## Running samples in Azure Notebooks - Jupyter based notebooks in the Azure cloud
## Running samples in Azure Notebooks - Jupyter based notebooks in the Azure cloud
1. [![Azure Notebooks](https://notebooks.azure.com/launch.png)](https://aka.ms/aml-clone-azure-notebooks)
[Import sample notebooks ](https://aka.ms/aml-clone-azure-notebooks) into Azure Notebooks.
1. Follow the instructions in the [../00.configuration](00.configuration.ipynb) notebook to create and connect to a workspace.
1. Open one of the sample notebooks.
**Make sure the Azure Notebook kernel is set to `Python 3.6`** when you open a notebook.
**Make sure the Azure Notebook kernel is set to `Python 3.6`** when you open a notebook.
![set kernel to Python 3.6](../images/python36.png)
<a name="databricks"></a>
## Running samples in Azure Databricks
**NOTE**: Please create your Azure Databricks cluster as v4.x (high concurrency preferred) with **Python 3** (dropdown).
**NOTE**: You should at least have contributor access to your Azure subcription to run the notebook.
- Please remove the previous SDK version if there is any and install the latest SDK by installing **azureml-sdk[automl_databricks]** as a PyPi library in Azure Databricks workspace.
- Download the sample notebook 16a.auto-ml-classification-local-azuredatabricks from [GitHub](https://github.com/Azure/MachineLearningNotebooks) and import into the Azure databricks workspace.
- Attach the notebook to the cluster.
<a name="localconda"></a>
## Running samples in a Local Conda environment
## Running samples in a Local Conda environment
To run these notebook on your own notebook server, use these installation instructions.
@@ -61,7 +48,7 @@ jupyter notebook
```
### 1. Install mini-conda from [here](https://conda.io/miniconda.html), choose Python 3.7 or higher.
### 1. Install mini-conda from [here](https://conda.io/miniconda.html), choose Python 3.7 or higher.
- **Note**: if you already have conda installed, you can keep using it but it should be version 4.4.10 or later (as shown by: conda -V). If you have a previous version installed, you can update it using the command: conda update conda.
There's no need to install mini-conda specifically.
@@ -70,23 +57,23 @@ There's no need to install mini-conda specifically.
### 3. Setup a new conda environment
The **automl/automl_setup** script creates a new conda environment, installs the necessary packages, configures the widget and starts a jupyter notebook.
It takes the conda environment name as an optional parameter. The default conda environment name is azure_automl. The exact command depends on the operating system. See the specific sections below for Windows, Mac and Linux. It can take about 10 minutes to execute.
It takes the conda environment name as an optional parameter. The default conda environment name is azure_automl. The exact command depends on the operating system. It can take about 10 minutes to execute.
## Windows
Start an **Anaconda Prompt** window, cd to the **automl** folder where the sample notebooks were extracted and then run:
Start a conda command windows, cd to the **automl** folder where the sample notebooks were extracted and then run:
```
automl_setup
```
## Mac
Install "Command line developer tools" if it is not already installed (you can use the command: `xcode-select --install`).
Start a Terminal windows, cd to the **automl** folder where the sample notebooks were extracted and then run:
Start a Terminal windows, cd to the **automl** folder where the sample notebooks were extracted and then run:
```
bash automl_setup_mac.sh
```
## Linux
cd to the **automl** folder where the sample notebooks were extracted and then run:
cd to the **automl** folder where the sample notebooks were extracted and then run:
```
bash automl_setup_linux.sh
@@ -101,7 +88,7 @@ bash automl_setup_linux.sh
- Follow the instructions in the individual notebooks to explore various features in AutoML
<a name="samples"></a>
# Automated ML SDK Sample Notebooks
# Automated ML SDK Sample Notebooks
- [00.configuration.ipynb](00.configuration.ipynb)
- Register Machine Learning Services Resource Provider
- Create new Azure ML Workspace
@@ -109,12 +96,12 @@ bash automl_setup_linux.sh
- [01.auto-ml-classification.ipynb](01.auto-ml-classification.ipynb)
- Dataset: scikit learn's [digit dataset](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_digits.html#sklearn.datasets.load_digits)
- Simple example of using Auto ML for classification
- Simple example of using Auto ML for classification
- Uses local compute for training
- [02.auto-ml-regression.ipynb](02.auto-ml-regression.ipynb)
- Dataset: scikit learn's [diabetes dataset](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_diabetes.html)
- Simple example of using Auto ML for regression
- Simple example of using Auto ML for regression
- Uses local compute for training
- [03.auto-ml-remote-execution.ipynb](03.auto-ml-remote-execution.ipynb)
@@ -139,7 +126,7 @@ bash automl_setup_linux.sh
- Dataset: [Burning Man 2016 dataset](https://innovate.burningman.org/datasets-page/)
- handling text data with preprocess flag
- Reading data from a blob store for remote executions
- using pandas dataframes for reading data
- using pandas dataframes for reading data
- [05.auto-ml-missing-data-blacklist-early-termination.ipynb](05.auto-ml-missing-data-blacklist-early-termination.ipynb)
- Dataset: scikit learn's [digit dataset](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_digits.html#sklearn.datasets.load_digits)
@@ -148,7 +135,7 @@ bash automl_setup_linux.sh
- Handling Missing Data in the input
- [06.auto-ml-sparse-data-train-test-split.ipynb](06.auto-ml-sparse-data-train-test-split.ipynb)
- Dataset: Scikit learn's [20newsgroup](http://scikit-learn.org/stable/datasets/twenty_newsgroups.html)
- Dataset: Scikit learn's [20newsgroup](http://scikit-learn.org/stable/datasets/twenty_newsgroups.html)
- Handle sparse datasets
- Specify custom train and validation set
@@ -164,7 +151,7 @@ bash automl_setup_linux.sh
- [09.auto-ml-classification-with-deployment.ipynb](09.auto-ml-classification-with-deployment.ipynb)
- Dataset: scikit learn's [digit dataset](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_digits.html#sklearn.datasets.load_digits)
- Simple example of using Auto ML for classification
- Simple example of using Auto ML for classification
- Registering the model
- Creating Image and creating aci service
- Testing the aci service
@@ -185,10 +172,10 @@ bash automl_setup_linux.sh
- Using DataPrep for reading data
- [14.auto-ml-model-explanation.ipynb](14.auto-ml-model-explanation.ipynb)
- Dataset: sklearn's [iris dataset](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_iris.html)
- Dataset: seaborn's [iris dataset](https://seaborn.pydata.org/generated/seaborn.load_dataset.html)
- Explaining the AutoML classification pipeline
- Visualizing feature importance in widget
- [15a.auto-ml-classification-ensemble.ipynb](15a.auto-ml-classification-ensemble.ipynb)
- Dataset: scikit learn's [digit dataset](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_digits.html#sklearn.datasets.load_digits)
- Enables an extra iteration for generating an Ensemble of models
@@ -199,21 +186,8 @@ bash automl_setup_linux.sh
- Enables an extra iteration for generating an Ensemble of models
- Uses remote Linux DSVM for training
- [16a.auto-ml-classification-local-azuredatabricks.ipynb](16a.auto-ml-classification-local-azuredatabricks.ipynb)
- Dataset: scikit learn's [digit dataset](https://innovate.burningman.org/datasets-page/)
- Example of using AutoML for classification using Azure Databricks as the platform for training
- [17.auto-ml-classification_with_tensorflow.ipynb](17.auto-ml-classification_with_tensorflow.ipynb)
- Dataset: scikit learn's [digit dataset](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_digits.html#sklearn.datasets.load_digits)
- Simple example of using Auto ML for classification with whitelisting tensorflow models.checkout
- Uses local compute for training
- [18.auto-ml-timeseries.ipynb](18.auto-ml-timeseries.ipynb)
- Dataset: NYC energy demanding data
- Example of using AutoML for timeseries data training
<a name="documentation"></a>
# Documentation
# Documentation
## Table of Contents
1. [Automated ML Settings ](#automlsettings)
1. [Cross validation split options](#cvsplits)
@@ -221,51 +195,22 @@ bash automl_setup_linux.sh
1. [Data pre-processing and featurization](#preprocessing)
<a name="automlsettings"></a>
## Automated ML Settings
## Automated ML Settings
|Property|Description|Default|
|-|-|-|
|**primary_metric**|This is the metric that you want to optimize.<br><br> Classification supports the following primary metrics <br><i>accuracy</i><br><i>AUC_weighted</i><br><i>average_precision_score_weighted</i><br><i>norm_macro_recall</i><br><i>precision_score_weighted</i><br><br> Regression supports the following primary metrics <br><i>spearman_correlation</i><br><i>normalized_root_mean_squared_error</i><br><i>r2_score</i><br><i>normalized_mean_absolute_error</i><br><i>normalized_root_mean_squared_log_error</i>| Classification: accuracy <br><br> Regression: spearman_correlation
|**iteration_timeout_minutes**|Time limit in minutes for each iteration|None|
|**primary_metric**|This is the metric that you want to optimize.<br><br> Classification supports the following primary metrics <br><i>accuracy</i><br><i>AUC_weighted</i><br><i>balanced_accuracy</i><br><i>average_precision_score_weighted</i><br><i>precision_score_weighted</i><br><br> Regression supports the following primary metrics <br><i>spearman_correlation</i><br><i>normalized_root_mean_squared_error</i><br><i>r2_score</i><br><i>normalized_mean_absolute_error</i><br><i>normalized_root_mean_squared_log_error</i>| Classification: accuracy <br><br> Regression: spearman_correlation
|**max_time_sec**|Time limit in seconds for each iteration|None|
|**iterations**|Number of iterations. In each iteration trains the data with a specific pipeline. To get the best result, use at least 100. |100|
|**n_cross_validations**|Number of cross validation splits|None|
|**validation_size**|Size of validation set as percentage of all training samples|None|
|**max_concurrent_iterations**|Max number of iterations that would be executed in parallel|1|
|**concurrent_iterations**|Max number of iterations that would be executed in parallel|1|
|**preprocess**|*True/False* <br>Setting this to *True* enables preprocessing <br>on the input to handle missing data, and perform some common feature extraction<br>*Note: If input data is Sparse you cannot use preprocess=True*|False|
|**max_cores_per_iteration**| Indicates how many cores on the compute target would be used to train a single pipeline.<br> You can set it to *-1* to use all cores|1|
|**experiment_exit_score**|*double* value indicating the target for *primary_metric*. <br> Once the target is surpassed the run terminates|None|
|**blacklist_models**|*Array* of *strings* indicating models to ignore for Auto ML from the list of models.|None|
|**whitelist_models**|*Array* of *strings* use only models listed for Auto ML from the list of models..|None|
<a name="cvsplits"></a>
## List of models for white list/blacklist
**Classification**
<br><i>LogisticRegression</i>
<br><i>SGD</i>
<br><i>MultinomialNaiveBayes</i>
<br><i>BernoulliNaiveBayes</i>
<br><i>SVM</i>
<br><i>LinearSVM</i>
<br><i>KNN</i>
<br><i>DecisionTree</i>
<br><i>RandomForest</i>
<br><i>ExtremeRandomTrees</i>
<br><i>LightGBM</i>
<br><i>GradientBoosting</i>
<br><i>TensorFlowDNN</i>
<br><i>TensorFlowLinearClassifier</i>
<br><br>**Regression**
<br><i>ElasticNet</i>
<br><i>GradientBoosting</i>
<br><i>DecisionTree</i>
<br><i>KNN</i>
<br><i>LassoLars</i>
<br><i>SGD</i>
<br><i>RandomForest</i>
<br><i>ExtremeRandomTrees</i>
<br><i>LightGBM</i>
<br><i>TensorFlowLinearRegressor</i>
<br><i>TensorFlowDNN</i>
|**exit_score**|*double* value indicating the target for *primary_metric*. <br> Once the target is surpassed the run terminates|None|
|**blacklist_algos**|*Array* of *strings* indicating pipelines to ignore for Auto ML.<br><br> Allowed values for **Classification**<br><i>LogisticRegression</i><br><i>SGDClassifierWrapper</i><br><i>NBWrapper</i><br><i>BernoulliNB</i><br><i>SVCWrapper</i><br><i>LinearSVMWrapper</i><br><i>KNeighborsClassifier</i><br><i>DecisionTreeClassifier</i><br><i>RandomForestClassifier</i><br><i>ExtraTreesClassifier</i><br><i>LightGBMClassifier</i><br><br>Allowed values for **Regression**<br><i>ElasticNet</i><br><i>GradientBoostingRegressor</i><br><i>DecisionTreeRegressor</i><br><i>KNeighborsRegressor</i><br><i>LassoLars</i><br><i>SGDRegressor</i><br><i>RandomForestRegressor</i><br><i>ExtraTreesRegressor</i>|None|
<a name="cvsplits"></a>
## Cross validation split options
### K-Folds Cross Validation
Use *n_cross_validations* setting to specify the number of cross validations. The training data set will be randomly split into *n_cross_validations* folds of equal size. During each cross validation round, one of the folds will be used for validation of the model trained on the remaining folds. This process repeats for *n_cross_validations* rounds until each fold is used once as validation set. Finally, the average scores accross all *n_cross_validations* rounds will be reported, and the corresponding model will be retrained on the whole training data set.
@@ -277,7 +222,7 @@ Use *validation_size* to specify the percentage of the training data set that sh
You can specify seperate train and validation set either through the get_data() or directly to the fit method.
<a name="getdata"></a>
## get_data() syntax
## get_data() syntax
The *get_data()* function can be used to return a dictionary with these values:
|Key|Type|Dependency|Mutually Exclusive with|Description|
@@ -294,7 +239,7 @@ The *get_data()* function can be used to return a dictionary with these values:
|cv_splits_indices|Array of integers|data_train||*Optional* List of indexes to split the data for cross validation|
<a name="preprocessing"></a>
## Data pre-processing and featurization
## Data pre-processing and featurization
If you use `preprocess=True`, the following data preprocessing steps are performed automatically for you:
1. Dropping high cardinality or no variance features
@@ -309,7 +254,7 @@ If you use `preprocess=True`, the following data preprocessing steps are perform
- Numeric features with very few unique values are transformed into categorical features.
<a name="pythoncommand"></a>
# Running using python command
# Running using python command
Jupyter notebook provides a File / Download as / Python (.py) option for saving the notebook as a Python file.
You can then run this file using the python command.
However, on Windows the file needs to be modified before it can be run.
@@ -320,12 +265,13 @@ The following condition must be added to the main code in the file:
The main code of the file must be indented so that it is under this condition.
<a name="troubleshooting"></a>
# Troubleshooting
# Troubleshooting
## Iterations fail and the log contains "MemoryError"
This can be caused by insufficient memory on the DSVM. AutoML loads all training data into memory. So, the available memory should be more than the training data size.
If you are using a remote DSVM, memory is needed for each concurrent iteration. The max_concurrent_iterations setting specifies the maximum concurrent iterations. For example, if the training data size is 8Gb and max_concurrent_iterations is set to 10, the minimum memory required is at least 80Gb.
To resolve this issue, allocate a DSVM with more memory or reduce the value specified for max_concurrent_iterations.
If you are using a remote DSVM, memory is needed for each concurrent iteration. The concurrent_iterations setting specifies the maximum concurrent iterations. For example, if the training data size is 8Gb and concurrent_iterations is set to 10, the minimum memory required is at least 80Gb.
To resolve this issue, allocate a DSVM with more memory or reduce the value specified for concurrent_iterations.
## Iterations show as "Not Responding" in the RunDetails widget.
This can be caused by too many concurrent iterations for a remote DSVM. Each concurrent iteration usually takes 100% of a core when it is running. Some iterations can use multiple cores. So, the max_concurrent_iterations setting should always be less than the number of cores of the DSVM.
To resolve this issue, try reducing the value specified for the max_concurrent_iterations setting.
This can be caused by too many concurrent iterations for a remote DSVM. Each concurrent iteration usually takes 100% of a core when it is running. Some iterations can use multiple cores. So, the concurrent_iterations setting should always be less than the number of cores of the DSVM.
To resolve this issue, try reducing the value specified for the concurrent_iterations setting.

19
automl/automl_env.yml Normal file
View File

@@ -0,0 +1,19 @@
name: azure_automl
dependencies:
# The python interpreter version.
# Currently Azure ML only supports 3.5.2 and later.
- python=3.6
- nb_conda
- matplotlib
- numpy>=1.11.0,<1.15.0
- cython
- urllib3<1.24
- scipy>=0.19.0,<0.20.0
- scikit-learn>=0.18.0,<=0.19.1
- pandas>=0.22.0,<0.23.0
- pip:
# Required packages for AzureML execution, history, and data preparation.
- azureml-sdk[automl,notebooks]
- pandas_ml

17
automl/automl_env_mac.yml Normal file
View File

@@ -0,0 +1,17 @@
name: azure_automl
dependencies:
# The python interpreter version.
# Currently Azure ML only supports 3.5.2 and later.
- python=3.6
- nb_conda
- matplotlib
- numpy>=1.15.3
- cython
- urllib3<1.24
- scipy>=0.19.0,<0.20.0
- scikit-learn>=0.18.0,<=0.19.1
- pandas>=0.22.0,<0.23.0
- pip:
# Required packages for AzureML execution, history, and data preparation.
- azureml-sdk[automl,notebooks]
- pandas_ml

50
automl/automl_setup.cmd Normal file
View File

@@ -0,0 +1,50 @@
@echo off
set conda_env_name=%1
set automl_env_file=%2
set PIP_NO_WARN_SCRIPT_LOCATION=0
IF "%conda_env_name%"=="" SET conda_env_name="azure_automl"
IF "%automl_env_file%"=="" SET automl_env_file="automl_env.yml"
IF NOT EXIST %automl_env_file% GOTO YmlMissing
call conda activate %conda_env_name% 2>nul:
if not errorlevel 1 (
echo Upgrading azureml-sdk[automl] in existing conda environment %conda_env_name%
call pip install --upgrade azureml-sdk[automl,notebooks]
if errorlevel 1 goto ErrorExit
) else (
call conda env create -f %automl_env_file% -n %conda_env_name%
)
call conda activate %conda_env_name% 2>nul:
if errorlevel 1 goto ErrorExit
call pip install psutil
call jupyter nbextension install --py azureml.train.widgets --user
if errorlevel 1 goto ErrorExit
call jupyter nbextension enable --py azureml.train.widgets --user
if errorlevel 1 goto ErrorExit
echo.
echo.
echo ***************************************
echo * AutoML setup completed successfully *
echo ***************************************
echo.
echo Starting jupyter notebook - please run notebook 00.configuration
echo.
jupyter notebook --log-level=50
goto End
:YmlMissing
echo File %automl_env_file% not found.
:ErrorExit
echo Install failed
:End

47
automl/automl_setup_linux.sh Normal file
View File

@@ -0,0 +1,47 @@
#!/bin/bash
CONDA_ENV_NAME=$1
AUTOML_ENV_FILE=$2
PIP_NO_WARN_SCRIPT_LOCATION=0
if [ "$CONDA_ENV_NAME" == "" ]
then
CONDA_ENV_NAME="azure_automl"
fi
if [ "$AUTOML_ENV_FILE" == "" ]
then
AUTOML_ENV_FILE="automl_env.yml"
fi
if [ ! -f $AUTOML_ENV_FILE ]; then
echo "File $AUTOML_ENV_FILE not found"
exit 1
fi
if source activate $CONDA_ENV_NAME 2> /dev/null
then
echo "Upgrading azureml-sdk[automl] in existing conda environment" $CONDA_ENV_NAME
pip install --upgrade azureml-sdk[automl,notebooks]
else
conda env create -f $AUTOML_ENV_FILE -n $CONDA_ENV_NAME &&
source activate $CONDA_ENV_NAME &&
jupyter nbextension install --py azureml.train.widgets --user &&
jupyter nbextension enable --py azureml.train.widgets --user &&
echo "" &&
echo "" &&
echo "***************************************" &&
echo "* AutoML setup completed successfully *" &&
echo "***************************************" &&
echo "" &&
echo "Starting jupyter notebook - please run notebook 00.configuration" &&
echo "" &&
jupyter notebook --log-level=50
fi
if [ $? -gt 0 ]
then
echo "Installation failed"
fi

47
automl/automl_setup_mac.sh Normal file
View File

@@ -0,0 +1,47 @@
#!/bin/bash
CONDA_ENV_NAME=$1
AUTOML_ENV_FILE=$2
PIP_NO_WARN_SCRIPT_LOCATION=0
if [ "$CONDA_ENV_NAME" == "" ]
then
CONDA_ENV_NAME="azure_automl"
fi
if [ "$AUTOML_ENV_FILE" == "" ]
then
AUTOML_ENV_FILE="automl_env_mac.yml"
fi
if [ ! -f $AUTOML_ENV_FILE ]; then
echo "File $AUTOML_ENV_FILE not found"
exit 1
fi
if source activate $CONDA_ENV_NAME 2> /dev/null
then
echo "Upgrading azureml-sdk[automl] in existing conda environment" $CONDA_ENV_NAME
pip install --upgrade azureml-sdk[automl,notebooks]
else
conda env create -f $AUTOML_ENV_FILE -n $CONDA_ENV_NAME &&
source activate $CONDA_ENV_NAME &&
conda install lightgbm -c conda-forge -y &&
jupyter nbextension install --py azureml.train.widgets --user &&
jupyter nbextension enable --py azureml.train.widgets --user &&
pip install numpy==1.15.3
echo "" &&
echo "" &&
echo "***************************************" &&
echo "* AutoML setup completed successfully *" &&
echo "***************************************" &&
echo "" &&
echo "Starting jupyter notebook - please run notebook 00.configuration" &&
echo "" &&
jupyter notebook --log-level=50
fi
if [ $? -gt 0 ]
then
echo "Installation failed"
fi

49206
automl/nyc_energy.csv
View File

File diff suppressed because it is too large Load Diff

1
databricks/01.Installation_and_Configuration.ipynb Normal file
View File

@@ -0,0 +1 @@
{"cells":[{"cell_type":"markdown","source":["Azure ML & Azure Databricks notebooks by Parashar Shah.\n\nCopyright (c) Microsoft Corporation. All rights reserved.\n\nLicensed under the MIT License."],"metadata":{}},{"cell_type":"markdown","source":["Please ensure you have run this notebook before proceeding."],"metadata":{}},{"cell_type":"markdown","source":["Now we support installing AML SDK as library from GUI. When attaching a library follow this https://docs.databricks.com/user-guide/libraries.html and add the below string as your PyPi package (during private preview). You can select the option to attach the library to all clusters or just one cluster.\n\nProvide this full string to install the SDK:\n\nazureml-sdk[databricks]"],"metadata":{}},{"cell_type":"code","source":["import azureml.core\n\n# Check core SDK version number - based on build number of preview/master.\nprint(\"SDK version:\", azureml.core.VERSION)"],"metadata":{},"outputs":[],"execution_count":4},{"cell_type":"code","source":["subscription_id = \"<your-subscription-id>\"\nresource_group = \"<your-existing-resource-group>\"\nworkspace_name = \"<a-new-or-existing-workspace; it is unrelated to Databricks workspace>\"\nworkspace_region = \"<your-resource group-region>\""],"metadata":{},"outputs":[],"execution_count":5},{"cell_type":"code","source":["# import the Workspace class and check the azureml SDK version\n# exist_ok checks if workspace exists or not.\n\nfrom azureml.core import Workspace\n\nws = Workspace.create(name = workspace_name,\n subscription_id = subscription_id,\n resource_group = resource_group, \n location = workspace_region,\n exist_ok=True)\n\nws.get_details()"],"metadata":{},"outputs":[],"execution_count":6},{"cell_type":"code","source":["ws = Workspace(workspace_name = workspace_name,\n subscription_id = subscription_id,\n resource_group = resource_group)\n\n# persist the subscription id, resource group name, and workspace name in aml_config/config.json.\nws.write_config()"],"metadata":{},"outputs":[],"execution_count":7},{"cell_type":"code","source":["%sh\ncat /databricks/driver/aml_config/config.json"],"metadata":{},"outputs":[],"execution_count":8},{"cell_type":"code","source":["# import the Workspace class and check the azureml SDK version\nfrom azureml.core import Workspace\n\nws = Workspace.from_config()\nprint('Workspace name: ' + ws.name, \n 'Azure region: ' + ws.location, \n 'Subscription id: ' + ws.subscription_id, \n 'Resource group: ' + ws.resource_group, sep = '\\n')"],"metadata":{},"outputs":[],"execution_count":9},{"cell_type":"code","source":["dbutils.notebook.exit(\"success\")"],"metadata":{},"outputs":[],"execution_count":10},{"cell_type":"code","source":[""],"metadata":{},"outputs":[],"execution_count":11}],"metadata":{"name":"01.Installation_and_Configuration","notebookId":3874566296719377},"nbformat":4,"nbformat_minor":0}

1
databricks/02.Ingest_data.ipynb Normal file
View File

@@ -0,0 +1 @@
{"cells":[{"cell_type":"markdown","source":["Azure ML & Azure Databricks notebooks by Parashar Shah.\n\nCopyright (c) Microsoft Corporation. All rights reserved.\n\nLicensed under the MIT License."],"metadata":{}},{"cell_type":"markdown","source":["Please ensure you have run all previous notebooks in sequence before running this."],"metadata":{}},{"cell_type":"markdown","source":["#Data Ingestion"],"metadata":{}},{"cell_type":"code","source":["import os\nimport urllib"],"metadata":{},"outputs":[],"execution_count":4},{"cell_type":"code","source":["# Download AdultCensusIncome.csv from Azure CDN. This file has 32,561 rows.\nbasedataurl = \"https://amldockerdatasets.azureedge.net\"\ndatafile = \"AdultCensusIncome.csv\"\ndatafile_dbfs = os.path.join(\"/dbfs\", datafile)\n\nif os.path.isfile(datafile_dbfs):\n print(\"found {} at {}\".format(datafile, datafile_dbfs))\nelse:\n print(\"downloading {} to {}\".format(datafile, datafile_dbfs))\n urllib.request.urlretrieve(os.path.join(basedataurl, datafile), datafile_dbfs)"],"metadata":{},"outputs":[],"execution_count":5},{"cell_type":"code","source":["# Create a Spark dataframe out of the csv file.\ndata_all = sqlContext.read.format('csv').options(header='true', inferSchema='true', ignoreLeadingWhiteSpace='true', ignoreTrailingWhiteSpace='true').load(datafile)\nprint(\"({}, {})\".format(data_all.count(), len(data_all.columns)))\ndata_all.printSchema()"],"metadata":{},"outputs":[],"execution_count":6},{"cell_type":"code","source":["#renaming columns\ncolumns_new = [col.replace(\"-\", \"_\") for col in data_all.columns]\ndata_all = data_all.toDF(*columns_new)\ndata_all.printSchema()"],"metadata":{},"outputs":[],"execution_count":7},{"cell_type":"code","source":["display(data_all.limit(5))"],"metadata":{},"outputs":[],"execution_count":8},{"cell_type":"markdown","source":["#Data Preparation"],"metadata":{}},{"cell_type":"code","source":["# Choose feature columns and the label column.\nlabel = \"income\"\nxvals_all = set(data_all.columns) - {label}\n\n#dbutils.widgets.remove(\"xvars_multiselect\")\ndbutils.widgets.removeAll()\n\ndbutils.widgets.multiselect('xvars_multiselect', 'hours_per_week', xvals_all)\nxvars_multiselect = dbutils.widgets.get(\"xvars_multiselect\")\nxvars = xvars_multiselect.split(\",\")\n\nprint(\"label = {}\".format(label))\nprint(\"features = {}\".format(xvars))\n\ndata = data_all.select([*xvars, label])\n\n# Split data into train and test.\ntrain, test = data.randomSplit([0.75, 0.25], seed=123)\n\nprint(\"train ({}, {})\".format(train.count(), len(train.columns)))\nprint(\"test ({}, {})\".format(test.count(), len(test.columns)))"],"metadata":{},"outputs":[],"execution_count":10},{"cell_type":"markdown","source":["#Data Persistence"],"metadata":{}},{"cell_type":"code","source":["# Write the train and test data sets to intermediate storage\ntrain_data_path = \"AdultCensusIncomeTrain\"\ntest_data_path = \"AdultCensusIncomeTest\"\n\ntrain_data_path_dbfs = os.path.join(\"/dbfs\", \"AdultCensusIncomeTrain\")\ntest_data_path_dbfs = os.path.join(\"/dbfs\", \"AdultCensusIncomeTest\")\n\ntrain.write.mode('overwrite').parquet(train_data_path)\ntest.write.mode('overwrite').parquet(test_data_path)\nprint(\"train and test datasets saved to {} and {}\".format(train_data_path_dbfs, test_data_path_dbfs))"],"metadata":{},"outputs":[],"execution_count":12},{"cell_type":"code","source":["dbutils.notebook.exit(\"success\")"],"metadata":{},"outputs":[],"execution_count":13},{"cell_type":"code","source":[""],"metadata":{},"outputs":[],"execution_count":14}],"metadata":{"name":"02.Ingest_data","notebookId":3874566296719393},"nbformat":4,"nbformat_minor":0}

1
databricks/03a.Build_model.ipynb Normal file
View File

File diff suppressed because one or more lines are too long

1
databricks/03b.Build_model_runHistory.ipynb Normal file
View File

File diff suppressed because one or more lines are too long

1
databricks/04.Deploy_to_ACI.ipynb Normal file
View File

File diff suppressed because one or more lines are too long

1
databricks/05.Deploy_to_AKS_existingImage.ipynb Normal file
View File

@@ -0,0 +1 @@
{"cells":[{"cell_type":"markdown","source":["Azure ML & Azure Databricks notebooks by Parashar Shah.\n\nCopyright (c) Microsoft Corporation. All rights reserved.\n\nLicensed under the MIT License."],"metadata":{}},{"cell_type":"markdown","source":["Please ensure you have run all previous notebooks in sequence before running this. This notebook uses image from ACI notebook for deploying to AKS."],"metadata":{}},{"cell_type":"code","source":["from azureml.core import Workspace\nimport azureml.core\n\n# Check core SDK version number\nprint(\"SDK version:\", azureml.core.VERSION)\n\n#'''\nws = Workspace.from_config()\nprint('Workspace name: ' + ws.name, \n 'Azure region: ' + ws.location, \n 'Subscription id: ' + ws.subscription_id, \n 'Resource group: ' + ws.resource_group, sep = '\\n')\n#'''"],"metadata":{},"outputs":[],"execution_count":3},{"cell_type":"code","source":["# List images by ws\n\nfrom azureml.core.image import ContainerImage\nfor i in ContainerImage.list(workspace = ws):\n print('{}(v.{} [{}]) stored at {} with build log {}'.format(i.name, i.version, i.creation_state, i.image_location, i.image_build_log_uri))"],"metadata":{},"outputs":[],"execution_count":4},{"cell_type":"code","source":["from azureml.core.image import Image\nmyimage = Image(workspace=ws, id=\"aciws:25\")"],"metadata":{},"outputs":[],"execution_count":5},{"cell_type":"code","source":["#create AKS compute\n#it may take 20-25 minutes to create a new cluster\n\nfrom azureml.core.compute import AksCompute, ComputeTarget\n\n# Use the default configuration (can also provide parameters to customize)\nprov_config = AksCompute.provisioning_configuration()\n\naks_name = 'ps-aks-clus2' \n\n# Create the cluster\naks_target = ComputeTarget.create(workspace = ws, \n name = aks_name, \n provisioning_configuration = prov_config)\n\naks_target.wait_for_completion(show_output = True)\n\nprint(aks_target.provisioning_state)\nprint(aks_target.provisioning_errors)"],"metadata":{},"outputs":[],"execution_count":6},{"cell_type":"code","source":["from azureml.core.webservice import Webservice\nhelp( Webservice.deploy_from_image)"],"metadata":{},"outputs":[],"execution_count":7},{"cell_type":"code","source":["from azureml.core.webservice import Webservice, AksWebservice\nfrom azureml.core.image import ContainerImage\n\n#Set the web service configuration (using default here)\naks_config = AksWebservice.deploy_configuration()\n\n#unique service name\nservice_name ='ps-aks-service'\n\n# Webservice creation using single command, there is a variant to use image directly as well.\naks_service = Webservice.deploy_from_image(\n workspace=ws, \n name=service_name,\n deployment_config = aks_config,\n image = myimage,\n deployment_target = aks_target\n )\n\naks_service.wait_for_deployment(show_output=True)"],"metadata":{},"outputs":[],"execution_count":8},{"cell_type":"code","source":["#for using the Web HTTP API \nprint(aks_service.scoring_uri)\nprint(aks_service.get_keys())"],"metadata":{},"outputs":[],"execution_count":9},{"cell_type":"code","source":["import json\n\n#get the some sample data\ntest_data_path = \"AdultCensusIncomeTest\"\ntest = spark.read.parquet(test_data_path).limit(5)\n\ntest_json = json.dumps(test.toJSON().collect())\n\nprint(test_json)"],"metadata":{},"outputs":[],"execution_count":10},{"cell_type":"code","source":["#using data defined above predict if income is >50K (1) or <=50K (0)\naks_service.run(input_data=test_json)"],"metadata":{},"outputs":[],"execution_count":11},{"cell_type":"code","source":["#comment to not delete the web service\naks_service.delete()\n#image.delete()\n#model.delete()\n#aks_target.delete()"],"metadata":{},"outputs":[],"execution_count":12},{"cell_type":"code","source":[""],"metadata":{},"outputs":[],"execution_count":13}],"metadata":{"name":"04.DeploytoACI","notebookId":3874566296719318},"nbformat":4,"nbformat_minor":0}

BIN
databricks/Databricks_AMLSDK_github.dbc Normal file
View File

Binary file not shown.

29
databricks/readme.md Normal file
View File

@@ -0,0 +1,29 @@
# Azure Databricks - Azure Machine Learning SDK Sample Notebooks
**NOTE**: With the latest version of Azure Machine Learning SDK, there are some API changes due to which previous version of notebooks will not work.
Please remove the previous SDK version and install the latest SDK by installing **azureml-sdk[databricks]** as a PyPi library in Azure Databricks workspace.
**NOTE**: Please create your Azure Databricks cluster as v4.x (high concurrency preferred) with **Python 3** (dropdown).
**NOTE**: Some packages like psutil upgrade libs that can cause a conflict, please install such packages by freezing lib version. Eg. "psutil **cryptography==1.5 pyopenssl==16.0.0 ipython=2.2.0**" to avoid install error. This issue is related to Databricks and not related to AML SDK.
**NOTE**: You should at least have contributor access to your Azure subcription to run some of the notebooks.
The iPython Notebooks have to be run sequentially after making changes based on your subscription. The corresponding DBC archive contains all the notebooks and can be imported into your Databricks workspace. You can the run notebooks after importing .dbc instead of downloading individually.
This set of notebooks are related to Income prediction experiment based on this [dataset](https://archive.ics.uci.edu/ml/datasets/adult) and demonstrate how to data prep, train and operationalize a Spark ML model with Azure ML Python SDK from within Azure Databricks. For details on SDK concepts, please refer to [notebooks](https://github.com/Azure/MachineLearningNotebooks)
(Recommended) [Azure Databricks AML SDK notebooks](Databricks_AMLSDK_github.dbc) A single DBC package to import all notebooks in your Azure Databricks workspace.
01. [Installation and Configuration](01.Installation_and_Configuration.ipynb): Install the Azure ML Python SDK and Initialize an Azure ML Workspace and save the Workspace configuration file.
02. [Ingest data](02.Ingest_data.ipynb): Download the Adult Census Income dataset and split it into train and test sets.
03. [Build model](03a.Build_model.ipynb): Train a binary classification model in Azure Databricks with a Spark ML Pipeline.
04. [Build model with Run History](03b.Build_model_runHistory.ipynb): Train model and also capture run history (tracking) with Azure ML Python SDK.
05. [Deploy to ACI](04.Deploy_to_ACI.ipynb): Deploy model to Azure Container Instance (ACI) with Azure ML Python SDK.
06. [Deploy to AKS](04.Deploy_to_AKS_existingImage.ipynb): Deploy model to Azure Kubernetis Service (AKS) with Azure ML Python SDK from an existing Image with model, conda and score file.
Copyright (c) Microsoft Corporation. All rights reserved.
All notebooks in this folder are licensed under the MIT License.
Apache®, Apache Spark, and Spark® are either registered trademarks or trademarks of the Apache Software Foundation in the United States and/or other countries.

2985
dataprep/getting-started.ipynb Normal file
View File

File diff suppressed because it is too large Load Diff

16
getting-started/README.md
View File

@@ -1,16 +0,0 @@
## Examples to get started with Azure Machine Learning service
Learn how to use Azure Machine Learning services for experimentation and model management.
As a pre-requisite, run the [configuration Notebook](../configuration.ipynb) notebook first to set up your Azure ML Workspace. Then, run the notebooks in following recommended order.
* [train-within-notebook](train-within-notebook/train-within-notebook.ipynb): Train a model hile tracking run history, and learn how to deploy the model as web service to Azure Container Instance.
* [train-on-local](train-on-local/train-on-local.ipynb): Learn how to submit a run and use Azure ML managed run configuration.
* [train-on-aci](train-on-aci/train-on-aci.ipynb): Submit a remote run on serverless Docker-based compute.
* [train-on-remote-vm](train-on-remote-vm/train-on-remote-vm.ipynb): Use Data Science Virtual Machine as a target for remote runs.
* [logging-api](logging-api/logging-api.ipynb): Learn about the details of logging metrics to run history.
* [register-model-create-image-deploy-service](register-model-create-image-deploy-service/register-model-create-image-deploy-service.ipynb): Learn about the details of model management.
* [production-deploy-to-aks](production-deploy-to-aks/production-deploy-to-aks.ipynb) Deploy a model to production at scale on Azure Kubernetes Service.
* [enable-data-collection-for-models-in-aks](enable-data-collection-for-models-in-aks/enable-data-collection-for-models-in-aks.ipynb) Learn about data collection APIs for deployed model.
* [enable-app-insights-in-production-service](enable-app-insights-in-production-serviceenable-app-insights-in-production-service.ipynb) Learn how to use App Insights with production web service.

415
getting-started/enable-app-insights-in-production-service/enable-app-insights-in-production-service.ipynb
View File

@@ -1,415 +0,0 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Enabling App Insights for Services in Production\n",
"With this notebook, you can learn how to enable App Insights for standard service monitoring, plus, we provide examples for doing custom logging within a scoring files in a model. \n",
"\n",
"\n",
"## What does Application Insights monitor?\n",
"It monitors request rates, response times, failure rates, etc. For more information visit [App Insights docs.](https://docs.microsoft.com/en-us/azure/application-insights/app-insights-overview)\n",
"\n",
"\n",
"## What is different compared to standard production deployment process?\n",
"If you want to enable generic App Insights for a service run:\n",
"```python\n",
"aks_service= Webservice(ws, \"aks-w-dc2\")\n",
"aks_service.update(enable_app_insights=True)```\n",
"Where \"aks-w-dc2\" is your service name. You can also do this from the Azure Portal under your Workspace--> deployments--> Select deployment--> Edit--> Advanced Settings--> Select \"Enable AppInsights diagnostics\"\n",
"\n",
"If you want to log custom traces, you will follow the standard deplyment process for AKS and you will:\n",
"1. Update scoring file.\n",
"2. Update aks configuration.\n",
"3. Build new image and deploy it. "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 1. Import your dependencies"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Workspace, Run\n",
"from azureml.core.compute import AksCompute, ComputeTarget\n",
"from azureml.core.webservice import Webservice, AksWebservice\n",
"from azureml.core.image import Image\n",
"from azureml.core.model import Model\n",
"\n",
"import azureml.core\n",
"print(azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 2. Set up your configuration and create a workspace\n",
"Follow Notebook 00 instructions to do this.\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"print(ws.name, ws.resource_group, ws.location, ws.subscription_id, sep = '\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 3. Register Model\n",
"Register an existing trained model, add descirption and tags."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#Register the model\n",
"from azureml.core.model import Model\n",
"model = Model.register(model_path = \"sklearn_regression_model.pkl\", # this points to a local file\n",
" model_name = \"sklearn_regression_model.pkl\", # this is the name the model is registered as\n",
" tags = {'area': \"diabetes\", 'type': \"regression\"},\n",
" description = \"Ridge regression model to predict diabetes\",\n",
" workspace = ws)\n",
"\n",
"print(model.name, model.description, model.version)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 4. *Update your scoring file with custom print statements*\n",
"Here is an example:\n",
"### a. In your init function add:\n",
"```python\n",
"print (\"model initialized\" + time.strftime(\"%H:%M:%S\"))```\n",
"\n",
"### b. In your run function add:\n",
"```python\n",
"print (\"saving input data\" + time.strftime(\"%H:%M:%S\"))\n",
"print (\"saving prediction data\" + time.strftime(\"%H:%M:%S\"))```"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile score.py\n",
"import pickle\n",
"import json\n",
"import numpy \n",
"from sklearn.externals import joblib\n",
"from sklearn.linear_model import Ridge\n",
"from azureml.core.model import Model\n",
"from azureml.monitoring import ModelDataCollector\n",
"import time\n",
"\n",
"def init():\n",
" global model\n",
" #Print statement for appinsights custom traces:\n",
" print (\"model initialized\" + time.strftime(\"%H:%M:%S\"))\n",
" \n",
" # note here \"sklearn_regression_model.pkl\" is the name of the model registered under the workspace\n",
" # this call should return the path to the model.pkl file on the local disk.\n",
" model_path = Model.get_model_path(model_name = 'sklearn_regression_model.pkl')\n",
" \n",
" # deserialize the model file back into a sklearn model\n",
" model = joblib.load(model_path)\n",
" \n",
" global inputs_dc, prediction_dc\n",
" \n",
" # this setup will help us save our inputs under the \"inputs\" path in our Azure Blob\n",
" inputs_dc = ModelDataCollector(model_name=\"sklearn_regression_model\", identifier=\"inputs\", feature_names=[\"feat1\", \"feat2\"]) \n",
" \n",
" # this setup will help us save our ipredictions under the \"predictions\" path in our Azure Blob\n",
" prediction_dc = ModelDataCollector(\"sklearn_regression_model\", identifier=\"predictions\", feature_names=[\"prediction1\", \"prediction2\"]) \n",
" \n",
"# note you can pass in multiple rows for scoring\n",
"def run(raw_data):\n",
" global inputs_dc, prediction_dc\n",
" try:\n",
" data = json.loads(raw_data)['data']\n",
" data = numpy.array(data)\n",
" result = model.predict(data)\n",
" \n",
" #Print statement for appinsights custom traces:\n",
" print (\"saving input data\" + time.strftime(\"%H:%M:%S\"))\n",
" \n",
" #this call is saving our input data into our blob\n",
" inputs_dc.collect(data) \n",
" #this call is saving our prediction data into our blob\n",
" prediction_dc.collect(result)\n",
" \n",
" #Print statement for appinsights custom traces:\n",
" print (\"saving prediction data\" + time.strftime(\"%H:%M:%S\"))\n",
" # you can return any data type as long as it is JSON-serializable\n",
" return result.tolist()\n",
" except Exception as e:\n",
" error = str(e)\n",
" print (error + time.strftime(\"%H:%M:%S\"))\n",
" return error"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 5. *Create myenv.yml file*"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.conda_dependencies import CondaDependencies \n",
"\n",
"myenv = CondaDependencies.create(conda_packages=['numpy','scikit-learn'])\n",
"\n",
"with open(\"myenv.yml\",\"w\") as f:\n",
" f.write(myenv.serialize_to_string())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 6. Create your new Image"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.image import ContainerImage\n",
"\n",
"image_config = ContainerImage.image_configuration(execution_script = \"score.py\",\n",
" runtime = \"python\",\n",
" conda_file = \"myenv.yml\",\n",
" description = \"Image with ridge regression model\",\n",
" tags = {'area': \"diabetes\", 'type': \"regression\"}\n",
" )\n",
"\n",
"image = ContainerImage.create(name = \"myimage1\",\n",
" # this is the model object\n",
" models = [model],\n",
" image_config = image_config,\n",
" workspace = ws)\n",
"\n",
"image.wait_for_creation(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 7. Deploy to AKS service"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create AKS compute if you haven't done so (Notebook 11)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Use the default configuration (can also provide parameters to customize)\n",
"prov_config = AksCompute.provisioning_configuration()\n",
"\n",
"aks_name = 'my-aks-test2' \n",
"# Create the cluster\n",
"aks_target = ComputeTarget.create(workspace = ws, \n",
" name = aks_name, \n",
" provisioning_configuration = prov_config)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%time\n",
"aks_target.wait_for_completion(show_output = True)\n",
"print(aks_target.provisioning_state)\n",
"print(aks_target.provisioning_errors)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"If you already have a cluster you can attach the service to it:"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"```python \n",
"%%time\n",
"resource_id = '/subscriptions/<subscriptionid>/resourcegroups/<resourcegroupname>/providers/Microsoft.ContainerService/managedClusters/<aksservername>'\n",
"create_name= 'myaks4'\n",
"attach_config = AksCompute.attach_configuration(resource_id=resource_id)\n",
"aks_target = ComputeTarget.attach(workspace = ws, \n",
" name = create_name, \n",
" attach_configuration=attach_config)\n",
"## Wait for the operation to complete\n",
"aks_target.wait_for_provisioning(True)```"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### a. *Activate App Insights through updating AKS Webservice configuration*\n",
"In order to enable App Insights in your service you will need to update your AKS configuration file:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#Set the web service configuration\n",
"aks_config = AksWebservice.deploy_configuration(enable_app_insights=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### b. Deploy your service"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%time\n",
"aks_service_name ='aks-w-dc3'\n",
"\n",
"aks_service = Webservice.deploy_from_image(workspace = ws, \n",
" name = aks_service_name,\n",
" image = image,\n",
" deployment_config = aks_config,\n",
" deployment_target = aks_target\n",
" )\n",
"aks_service.wait_for_deployment(show_output = True)\n",
"print(aks_service.state)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 8. Test your service "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%time\n",
"import json\n",
"\n",
"test_sample = json.dumps({'data': [\n",
" [1,28,13,45,54,6,57,8,8,10], \n",
" [101,9,8,37,6,45,4,3,2,41]\n",
"]})\n",
"test_sample = bytes(test_sample,encoding='utf8')\n",
"\n",
"prediction = aks_service.run(input_data=test_sample)\n",
"print(prediction)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 9. See your service telemetry in App Insights\n",
"1. Go to the [Azure Portal](https://portal.azure.com/)\n",
"2. All resources--> Select the subscription/resource group where you created your Workspace--> Select the App Insights type\n",
"3. Click on the AppInsights resource. You'll see a highlevel dashboard with information on Requests, Server response time and availability.\n",
"4. Click on the top banner \"Analytics\"\n",
"5. In the \"Schema\" section select \"traces\" and run your query.\n",
"6. Voila! All your custom traces should be there."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Disable App Insights"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"aks_service.update(enable_app_insights=False)"
]
}
],
"metadata": {
"authors": [
{
"name": "marthalc"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

328
getting-started/logging-api/logging-api.ipynb
View File

@@ -1,328 +0,0 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# 06. Logging APIs\n",
"This notebook showcase various ways to use the Azure Machine Learning service run logging APIs, and view the results in the Azure portal."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"Make sure you go through the [00. Installation and Configuration](../../00.configuration.ipynb) Notebook first if you haven't. Also make sure you have tqdm and matplotlib installed in the current kernel.\n",
"\n",
"```\n",
"(myenv) $ conda install -y tqdm matplotlib\n",
"```"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Validate Azure ML SDK installation and get version number for debugging purposes"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"install"
]
},
"outputs": [],
"source": [
"from azureml.core import Experiment, Run, Workspace\n",
"import azureml.core\n",
"import numpy as np\n",
"\n",
"# Check core SDK version number\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize Workspace\n",
"\n",
"Initialize a workspace object from persisted configuration."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"create workspace"
]
},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"print('Workspace name: ' + ws.name, \n",
" 'Azure region: ' + ws.location, \n",
" 'Subscription id: ' + ws.subscription_id, \n",
" 'Resource group: ' + ws.resource_group, sep='\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Set experiment\n",
"Create a new experiment (or get the one with such name)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"exp = Experiment(workspace=ws, name='logging-api-test')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Log metrics\n",
"We will start a run, and use the various logging APIs to record different types of metrics during the run."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from tqdm import tqdm\n",
"\n",
"# start logging for the run\n",
"run = exp.start_logging()\n",
"\n",
"# log a string value\n",
"run.log(name='Name', value='Logging API run')\n",
"\n",
"# log a numerical value\n",
"run.log(name='Magic Number', value=42)\n",
"\n",
"# Log a list of values. Note this will generate a single-variable line chart.\n",
"run.log_list(name='Fibonacci', value=[0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89])\n",
"\n",
"# create a dictionary to hold a table of values\n",
"sines = {}\n",
"sines['angle'] = []\n",
"sines['sine'] = []\n",
"\n",
"for i in tqdm(range(-10, 10)):\n",
" # log a metric value repeatedly, this will generate a single-variable line chart.\n",
" run.log(name='Sigmoid', value=1 / (1 + np.exp(-i)))\n",
" angle = i / 2.0\n",
" \n",
" # log a 2 (or more) values as a metric repeatedly. This will generate a 2-variable line chart if you have 2 numerical columns.\n",
" run.log_row(name='Cosine Wave', angle=angle, cos=np.cos(angle))\n",
" \n",
" sines['angle'].append(angle)\n",
" sines['sine'].append(np.sin(angle))\n",
"\n",
"# log a dictionary as a table, this will generate a 2-variable chart if you have 2 numerical columns\n",
"run.log_table(name='Sine Wave', value=sines)\n",
"\n",
"run.complete()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Even after the run is marked completed, you can still log things."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Log an image\n",
"This is how to log a _matplotlib_ pyplot object."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%matplotlib inline\n",
"import matplotlib.pyplot as plt\n",
"angle = np.linspace(-3, 3, 50)\n",
"plt.plot(angle, np.tanh(angle), label='tanh')\n",
"plt.legend(fontsize=12)\n",
"plt.title('Hyperbolic Tangent', fontsize=16)\n",
"plt.grid(True)\n",
"\n",
"run.log_image(name='Hyperbolic Tangent', plot=plt)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Upload a file"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can also upload an abitrary file. First, let's create a dummy file locally."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile myfile.txt\n",
"\n",
"This is a dummy file."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now let's upload this file into the run record as a run artifact, and display the properties after the upload."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"props = run.upload_file(name='myfile_in_the_cloud.txt', path_or_stream='./myfile.txt')\n",
"props.serialize()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Examine the run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now let's take a look at the run detail page in Azure portal. Make sure you checkout the various charts and plots generated/uploaded."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can get all the metrics in that run back."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.get_metrics()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can also see the files uploaded for this run."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.get_file_names()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can also download all the files locally."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"os.makedirs('files', exist_ok=True)\n",
"\n",
"for f in run.get_file_names():\n",
" dest = os.path.join('files', f.split('/')[-1])\n",
" print('Downloading file {} to {}...'.format(f, dest))\n",
" run.download_file(f, dest) "
]
}
],
"metadata": {
"authors": [
{
"name": "haining"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

343
getting-started/production-deploy-to-aks/production-deploy-to-aks.ipynb
View File

@@ -1,343 +0,0 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Deploying a web service to Azure Kubernetes Service (AKS)\n",
"This notebook shows the steps for deploying a service: registering a model, creating an image, provisioning a cluster (one time action), and deploying a service to it. \n",
"We then test and delete the service, image and model."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Workspace\n",
"from azureml.core.compute import AksCompute, ComputeTarget\n",
"from azureml.core.webservice import Webservice, AksWebservice\n",
"from azureml.core.image import Image\n",
"from azureml.core.model import Model"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import azureml.core\n",
"print(azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Get workspace\n",
"Load existing workspace from the config file info."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.workspace import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print(ws.name, ws.resource_group, ws.location, ws.subscription_id, sep = '\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Register the model\n",
"Register an existing trained model, add descirption and tags."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#Register the model\n",
"from azureml.core.model import Model\n",
"model = Model.register(model_path = \"sklearn_regression_model.pkl\", # this points to a local file\n",
" model_name = \"sklearn_regression_model.pkl\", # this is the name the model is registered as\n",
" tags = {'area': \"diabetes\", 'type': \"regression\"},\n",
" description = \"Ridge regression model to predict diabetes\",\n",
" workspace = ws)\n",
"\n",
"print(model.name, model.description, model.version)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Create an image\n",
"Create an image using the registered model the script that will load and run the model."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile score.py\n",
"import pickle\n",
"import json\n",
"import numpy\n",
"from sklearn.externals import joblib\n",
"from sklearn.linear_model import Ridge\n",
"from azureml.core.model import Model\n",
"\n",
"def init():\n",
" global model\n",
" # note here \"sklearn_regression_model.pkl\" is the name of the model registered under\n",
" # this is a different behavior than before when the code is run locally, even though the code is the same.\n",
" model_path = Model.get_model_path('sklearn_regression_model.pkl')\n",
" # deserialize the model file back into a sklearn model\n",
" model = joblib.load(model_path)\n",
"\n",
"# note you can pass in multiple rows for scoring\n",
"def run(raw_data):\n",
" try:\n",
" data = json.loads(raw_data)['data']\n",
" data = numpy.array(data)\n",
" result = model.predict(data)\n",
" # you can return any data type as long as it is JSON-serializable\n",
" return result.tolist()\n",
" except Exception as e:\n",
" error = str(e)\n",
" return error"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.conda_dependencies import CondaDependencies \n",
"\n",
"myenv = CondaDependencies.create(conda_packages=['numpy','scikit-learn'])\n",
"\n",
"with open(\"myenv.yml\",\"w\") as f:\n",
" f.write(myenv.serialize_to_string())"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.image import ContainerImage\n",
"\n",
"image_config = ContainerImage.image_configuration(execution_script = \"score.py\",\n",
" runtime = \"python\",\n",
" conda_file = \"myenv.yml\",\n",
" description = \"Image with ridge regression model\",\n",
" tags = {'area': \"diabetes\", 'type': \"regression\"}\n",
" )\n",
"\n",
"image = ContainerImage.create(name = \"myimage1\",\n",
" # this is the model object\n",
" models = [model],\n",
" image_config = image_config,\n",
" workspace = ws)\n",
"\n",
"image.wait_for_creation(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Provision the AKS Cluster\n",
"This is a one time setup. You can reuse this cluster for multiple deployments after it has been created. If you delete the cluster or the resource group that contains it, then you would have to recreate it."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Use the default configuration (can also provide parameters to customize)\n",
"prov_config = AksCompute.provisioning_configuration()\n",
"\n",
"aks_name = 'my-aks-9' \n",
"# Create the cluster\n",
"aks_target = ComputeTarget.create(workspace = ws, \n",
" name = aks_name, \n",
" provisioning_configuration = prov_config)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%time\n",
"aks_target.wait_for_completion(show_output = True)\n",
"print(aks_target.provisioning_state)\n",
"print(aks_target.provisioning_errors)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Optional step: Attach existing AKS cluster\n",
"\n",
"If you have existing AKS cluster in your Azure subscription, you can attach it to the Workspace."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"'''\n",
"# Use the default configuration (can also provide parameters to customize)\n",
"resource_id = '/subscriptions/92c76a2f-0e1c-4216-b65e-abf7a3f34c1e/resourcegroups/raymondsdk0604/providers/Microsoft.ContainerService/managedClusters/my-aks-0605d37425356b7d01'\n",
"\n",
"create_name='my-existing-aks' \n",
"# Create the cluster\n",
"attach_config = AksCompute.attach_configuration(resource_id=resource_id)\n",
"aks_target = ComputeTarget.attach(workspace=ws, name=create_name, attach_configuration=attach_config)\n",
"# Wait for the operation to complete\n",
"aks_target.wait_for_completion(True)\n",
"'''"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Deploy web service to AKS"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#Set the web service configuration (using default here)\n",
"aks_config = AksWebservice.deploy_configuration()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%time\n",
"aks_service_name ='aks-service-1'\n",
"\n",
"aks_service = Webservice.deploy_from_image(workspace = ws, \n",
" name = aks_service_name,\n",
" image = image,\n",
" deployment_config = aks_config,\n",
" deployment_target = aks_target)\n",
"aks_service.wait_for_deployment(show_output = True)\n",
"print(aks_service.state)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Test the web service\n",
"We test the web sevice by passing data."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%time\n",
"import json\n",
"\n",
"test_sample = json.dumps({'data': [\n",
" [1,2,3,4,5,6,7,8,9,10], \n",
" [10,9,8,7,6,5,4,3,2,1]\n",
"]})\n",
"test_sample = bytes(test_sample,encoding = 'utf8')\n",
"\n",
"prediction = aks_service.run(input_data = test_sample)\n",
"print(prediction)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Clean up\n",
"Delete the service, image and model."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%time\n",
"aks_service.delete()\n",
"image.delete()\n",
"model.delete()"
]
}
],
"metadata": {
"authors": [
{
"name": "raymondl"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

420
getting-started/register-model-create-image-deploy-service/register-model-create-image-deploy-service.ipynb
View File

@@ -1,420 +0,0 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 10. Register Model, Create Image and Deploy Service\n",
"\n",
"This example shows how to deploy a web service in step-by-step fashion:\n",
"\n",
" 1. Register model\n",
" 2. Query versions of models and select one to deploy\n",
" 3. Create Docker image\n",
" 4. Query versions of images\n",
" 5. Deploy the image as web service\n",
" \n",
"**IMPORTANT**:\n",
" * This notebook requires you to first complete \"01.SDK-101-Train-and-Deploy-to-ACI.ipynb\" Notebook\n",
" \n",
"The 101 Notebook taught you how to deploy a web service directly from model in one step. This Notebook shows a more advanced approach that gives you more control over model versions and Docker image versions. "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"Make sure you go through the [00. Installation and Configuration](00.configuration.ipynb) Notebook first if you haven't."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize Workspace\n",
"\n",
"Initialize a workspace object from persisted configuration."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"create workspace"
]
},
"outputs": [],
"source": [
"from azureml.core import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print(ws.name, ws.resource_group, ws.location, ws.subscription_id, sep = '\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Register Model"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can add tags and descriptions to your models. Note you need to have a `sklearn_linreg_model.pkl` file in the current directory. This file is generated by the 01 notebook. The below call registers that file as a model with the same name `sklearn_linreg_model.pkl` in the workspace.\n",
"\n",
"Using tags, you can track useful information such as the name and version of the machine learning library used to train the model. Note that tags must be alphanumeric."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"register model from file"
]
},
"outputs": [],
"source": [
"from azureml.core.model import Model\n",
"import sklearn\n",
"\n",
"library_version = \"sklearn\"+sklearn.__version__.replace(\".\",\"x\")\n",
"\n",
"model = Model.register(model_path = \"sklearn_regression_model.pkl\",\n",
" model_name = \"sklearn_regression_model.pkl\",\n",
" tags = {'area': \"diabetes\", 'type': \"regression\", 'version': library_version},\n",
" description = \"Ridge regression model to predict diabetes\",\n",
" workspace = ws)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can explore the registered models within your workspace and query by tag. Models are versioned. If you call the register_model command many times with same model name, you will get multiple versions of the model with increasing version numbers."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"register model from file"
]
},
"outputs": [],
"source": [
"regression_models = Model.list(workspace=ws, tags=['area'])\n",
"for m in regression_models:\n",
" print(\"Name:\", m.name,\"\\tVersion:\", m.version, \"\\tDescription:\", m.description, m.tags)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can pick a specific model to deploy"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(model.name, model.description, model.version, sep = '\\t')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create Docker Image"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Show `score.py`. Note that the `sklearn_regression_model.pkl` in the `get_model_path` call is referring to a model named `sklearn_linreg_model.pkl` registered under the workspace. It is NOT referenceing the local file."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile score.py\n",
"import pickle\n",
"import json\n",
"import numpy\n",
"from sklearn.externals import joblib\n",
"from sklearn.linear_model import Ridge\n",
"from azureml.core.model import Model\n",
"\n",
"def init():\n",
" global model\n",
" # note here \"sklearn_regression_model.pkl\" is the name of the model registered under\n",
" # this is a different behavior than before when the code is run locally, even though the code is the same.\n",
" model_path = Model.get_model_path('sklearn_regression_model.pkl')\n",
" # deserialize the model file back into a sklearn model\n",
" model = joblib.load(model_path)\n",
"\n",
"# note you can pass in multiple rows for scoring\n",
"def run(raw_data):\n",
" try:\n",
" data = json.loads(raw_data)['data']\n",
" data = numpy.array(data)\n",
" result = model.predict(data)\n",
" # you can return any datatype as long as it is JSON-serializable\n",
" return result.tolist()\n",
" except Exception as e:\n",
" error = str(e)\n",
" return error"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.conda_dependencies import CondaDependencies \n",
"\n",
"myenv = CondaDependencies.create(conda_packages=['numpy','scikit-learn'])\n",
"\n",
"with open(\"myenv.yml\",\"w\") as f:\n",
" f.write(myenv.serialize_to_string())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Note that following command can take few minutes. \n",
"\n",
"You can add tags and descriptions to images. Also, an image can contain multiple models."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"create image"
]
},
"outputs": [],
"source": [
"from azureml.core.image import Image, ContainerImage\n",
"\n",
"image_config = ContainerImage.image_configuration(runtime= \"python\",\n",
" execution_script=\"score.py\",\n",
" conda_file=\"myenv.yml\",\n",
" tags = {'area': \"diabetes\", 'type': \"regression\"},\n",
" description = \"Image with ridge regression model\")\n",
"\n",
"image = Image.create(name = \"myimage1\",\n",
" # this is the model object \n",
" models = [model],\n",
" image_config = image_config, \n",
" workspace = ws)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"create image"
]
},
"outputs": [],
"source": [
"image.wait_for_creation(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"List images by tag and find out the detailed build log for debugging."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"create image"
]
},
"outputs": [],
"source": [
"for i in Image.list(workspace = ws,tags = [\"area\"]):\n",
" print('{}(v.{} [{}]) stored at {} with build log {}'.format(i.name, i.version, i.creation_state, i.image_location, i.image_build_log_uri))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Deploy image as web service on Azure Container Instance\n",
"\n",
"Note that the service creation can take few minutes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"from azureml.core.webservice import AciWebservice\n",
"\n",
"aciconfig = AciWebservice.deploy_configuration(cpu_cores = 1, \n",
" memory_gb = 1, \n",
" tags = {'area': \"diabetes\", 'type': \"regression\"}, \n",
" description = 'Predict diabetes using regression model')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"from azureml.core.webservice import Webservice\n",
"\n",
"aci_service_name = 'my-aci-service-2'\n",
"print(aci_service_name)\n",
"aci_service = Webservice.deploy_from_image(deployment_config = aciconfig,\n",
" image = image,\n",
" name = aci_service_name,\n",
" workspace = ws)\n",
"aci_service.wait_for_deployment(True)\n",
"print(aci_service.state)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test web service"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Call the web service with some dummy input data to get a prediction."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"import json\n",
"\n",
"test_sample = json.dumps({'data': [\n",
" [1,2,3,4,5,6,7,8,9,10], \n",
" [10,9,8,7,6,5,4,3,2,1]\n",
"]})\n",
"test_sample = bytes(test_sample,encoding = 'utf8')\n",
"\n",
"prediction = aci_service.run(input_data=test_sample)\n",
"print(prediction)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Delete ACI to clean up"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"aci_service.delete()"
]
}
],
"metadata": {
"authors": [
{
"name": "raymondl"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

BIN
getting-started/register-model-create-image-deploy-service/sklearn_regression_model.pkl
View File

Binary file not shown.

289
getting-started/train-on-aci/train-on-aci.ipynb
View File

@@ -1,289 +0,0 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# 03. Train on Azure Container Instance\n",
"\n",
"* Create Workspace\n",
"* Create `train.py` in the project folder.\n",
"* Configure an ACI (Azure Container Instance) run\n",
"* Execute in ACI"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"Make sure you go through the [00. Installation and Configuration](00.configuration.ipynb) Notebook first if you haven't."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize Workspace\n",
"\n",
"Initialize a workspace object from persisted configuration"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"create workspace"
]
},
"outputs": [],
"source": [
"from azureml.core import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print(ws.name, ws.resource_group, ws.location, ws.subscription_id, sep = '\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create An Experiment\n",
"\n",
"**Experiment** is a logical container in an Azure ML Workspace. It hosts run records which can include run metrics and output artifacts from your experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Experiment\n",
"experiment_name = 'train-on-aci'\n",
"experiment = Experiment(workspace = ws, name = experiment_name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Remote execution on ACI\n",
"\n",
"The training script `train.py` is already created for you. Let's have a look."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"with open('./train.py', 'r') as f:\n",
" print(f.read())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure for using ACI\n",
"Linux-based ACI is available in `West US`, `East US`, `West Europe`, `North Europe`, `West US 2`, `Southeast Asia`, `Australia East`, `East US 2`, and `Central US` regions. See details [here](https://docs.microsoft.com/en-us/azure/container-instances/container-instances-quotas#region-availability)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"configure run"
]
},
"outputs": [],
"source": [
"from azureml.core.runconfig import RunConfiguration\n",
"from azureml.core.conda_dependencies import CondaDependencies\n",
"\n",
"# create a new runconfig object\n",
"run_config = RunConfiguration()\n",
"\n",
"# signal that you want to use ACI to execute script.\n",
"run_config.target = \"containerinstance\"\n",
"\n",
"# ACI container group is only supported in certain regions, which can be different than the region the Workspace is in.\n",
"run_config.container_instance.region = 'eastus2'\n",
"\n",
"# set the ACI CPU and Memory \n",
"run_config.container_instance.cpu_cores = 1\n",
"run_config.container_instance.memory_gb = 2\n",
"\n",
"# enable Docker \n",
"run_config.environment.docker.enabled = True\n",
"\n",
"# set Docker base image to the default CPU-based image\n",
"run_config.environment.docker.base_image = azureml.core.runconfig.DEFAULT_CPU_IMAGE\n",
"\n",
"# use conda_dependencies.yml to create a conda environment in the Docker image for execution\n",
"run_config.environment.python.user_managed_dependencies = False\n",
"\n",
"# auto-prepare the Docker image when used for execution (if it is not already prepared)\n",
"run_config.auto_prepare_environment = True\n",
"\n",
"# specify CondaDependencies obj\n",
"run_config.environment.python.conda_dependencies = CondaDependencies.create(conda_packages=['scikit-learn'])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Submit the Experiment\n",
"Finally, run the training job on the ACI"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"remote run",
"aci"
]
},
"outputs": [],
"source": [
"%%time \n",
"from azureml.core.script_run_config import ScriptRunConfig\n",
"\n",
"script_run_config = ScriptRunConfig(source_directory='./',\n",
" script='train.py',\n",
" run_config=run_config)\n",
"\n",
"run = experiment.submit(script_run_config)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"query history"
]
},
"outputs": [],
"source": [
"# Show run details\n",
"run"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"remote run",
"aci"
]
},
"outputs": [],
"source": [
"%%time\n",
"# Shows output of the run on stdout.\n",
"run.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"get metrics"
]
},
"outputs": [],
"source": [
"# get all metris logged in the run\n",
"run.get_metrics()\n",
"metrics = run.get_metrics()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import numpy as np\n",
"print('When alpha is {1:0.2f}, we have min MSE {0:0.2f}.'.format(\n",
" min(metrics['mse']), \n",
" metrics['alpha'][np.argmin(metrics['mse'])]\n",
"))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# show all the files stored within the run record\n",
"run.get_file_names()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now you can take a model produced here, register it and then deploy as a web service."
]
}
],
"metadata": {
"authors": [
{
"name": "roastala"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

477
getting-started/train-on-local/train-on-local.ipynb
View File

@@ -1,477 +0,0 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# 02. Train locally\n",
"* Create or load workspace.\n",
"* Create scripts locally.\n",
"* Create `train.py` in a folder, along with a `my.lib` file.\n",
"* Configure & execute a local run in a user-managed Python environment.\n",
"* Configure & execute a local run in a system-managed Python environment.\n",
"* Configure & execute a local run in a Docker environment.\n",
"* Query run metrics to find the best model\n",
"* Register model for operationalization."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"Make sure you go through the [00. Installation and Configuration](00.configuration.ipynb) Notebook first if you haven't."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize Workspace\n",
"\n",
"Initialize a workspace object from persisted configuration."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.workspace import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print(ws.name, ws.resource_group, ws.location, ws.subscription_id, sep='\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create An Experiment\n",
"**Experiment** is a logical container in an Azure ML Workspace. It hosts run records which can include run metrics and output artifacts from your experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Experiment\n",
"experiment_name = 'train-on-local'\n",
"exp = Experiment(workspace=ws, name=experiment_name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## View `train.py`\n",
"\n",
"`train.py` is already created for you."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"with open('./train.py', 'r') as f:\n",
" print(f.read())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Note `train.py` also references a `mylib.py` file."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"with open('./mylib.py', 'r') as f:\n",
" print(f.read())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure & Run\n",
"### User-managed environment\n",
"Below, we use a user-managed run, which means you are responsible to ensure all the necessary packages are available in the Python environment you choose to run the script."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.runconfig import RunConfiguration\n",
"\n",
"# Editing a run configuration property on-fly.\n",
"run_config_user_managed = RunConfiguration()\n",
"\n",
"run_config_user_managed.environment.python.user_managed_dependencies = True\n",
"\n",
"# You can choose a specific Python environment by pointing to a Python path \n",
"#run_config.environment.python.interpreter_path = '/home/johndoe/miniconda3/envs/sdk2/bin/python'"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Submit script to run in the user-managed environment\n",
"Note whole script folder is submitted for execution, including the `mylib.py` file."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import ScriptRunConfig\n",
"\n",
"src = ScriptRunConfig(source_directory='./', script='train.py', run_config=run_config_user_managed)\n",
"run = exp.submit(src)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Get run history details"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Block to wait till run finishes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### System-managed environment\n",
"You can also ask the system to build a new conda environment and execute your scripts in it. The environment is built once and will be reused in subsequent executions as long as the conda dependencies remain unchanged. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.runconfig import RunConfiguration\n",
"from azureml.core.conda_dependencies import CondaDependencies\n",
"\n",
"run_config_system_managed = RunConfiguration()\n",
"\n",
"run_config_system_managed.environment.python.user_managed_dependencies = False\n",
"run_config_system_managed.auto_prepare_environment = True\n",
"\n",
"# Specify conda dependencies with scikit-learn\n",
"cd = CondaDependencies.create(conda_packages=['scikit-learn'])\n",
"run_config_system_managed.environment.python.conda_dependencies = cd"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Submit script to run in the system-managed environment\n",
"A new conda environment is built based on the conda dependencies object. If you are running this for the first time, this might take up to 5 mninutes. But this conda environment is reused so long as you don't change the conda dependencies."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"src = ScriptRunConfig(source_directory=\"./\", script='train.py', run_config=run_config_system_managed)\n",
"run = exp.submit(src)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Get run history details"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Block and wait till run finishes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.wait_for_completion(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Docker-based execution\n",
"**IMPORTANT**: You must have Docker engine installed locally in order to use this execution mode. If your kernel is already running in a Docker container, such as **Azure Notebooks**, this mode will **NOT** work.\n",
"\n",
"You can also ask the system to pull down a Docker image and execute your scripts in it."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run_config_docker = RunConfiguration()\n",
"run_config_docker.environment.python.user_managed_dependencies = False\n",
"run_config_docker.auto_prepare_environment = True\n",
"run_config_docker.environment.docker.enabled = True\n",
"run_config_docker.environment.docker.base_image = azureml.core.runconfig.DEFAULT_CPU_IMAGE\n",
"\n",
"# Specify conda dependencies with scikit-learn\n",
"cd = CondaDependencies.create(conda_packages=['scikit-learn'])\n",
"run_config_docker.environment.python.conda_dependencies = cd\n",
"\n",
"src = ScriptRunConfig(source_directory=\"./\", script='train.py', run_config=run_config_docker)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Submit script to run in the system-managed environment\n",
"A new conda environment is built based on the conda dependencies object. If you are running this for the first time, this might take up to 5 mninutes. But this conda environment is reused so long as you don't change the conda dependencies.\n",
"\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import subprocess\n",
"\n",
"# Check if Docker is installed and Linux containers are enables\n",
"if subprocess.run(\"docker -v\", shell=True) == 0:\n",
" out = subprocess.check_output(\"docker system info\", shell=True, encoding=\"ascii\").split(\"\\n\")\n",
" if not \"OSType: linux\" in out:\n",
" print(\"Switch Docker engine to use Linux containers.\")\n",
" else:\n",
" run = exp.submit(src)\n",
"else:\n",
" print(\"Docker engine not installed.\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#Get run history details\n",
"run"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Query run metrics"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"query history",
"get metrics"
]
},
"outputs": [],
"source": [
"# get all metris logged in the run\n",
"run.get_metrics()\n",
"metrics = run.get_metrics()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Let's find the model that has the lowest MSE value logged."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import numpy as np\n",
"\n",
"best_alpha = metrics['alpha'][np.argmin(metrics['mse'])]\n",
"\n",
"print('When alpha is {1:0.2f}, we have min MSE {0:0.2f}.'.format(\n",
" min(metrics['mse']), \n",
" best_alpha\n",
"))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can also list all the files that are associated with this run record"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.get_file_names()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We know the model `ridge_0.40.pkl` is the best performing model from the eariler queries. So let's register it with the workspace."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# supply a model name, and the full path to the serialized model file.\n",
"model = run.register_model(model_name='best_ridge_model', model_path='./outputs/ridge_0.40.pkl')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(model.name, model.version, model.url)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now you can deploy this model following the example in the 01 notebook."
]
}
],
"metadata": {
"authors": [
{
"name": "roastala"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

630
getting-started/train-on-remote-vm/train-on-remote-vm.ipynb
View File

@@ -1,630 +0,0 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# 04. Train in a remote Linux VM\n",
"* Create Workspace\n",
"* Create `train.py` file\n",
"* Create (or attach) DSVM as compute resource.\n",
"* Upoad data files into default datastore\n",
"* Configure & execute a run in a few different ways\n",
" - Use system-built conda\n",
" - Use existing Python environment\n",
" - Use Docker \n",
"* Find the best model in the run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"Make sure you go through the [00. Installation and Configuration](00.configuration.ipynb) Notebook first if you haven't."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize Workspace\n",
"\n",
"Initialize a workspace object from persisted configuration."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print(ws.name, ws.resource_group, ws.location, ws.subscription_id, sep='\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create Experiment\n",
"\n",
"**Experiment** is a logical container in an Azure ML Workspace. It hosts run records which can include run metrics and output artifacts from your experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"experiment_name = 'train-on-remote-vm'\n",
"\n",
"from azureml.core import Experiment\n",
"exp = Experiment(workspace=ws, name=experiment_name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Let's also create a local folder to hold the training script."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"script_folder = './vm-run'\n",
"os.makedirs(script_folder, exist_ok=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Upload data files into datastore\n",
"Every workspace comes with a default datastore (and you can register more) which is backed by the Azure blob storage account associated with the workspace. We can use it to transfer data from local to the cloud, and access it from the compute target."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# get the default datastore\n",
"ds = ws.get_default_datastore()\n",
"print(ds.name, ds.datastore_type, ds.account_name, ds.container_name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Load diabetes data from `scikit-learn` and save it as 2 local files."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.datasets import load_diabetes\n",
"import numpy as np\n",
"\n",
"training_data = load_diabetes()\n",
"np.save(file='./features.npy', arr=training_data['data'])\n",
"np.save(file='./labels.npy', arr=training_data['target'])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now let's upload the 2 files into the default datastore under a path named `diabetes`:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ds.upload_files(['./features.npy', './labels.npy'], target_path='diabetes', overwrite=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## View `train.py`\n",
"\n",
"For convenience, we created a training script for you. It is printed below as a text, but you can also run `%pfile ./train.py` in a cell to show the file. Please pay special attention on how we are loading the features and labels from files in the `data_folder` path, which is passed in as an argument of the training script (shown later)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# copy train.py into the script folder\n",
"import shutil\n",
"shutil.copy('./train.py', os.path.join(script_folder, 'train.py'))\n",
"\n",
"with open(os.path.join(script_folder, './train.py'), 'r') as training_script:\n",
" print(training_script.read())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create Linux DSVM as a compute target\n",
"\n",
"**Note**: If creation fails with a message about Marketplace purchase eligibilty, go to portal.azure.com, start creating DSVM there, and select \"Want to create programmatically\" to enable programmatic creation. Once you've enabled it, you can exit without actually creating VM.\n",
" \n",
"**Note**: By default SSH runs on port 22 and you don't need to specify it. But if for security reasons you switch to a different port (such as 5022), you can specify the port number in the provisioning configuration object."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import DsvmCompute\n",
"from azureml.core.compute_target import ComputeTargetException\n",
"\n",
"compute_target_name = 'cpudsvm'\n",
"\n",
"try:\n",
" dsvm_compute = DsvmCompute(workspace=ws, name=compute_target_name)\n",
" print('found existing:', dsvm_compute.name)\n",
"except ComputeTargetException:\n",
" print('creating new.')\n",
" dsvm_config = DsvmCompute.provisioning_configuration(vm_size=\"Standard_D2_v2\")\n",
" dsvm_compute = DsvmCompute.create(ws, name=compute_target_name, provisioning_configuration=dsvm_config)\n",
" dsvm_compute.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Attach an existing Linux DSVM\n",
"You can also attach an existing Linux VM as a compute target. To create one, you can use Azure CLI command:\n",
"\n",
"```\n",
"az vm create -n cpudsvm -l eastus2 -g <my-resource-group> --size Standard_D2_v2 --image microsoft-dsvm:linux-data-science-vm-ubuntu:linuxdsvmubuntu:latest --generate-ssh-keys\n",
"```\n",
"\n",
"The ```--generate-ssh-keys``` automatically places the ssh keys to standard location, typically to ~/.ssh folder. The default port is 22."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"'''\n",
"from azureml.core.compute import ComputeTarget, RemoteCompute \n",
"attach_config = RemoteCompute.attach_configuration(username='<my_username>',\n",
" address='<ip_adress_or_fqdn>',\n",
" ssh_port=22,\n",
" private_key_file='./.ssh/id_rsa')\n",
"attached_dsvm_compute = ComputeTarget.attach(workspace=ws,\n",
" name='attached_vm',\n",
" attach_configuration=attach_config)\n",
"attached_dsvm_compute.wait_for_completion(show_output=True)\n",
"'''\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure & Run\n",
"First let's create a `DataReferenceConfiguration` object to inform the system what data folder to download to the copmute target."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.runconfig import DataReferenceConfiguration\n",
"dr = DataReferenceConfiguration(datastore_name=ds.name, \n",
" path_on_datastore='diabetes', \n",
" mode='download', # download files from datastore to compute target\n",
" overwrite=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now we can try a few different ways to run the training script in the VM."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Conda run\n",
"You can ask the system to build a conda environment based on your dependency specification, and submit your script to run there. Once the environment is built, and if you don't change your dependencies, it will be reused in subsequent runs."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.runconfig import RunConfiguration\n",
"from azureml.core.conda_dependencies import CondaDependencies\n",
"\n",
"# create a new RunConfig object\n",
"conda_run_config = RunConfiguration(framework=\"python\")\n",
"\n",
"# Set compute target to the Linux DSVM\n",
"conda_run_config.target = dsvm_compute.name\n",
"\n",
"# set the data reference of the run configuration\n",
"conda_run_config.data_references = {ds.name: dr}\n",
"\n",
"# specify CondaDependencies obj\n",
"conda_run_config.environment.python.conda_dependencies = CondaDependencies.create(conda_packages=['scikit-learn'])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Run\n",
"from azureml.core import ScriptRunConfig\n",
"\n",
"src = ScriptRunConfig(source_directory=script_folder, \n",
" script='train.py', \n",
" run_config=conda_run_config, \n",
" # pass the datastore reference as a parameter to the training script\n",
" arguments=['--data-folder', str(ds.as_download())] \n",
" ) \n",
"run = exp.submit(config=src)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Show the run object. You can navigate to the Azure portal to see detailed information about the run."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Native VM run\n",
"You can also configure to use an exiting Python environment in the VM to execute the script without asking the system to create a conda environment for you."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# create a new RunConfig object\n",
"vm_run_config = RunConfiguration(framework=\"python\")\n",
"\n",
"# Set compute target to the Linux DSVM\n",
"vm_run_config.target = dsvm_compute.name\n",
"\n",
"# set the data reference of the run coonfiguration\n",
"conda_run_config.data_references = {ds.name: dr}\n",
"\n",
"# Let system know that you will configure the Python environment yourself.\n",
"vm_run_config.environment.python.user_managed_dependencies = True"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The below run will likely fail because `train.py` needs dependency `azureml`, `scikit-learn` and others, which are not found in that Python environment. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"src = ScriptRunConfig(source_directory=script_folder, \n",
" script='train.py', \n",
" run_config=vm_run_config,\n",
" # pass the datastore reference as a parameter to the training script\n",
" arguments=['--data-folder', str(ds.as_download())])\n",
"run = exp.submit(config=src)\n",
"run.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can choose to SSH into the VM and install Azure ML SDK, and any other missing dependencies, in that Python environment. For demonstration purposes, we simply are going to create another script `train2.py` that doesn't have azureml dependencies, and submit it instead."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile $script_folder/train2.py\n",
"\n",
"print('####################################')\n",
"print('Hello World (without Azure ML SDK)!')\n",
"print('####################################')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now let's try again. And this time it should work fine."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"src = ScriptRunConfig(source_directory=script_folder, \n",
" script='train2.py', \n",
" run_config=vm_run_config)\n",
"run = exp.submit(config=src)\n",
"run.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Note even in this case you get a run record with some basic statistics."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Configure a Docker run with new conda environment on the VM\n",
"You can execute in a Docker container in the VM. If you choose this option, the system will pull down a base Docker image, build a new conda environment in it if you ask for (you can also skip this if you are using a customer Docker image when a preconfigured Python environment), start a container, and run your script in there. This image is also uploaded into your ACR (Azure Container Registry) assoicated with your workspace, an reused if your dependencies don't change in the subsequent runs."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.runconfig import RunConfiguration\n",
"from azureml.core.conda_dependencies import CondaDependencies\n",
"\n",
"\n",
"# Load the \"cpu-dsvm.runconfig\" file (created by the above attach operation) in memory\n",
"docker_run_config = RunConfiguration(framework=\"python\")\n",
"\n",
"# Set compute target to the Linux DSVM\n",
"docker_run_config.target = dsvm_compute.name\n",
"\n",
"# Use Docker in the remote VM\n",
"docker_run_config.environment.docker.enabled = True\n",
"\n",
"# Use CPU base image from DockerHub\n",
"docker_run_config.environment.docker.base_image = azureml.core.runconfig.DEFAULT_CPU_IMAGE\n",
"print('Base Docker image is:', docker_run_config.environment.docker.base_image)\n",
"\n",
"# set the data reference of the run coonfiguration\n",
"docker_run_config.data_references = {ds.name: dr}\n",
"\n",
"# specify CondaDependencies obj\n",
"docker_run_config.environment.python.conda_dependencies = CondaDependencies.create(conda_packages=['scikit-learn'])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Submit the Experiment\n",
"Submit script to run in the Docker image in the remote VM. If you run this for the first time, the system will download the base image, layer in packages specified in the `conda_dependencies.yml` file on top of the base image, create a container and then execute the script in the container."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"src = ScriptRunConfig(source_directory=script_folder, \n",
" script='train.py', \n",
" run_config=docker_run_config,\n",
" # pass the datastore reference as a parameter to the training script\n",
" arguments=['--data-folder', str(ds.as_download())])\n",
"run = exp.submit(config=src)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### View run history details"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Find the best model"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now we have tried various execution modes, we can find the best model from the last run."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# get all metris logged in the run\n",
"run.get_metrics()\n",
"metrics = run.get_metrics()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# find the index where MSE is the smallest\n",
"indices = list(range(0, len(metrics['mse'])))\n",
"min_mse_index = min(indices, key=lambda x: metrics['mse'][x])\n",
"\n",
"print('When alpha is {1:0.2f}, we have min MSE {0:0.2f}.'.format(\n",
" metrics['mse'][min_mse_index], \n",
" metrics['alpha'][min_mse_index]\n",
"))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Clean up compute resource\n",
"\n",
"Use ```detach()``` to detach an existing DSVM from Workspace without deleting it. Use ```delete()``` if you created a new ```DsvmCompute``` and want to delete it."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# dsvm_compute.detach()\n",
"# dsvm_compute.delete()"
]
}
],
"metadata": {
"authors": [
{
"name": "haining"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

808
getting-started/train-within-notebook/train-within-notebook.ipynb
View File

@@ -1,808 +0,0 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# 01. Train in the Notebook & Deploy Model to ACI\n",
"\n",
"* Load workspace\n",
"* Train a simple regression model directly in the Notebook python kernel\n",
"* Record run history\n",
"* Find the best model in run history and download it.\n",
"* Deploy the model as an Azure Container Instance (ACI)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"1. Make sure you go through the [00. Installation and Configuration](../../00.configuration.ipynb) Notebook first if you haven't. \n",
"\n",
"2. Install following pre-requisite libraries to your conda environment and restart notebook.\n",
"```shell\n",
"(myenv) $ conda install -y matplotlib tqdm scikit-learn\n",
"```\n",
"\n",
"3. Check that ACI is registered for your Azure Subscription. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"!az provider show -n Microsoft.ContainerInstance -o table"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"If ACI is not registered, run following command to register it. Note that you have to be a subscription owner, or this command will fail."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"!az provider register -n Microsoft.ContainerInstance"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Validate Azure ML SDK installation and get version number for debugging purposes"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"install"
]
},
"outputs": [],
"source": [
"from azureml.core import Experiment, Run, Workspace\n",
"import azureml.core\n",
"\n",
"# Check core SDK version number\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize Workspace\n",
"\n",
"Initialize a workspace object from persisted configuration."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"create workspace"
]
},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"print('Workspace name: ' + ws.name, \n",
" 'Azure region: ' + ws.location, \n",
" 'Subscription id: ' + ws.subscription_id, \n",
" 'Resource group: ' + ws.resource_group, sep='\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Set experiment name\n",
"Choose a name for experiment."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"experiment_name = 'train-in-notebook'"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Start a training run in local Notebook"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# load diabetes dataset, a well-known small dataset that comes with scikit-learn\n",
"from sklearn.datasets import load_diabetes\n",
"from sklearn.linear_model import Ridge\n",
"from sklearn.metrics import mean_squared_error\n",
"from sklearn.model_selection import train_test_split\n",
"from sklearn.externals import joblib\n",
"\n",
"X, y = load_diabetes(return_X_y = True)\n",
"columns = ['age', 'gender', 'bmi', 'bp', 's1', 's2', 's3', 's4', 's5', 's6']\n",
"X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=0)\n",
"data = {\n",
" \"train\":{\"X\": X_train, \"y\": y_train}, \n",
" \"test\":{\"X\": X_test, \"y\": y_test}\n",
"}"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Train a simple Ridge model\n",
"Train a very simple Ridge regression model in scikit-learn, and save it as a pickle file."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"reg = Ridge(alpha = 0.03)\n",
"reg.fit(X=data['train']['X'], y=data['train']['y'])\n",
"preds = reg.predict(data['test']['X'])\n",
"print('Mean Squared Error is', mean_squared_error(data['test']['y'], preds))\n",
"joblib.dump(value=reg, filename='model.pkl');"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Add experiment tracking\n",
"Now, let's add Azure ML experiment logging, and upload persisted model into run record as well."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"local run",
"outputs upload"
]
},
"outputs": [],
"source": [
"experiment = Experiment(workspace=ws, name=experiment_name)\n",
"run = experiment.start_logging()\n",
"\n",
"run.tag(\"Description\",\"My first run!\")\n",
"run.log('alpha', 0.03)\n",
"reg = Ridge(alpha=0.03)\n",
"reg.fit(data['train']['X'], data['train']['y'])\n",
"preds = reg.predict(data['test']['X'])\n",
"run.log('mse', mean_squared_error(data['test']['y'], preds))\n",
"joblib.dump(value=reg, filename='model.pkl')\n",
"run.upload_file(name='outputs/model.pkl', path_or_stream='./model.pkl')\n",
"\n",
"run.complete()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We can browse to the recorded run. Please make sure you use Chrome to navigate the run history page."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Simple parameter sweep\n",
"Sweep over alpha values of a sklearn ridge model, and capture metrics and trained model in the Azure ML experiment."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import numpy as np\n",
"import os\n",
"from tqdm import tqdm\n",
"\n",
"model_name = \"model.pkl\"\n",
"\n",
"# list of numbers from 0 to 1.0 with a 0.05 interval\n",
"alphas = np.arange(0.0, 1.0, 0.05)\n",
"\n",
"# try a bunch of alpha values in a Linear Regression (Ridge) model\n",
"for alpha in tqdm(alphas):\n",
" # create a bunch of runs, each train a model with a different alpha value\n",
" with experiment.start_logging() as run:\n",
" # Use Ridge algorithm to build a regression model\n",
" reg = Ridge(alpha=alpha)\n",
" reg.fit(X=data[\"train\"][\"X\"], y=data[\"train\"][\"y\"])\n",
" preds = reg.predict(X=data[\"test\"][\"X\"])\n",
" mse = mean_squared_error(y_true=data[\"test\"][\"y\"], y_pred=preds)\n",
"\n",
" # log alpha, mean_squared_error and feature names in run history\n",
" run.log(name=\"alpha\", value=alpha)\n",
" run.log(name=\"mse\", value=mse)\n",
" run.log_list(name=\"columns\", value=columns)\n",
"\n",
" with open(model_name, \"wb\") as file:\n",
" joblib.dump(value=reg, filename=file)\n",
" \n",
" # upload the serialized model into run history record\n",
" run.upload_file(name=\"outputs/\" + model_name, path_or_stream=model_name)\n",
"\n",
" # now delete the serialized model from local folder since it is already uploaded to run history \n",
" os.remove(path=model_name)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# now let's take a look at the experiment in Azure portal.\n",
"experiment"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Select best model from the experiment\n",
"Load all experiment run metrics recursively from the experiment into a dictionary object."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"runs = {}\n",
"run_metrics = {}\n",
"\n",
"for r in tqdm(experiment.get_runs()):\n",
" metrics = r.get_metrics()\n",
" if 'mse' in metrics.keys():\n",
" runs[r.id] = r\n",
" run_metrics[r.id] = metrics"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now find the run with the lowest Mean Squared Error value"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run_id = min(run_metrics, key = lambda k: run_metrics[k]['mse'])\n",
"best_run = runs[best_run_id]\n",
"print('Best run is:', best_run_id)\n",
"print('Metrics:', run_metrics[best_run_id])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can add tags to your runs to make them easier to catalog"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"query history"
]
},
"outputs": [],
"source": [
"best_run.tag(key=\"Description\", value=\"The best one\")\n",
"best_run.get_tags()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Plot MSE over alpha\n",
"\n",
"Let's observe the best model visually by plotting the MSE values over alpha values:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%matplotlib inline\n",
"import matplotlib\n",
"import matplotlib.pyplot as plt\n",
"\n",
"best_alpha = run_metrics[best_run_id]['alpha']\n",
"min_mse = run_metrics[best_run_id]['mse']\n",
"\n",
"alpha_mse = np.array([(run_metrics[k]['alpha'], run_metrics[k]['mse']) for k in run_metrics.keys()])\n",
"sorted_alpha_mse = alpha_mse[alpha_mse[:,0].argsort()]\n",
"\n",
"plt.plot(sorted_alpha_mse[:,0], sorted_alpha_mse[:,1], 'r--')\n",
"plt.plot(sorted_alpha_mse[:,0], sorted_alpha_mse[:,1], 'bo')\n",
"\n",
"plt.xlabel('alpha', fontsize = 14)\n",
"plt.ylabel('mean squared error', fontsize = 14)\n",
"plt.title('MSE over alpha', fontsize = 16)\n",
"\n",
"# plot arrow\n",
"plt.arrow(x = best_alpha, y = min_mse + 39, dx = 0, dy = -26, ls = '-', lw = 0.4,\n",
" width = 0, head_width = .03, head_length = 8)\n",
"\n",
"# plot \"best run\" text\n",
"plt.text(x = best_alpha - 0.08, y = min_mse + 50, s = 'Best Run', fontsize = 14)\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Register the best model"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Find the model file saved in the run record of best run."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"query history"
]
},
"outputs": [],
"source": [
"for f in best_run.get_file_names():\n",
" print(f)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now we can register this model in the model registry of the workspace"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"register model from history"
]
},
"outputs": [],
"source": [
"model = best_run.register_model(model_name='best_model', model_path='outputs/model.pkl')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Verify that the model has been registered properly. If you have done this several times you'd see the version number auto-increases each time."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"register model from history"
]
},
"outputs": [],
"source": [
"from azureml.core.model import Model\n",
"models = Model.list(workspace=ws, name='best_model')\n",
"for m in models:\n",
" print(m.name, m.version)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can also download the registered model. Afterwards, you should see a `model.pkl` file in the current directory. You can then use it for local testing if you'd like."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"download file"
]
},
"outputs": [],
"source": [
"# remove the model file if it is already on disk\n",
"if os.path.isfile('model.pkl'): \n",
" os.remove('model.pkl')\n",
"# download the model\n",
"model.download(target_dir=\"./\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Scoring script\n",
"\n",
"Now we are ready to build a Docker image and deploy the model in it as a web service. The first step is creating the scoring script. For convenience, we have created the scoring script for you. It is printed below as text, but you can also run `%pfile ./score.py` in a cell to show the file.\n",
"\n",
"Tbe scoring script consists of two functions: `init` that is used to load the model to memory when starting the container, and `run` that makes the prediction when web service is called. Please pay special attention to how the model is loaded in the `init()` function. When Docker image is built for this model, the actual model file is downloaded and placed on disk, and `get_model_path` function returns the local path where the model is placed."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"with open('./score.py', 'r') as scoring_script:\n",
" print(scoring_script.read())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create environment dependency file\n",
"\n",
"We need a environment dependency file `myenv.yml` to specify which libraries are needed by the scoring script when building the Docker image for web service deployment. We can manually create this file, or we can use the `CondaDependencies` API to automatically create this file."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.conda_dependencies import CondaDependencies \n",
"\n",
"myenv = CondaDependencies.create(conda_packages=[\"scikit-learn\"])\n",
"print(myenv.serialize_to_string())\n",
"\n",
"with open(\"myenv.yml\",\"w\") as f:\n",
" f.write(myenv.serialize_to_string())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Deploy web service into an Azure Container Instance\n",
"The deployment process takes the registered model and your scoring scrip, and builds a Docker image. It then deploys the Docker image into Azure Container Instance as a running container with an HTTP endpoint readying for scoring calls. Read more about [Azure Container Instance](https://azure.microsoft.com/en-us/services/container-instances/).\n",
"\n",
"Note ACI is great for quick and cost-effective dev/test deployment scenarios. For production workloads, please use [Azure Kubernentes Service (AKS)](https://azure.microsoft.com/en-us/services/kubernetes-service/) instead. Please follow in struction in [this notebook](11.production-deploy-to-aks.ipynb) to see how that can be done from Azure ML.\n",
" \n",
"** Note: ** The web service creation can take 6-7 minutes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"from azureml.core.webservice import AciWebservice, Webservice\n",
"\n",
"aciconfig = AciWebservice.deploy_configuration(cpu_cores=1, \n",
" memory_gb=1, \n",
" tags={'sample name': 'AML 101'}, \n",
" description='This is a great example.')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Note the below `WebService.deploy_from_model()` function takes a model object registered under the workspace. It then bakes the model file in the Docker image so it can be looked-up using the `Model.get_model_path()` function in `score.py`. \n",
"\n",
"If you have a local model file instead of a registered model object, you can also use the `WebService.deploy()` function which would register the model and then deploy."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"from azureml.core.image import ContainerImage\n",
"image_config = ContainerImage.image_configuration(execution_script=\"score.py\", \n",
" runtime=\"python\", \n",
" conda_file=\"myenv.yml\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"%%time\n",
"# this will take 5-10 minutes to finish\n",
"# you can also use \"az container list\" command to find the ACI being deployed\n",
"service = Webservice.deploy_from_model(name='my-aci-svc',\n",
" deployment_config=aciconfig,\n",
" models=[model],\n",
" image_config=image_config,\n",
" workspace=ws)\n",
"\n",
"service.wait_for_deployment(show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"## Test web service"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"print('web service is hosted in ACI:', service.scoring_uri)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Use the `run` API to call the web service with one row of data to get a prediction."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"import json\n",
"# score the first row from the test set.\n",
"test_samples = json.dumps({\"data\": X_test[0:1, :].tolist()})\n",
"service.run(input_data = test_samples)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Feed the entire test set and calculate the errors (residual values)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"# score the entire test set.\n",
"test_samples = json.dumps({'data': X_test.tolist()})\n",
"\n",
"result = service.run(input_data = test_samples)\n",
"residual = result - y_test"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can also send raw HTTP request to test the web service."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"import requests\n",
"import json\n",
"\n",
"# 2 rows of input data, each with 10 made-up numerical features\n",
"input_data = \"{\\\"data\\\": [[1, 2, 3, 4, 5, 6, 7, 8, 9, 10], [10, 9, 8, 7, 6, 5, 4, 3, 2, 1]]}\"\n",
"\n",
"headers = {'Content-Type':'application/json'}\n",
"\n",
"# for AKS deployment you'd need to the service key in the header as well\n",
"# api_key = service.get_key()\n",
"# headers = {'Content-Type':'application/json', 'Authorization':('Bearer '+ api_key)} \n",
"\n",
"resp = requests.post(service.scoring_uri, input_data, headers = headers)\n",
"print(resp.text)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Residual graph\n",
"Plot a residual value graph to chart the errors on the entire test set. Observe the nice bell curve."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"f, (a0, a1) = plt.subplots(1, 2, gridspec_kw={'width_ratios':[3, 1], 'wspace':0, 'hspace': 0})\n",
"f.suptitle('Residual Values', fontsize = 18)\n",
"\n",
"f.set_figheight(6)\n",
"f.set_figwidth(14)\n",
"\n",
"a0.plot(residual, 'bo', alpha=0.4);\n",
"a0.plot([0,90], [0,0], 'r', lw=2)\n",
"a0.set_ylabel('residue values', fontsize=14)\n",
"a0.set_xlabel('test data set', fontsize=14)\n",
"\n",
"a1.hist(residual, orientation='horizontal', color='blue', bins=10, histtype='step');\n",
"a1.hist(residual, orientation='horizontal', color='blue', alpha=0.2, bins=10);\n",
"a1.set_yticklabels([])\n",
"\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Delete ACI to clean up"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Deleting ACI is super fast!"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"deploy service",
"aci"
]
},
"outputs": [],
"source": [
"%%time\n",
"service.delete()"
]
}
],
"metadata": {
"authors": [
{
"name": "roastala"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

124
onnx/mnist.py Normal file
View File

@@ -0,0 +1,124 @@
# This is a modified version of https://github.com/pytorch/examples/blob/master/mnist/main.py which is
# licensed under BSD 3-Clause (https://github.com/pytorch/examples/blob/master/LICENSE)
from __future__ import print_function
import argparse
import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
from torchvision import datasets, transforms
import os
class Net(nn.Module):
def __init__(self):
super(Net, self).__init__()
self.conv1 = nn.Conv2d(1, 10, kernel_size=5)
self.conv2 = nn.Conv2d(10, 20, kernel_size=5)
self.conv2_drop = nn.Dropout2d()
self.fc1 = nn.Linear(320, 50)
self.fc2 = nn.Linear(50, 10)
def forward(self, x):
x = F.relu(F.max_pool2d(self.conv1(x), 2))
x = F.relu(F.max_pool2d(self.conv2_drop(self.conv2(x)), 2))
x = x.view(-1, 320)
x = F.relu(self.fc1(x))
x = F.dropout(x, training=self.training)
x = self.fc2(x)
return F.log_softmax(x, dim=1)
def train(args, model, device, train_loader, optimizer, epoch, output_dir):
model.train()
for batch_idx, (data, target) in enumerate(train_loader):
data, target = data.to(device), target.to(device)
optimizer.zero_grad()
output = model(data)
loss = F.nll_loss(output, target)
loss.backward()
optimizer.step()
if batch_idx % args.log_interval == 0:
print('Train Epoch: {} [{}/{} ({:.0f}%)]\tLoss: {:.6f}'.format(
epoch, batch_idx * len(data), len(train_loader.dataset),
100. * batch_idx / len(train_loader), loss.item()))
def test(args, model, device, test_loader):
model.eval()
test_loss = 0
correct = 0
with torch.no_grad():
for data, target in test_loader:
data, target = data.to(device), target.to(device)
output = model(data)
test_loss += F.nll_loss(output, target, size_average=False, reduce=True).item() # sum up batch loss
pred = output.max(1, keepdim=True)[1] # get the index of the max log-probability
correct += pred.eq(target.view_as(pred)).sum().item()
test_loss /= len(test_loader.dataset)
print('\nTest set: Average loss: {:.4f}, Accuracy: {}/{} ({:.0f}%)\n'.format(
test_loss, correct, len(test_loader.dataset),
100. * correct / len(test_loader.dataset)))
def main():
# Training settings
parser = argparse.ArgumentParser(description='PyTorch MNIST Example')
parser.add_argument('--batch-size', type=int, default=64, metavar='N',
help='input batch size for training (default: 64)')
parser.add_argument('--test-batch-size', type=int, default=1000, metavar='N',
help='input batch size for testing (default: 1000)')
parser.add_argument('--epochs', type=int, default=10, metavar='N',
help='number of epochs to train (default: 10)')
parser.add_argument('--lr', type=float, default=0.01, metavar='LR',
help='learning rate (default: 0.01)')
parser.add_argument('--momentum', type=float, default=0.5, metavar='M',
help='SGD momentum (default: 0.5)')
parser.add_argument('--no-cuda', action='store_true', default=False,
help='disables CUDA training')
parser.add_argument('--seed', type=int, default=1, metavar='S',
help='random seed (default: 1)')
parser.add_argument('--log-interval', type=int, default=10, metavar='N',
help='how many batches to wait before logging training status')
parser.add_argument('--output-dir', type=str, default='outputs')
args = parser.parse_args()
use_cuda = not args.no_cuda and torch.cuda.is_available()
torch.manual_seed(args.seed)
device = torch.device("cuda" if use_cuda else "cpu")
output_dir = args.output_dir
os.makedirs(output_dir, exist_ok=True)
kwargs = {'num_workers': 1, 'pin_memory': True} if use_cuda else {}
train_loader = torch.utils.data.DataLoader(
datasets.MNIST('data', train=True, download=True,
transform=transforms.Compose([transforms.ToTensor(),
transforms.Normalize((0.1307,), (0.3081,))])
),
batch_size=args.batch_size, shuffle=True, **kwargs)
test_loader = torch.utils.data.DataLoader(
datasets.MNIST('data', train=False,
transform=transforms.Compose([transforms.ToTensor(),
transforms.Normalize((0.1307,), (0.3081,))])
),
batch_size=args.test_batch_size, shuffle=True, **kwargs)
model = Net().to(device)
optimizer = optim.SGD(model.parameters(), lr=args.lr, momentum=args.momentum)
for epoch in range(1, args.epochs + 1):
train(args, model, device, train_loader, optimizer, epoch, output_dir)
test(args, model, device, test_loader)
# save model
dummy_input = torch.randn(1, 1, 28, 28, device=device)
model_path = os.path.join(output_dir, 'mnist.onnx')
torch.onnx.export(model, dummy_input, model_path)
if __name__ == '__main__':
main()

864
onnx/onnx-convert-aml-deploy-tinyyolo.ipynb
View File

@@ -1,435 +1,435 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved. \n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# YOLO Real-time Object Detection using ONNX on AzureML\n",
"\n",
"This example shows how to convert the TinyYOLO model from CoreML to ONNX and operationalize it as a web service using Azure Machine Learning services and the ONNX Runtime.\n",
"\n",
"## What is ONNX\n",
"ONNX is an open format for representing machine learning and deep learning models. ONNX enables open and interoperable AI by enabling data scientists and developers to use the tools of their choice without worrying about lock-in and flexibility to deploy to a variety of platforms. ONNX is developed and supported by a community of partners including Microsoft, Facebook, and Amazon. For more information, explore the [ONNX website](http://onnx.ai).\n",
"\n",
"## YOLO Details\n",
"You Only Look Once (YOLO) is a state-of-the-art, real-time object detection system. For more information about YOLO, please visit the [YOLO website](https://pjreddie.com/darknet/yolo/)."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"\n",
"To make the best use of your time, make sure you have done the following:\n",
"\n",
"* Understand the [architecture and terms](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture) introduced by Azure Machine Learning\n",
"* Go through the [00.configuration.ipynb](../00.configuration.ipynb) notebook to:\n",
" * install the AML SDK\n",
" * create a workspace and its configuration file (config.json)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Install necessary packages\n",
"\n",
"You'll need to run the following commands to use this tutorial:\n",
"\n",
"```sh\n",
"pip install onnxmltools\n",
"pip install coremltools # use this on Linux and Mac\n",
"pip install git+https://github.com/apple/coremltools # use this on Windows\n",
"```"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Convert model to ONNX\n",
"\n",
"First we download the CoreML model. We use the CoreML model listed at https://coreml.store/tinyyolo. This may take a few minutes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import urllib.request\n",
"\n",
"onnx_model_url = \"https://s3-us-west-2.amazonaws.com/coreml-models/TinyYOLO.mlmodel\"\n",
"urllib.request.urlretrieve(onnx_model_url, filename=\"TinyYOLO.mlmodel\")\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Then we use ONNXMLTools to convert the model."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import onnxmltools\n",
"import coremltools\n",
"\n",
"# Load a CoreML model\n",
"coreml_model = coremltools.utils.load_spec('TinyYOLO.mlmodel')\n",
"\n",
"# Convert from CoreML into ONNX\n",
"onnx_model = onnxmltools.convert_coreml(coreml_model, 'TinyYOLOv2')\n",
"\n",
"# Save ONNX model\n",
"onnxmltools.utils.save_model(onnx_model, 'tinyyolov2.onnx')\n",
"\n",
"import os\n",
"print(os.path.getsize('tinyyolov2.onnx'))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Deploying as a web service with Azure ML\n",
"\n",
"### Load Azure ML workspace\n",
"\n",
"We begin by instantiating a workspace object from the existing workspace created earlier in the configuration notebook."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print(ws.name, ws.location, ws.resource_group, sep = '\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Registering your model with Azure ML\n",
"\n",
"Now we upload the model and register it in the workspace."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.model import Model\n",
"\n",
"model = Model.register(model_path = \"tinyyolov2.onnx\",\n",
" model_name = \"tinyyolov2\",\n",
" tags = {\"onnx\": \"demo\"},\n",
" description = \"TinyYOLO\",\n",
" workspace = ws)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Displaying your registered models\n",
"\n",
"You can optionally list out all the models that you have registered in this workspace."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"models = ws.models\n",
"for name, m in models.items():\n",
" print(\"Name:\", name,\"\\tVersion:\", m.version, \"\\tDescription:\", m.description, m.tags)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Write scoring file\n",
"\n",
"We are now going to deploy our ONNX model on Azure ML using the ONNX Runtime. We begin by writing a score.py file that will be invoked by the web service call. The `init()` function is called once when the container is started so we load the model using the ONNX Runtime into a global session object."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile score.py\n",
"import json\n",
"import time\n",
"import sys\n",
"import os\n",
"from azureml.core.model import Model\n",
"import numpy as np # we're going to use numpy to process input and output data\n",
"import onnxruntime # to inference ONNX models, we use the ONNX Runtime\n",
"\n",
"def init():\n",
" global session\n",
" model = Model.get_model_path(model_name = 'tinyyolov2')\n",
" session = onnxruntime.InferenceSession(model)\n",
"\n",
"def preprocess(input_data_json):\n",
" # convert the JSON data into the tensor input\n",
" return np.array(json.loads(input_data_json)['data']).astype('float32')\n",
"\n",
"def postprocess(result):\n",
" return np.array(result).tolist()\n",
"\n",
"def run(input_data_json):\n",
" try:\n",
" start = time.time() # start timer\n",
" input_data = preprocess(input_data_json)\n",
" input_name = session.get_inputs()[0].name # get the id of the first input of the model \n",
" result = session.run([], {input_name: input_data})\n",
" end = time.time() # stop timer\n",
" return {\"result\": postprocess(result),\n",
" \"time\": end - start}\n",
" except Exception as e:\n",
" result = str(e)\n",
" return {\"error\": result}"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create container image\n",
"First we create a YAML file that specifies which dependencies we would like to see in our container."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.conda_dependencies import CondaDependencies \n",
"\n",
"myenv = CondaDependencies.create(pip_packages=[\"numpy\",\"onnxruntime\",\"azureml-core\"])\n",
"\n",
"with open(\"myenv.yml\",\"w\") as f:\n",
" f.write(myenv.serialize_to_string())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Then we have Azure ML create the container. This step will likely take a few minutes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.image import ContainerImage\n",
"\n",
"image_config = ContainerImage.image_configuration(execution_script = \"score.py\",\n",
" runtime = \"python\",\n",
" conda_file = \"myenv.yml\",\n",
" description = \"TinyYOLO ONNX Demo\",\n",
" tags = {\"demo\": \"onnx\"}\n",
" )\n",
"\n",
"\n",
"image = ContainerImage.create(name = \"onnxyolo\",\n",
" models = [model],\n",
" image_config = image_config,\n",
" workspace = ws)\n",
"\n",
"image.wait_for_creation(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"In case you need to debug your code, the next line of code accesses the log file."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(image.image_build_log_uri)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We're all set! Let's get our model chugging.\n",
"\n",
"### Deploy the container image"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.webservice import AciWebservice\n",
"\n",
"aciconfig = AciWebservice.deploy_configuration(cpu_cores = 1, \n",
" memory_gb = 1, \n",
" tags = {'demo': 'onnx'}, \n",
" description = 'web service for TinyYOLO ONNX model')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The following cell will likely take a few minutes to run as well."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.webservice import Webservice\n",
"from random import randint\n",
"\n",
"aci_service_name = 'onnx-tinyyolo'+str(randint(0,100))\n",
"print(\"Service\", aci_service_name)\n",
"\n",
"aci_service = Webservice.deploy_from_image(deployment_config = aciconfig,\n",
" image = image,\n",
" name = aci_service_name,\n",
" workspace = ws)\n",
"\n",
"aci_service.wait_for_deployment(True)\n",
"print(aci_service.state)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"In case the deployment fails, you can check the logs. Make sure to delete your aci_service before trying again."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"if aci_service.state != 'Healthy':\n",
" # run this command for debugging.\n",
" print(aci_service.get_logs())\n",
" aci_service.delete()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Success!\n",
"\n",
"If you've made it this far, you've deployed a working web service that does object detection using an ONNX model. You can get the URL for the webservice with the code below."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(aci_service.scoring_uri)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"When you are eventually done using the web service, remember to delete it."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#aci_service.delete()"
]
}
],
"metadata": {
"authors": [
{
"name": "onnx"
}
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved. \n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# YOLO Real-time Object Detection using ONNX on AzureML\n",
"\n",
"This example shows how to convert the TinyYOLO model from CoreML to ONNX and operationalize it as a web service using Azure Machine Learning services and the ONNX Runtime.\n",
"\n",
"## What is ONNX\n",
"ONNX is an open format for representing machine learning and deep learning models. ONNX enables open and interoperable AI by enabling data scientists and developers to use the tools of their choice without worrying about lock-in and flexibility to deploy to a variety of platforms. ONNX is developed and supported by a community of partners including Microsoft, Facebook, and Amazon. For more information, explore the [ONNX website](http://onnx.ai).\n",
"\n",
"## YOLO Details\n",
"You Only Look Once (YOLO) is a state-of-the-art, real-time object detection system. For more information about YOLO, please visit the [YOLO website](https://pjreddie.com/darknet/yolo/)."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"\n",
"To make the best use of your time, make sure you have done the following:\n",
"\n",
"* Understand the [architecture and terms](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture) introduced by Azure Machine Learning\n",
"* Go through the [00.configuration.ipynb](../00.configuration.ipynb) notebook to:\n",
" * install the AML SDK\n",
" * create a workspace and its configuration file (config.json)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Install necessary packages\n",
"\n",
"You'll need to run the following commands to use this tutorial:\n",
"\n",
"```sh\n",
"pip install onnxmltools\n",
"pip install coremltools # use this on Linux and Mac\n",
"pip install git+https://github.com/apple/coremltools # use this on Windows\n",
"```"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Convert model to ONNX\n",
"\n",
"First we download the CoreML model. We use the CoreML model listed at https://coreml.store/tinyyolo. This may take a few minutes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import urllib.request\n",
"\n",
"onnx_model_url = \"https://s3-us-west-2.amazonaws.com/coreml-models/TinyYOLO.mlmodel\"\n",
"urllib.request.urlretrieve(onnx_model_url, filename=\"TinyYOLO.mlmodel\")\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Then we use ONNXMLTools to convert the model."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import onnxmltools\n",
"import coremltools\n",
"\n",
"# Load a CoreML model\n",
"coreml_model = coremltools.utils.load_spec('TinyYOLO.mlmodel')\n",
"\n",
"# Convert from CoreML into ONNX\n",
"onnx_model = onnxmltools.convert_coreml(coreml_model, 'TinyYOLOv2')\n",
"\n",
"# Save ONNX model\n",
"onnxmltools.utils.save_model(onnx_model, 'tinyyolov2.onnx')\n",
"\n",
"import os\n",
"print(os.path.getsize('tinyyolov2.onnx'))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Deploying as a web service with Azure ML\n",
"\n",
"### Load Azure ML workspace\n",
"\n",
"We begin by instantiating a workspace object from the existing workspace created earlier in the configuration notebook."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print(ws.name, ws.location, ws.resource_group, sep = '\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Registering your model with Azure ML\n",
"\n",
"Now we upload the model and register it in the workspace."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.model import Model\n",
"\n",
"model = Model.register(model_path = \"tinyyolov2.onnx\",\n",
" model_name = \"tinyyolov2\",\n",
" tags = {\"onnx\": \"demo\"},\n",
" description = \"TinyYOLO\",\n",
" workspace = ws)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Displaying your registered models\n",
"\n",
"You can optionally list out all the models that you have registered in this workspace."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"models = ws.models\n",
"for name, m in models.items():\n",
" print(\"Name:\", name,\"\\tVersion:\", m.version, \"\\tDescription:\", m.description, m.tags)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Write scoring file\n",
"\n",
"We are now going to deploy our ONNX model on Azure ML using the ONNX Runtime. We begin by writing a score.py file that will be invoked by the web service call. The `init()` function is called once when the container is started so we load the model using the ONNX Runtime into a global session object."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile score.py\n",
"import json\n",
"import time\n",
"import sys\n",
"import os\n",
"from azureml.core.model import Model\n",
"import numpy as np # we're going to use numpy to process input and output data\n",
"import onnxruntime # to inference ONNX models, we use the ONNX Runtime\n",
"\n",
"def init():\n",
" global session\n",
" model = Model.get_model_path(model_name = 'tinyyolov2')\n",
" session = onnxruntime.InferenceSession(model)\n",
"\n",
"def preprocess(input_data_json):\n",
" # convert the JSON data into the tensor input\n",
" return np.array(json.loads(input_data_json)['data']).astype('float32')\n",
"\n",
"def postprocess(result):\n",
" return np.array(result).tolist()\n",
"\n",
"def run(input_data_json):\n",
" try:\n",
" start = time.time() # start timer\n",
" input_data = preprocess(input_data_json)\n",
" input_name = session.get_inputs()[0].name # get the id of the first input of the model \n",
" result = session.run([], {input_name: input_data})\n",
" end = time.time() # stop timer\n",
" return {\"result\": postprocess(result),\n",
" \"time\": end - start}\n",
" except Exception as e:\n",
" result = str(e)\n",
" return {\"error\": result}"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create container image\n",
"First we create a YAML file that specifies which dependencies we would like to see in our container."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.conda_dependencies import CondaDependencies \n",
"\n",
"myenv = CondaDependencies.create(pip_packages=[\"numpy\",\"onnxruntime\",\"azureml-core\"])\n",
"\n",
"with open(\"myenv.yml\",\"w\") as f:\n",
" f.write(myenv.serialize_to_string())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Then we have Azure ML create the container. This step will likely take a few minutes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.image import ContainerImage\n",
"\n",
"image_config = ContainerImage.image_configuration(execution_script = \"score.py\",\n",
" runtime = \"python\",\n",
" conda_file = \"myenv.yml\",\n",
" description = \"TinyYOLO ONNX Demo\",\n",
" tags = {\"demo\": \"onnx\"}\n",
" )\n",
"\n",
"\n",
"image = ContainerImage.create(name = \"onnxyolo\",\n",
" models = [model],\n",
" image_config = image_config,\n",
" workspace = ws)\n",
"\n",
"image.wait_for_creation(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"In case you need to debug your code, the next line of code accesses the log file."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(image.image_build_log_uri)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We're all set! Let's get our model chugging.\n",
"\n",
"### Deploy the container image"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.webservice import AciWebservice\n",
"\n",
"aciconfig = AciWebservice.deploy_configuration(cpu_cores = 1, \n",
" memory_gb = 1, \n",
" tags = {'demo': 'onnx'}, \n",
" description = 'web service for TinyYOLO ONNX model')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The following cell will likely take a few minutes to run as well."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.webservice import Webservice\n",
"from random import randint\n",
"\n",
"aci_service_name = 'onnx-tinyyolo'+str(randint(0,100))\n",
"print(\"Service\", aci_service_name)\n",
"\n",
"aci_service = Webservice.deploy_from_image(deployment_config = aciconfig,\n",
" image = image,\n",
" name = aci_service_name,\n",
" workspace = ws)\n",
"\n",
"aci_service.wait_for_deployment(True)\n",
"print(aci_service.state)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"In case the deployment fails, you can check the logs. Make sure to delete your aci_service before trying again."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"if aci_service.state != 'Healthy':\n",
" # run this command for debugging.\n",
" print(aci_service.get_logs())\n",
" aci_service.delete()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Success!\n",
"\n",
"If you've made it this far, you've deployed a working web service that does object detection using an ONNX model. You can get the URL for the webservice with the code below."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(aci_service.scoring_uri)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"When you are eventually done using the web service, remember to delete it."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#aci_service.delete()"
]
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
"metadata": {
"authors": [
{
"name": "onnx"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.5.6"
}
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.5.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
"nbformat": 4,
"nbformat_minor": 2
}

1610
onnx/onnx-inference-facial-expression-recognition-deploy.ipynb
View File

File diff suppressed because it is too large Load Diff

1576
onnx/onnx-inference-mnist-deploy.ipynb
View File

File diff suppressed because it is too large Load Diff

832
onnx/onnx-modelzoo-aml-deploy-resnet50.ipynb
View File

@@ -1,419 +1,419 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved. \n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# ResNet50 Image Classification using ONNX and AzureML\n",
"\n",
"This example shows how to deploy the ResNet50 ONNX model as a web service using Azure Machine Learning services and the ONNX Runtime.\n",
"\n",
"## What is ONNX\n",
"ONNX is an open format for representing machine learning and deep learning models. ONNX enables open and interoperable AI by enabling data scientists and developers to use the tools of their choice without worrying about lock-in and flexibility to deploy to a variety of platforms. ONNX is developed and supported by a community of partners including Microsoft, Facebook, and Amazon. For more information, explore the [ONNX website](http://onnx.ai).\n",
"\n",
"## ResNet50 Details\n",
"ResNet classifies the major object in an input image into a set of 1000 pre-defined classes. For more information about the ResNet50 model and how it was created can be found on the [ONNX Model Zoo github](https://github.com/onnx/models/tree/master/models/image_classification/resnet). "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"\n",
"To make the best use of your time, make sure you have done the following:\n",
"\n",
"* Understand the [architecture and terms](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture) introduced by Azure Machine Learning\n",
"* Go through the [00.configuration.ipynb](../00.configuration.ipynb) notebook to:\n",
" * install the AML SDK\n",
" * create a workspace and its configuration file (config.json)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Download pre-trained ONNX model from ONNX Model Zoo.\n",
"\n",
"Download the [ResNet50v2 model and test data](https://s3.amazonaws.com/onnx-model-zoo/resnet/resnet50v2/resnet50v2.tar.gz) and extract it in the same folder as this tutorial notebook.\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import urllib.request\n",
"\n",
"onnx_model_url = \"https://s3.amazonaws.com/onnx-model-zoo/resnet/resnet50v2/resnet50v2.tar.gz\"\n",
"urllib.request.urlretrieve(onnx_model_url, filename=\"resnet50v2.tar.gz\")\n",
"\n",
"!tar xvzf resnet50v2.tar.gz"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Deploying as a web service with Azure ML"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Load your Azure ML workspace\n",
"\n",
"We begin by instantiating a workspace object from the existing workspace created earlier in the configuration notebook."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print(ws.name, ws.location, ws.resource_group, sep = '\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Register your model with Azure ML\n",
"\n",
"Now we upload the model and register it in the workspace."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.model import Model\n",
"\n",
"model = Model.register(model_path = \"resnet50v2/resnet50v2.onnx\",\n",
" model_name = \"resnet50v2\",\n",
" tags = {\"onnx\": \"demo\"},\n",
" description = \"ResNet50v2 from ONNX Model Zoo\",\n",
" workspace = ws)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Displaying your registered models\n",
"\n",
"You can optionally list out all the models that you have registered in this workspace."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"models = ws.models\n",
"for name, m in models.items():\n",
" print(\"Name:\", name,\"\\tVersion:\", m.version, \"\\tDescription:\", m.description, m.tags)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Write scoring file\n",
"\n",
"We are now going to deploy our ONNX model on Azure ML using the ONNX Runtime. We begin by writing a score.py file that will be invoked by the web service call. The `init()` function is called once when the container is started so we load the model using the ONNX Runtime into a global session object."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile score.py\n",
"import json\n",
"import time\n",
"import sys\n",
"import os\n",
"from azureml.core.model import Model\n",
"import numpy as np # we're going to use numpy to process input and output data\n",
"import onnxruntime # to inference ONNX models, we use the ONNX Runtime\n",
"\n",
"def softmax(x):\n",
" x = x.reshape(-1)\n",
" e_x = np.exp(x - np.max(x))\n",
" return e_x / e_x.sum(axis=0)\n",
"\n",
"def init():\n",
" global session\n",
" model = Model.get_model_path(model_name = 'resnet50v2')\n",
" session = onnxruntime.InferenceSession(model, None)\n",
"\n",
"def preprocess(input_data_json):\n",
" # convert the JSON data into the tensor input\n",
" img_data = np.array(json.loads(input_data_json)['data']).astype('float32')\n",
" \n",
" #normalize\n",
" mean_vec = np.array([0.485, 0.456, 0.406])\n",
" stddev_vec = np.array([0.229, 0.224, 0.225])\n",
" norm_img_data = np.zeros(img_data.shape).astype('float32')\n",
" for i in range(img_data.shape[0]):\n",
" norm_img_data[i,:,:] = (img_data[i,:,:]/255 - mean_vec[i]) / stddev_vec[i]\n",
"\n",
" return norm_img_data\n",
"\n",
"def postprocess(result):\n",
" return softmax(np.array(result)).tolist()\n",
"\n",
"def run(input_data_json):\n",
" try:\n",
" start = time.time()\n",
" # load in our data which is expected as NCHW 224x224 image\n",
" input_data = preprocess(input_data_json)\n",
" input_name = session.get_inputs()[0].name # get the id of the first input of the model \n",
" result = session.run([], {input_name: input_data})\n",
" end = time.time() # stop timer\n",
" return {\"result\": postprocess(result),\n",
" \"time\": end - start}\n",
" except Exception as e:\n",
" result = str(e)\n",
" return {\"error\": result}"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create container image"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"First we create a YAML file that specifies which dependencies we would like to see in our container."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.conda_dependencies import CondaDependencies \n",
"\n",
"myenv = CondaDependencies.create(pip_packages=[\"numpy\",\"onnxruntime\",\"azureml-core\"])\n",
"\n",
"with open(\"myenv.yml\",\"w\") as f:\n",
" f.write(myenv.serialize_to_string())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Then we have Azure ML create the container. This step will likely take a few minutes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.image import ContainerImage\n",
"\n",
"image_config = ContainerImage.image_configuration(execution_script = \"score.py\",\n",
" runtime = \"python\",\n",
" conda_file = \"myenv.yml\",\n",
" description = \"ONNX ResNet50 Demo\",\n",
" tags = {\"demo\": \"onnx\"}\n",
" )\n",
"\n",
"\n",
"image = ContainerImage.create(name = \"onnxresnet50v2\",\n",
" models = [model],\n",
" image_config = image_config,\n",
" workspace = ws)\n",
"\n",
"image.wait_for_creation(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"In case you need to debug your code, the next line of code accesses the log file."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(image.image_build_log_uri)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We're all set! Let's get our model chugging.\n",
"\n",
"### Deploy the container image"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.webservice import AciWebservice\n",
"\n",
"aciconfig = AciWebservice.deploy_configuration(cpu_cores = 1, \n",
" memory_gb = 1, \n",
" tags = {'demo': 'onnx'}, \n",
" description = 'web service for ResNet50 ONNX model')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The following cell will likely take a few minutes to run as well."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.webservice import Webservice\n",
"from random import randint\n",
"\n",
"aci_service_name = 'onnx-demo-resnet50'+str(randint(0,100))\n",
"print(\"Service\", aci_service_name)\n",
"\n",
"aci_service = Webservice.deploy_from_image(deployment_config = aciconfig,\n",
" image = image,\n",
" name = aci_service_name,\n",
" workspace = ws)\n",
"\n",
"aci_service.wait_for_deployment(True)\n",
"print(aci_service.state)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"In case the deployment fails, you can check the logs. Make sure to delete your aci_service before trying again."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"if aci_service.state != 'Healthy':\n",
" # run this command for debugging.\n",
" print(aci_service.get_logs())\n",
" aci_service.delete()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Success!\n",
"\n",
"If you've made it this far, you've deployed a working web service that does image classification using an ONNX model. You can get the URL for the webservice with the code below."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(aci_service.scoring_uri)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"When you are eventually done using the web service, remember to delete it."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#aci_service.delete()"
]
}
],
"metadata": {
"authors": [
{
"name": "onnx"
}
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved. \n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# ResNet50 Image Classification using ONNX and AzureML\n",
"\n",
"This example shows how to deploy the ResNet50 ONNX model as a web service using Azure Machine Learning services and the ONNX Runtime.\n",
"\n",
"## What is ONNX\n",
"ONNX is an open format for representing machine learning and deep learning models. ONNX enables open and interoperable AI by enabling data scientists and developers to use the tools of their choice without worrying about lock-in and flexibility to deploy to a variety of platforms. ONNX is developed and supported by a community of partners including Microsoft, Facebook, and Amazon. For more information, explore the [ONNX website](http://onnx.ai).\n",
"\n",
"## ResNet50 Details\n",
"ResNet classifies the major object in an input image into a set of 1000 pre-defined classes. For more information about the ResNet50 model and how it was created can be found on the [ONNX Model Zoo github](https://github.com/onnx/models/tree/master/models/image_classification/resnet). "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"\n",
"To make the best use of your time, make sure you have done the following:\n",
"\n",
"* Understand the [architecture and terms](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture) introduced by Azure Machine Learning\n",
"* Go through the [00.configuration.ipynb](../00.configuration.ipynb) notebook to:\n",
" * install the AML SDK\n",
" * create a workspace and its configuration file (config.json)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Download pre-trained ONNX model from ONNX Model Zoo.\n",
"\n",
"Download the [ResNet50v2 model and test data](https://s3.amazonaws.com/onnx-model-zoo/resnet/resnet50v2/resnet50v2.tar.gz) and extract it in the same folder as this tutorial notebook.\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import urllib.request\n",
"\n",
"onnx_model_url = \"https://s3.amazonaws.com/onnx-model-zoo/resnet/resnet50v2/resnet50v2.tar.gz\"\n",
"urllib.request.urlretrieve(onnx_model_url, filename=\"resnet50v2.tar.gz\")\n",
"\n",
"!tar xvzf resnet50v2.tar.gz"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Deploying as a web service with Azure ML"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Load your Azure ML workspace\n",
"\n",
"We begin by instantiating a workspace object from the existing workspace created earlier in the configuration notebook."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print(ws.name, ws.location, ws.resource_group, sep = '\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Register your model with Azure ML\n",
"\n",
"Now we upload the model and register it in the workspace."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.model import Model\n",
"\n",
"model = Model.register(model_path = \"resnet50v2/resnet50v2.onnx\",\n",
" model_name = \"resnet50v2\",\n",
" tags = {\"onnx\": \"demo\"},\n",
" description = \"ResNet50v2 from ONNX Model Zoo\",\n",
" workspace = ws)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Displaying your registered models\n",
"\n",
"You can optionally list out all the models that you have registered in this workspace."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"models = ws.models\n",
"for name, m in models.items():\n",
" print(\"Name:\", name,\"\\tVersion:\", m.version, \"\\tDescription:\", m.description, m.tags)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Write scoring file\n",
"\n",
"We are now going to deploy our ONNX model on Azure ML using the ONNX Runtime. We begin by writing a score.py file that will be invoked by the web service call. The `init()` function is called once when the container is started so we load the model using the ONNX Runtime into a global session object."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile score.py\n",
"import json\n",
"import time\n",
"import sys\n",
"import os\n",
"from azureml.core.model import Model\n",
"import numpy as np # we're going to use numpy to process input and output data\n",
"import onnxruntime # to inference ONNX models, we use the ONNX Runtime\n",
"\n",
"def softmax(x):\n",
" x = x.reshape(-1)\n",
" e_x = np.exp(x - np.max(x))\n",
" return e_x / e_x.sum(axis=0)\n",
"\n",
"def init():\n",
" global session\n",
" model = Model.get_model_path(model_name = 'resnet50v2')\n",
" session = onnxruntime.InferenceSession(model, None)\n",
"\n",
"def preprocess(input_data_json):\n",
" # convert the JSON data into the tensor input\n",
" img_data = np.array(json.loads(input_data_json)['data']).astype('float32')\n",
" \n",
" #normalize\n",
" mean_vec = np.array([0.485, 0.456, 0.406])\n",
" stddev_vec = np.array([0.229, 0.224, 0.225])\n",
" norm_img_data = np.zeros(img_data.shape).astype('float32')\n",
" for i in range(img_data.shape[0]):\n",
" norm_img_data[i,:,:] = (img_data[i,:,:]/255 - mean_vec[i]) / stddev_vec[i]\n",
"\n",
" return norm_img_data\n",
"\n",
"def postprocess(result):\n",
" return softmax(np.array(result)).tolist()\n",
"\n",
"def run(input_data_json):\n",
" try:\n",
" start = time.time()\n",
" # load in our data which is expected as NCHW 224x224 image\n",
" input_data = preprocess(input_data_json)\n",
" input_name = session.get_inputs()[0].name # get the id of the first input of the model \n",
" result = session.run([], {input_name: input_data})\n",
" end = time.time() # stop timer\n",
" return {\"result\": postprocess(result),\n",
" \"time\": end - start}\n",
" except Exception as e:\n",
" result = str(e)\n",
" return {\"error\": result}"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create container image"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"First we create a YAML file that specifies which dependencies we would like to see in our container."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.conda_dependencies import CondaDependencies \n",
"\n",
"myenv = CondaDependencies.create(pip_packages=[\"numpy\",\"onnxruntime\",\"azureml-core\"])\n",
"\n",
"with open(\"myenv.yml\",\"w\") as f:\n",
" f.write(myenv.serialize_to_string())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Then we have Azure ML create the container. This step will likely take a few minutes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.image import ContainerImage\n",
"\n",
"image_config = ContainerImage.image_configuration(execution_script = \"score.py\",\n",
" runtime = \"python\",\n",
" conda_file = \"myenv.yml\",\n",
" description = \"ONNX ResNet50 Demo\",\n",
" tags = {\"demo\": \"onnx\"}\n",
" )\n",
"\n",
"\n",
"image = ContainerImage.create(name = \"onnxresnet50v2\",\n",
" models = [model],\n",
" image_config = image_config,\n",
" workspace = ws)\n",
"\n",
"image.wait_for_creation(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"In case you need to debug your code, the next line of code accesses the log file."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(image.image_build_log_uri)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We're all set! Let's get our model chugging.\n",
"\n",
"### Deploy the container image"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.webservice import AciWebservice\n",
"\n",
"aciconfig = AciWebservice.deploy_configuration(cpu_cores = 1, \n",
" memory_gb = 1, \n",
" tags = {'demo': 'onnx'}, \n",
" description = 'web service for ResNet50 ONNX model')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The following cell will likely take a few minutes to run as well."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.webservice import Webservice\n",
"from random import randint\n",
"\n",
"aci_service_name = 'onnx-demo-resnet50'+str(randint(0,100))\n",
"print(\"Service\", aci_service_name)\n",
"\n",
"aci_service = Webservice.deploy_from_image(deployment_config = aciconfig,\n",
" image = image,\n",
" name = aci_service_name,\n",
" workspace = ws)\n",
"\n",
"aci_service.wait_for_deployment(True)\n",
"print(aci_service.state)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"In case the deployment fails, you can check the logs. Make sure to delete your aci_service before trying again."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"if aci_service.state != 'Healthy':\n",
" # run this command for debugging.\n",
" print(aci_service.get_logs())\n",
" aci_service.delete()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Success!\n",
"\n",
"If you've made it this far, you've deployed a working web service that does image classification using an ONNX model. You can get the URL for the webservice with the code below."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(aci_service.scoring_uri)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"When you are eventually done using the web service, remember to delete it."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#aci_service.delete()"
]
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
"metadata": {
"authors": [
{
"name": "onnx"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.5.6"
}
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.5.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
"nbformat": 4,
"nbformat_minor": 2
}

667
onnx/onnx-train-pytorch-aml-deploy-mnist.ipynb Normal file
View File

File diff suppressed because one or more lines are too long

156
pipeline/00.pipeline-setup.ipynb
View File

@@ -1,81 +1,81 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Packages"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"!pip install pandas\n",
"!pip install requests"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Widgets\n",
"Install the following widgets to see the status of each run"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"!jupyter nbextension install --py --user azureml.widgets\n",
"!jupyter nbextension enable --py --user azureml.widgets"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"authors": [
{
"name": "hichando"
}
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Packages"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"!pip install pandas\n",
"!pip install requests"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Widgets\n",
"Install the following widgets to see the status of each run"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"!jupyter nbextension install --py --user azureml.train.widgets\n",
"!jupyter nbextension enable --py --user azureml.train.widgets"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
"metadata": {
"authors": [
{
"name": "hichando"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.3"
}
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.3"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
"nbformat": 4,
"nbformat_minor": 2
}

664
pipeline/pipeline-batch-scoring.ipynb Normal file
View File

@@ -0,0 +1,664 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"This notebook demonstrates how to run batch scoring job. __[Inception-V3 model](https://arxiv.org/abs/1512.00567)__ and unlabeled images from __[ImageNet](http://image-net.org/)__ dataset will be used. It registers a pretrained inception model in model registry then uses the model to do batch scoring on images in a blob container."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"Make sure you go through the [00. Installation and Configuration](./00.configuration.ipynb) Notebook first if you haven't.\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"from azureml.core import Workspace, Run, Experiment\n",
"\n",
"ws = Workspace.from_config()\n",
"print('Workspace name: ' + ws.name, \n",
" 'Azure region: ' + ws.location, \n",
" 'Subscription id: ' + ws.subscription_id, \n",
" 'Resource group: ' + ws.resource_group, sep = '\\n')\n",
"\n",
"# Also create a Project and attach to Workspace\n",
"scripts_folder = \"scripts\"\n",
"\n",
"if not os.path.isdir(scripts_folder):\n",
" os.mkdir(scripts_folder)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import BatchAiCompute, ComputeTarget\n",
"from azureml.core.datastore import Datastore\n",
"from azureml.data.data_reference import DataReference\n",
"from azureml.pipeline.core import Pipeline, PipelineData\n",
"from azureml.pipeline.steps import PythonScriptStep\n",
"from azureml.core.runconfig import CondaDependencies, RunConfiguration"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create and attach Compute targets\n",
"Use the below code to create and attach Compute targets. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"\n",
"# choose a name for your cluster\n",
"batchai_cluster_name = os.environ.get(\"BATCHAI_CLUSTER_NAME\", \"gpu-cluster\")\n",
"cluster_min_nodes = os.environ.get(\"BATCHAI_CLUSTER_MIN_NODES\", 0)\n",
"cluster_max_nodes = os.environ.get(\"BATCHAI_CLUSTER_MAX_NODES\", 1)\n",
"vm_size = os.environ.get(\"BATCHAI_CLUSTER_SKU\", \"STANDARD_NC6\")\n",
"autoscale_enabled = os.environ.get(\"BATCHAI_CLUSTER_AUTOSCALE_ENABLED\", True)\n",
"\n",
"\n",
"if batchai_cluster_name in ws.compute_targets:\n",
" compute_target = ws.compute_targets[batchai_cluster_name]\n",
" if compute_target and type(compute_target) is BatchAiCompute:\n",
" print('found compute target. just use it. ' + batchai_cluster_name)\n",
"else:\n",
" print('creating a new compute target...')\n",
" provisioning_config = BatchAiCompute.provisioning_configuration(vm_size = vm_size, # NC6 is GPU-enabled\n",
" vm_priority = 'lowpriority', # optional\n",
" autoscale_enabled = autoscale_enabled,\n",
" cluster_min_nodes = cluster_min_nodes, \n",
" cluster_max_nodes = cluster_max_nodes)\n",
"\n",
" # create the cluster\n",
" compute_target = ComputeTarget.create(ws, batchai_cluster_name, provisioning_config)\n",
" \n",
" # can poll for a minimum number of nodes and for a specific timeout. \n",
" # if no min node count is provided it will use the scale settings for the cluster\n",
" compute_target.wait_for_completion(show_output=True, min_node_count=None, timeout_in_minutes=20)\n",
" \n",
" # For a more detailed view of current BatchAI cluster status, use the 'status' property \n",
" print(compute_target.status.serialize())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Python scripts to run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Python scripts that run the batch scoring. `batchai_score.py` takes input images in `dataset_path`, pretrained models in `model_dir` and outputs a `results-label.txt` to `output_dir`."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile $scripts_folder/batchai_score.py\n",
"import os\n",
"import argparse\n",
"import datetime,time\n",
"import tensorflow as tf\n",
"from math import ceil\n",
"import numpy as np\n",
"import shutil\n",
"from tensorflow.contrib.slim.python.slim.nets import inception_v3\n",
"from azureml.core.model import Model\n",
"\n",
"slim = tf.contrib.slim\n",
"\n",
"parser = argparse.ArgumentParser(description=\"Start a tensorflow model serving\")\n",
"parser.add_argument('--model_name', dest=\"model_name\", required=True)\n",
"parser.add_argument('--label_dir', dest=\"label_dir\", required=True)\n",
"parser.add_argument('--dataset_path', dest=\"dataset_path\", required=True)\n",
"parser.add_argument('--output_dir', dest=\"output_dir\", required=True)\n",
"parser.add_argument('--batch_size', dest=\"batch_size\", type=int, required=True)\n",
"\n",
"args = parser.parse_args()\n",
"\n",
"image_size = 299\n",
"num_channel = 3\n",
"\n",
"# create output directory if it does not exist\n",
"os.makedirs(args.output_dir, exist_ok=True)\n",
"\n",
"def get_class_label_dict(label_file):\n",
" label = []\n",
" proto_as_ascii_lines = tf.gfile.GFile(label_file).readlines()\n",
" for l in proto_as_ascii_lines:\n",
" label.append(l.rstrip())\n",
" return label\n",
"\n",
"\n",
"class DataIterator:\n",
" def __init__(self, data_dir):\n",
" self.file_paths = []\n",
" image_list = os.listdir(data_dir)\n",
" total_size = len(image_list)\n",
" self.file_paths = [data_dir + '/' + file_name.rstrip() for file_name in image_list ]\n",
"\n",
" self.labels = [1 for file_name in self.file_paths]\n",
"\n",
" @property\n",
" def size(self):\n",
" return len(self.labels)\n",
"\n",
" def input_pipeline(self, batch_size):\n",
" images_tensor = tf.convert_to_tensor(self.file_paths, dtype=tf.string)\n",
" labels_tensor = tf.convert_to_tensor(self.labels, dtype=tf.int64)\n",
" input_queue = tf.train.slice_input_producer([images_tensor, labels_tensor], shuffle=False)\n",
" labels = input_queue[1]\n",
" images_content = tf.read_file(input_queue[0])\n",
"\n",
" image_reader = tf.image.decode_jpeg(images_content, channels=num_channel, name=\"jpeg_reader\")\n",
" float_caster = tf.cast(image_reader, tf.float32)\n",
" new_size = tf.constant([image_size, image_size], dtype=tf.int32)\n",
" images = tf.image.resize_images(float_caster, new_size)\n",
" images = tf.divide(tf.subtract(images, [0]), [255])\n",
"\n",
" image_batch, label_batch = tf.train.batch([images, labels], batch_size=batch_size, capacity=5 * batch_size)\n",
" return image_batch\n",
"\n",
"def main(_):\n",
" start_time = datetime.datetime.now()\n",
" label_file_name = os.path.join(args.label_dir, \"labels.txt\")\n",
" label_dict = get_class_label_dict(label_file_name)\n",
" classes_num = len(label_dict)\n",
" test_feeder = DataIterator(data_dir=args.dataset_path)\n",
" total_size = len(test_feeder.labels)\n",
" count = 0\n",
" # get model from model registry\n",
" model_path = Model.get_model_path(args.model_name)\n",
" with tf.Session() as sess:\n",
" test_images = test_feeder.input_pipeline(batch_size=args.batch_size)\n",
" with slim.arg_scope(inception_v3.inception_v3_arg_scope()):\n",
" input_images = tf.placeholder(tf.float32, [args.batch_size, image_size, image_size, num_channel])\n",
" logits, _ = inception_v3.inception_v3(input_images,\n",
" num_classes=classes_num,\n",
" is_training=False)\n",
" probabilities = tf.argmax(logits, 1)\n",
"\n",
" sess.run(tf.global_variables_initializer())\n",
" sess.run(tf.local_variables_initializer())\n",
" coord = tf.train.Coordinator()\n",
" threads = tf.train.start_queue_runners(sess=sess, coord=coord)\n",
" saver = tf.train.Saver()\n",
" saver.restore(sess, model_path)\n",
" out_filename = os.path.join(args.output_dir, \"result-labels.txt\")\n",
" with open(out_filename, \"w\") as result_file:\n",
" i = 0\n",
" while count < total_size and not coord.should_stop():\n",
" test_images_batch = sess.run(test_images)\n",
" file_names_batch = test_feeder.file_paths[i*args.batch_size: min(test_feeder.size, (i+1)*args.batch_size)]\n",
" results = sess.run(probabilities, feed_dict={input_images: test_images_batch})\n",
" new_add = min(args.batch_size, total_size-count)\n",
" count += new_add\n",
" i += 1\n",
" for j in range(new_add):\n",
" result_file.write(os.path.basename(file_names_batch[j]) + \": \" + label_dict[results[j]] + \"\\n\")\n",
" result_file.flush()\n",
" coord.request_stop()\n",
" coord.join(threads)\n",
" \n",
" # copy the file to artifacts\n",
" shutil.copy(out_filename, \"./outputs/\")\n",
" # Move the processed data out of the blob so that the next run can process the data.\n",
"\n",
"if __name__ == \"__main__\":\n",
" tf.app.run()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prepare Model and Input data"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Download Model\n",
"\n",
"Download and extract model from http://download.tensorflow.org/models/inception_v3_2016_08_28.tar.gz to `\"models\"`"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# create directory for model\n",
"model_dir = 'models'\n",
"if not os.path.isdir(model_dir):\n",
" os.mkdir(model_dir)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import tarfile\n",
"import urllib.request\n",
"\n",
"url=\"http://download.tensorflow.org/models/inception_v3_2016_08_28.tar.gz\"\n",
"response = urllib.request.urlretrieve(url, \"model.tar.gz\")\n",
"tar = tarfile.open(\"model.tar.gz\", \"r:gz\")\n",
"tar.extractall(model_dir)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create a datastore that points to blob container containing sample images\n",
"\n",
"We have created a public blob container `sampledata` on an account named `pipelinedata` containing images from ImageNet evaluation set. In the next step, we create a datastore with name `images_datastore` that points to this container. The `overwrite=True` step overwrites any datastore that was created previously with that name. \n",
"\n",
"This step can be changed to point to your blob container by providing an additional `account_key` parameter with `account_name`. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"account_name = \"pipelinedata\"\n",
"sample_data = Datastore.register_azure_blob_container(ws, datastore_name=\"images_datastore\", container_name=\"sampledata\", \n",
" account_name=account_name, \n",
" overwrite=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Output datastore"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We write the outputs to the default datastore"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"default_ds = ws.get_default_datastore()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Specify where the data is stored or will be written to"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.conda_dependencies import CondaDependencies\n",
"from azureml.data.data_reference import DataReference\n",
"from azureml.pipeline.core import Pipeline, PipelineData\n",
"from azureml.core import Datastore\n",
"from azureml.core import Experiment"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"input_images = DataReference(datastore=sample_data, \n",
" data_reference_name=\"input_images\",\n",
" path_on_datastore=\"batchscoring/images\",\n",
" mode=\"download\"\n",
" )\n",
"model_dir = DataReference(datastore=sample_data, \n",
" data_reference_name=\"input_model\",\n",
" path_on_datastore=\"batchscoring/models\",\n",
" mode=\"download\" \n",
" )\n",
"label_dir = DataReference(datastore=sample_data, \n",
" data_reference_name=\"input_labels\",\n",
" path_on_datastore=\"batchscoring/labels\",\n",
" mode=\"download\" \n",
" )\n",
"output_dir = PipelineData(name=\"scores\", \n",
" datastore=default_ds, \n",
" output_path_on_compute=\"batchscoring/results\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Register the model with Workspace"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import shutil\n",
"from azureml.core.model import Model\n",
"\n",
"# register downloaded model \n",
"model = Model.register(model_path = \"models/inception_v3.ckpt\",\n",
" model_name = \"inception\", # this is the name the model is registered as\n",
" tags = {'pretrained': \"inception\"},\n",
" description = \"Imagenet trained tensorflow inception\",\n",
" workspace = ws)\n",
"# remove the downloaded dir after registration if you wish\n",
"shutil.rmtree(\"models\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Specify environment to run the script"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"cd = CondaDependencies.create(pip_packages=[\"tensorflow-gpu==1.4.0\", \"azureml-defaults\"])\n",
"\n",
"# Runconfig\n",
"batchai_run_config = RunConfiguration(conda_dependencies=cd)\n",
"batchai_run_config.environment.docker.enabled = True\n",
"batchai_run_config.environment.docker.gpu_support = True\n",
"batchai_run_config.environment.docker.base_image = \"microsoft/mmlspark:gpu-0.12\"\n",
"batchai_run_config.environment.spark.precache_packages = False"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Steps to run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"A subset of the parameters to the python script can be given as input when we re-run a `PublishedPipeline`. In the current example, we define `batch_size` taken by the script as such parameter."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.pipeline.core.graph import PipelineParameter\n",
"batch_size_param = PipelineParameter(name=\"param_batch_size\", default_value=20)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"inception_model_name = \"inception_v3.ckpt\"\n",
"\n",
"batch_score_step = PythonScriptStep(\n",
" name=\"batch ai scoring\",\n",
" script_name=\"batchai_score.py\",\n",
" arguments=[\"--dataset_path\", input_images, \n",
" \"--model_name\", \"inception\",\n",
" \"--label_dir\", label_dir, \n",
" \"--output_dir\", output_dir, \n",
" \"--batch_size\", batch_size_param],\n",
" target=compute_target,\n",
" inputs=[input_images, label_dir],\n",
" outputs=[output_dir],\n",
" runconfig=batchai_run_config,\n",
" source_directory=scripts_folder\n",
")"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"pipeline = Pipeline(workspace=ws, steps=[batch_score_step])\n",
"pipeline_run = Experiment(ws, 'batch_scoring').submit(pipeline, pipeline_params={\"param_batch_size\": 20})"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Monitor run"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.widgets import RunDetails\n",
"RunDetails(pipeline_run).show()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"pipeline_run.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Download and review output"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"step_run = list(pipeline_run.get_children())[0]\n",
"step_run.download_file(\"./outputs/result-labels.txt\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import pandas as pd\n",
"df = pd.read_csv(\"result-labels.txt\", delimiter=\":\", header=None)\n",
"df.columns = [\"Filename\", \"Prediction\"]\n",
"df.head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Publish a pipeline and rerun using a REST call"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create a published pipeline"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"published_pipeline = pipeline_run.publish_pipeline(\n",
" name=\"Inception v3 scoring\", description=\"Batch scoring using Inception v3 model\", version=\"1.0\")\n",
"\n",
"published_id = published_pipeline.id"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Rerun using REST call"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Get AAD token"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.authentication import AzureCliAuthentication\n",
"import requests\n",
"\n",
"cli_auth = AzureCliAuthentication()\n",
"aad_token = cli_auth.get_authentication_header()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Run published pipeline using its REST endpoint"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.pipeline.core import PublishedPipeline\n",
"\n",
"rest_endpoint = published_pipeline.endpoint\n",
"# specify batch size when running the pipeline\n",
"response = requests.post(rest_endpoint, \n",
" headers=aad_token, \n",
" json={\"ExperimentName\": \"batch_scoring\",\n",
" \"ParameterAssignments\": {\"param_batch_size\": 50}})\n",
"run_id = response.json()[\"Id\"]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Monitor the new run"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.pipeline.core.run import PipelineRun\n",
"published_pipeline_run = PipelineRun(ws.experiments[\"batch_scoring\"], run_id)\n",
"\n",
"RunDetails(published_pipeline_run).show()"
]
}
],
"metadata": {
"authors": [
{
"name": "hichando"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

625
project-brainwave/project-brainwave-custom-weights.ipynb Normal file
View File

@@ -0,0 +1,625 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Model Development with Custom Weights"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"This example shows how to retrain a model with custom weights and fine-tune the model with quantization, then deploy the model running on FPGA. Only Windows is supported. We use TensorFlow and Keras to build our model. We are going to use transfer learning, with ResNet50 as a featurizer. We don't use the last layer of ResNet50 in this case and instead add our own classification layer using Keras.\n",
"\n",
"The custom wegiths are trained with ImageNet on ResNet50. We will use the Kaggle Cats and Dogs dataset to retrain and fine-tune the model. The dataset can be downloaded [here](https://www.microsoft.com/en-us/download/details.aspx?id=54765). Download the zip and extract to a directory named 'catsanddogs' under your user directory (\"~/catsanddogs\"). \n",
"\n",
"Please set up your environment as described in the [quick start](project-brainwave-quickstart.ipynb)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"import sys\n",
"import tensorflow as tf\n",
"import numpy as np\n",
"from keras import backend as K"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Setup Environment\n",
"After you train your model in float32, you'll write the weights to a place on disk. We also need a location to store the models that get downloaded."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"custom_weights_dir = os.path.expanduser(\"~/custom-weights\")\n",
"saved_model_dir = os.path.expanduser(\"~/models\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prepare Data\n",
"Load the files we are going to use for training and testing. By default this notebook uses only a very small subset of the Cats and Dogs dataset. That makes it run relatively quickly."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import glob\n",
"import imghdr\n",
"datadir = os.path.expanduser(\"~/catsanddogs\")\n",
"\n",
"cat_files = glob.glob(os.path.join(datadir, 'PetImages', 'Cat', '*.jpg'))\n",
"dog_files = glob.glob(os.path.join(datadir, 'PetImages', 'Dog', '*.jpg'))\n",
"\n",
"# Limit the data set to make the notebook execute quickly.\n",
"cat_files = cat_files[:64]\n",
"dog_files = dog_files[:64]\n",
"\n",
"# The data set has a few images that are not jpeg. Remove them.\n",
"cat_files = [f for f in cat_files if imghdr.what(f) == 'jpeg']\n",
"dog_files = [f for f in dog_files if imghdr.what(f) == 'jpeg']\n",
"\n",
"if(not len(cat_files) or not len(dog_files)):\n",
" print(\"Please download the Kaggle Cats and Dogs dataset form https://www.microsoft.com/en-us/download/details.aspx?id=54765 and extract the zip to \" + datadir) \n",
" raise ValueError(\"Data not found\")\n",
"else:\n",
" print(cat_files[0])\n",
" print(dog_files[0])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Construct a numpy array as labels\n",
"image_paths = cat_files + dog_files\n",
"total_files = len(cat_files) + len(dog_files)\n",
"labels = np.zeros(total_files)\n",
"labels[len(cat_files):] = 1"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Split images data as training data and test data\n",
"from sklearn.model_selection import train_test_split\n",
"onehot_labels = np.array([[0,1] if i else [1,0] for i in labels])\n",
"img_train, img_test, label_train, label_test = train_test_split(image_paths, onehot_labels, random_state=42, shuffle=True)\n",
"\n",
"print(len(img_train), len(img_test), label_train.shape, label_test.shape)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Construct Model\n",
"We use ResNet50 for the featuirzer and build our own classifier using Keras layers. We train the featurizer and the classifier as one model. The weights trained on ImageNet are used as the starting point for the retraining of our featurizer. The weights are loaded from tensorflow chkeckpoint files."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Before passing image dataset to the ResNet50 featurizer, we need to preprocess the input file to get it into the form expected by ResNet50. ResNet50 expects float tensors representing the images in BGR, channel last order. We've provided a default implementation of the preprocessing that you can use."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import azureml.contrib.brainwave.models.utils as utils\n",
"\n",
"def preprocess_images():\n",
" # Convert images to 3D tensors [width,height,channel] - channels are in BGR order.\n",
" in_images = tf.placeholder(tf.string)\n",
" image_tensors = utils.preprocess_array(in_images)\n",
" return in_images, image_tensors"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We use Keras layer APIs to construct the classifier. Because we're using the tensorflow backend, we can train this classifier in one session with our Resnet50 model."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"def construct_classifier(in_tensor):\n",
" from keras.layers import Dropout, Dense, Flatten\n",
" K.set_session(tf.get_default_session())\n",
" \n",
" FC_SIZE = 1024\n",
" NUM_CLASSES = 2\n",
"\n",
" x = Dropout(0.2, input_shape=(1, 1, 2048,))(in_tensor)\n",
" x = Dense(FC_SIZE, activation='relu', input_dim=(1, 1, 2048,))(x)\n",
" x = Flatten()(x)\n",
" preds = Dense(NUM_CLASSES, activation='softmax', input_dim=FC_SIZE, name='classifier_output')(x)\n",
" return preds"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now every component of the model is defined, we can construct the model. Constructing the model with the project brainwave models is two steps - first we import the graph definition, then we restore the weights of the model into a tensorflow session. Because the quantized graph defintion and the float32 graph defintion share the same node names in the graph definitions, we can initally train the weights in float32, and then reload them with the quantized operations (which take longer) to fine-tune the model."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"def construct_model(quantized, starting_weights_directory = None):\n",
" from azureml.contrib.brainwave.models import Resnet50, QuantizedResnet50\n",
" \n",
" # Convert images to 3D tensors [width,height,channel]\n",
" in_images, image_tensors = preprocess_images()\n",
"\n",
" # Construct featurizer using quantized or unquantized ResNet50 model\n",
" if not quantized:\n",
" featurizer = Resnet50(saved_model_dir)\n",
" else:\n",
" featurizer = QuantizedResnet50(saved_model_dir, custom_weights_directory = starting_weights_directory)\n",
"\n",
"\n",
" features = featurizer.import_graph_def(input_tensor=image_tensors)\n",
" # Construct classifier\n",
" preds = construct_classifier(features)\n",
" \n",
" # Initialize weights\n",
" sess = tf.get_default_session()\n",
" tf.global_variables_initializer().run()\n",
"\n",
" featurizer.restore_weights(sess)\n",
"\n",
" return in_images, image_tensors, features, preds, featurizer"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train Model\n",
"First we train the model with custom weights but without quantization. Training is done with native float precision (32-bit floats). We load the traing data set and batch the training with 10 epochs. When the performance reaches desired level or starts decredation, we stop the training iteration and save the weights as tensorflow checkpoint files. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"def read_files(files):\n",
" \"\"\" Read files to array\"\"\"\n",
" contents = []\n",
" for path in files:\n",
" with open(path, 'rb') as f:\n",
" contents.append(f.read())\n",
" return contents"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"def train_model(preds, in_images, img_train, label_train, is_retrain = False, train_epoch = 10):\n",
" \"\"\" training model \"\"\"\n",
" from keras.objectives import binary_crossentropy\n",
" from tqdm import tqdm\n",
" \n",
" learning_rate = 0.001 if is_retrain else 0.01\n",
" \n",
" # Specify the loss function\n",
" in_labels = tf.placeholder(tf.float32, shape=(None, 2)) \n",
" cross_entropy = tf.reduce_mean(binary_crossentropy(in_labels, preds))\n",
" optimizer = tf.train.GradientDescentOptimizer(learning_rate).minimize(cross_entropy)\n",
"\n",
" def chunks(a, b, n):\n",
" \"\"\"Yield successive n-sized chunks from a and b.\"\"\"\n",
" if (len(a) != len(b)):\n",
" print(\"a and b are not equal in chunks(a,b,n)\")\n",
" raise ValueError(\"Parameter error\")\n",
"\n",
" for i in range(0, len(a), n):\n",
" yield a[i:i + n], b[i:i + n]\n",
"\n",
" chunk_size = 16\n",
" chunk_num = len(label_train) / chunk_size\n",
"\n",
" sess = tf.get_default_session()\n",
" for epoch in range(train_epoch):\n",
" avg_loss = 0\n",
" for img_chunk, label_chunk in tqdm(chunks(img_train, label_train, chunk_size)):\n",
" contents = read_files(img_chunk)\n",
" _, loss = sess.run([optimizer, cross_entropy],\n",
" feed_dict={in_images: contents,\n",
" in_labels: label_chunk,\n",
" K.learning_phase(): 1})\n",
" avg_loss += loss / chunk_num\n",
" print(\"Epoch:\", (epoch + 1), \"loss = \", \"{:.3f}\".format(avg_loss))\n",
" \n",
" # Reach desired performance\n",
" if (avg_loss < 0.001):\n",
" break"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"def test_model(preds, in_images, img_test, label_test):\n",
" \"\"\"Test the model\"\"\"\n",
" from keras.metrics import categorical_accuracy\n",
"\n",
" in_labels = tf.placeholder(tf.float32, shape=(None, 2))\n",
" accuracy = tf.reduce_mean(categorical_accuracy(in_labels, preds))\n",
" contents = read_files(img_test)\n",
"\n",
" accuracy = accuracy.eval(feed_dict={in_images: contents,\n",
" in_labels: label_test,\n",
" K.learning_phase(): 0})\n",
" return accuracy"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Launch the training\n",
"tf.reset_default_graph()\n",
"sess = tf.Session(graph=tf.get_default_graph())\n",
"\n",
"with sess.as_default():\n",
" in_images, image_tensors, features, preds, featurizer = construct_model(quantized=False)\n",
" train_model(preds, in_images, img_train, label_train, is_retrain=False, train_epoch=10) \n",
" accuracy = test_model(preds, in_images, img_test, label_test) \n",
" print(\"Accuracy:\", accuracy)\n",
" featurizer.save_weights(custom_weights_dir + \"/rn50\", tf.get_default_session())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Test Model\n",
"After training, we evaluate the trained model's accuracy on test dataset with quantization. So that we know the model's performance if it is deployed on the FPGA."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"tf.reset_default_graph()\n",
"sess = tf.Session(graph=tf.get_default_graph())\n",
"\n",
"with sess.as_default():\n",
" print(\"Testing trained model with quantization\")\n",
" in_images, image_tensors, features, preds, quantized_featurizer = construct_model(quantized=True, starting_weights_directory=custom_weights_dir)\n",
" accuracy = test_model(preds, in_images, img_test, label_test) \n",
" print(\"Accuracy:\", accuracy)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Fine-Tune Model\n",
"Sometimes, the model's accuracy can drop significantly after quantization. In those cases, we need to retrain the model enabled with quantization to get better model accuracy."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"if (accuracy < 0.93):\n",
" with sess.as_default():\n",
" print(\"Fine-tuning model with quantization\")\n",
" train_model(preds, in_images, img_train, label_train, is_retrain=True, train_epoch=10)\n",
" accuracy = test_model(preds, in_images, img_test, label_test) \n",
" print(\"Accuracy:\", accuracy)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Service Definition\n",
"Like in the QuickStart notebook our service definition pipeline consists of three stages. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.contrib.brainwave.pipeline import ModelDefinition, TensorflowStage, BrainWaveStage\n",
"\n",
"model_def_path = os.path.join(saved_model_dir, 'model_def.zip')\n",
"\n",
"model_def = ModelDefinition()\n",
"model_def.pipeline.append(TensorflowStage(sess, in_images, image_tensors))\n",
"model_def.pipeline.append(BrainWaveStage(sess, quantized_featurizer))\n",
"model_def.pipeline.append(TensorflowStage(sess, features, preds))\n",
"model_def.save(model_def_path)\n",
"print(model_def_path)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Deploy\n",
"Go to our [GitHub repo](https://aka.ms/aml-real-time-ai) \"docs\" folder to learn how to create a Model Management Account and find the required information below."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Workspace\n",
"\n",
"ws = Workspace.from_config()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The first time the code below runs it will create a new service running your model. If you want to change the model you can make changes above in this notebook and save a new service definition. Then this code will update the running service in place to run the new model."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.model import Model\n",
"from azureml.core.image import Image\n",
"from azureml.core.webservice import Webservice\n",
"from azureml.contrib.brainwave import BrainwaveWebservice, BrainwaveImage\n",
"from azureml.exceptions import WebserviceException\n",
"\n",
"model_name = \"catsanddogs-resnet50-model\"\n",
"image_name = \"catsanddogs-resnet50-image\"\n",
"service_name = \"modelbuild-service\"\n",
"\n",
"registered_model = Model.register(ws, model_def_path, model_name)\n",
"\n",
"image_config = BrainwaveImage.image_configuration()\n",
"deployment_config = BrainwaveWebservice.deploy_configuration()\n",
" \n",
"try:\n",
" service = Webservice(ws, service_name)\n",
" service.delete()\n",
" service = Webservice.deploy_from_model(ws, service_name, [registered_model], image_config, deployment_config)\n",
" service.wait_for_deployment(True)\n",
"except WebserviceException:\n",
" service = Webservice.deploy_from_model(ws, service_name, [registered_model], image_config, deployment_config)\n",
" service.wait_for_deployment(True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The service is now running in Azure and ready to serve requests. We can check the address and port."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(service.ipAddress + ':' + str(service.port))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Client\n",
"There is a simple test client at amlrealtimeai.PredictionClient which can be used for testing. We'll use this client to score an image with our new service."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.contrib.brainwave.client import PredictionClient\n",
"client = PredictionClient(service.ipAddress, service.port)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can adapt the client [code](../../pythonlib/amlrealtimeai/client.py) to meet your needs. There is also an example C# [client](../../sample-clients/csharp).\n",
"\n",
"The service provides an API that is compatible with TensorFlow Serving. There are instructions to download a sample client [here](https://www.tensorflow.org/serving/setup)."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Request\n",
"Let's see how our service does on a few images. It may get a few wrong."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Specify an image to classify\n",
"print('CATS')\n",
"for image_file in cat_files[:8]:\n",
" results = client.score_image(image_file)\n",
" result = 'CORRECT ' if results[0] > results[1] else 'WRONG '\n",
" print(result + str(results))\n",
"print('DOGS')\n",
"for image_file in dog_files[:8]:\n",
" results = client.score_image(image_file)\n",
" result = 'CORRECT ' if results[1] > results[0] else 'WRONG '\n",
" print(result + str(results))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Cleanup\n",
"Run the cell below to delete your service."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"service.delete()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Appendix"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"License for plot_confusion_matrix:\n",
"\n",
"New BSD License\n",
"\n",
"Copyright (c) 2007-2018 The scikit-learn developers.\n",
"All rights reserved.\n",
"\n",
"\n",
"Redistribution and use in source and binary forms, with or without\n",
"modification, are permitted provided that the following conditions are met:\n",
"\n",
" a. Redistributions of source code must retain the above copyright notice,\n",
" this list of conditions and the following disclaimer.\n",
" b. Redistributions in binary form must reproduce the above copyright\n",
" notice, this list of conditions and the following disclaimer in the\n",
" documentation and/or other materials provided with the distribution.\n",
" c. Neither the name of the Scikit-learn Developers nor the names of\n",
" its contributors may be used to endorse or promote products\n",
" derived from this software without specific prior written\n",
" permission. \n",
"\n",
"\n",
"THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS \"AS IS\"\n",
"AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE\n",
"IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE\n",
"ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE FOR\n",
"ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL\n",
"DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR\n",
"SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER\n",
"CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT\n",
"LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY\n",
"OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH\n",
"DAMAGE.\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"authors": [
{
"name": "coverste"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.5.2"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

312
project-brainwave/project-brainwave-quickstart.ipynb Normal file
View File

@@ -0,0 +1,312 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Azure ML Hardware Accelerated Models Quickstart"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"This tutorial will show you how to deploy an image recognition service based on the ResNet 50 classifier in just a few minutes using the Azure Machine Learning Accelerated AI service. Get more help from our [documentation](https://aka.ms/aml-real-time-ai) or [forum](https://aka.ms/aml-forum).\n",
"\n",
"We will use an accelerated ResNet50 featurizer running on an FPGA. This functionality is powered by Project Brainwave, which handles translating deep neural networks (DNN) into an FPGA program.\n",
"\n",
"## Request Quota\n",
"**IMPORTANT:** You must [request quota](https://aka.ms/aml-real-time-ai-request) and be approved before you can successfully run this notebook. Notebook 00 will show you how to create a workspace which you can use to request quota."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Imports"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"import tensorflow as tf"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Image preprocessing\n",
"We'd like our service to accept JPEG images as input. However the input to ResNet50 is a tensor. So we need code that decodes JPEG images and does the preprocessing required by ResNet50. The Accelerated AI service can execute TensorFlow graphs as part of the service and we'll use that ability to do the image preprocessing. This code defines a TensorFlow graph that preprocesses an array of JPEG images (as strings) and produces a tensor that is ready to be featurized by ResNet50."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Input images as a two-dimensional tensor containing an arbitrary number of images represented a strings\n",
"import azureml.contrib.brainwave.models.utils as utils\n",
"in_images = tf.placeholder(tf.string)\n",
"image_tensors = utils.preprocess_array(in_images)\n",
"print(image_tensors.shape)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Featurizer\n",
"We use ResNet50 as a featurizer. In this step we initialize the model. This downloads a TensorFlow checkpoint of the quantized ResNet50."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.contrib.brainwave.models import QuantizedResnet50\n",
"model_path = os.path.expanduser('~/models')\n",
"model = QuantizedResnet50(model_path, is_frozen = True)\n",
"feature_tensor = model.import_graph_def(image_tensors)\n",
"print(model.version)\n",
"print(feature_tensor.name)\n",
"print(feature_tensor.shape)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Classifier\n",
"The model we downloaded includes a classifier which takes the output of the ResNet50 and identifies an image. This classifier is trained on the ImageNet dataset. We are going to use this classifier for our service. The next [notebook](project-brainwave-trainsfer-learning.ipynb) shows how to train a classifier for a different data set. The input to the classifier is a tensor matching the output of our ResNet50 featurizer."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"classifier_output = model.get_default_classifier(feature_tensor)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Service Definition\n",
"Now that we've definied the image preprocessing, featurizer, and classifier that we will execute on our service we can create a service definition. The service definition is a set of files generated from the model that allow us to deploy to the FPGA service. The service definition consists of a pipeline. The pipeline is a series of stages that are executed in order. We support TensorFlow stages, Keras stages, and BrainWave stages. The stages will be executed in order on the service, with the output of each stage input into the subsequent stage.\n",
"\n",
"To create a TensorFlow stage we specify a session containing the graph (in this case we are using the default graph) and the input and output tensors to this stage. We use this information to save the graph so that we can execute it on the service."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.contrib.brainwave.pipeline import ModelDefinition, TensorflowStage, BrainWaveStage\n",
"\n",
"save_path = os.path.expanduser('~/models/save')\n",
"model_def_path = os.path.join(save_path, 'model_def.zip')\n",
"\n",
"model_def = ModelDefinition()\n",
"with tf.Session() as sess:\n",
" model_def.pipeline.append(TensorflowStage(sess, in_images, image_tensors))\n",
" model_def.pipeline.append(BrainWaveStage(sess, model))\n",
" model_def.pipeline.append(TensorflowStage(sess, feature_tensor, classifier_output))\n",
" model_def.save(model_def_path)\n",
" print(model_def_path)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Deploy\n",
"Time to create a service from the service definition. You need a Workspace in the **East US 2** location. In the previous notebooks, you've created this Workspace. The code below will load that Workspace from a configuration file."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print(ws.name, ws.resource_group, ws.location, ws.subscription_id, sep = '\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Upload the model to the workspace."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.model import Model\n",
"model_name = \"resnet-50-rtai\"\n",
"registered_model = Model.register(ws, model_def_path, model_name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Create a service from the model that we registered. If this is a new service then we create it. If you already have a service with this name then the existing service will be updated to use this model."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.webservice import Webservice\n",
"from azureml.exceptions import WebserviceException\n",
"from azureml.contrib.brainwave import BrainwaveWebservice, BrainwaveImage\n",
"service_name = \"imagenet-infer\"\n",
"service = None\n",
"try:\n",
" service = Webservice(ws, service_name)\n",
"except WebserviceException:\n",
" image_config = BrainwaveImage.image_configuration()\n",
" deployment_config = BrainwaveWebservice.deploy_configuration()\n",
" service = Webservice.deploy_from_model(ws, service_name, [registered_model], image_config, deployment_config)\n",
" service.wait_for_deployment(True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Client\n",
"The service supports gRPC and the TensorFlow Serving \"predict\" API. We provide a client that can call the service to get predictions on aka.ms/rtai. You can also invoke the service like any other web service."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"To understand the results we need a mapping to the human readable imagenet classes"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import requests\n",
"classes_entries = requests.get(\"https://raw.githubusercontent.com/Lasagne/Recipes/master/examples/resnet50/imagenet_classes.txt\").text.splitlines()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We can now send an image to the service and get the predictions. Let's see if it can identify a snow leopard.\n",
"![title](snowleopardgaze.jpg)\n",
"Snow leopard in a zoo. Photo by Peter Bolliger.\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"results = service.run('snowleopardgaze.jpg')\n",
"# map results [class_id] => [confidence]\n",
"results = enumerate(results)\n",
"# sort results by confidence\n",
"sorted_results = sorted(results, key=lambda x: x[1], reverse=True)\n",
"# print top 5 results\n",
"for top in sorted_results[:5]:\n",
" print(classes_entries[top[0]], 'confidence:', top[1])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Cleanup\n",
"Run the cell below to delete your service."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"service.delete()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Congratulations! You've just created a service that does predictions using an FPGA. The next [notebook](project-brainwave-trainsfer-learning.ipynb) shows how to customize the service using transfer learning to classify different types of images."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"authors": [
{
"name": "coverste"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.5.2"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

572
project-brainwave/project-brainwave-transfer-learning.ipynb Normal file
View File

@@ -0,0 +1,572 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Model Development"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"This example shows how to build, train, evaluate and deploy a model running on FPGA. Only Windows is supported. We use TensorFlow and Keras to build our model. We are going to use transfer learning, with ResNet152 as a featurizer. We don't use the last layer of ResNet152 in this case and instead add and train our own classification layer.\n",
"\n",
"We will use the Kaggle Cats and Dogs dataset to train the classifier. The dataset can be downloaded [here](https://www.microsoft.com/en-us/download/details.aspx?id=54765). Download the zip and extract to a directory named 'catsanddogs' under your user directory (\"~/catsanddogs\").\n",
"\n",
"Please set up your environment as described in the [quick start](project-brainwave-quickstart.ipynb)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"import tensorflow as tf\n",
"import numpy as np"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Model Construction\n",
"Load the files we are going to use for training and testing. By default this notebook uses only a very small subset of the Cats and Dogs dataset. That makes it run quickly, but doesn't create a very accurate classifier. You can improve the classifier by using more of the dataset."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import glob\n",
"import imghdr\n",
"datadir = os.path.expanduser(\"~/catsanddogs\")\n",
"\n",
"cat_files = glob.glob(os.path.join(datadir, 'PetImages', 'Cat', '*.jpg'))\n",
"dog_files = glob.glob(os.path.join(datadir, 'PetImages', 'Dog', '*.jpg'))\n",
"\n",
"# Limit the data set to make the notebook execute quickly.\n",
"cat_files = cat_files[:64]\n",
"dog_files = dog_files[:64]\n",
"\n",
"# The data set has a few images that are not jpeg. Remove them.\n",
"cat_files = [f for f in cat_files if imghdr.what(f) == 'jpeg']\n",
"dog_files = [f for f in dog_files if imghdr.what(f) == 'jpeg']\n",
"\n",
"if(not len(cat_files) or not len(dog_files)):\n",
" print(\"Please download the Kaggle Cats and Dogs dataset form https://www.microsoft.com/en-us/download/details.aspx?id=54765 and extract the zip to \" + datadir) \n",
" raise ValueError(\"Data not found\")\n",
"else:\n",
" print(cat_files[0])\n",
" print(dog_files[0])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# constructing a numpy array as labels\n",
"image_paths = cat_files + dog_files\n",
"total_files = len(cat_files) + len(dog_files)\n",
"labels = np.zeros(total_files)\n",
"labels[len(cat_files):] = 1"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We need to preprocess the input file to get it into the form expected by ResNet152. We've provided a default implementation of the preprocessing that you can use."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Input images as a two-dimensional tensor containing an arbitrary number of images represented a strings\n",
"import azureml.contrib.brainwave.models.utils as utils\n",
"in_images = tf.placeholder(tf.string)\n",
"image_tensors = utils.preprocess_array(in_images)\n",
"print(image_tensors.shape)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Alternatively, if you would like to customize the preprocessing, you can write your own preprocessor using TensorFlow operations.\n",
"\n",
"The input to the classifier we are training is the set of features produced by ResNet50. To train the classifier we need to \n",
"featurize the images using ResNet50. You can also run the featurizer locally on CPU or GPU. We import the featurizer as frozen, so that we are only training the classifier."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.contrib.brainwave.models import QuantizedResnet152\n",
"model_path = os.path.expanduser('~/models')\n",
"bwmodel = QuantizedResnet152(model_path, is_frozen = True)\n",
"print(bwmodel.version)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Calling import_graph_def on the featurizer will create a service that runs the featurizer on FPGA."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"features = bwmodel.import_graph_def(input_tensor=image_tensors)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Pre-compute features\n",
"Load the data set and compute the features. These can be precomputed because they don't change during training. This can take a while to run on CPU."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from tqdm import tqdm\n",
"\n",
"def chunks(l, n):\n",
" \"\"\"Yield successive n-sized chunks from l.\"\"\"\n",
" for i in range(0, len(l), n):\n",
" yield l[i:i + n]\n",
"\n",
"def read_files(files):\n",
" contents = []\n",
" for path in files:\n",
" with open(path, 'rb') as f:\n",
" contents.append(f.read())\n",
" return contents\n",
" \n",
"feature_list = []\n",
"with tf.Session() as sess:\n",
" for chunk in tqdm(chunks(image_paths, 5)):\n",
" contents = read_files(chunk)\n",
" result = sess.run([features], feed_dict={in_images: contents})\n",
" feature_list.extend(result[0])\n",
"\n",
"feature_results = np.array(feature_list)\n",
"print(feature_results.shape)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Add and Train the classifier\n",
"We use Keras to define and train a simple classifier."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from keras.models import Sequential\n",
"from keras.layers import Dropout, Dense, Flatten\n",
"from keras import optimizers\n",
"\n",
"FC_SIZE = 1024\n",
"NUM_CLASSES = 2\n",
"\n",
"model = Sequential()\n",
"model.add(Dropout(0.2, input_shape=(1, 1, 2048,)))\n",
"model.add(Dense(FC_SIZE, activation='relu', input_dim=(1, 1, 2048,)))\n",
"model.add(Flatten())\n",
"model.add(Dense(NUM_CLASSES, activation='sigmoid', input_dim=FC_SIZE))\n",
"\n",
"model.compile(optimizer=optimizers.SGD(lr=1e-4,momentum=0.9), loss='binary_crossentropy', metrics=['accuracy'])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Prepare the train and test data."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.model_selection import train_test_split\n",
"onehot_labels = np.array([[0,1] if i else [1,0] for i in labels])\n",
"X_train, X_test, y_train, y_test = train_test_split(feature_results, onehot_labels, random_state=42, shuffle=True)\n",
"print(X_train.shape, X_test.shape, y_train.shape, y_test.shape)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Train the classifier."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"model.fit(X_train, y_train, epochs=16, batch_size=32)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Test the Classifier\n",
"Let's test the classifier and see how well it does. Since we only trained on a few images, we are not expecting to win a Kaggle competition, but it will likely get most of the images correct. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from numpy import argmax\n",
"\n",
"y_probs = model.predict(X_test)\n",
"y_prob_max = np.argmax(y_probs, 1)\n",
"y_test_max = np.argmax(y_test, 1)\n",
"print(y_prob_max)\n",
"print(y_test_max)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.metrics import confusion_matrix, roc_auc_score, accuracy_score, precision_score, recall_score, f1_score\n",
"import itertools\n",
"import matplotlib\n",
"from matplotlib import pyplot as plt\n",
"\n",
"# compute a bunch of classification metrics \n",
"def classification_metrics(y_true, y_pred, y_prob):\n",
" cm_dict = {}\n",
" cm_dict['Accuracy'] = accuracy_score(y_true, y_pred)\n",
" cm_dict['Precision'] = precision_score(y_true, y_pred)\n",
" cm_dict['Recall'] = recall_score(y_true, y_pred)\n",
" cm_dict['F1'] = f1_score(y_true, y_pred) \n",
" cm_dict['AUC'] = roc_auc_score(y_true, y_prob[:,0])\n",
" cm_dict['Confusion Matrix'] = confusion_matrix(y_true, y_pred).tolist()\n",
" return cm_dict\n",
"\n",
"def plot_confusion_matrix(cm, classes, normalize=False, title='Confusion matrix', cmap=plt.cm.Blues):\n",
" \"\"\"Plots a confusion matrix.\n",
" Source: http://scikit-learn.org/stable/auto_examples/model_selection/plot_confusion_matrix.html\n",
" New BSD License - see appendix\n",
" \"\"\"\n",
" cm_max = cm.max()\n",
" cm_min = cm.min()\n",
" if cm_min > 0: cm_min = 0\n",
" if normalize:\n",
" cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
" cm_max = 1\n",
" plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
" plt.title(title)\n",
" plt.colorbar()\n",
" tick_marks = np.arange(len(classes))\n",
" plt.xticks(tick_marks, classes, rotation=45)\n",
" plt.yticks(tick_marks, classes)\n",
" thresh = cm_max / 2.\n",
" plt.clim(cm_min, cm_max)\n",
"\n",
" for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
" plt.text(j, i,\n",
" round(cm[i, j], 3), # round to 3 decimals if they are float\n",
" horizontalalignment=\"center\",\n",
" color=\"white\" if cm[i, j] > thresh else \"black\")\n",
" plt.ylabel('True label')\n",
" plt.xlabel('Predicted label')\n",
" plt.show()\n",
" \n",
"cm_dict = classification_metrics(y_test_max, y_prob_max, y_probs)\n",
"for m in cm_dict:\n",
" print(m, cm_dict[m])\n",
"cm = np.asarray(cm_dict['Confusion Matrix'])\n",
"plot_confusion_matrix(cm, ['fail','pass'], normalize=False)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Service Definition\n",
"Like in the QuickStart notebook our service definition pipeline consists of three stages. Because the preprocessing and featurizing stage don't contain any variables, we can use a default session.\n",
"Here we use the Keras classifier as the final stage."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.contrib.brainwave.pipeline import ModelDefinition, TensorflowStage, BrainWaveStage, KerasStage\n",
"\n",
"model_def = ModelDefinition()\n",
"model_def.pipeline.append(TensorflowStage(tf.Session(), in_images, image_tensors))\n",
"model_def.pipeline.append(BrainWaveStage(tf.Session(), bwmodel))\n",
"model_def.pipeline.append(KerasStage(model))\n",
"\n",
"model_def_path = os.path.join(datadir, 'save', 'model_def')\n",
"model_def.save(model_def_path)\n",
"print(model_def_path)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Deploy"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.model import Model\n",
"from azureml.core import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print(ws.name, ws.resource_group, ws.location, ws.subscription_id, sep = '\\n')\n",
"model_name = \"catsanddogs-model\"\n",
"service_name = \"modelbuild-service\"\n",
"\n",
"registered_model = Model.register(ws, model_def_path, model_name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The first time the code below runs it will create a new service running your model. If you want to change the model you can make changes above in this notebook and save a new service definition. Then this code will update the running service in place to run the new model."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.webservice import Webservice\n",
"from azureml.exceptions import WebserviceException\n",
"from azureml.contrib.brainwave import BrainwaveWebservice, BrainwaveImage\n",
"try:\n",
" service = Webservice(ws, service_name)\n",
"except WebserviceException:\n",
" image_config = BrainwaveImage.image_configuration()\n",
" deployment_config = BrainwaveWebservice.deploy_configuration()\n",
" service = Webservice.deploy_from_model(ws, service_name, [registered_model], image_config, deployment_config)\n",
" service.wait_for_deployment(True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The service is now running in Azure and ready to serve requests. We can check the address and port."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(service.ipAddress + ':' + str(service.port))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Client\n",
"There is a simple test client at amlrealtimeai.PredictionClient which can be used for testing. We'll use this client to score an image with our new service."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.contrib.brainwave.client import PredictionClient\n",
"client = PredictionClient(service.ipAddress, service.port)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can adapt the client [code](../../pythonlib/amlrealtimeai/client.py) to meet your needs. There is also an example C# [client](../../sample-clients/csharp).\n",
"\n",
"The service provides an API that is compatible with TensorFlow Serving. There are instructions to download a sample client [here](https://www.tensorflow.org/serving/setup)."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Request\n",
"Let's see how our service does on a few images. It may get a few wrong."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Specify an image to classify\n",
"print('CATS')\n",
"for image_file in cat_files[:8]:\n",
" results = client.score_image(image_file)\n",
" result = 'CORRECT ' if results[0] > results[1] else 'WRONG '\n",
" print(result + str(results))\n",
"print('DOGS')\n",
"for image_file in dog_files[:8]:\n",
" results = client.score_image(image_file)\n",
" result = 'CORRECT ' if results[1] > results[0] else 'WRONG '\n",
" print(result + str(results))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Cleanup\n",
"Run the cell below to delete your service. In the [next notebook](project-brainwave-custom-weights.ipynb) you will learn how to retrain all the weights of one of the models"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"service.delete()\n",
" \n",
"registered_model.delete()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Appendix"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"License for plot_confusion_matrix:\n",
"\n",
"New BSD License\n",
"\n",
"Copyright (c) 2007\u00e2\u20ac\u201c2018 The scikit-learn developers.\n",
"All rights reserved.\n",
"\n",
"\n",
"Redistribution and use in source and binary forms, with or without\n",
"modification, are permitted provided that the following conditions are met:\n",
"\n",
" a. Redistributions of source code must retain the above copyright notice,\n",
" this list of conditions and the following disclaimer.\n",
" b. Redistributions in binary form must reproduce the above copyright\n",
" notice, this list of conditions and the following disclaimer in the\n",
" documentation and/or other materials provided with the distribution.\n",
" c. Neither the name of the Scikit-learn Developers nor the names of\n",
" its contributors may be used to endorse or promote products\n",
" derived from this software without specific prior written\n",
" permission. \n",
"\n",
"\n",
"THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS \"AS IS\"\n",
"AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE\n",
"IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE\n",
"ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE FOR\n",
"ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL\n",
"DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR\n",
"SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER\n",
"CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT\n",
"LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY\n",
"OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH\n",
"DAMAGE.\n"
]
}
],
"metadata": {
"authors": [
{
"name": "coverste"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.5.2"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

804
training/01.train-hyperparameter-tune-deploy-with-pytorch/01.train-hyperparameter-tune-deploy-with-pytorch.ipynb Normal file
View File

@@ -0,0 +1,804 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved. \n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# 01. Train, hyperparameter tune, and deploy with PyTorch\n",
"\n",
"In this tutorial, you will train, hyperparameter tune, and deploy a PyTorch model using the Azure Machine Learning (AML) Python SDK.\n",
"\n",
"This tutorial will train an image classification model using transfer learning, based on PyTorch's [Transfer Learning tutorial](https://pytorch.org/tutorials/beginner/transfer_learning_tutorial.html). The model is trained to classify ants and bees by first using a pretrained ResNet18 model that has been trained on the [ImageNet](http://image-net.org/index) dataset."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"* Understand the [architecture and terms](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture) introduced by Azure Machine Learning\n",
"* Go through the [00.configuration.ipynb](https://github.com/Azure/MachineLearningNotebooks/blob/master/00.configuration.ipynb) notebook to:\n",
" * install the AML SDK\n",
" * create a workspace and its configuration file (`config.json`)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"Diagnostics"
]
},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize workspace\n",
"Initialize a [Workspace](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture#workspace) object from the existing workspace you created in the Prerequisites step. `Workspace.from_config()` creates a workspace object from the details stored in `config.json`."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.workspace import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print('Workspace name: ' + ws.name, \n",
" 'Azure region: ' + ws.location, \n",
" 'Subscription id: ' + ws.subscription_id, \n",
" 'Resource group: ' + ws.resource_group, sep = '\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create a remote compute target\n",
"You will need to create a [compute target](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture#compute-target) to execute your training script on. In this tutorial, you create an [Azure Batch AI](https://docs.microsoft.com/azure/batch-ai/overview) cluster as your training compute resource. This code creates a cluster for you if it does not already exist in your workspace.\n",
"\n",
"**Creation of the cluster takes approximately 5 minutes.** If the cluster is already in your workspace this code will skip the cluster creation process."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import ComputeTarget, BatchAiCompute\n",
"from azureml.core.compute_target import ComputeTargetException\n",
"\n",
"# choose a name for your cluster\n",
"cluster_name = \"gpucluster\"\n",
"\n",
"try:\n",
" compute_target = ComputeTarget(workspace=ws, name=cluster_name)\n",
" print('Found existing compute target.')\n",
"except ComputeTargetException:\n",
" print('Creating a new compute target...')\n",
" compute_config = BatchAiCompute.provisioning_configuration(vm_size='STANDARD_NC6', \n",
" autoscale_enabled=True,\n",
" cluster_min_nodes=0, \n",
" cluster_max_nodes=4)\n",
"\n",
" # create the cluster\n",
" compute_target = ComputeTarget.create(ws, cluster_name, compute_config)\n",
"\n",
" compute_target.wait_for_completion(show_output=True)\n",
"\n",
" # Use the 'status' property to get a detailed status for the current cluster. \n",
" print(compute_target.status.serialize())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The above code creates a GPU cluster. If you instead want to create a CPU cluster, provide a different VM size to the `vm_size` parameter, such as `STANDARD_D2_V2`."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Upload training data\n",
"The dataset we will use consists of about 120 training images each for ants and bees, with 75 validation images for each class.\n",
"\n",
"First, download the dataset (located [here](https://download.pytorch.org/tutorial/hymenoptera_data.zip) as a zip file) locally to your current directory and extract the files. This will create a folder called `hymenoptera_data` with two subfolders `train` and `val` that contain the training and validation images, respectively. [Hymenoptera](https://en.wikipedia.org/wiki/Hymenoptera) is the order of insects that includes ants and bees."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"import urllib\n",
"from zipfile import ZipFile\n",
"\n",
"# download data\n",
"download_url = 'https://download.pytorch.org/tutorial/hymenoptera_data.zip'\n",
"data_file = './hymenoptera_data.zip'\n",
"urllib.request.urlretrieve(download_url, filename=data_file)\n",
"\n",
"# extract files\n",
"with ZipFile(data_file, 'r') as zip:\n",
" print('extracting files...')\n",
" zip.extractall()\n",
" print('done')\n",
" \n",
"# delete zip file\n",
"os.remove(data_file)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"To make the data accessible for remote training, you will need to upload the data from your local machine to the cloud. AML provides a convenient way to do so via a [Datastore](https://docs.microsoft.com/azure/machine-learning/service/how-to-access-data). The datastore provides a mechanism for you to upload/download data, and interact with it from your remote compute targets. \n",
"\n",
"**Note: If your data is already stored in Azure, or you download the data as part of your training script, you will not need to do this step.**\n",
"\n",
"Each workspace is associated with a default datastore. In this tutorial, we will upload the training data to this default datastore."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ds = ws.get_default_datastore()\n",
"print(ds.datastore_type, ds.account_name, ds.container_name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The following code will upload the training data to the path `./hymenoptera_data` on the default datastore."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ds.upload(src_dir='./hymenoptera_data', target_path='hymenoptera_data')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now let's get a reference to the path on the datastore with the training data. We can do so using the `path` method. In the next section, we can then pass this reference to our training script's `--data_dir` argument. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"path_on_datastore = 'hymenoptera_data'\n",
"ds_data = ds.path(path_on_datastore)\n",
"print(ds_data)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train model on the remote compute\n",
"Now that you have your data and training script prepared, you are ready to train on your remote compute cluster. You can take advantage of Azure compute to leverage GPUs to cut down your training time. "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create a project directory\n",
"Create a directory that will contain all the necessary code from your local machine that you will need access to on the remote resource. This includes the training script and any additional files your training script depends on."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"\n",
"project_folder = './pytorch-hymenoptera'\n",
"os.makedirs(project_folder, exist_ok=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Prepare training script\n",
"Now you will need to create your training script. In this tutorial, the training script is already provided for you at `pytorch_train.py`. In practice, you should be able to take any custom training script as is and run it with AML without having to modify your code.\n",
"\n",
"However, if you would like to use AML's [tracking and metrics](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture#metrics) capabilities, you will have to add a small amount of AML code inside your training script. \n",
"\n",
"In `pytorch_train.py`, we will log some metrics to our AML run. To do so, we will access the AML run object within the script:\n",
"```Python\n",
"from azureml.core.run import Run\n",
"run = Run.get_context()\n",
"```\n",
"Further within `pytorch_train.py`, we log the learning rate and momentum parameters, and the best validation accuracy the model achieves:\n",
"```Python\n",
"run.log('lr', np.float(learning_rate))\n",
"run.log('momentum', np.float(momentum))\n",
"\n",
"run.log('best_val_acc', np.float(best_acc))\n",
"```\n",
"These run metrics will become particularly important when we begin hyperparameter tuning our model in the \"Tune model hyperparameters\" section."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Once your script is ready, copy the training script `pytorch_train.py` into your project directory."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import shutil\n",
"shutil.copy('pytorch_train.py', project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create an experiment\n",
"Create an [Experiment](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture#experiment) to track all the runs in your workspace for this transfer learning PyTorch tutorial. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Experiment\n",
"\n",
"experiment_name = 'pytorch-hymenoptera'\n",
"experiment = Experiment(ws, name=experiment_name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create a PyTorch estimator\n",
"The AML SDK's PyTorch estimator enables you to easily submit PyTorch training jobs for both single-node and distributed runs. For more information on the PyTorch estimator, refer [here](https://docs.microsoft.com/azure/machine-learning/service/how-to-train-pytorch). The following code will define a single-node PyTorch job."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.dnn import PyTorch\n",
"\n",
"script_params = {\n",
" '--data_dir': ds_data,\n",
" '--num_epochs': 25,\n",
" '--output_dir': './outputs'\n",
"}\n",
"\n",
"estimator = PyTorch(source_directory=project_folder, \n",
" script_params=script_params,\n",
" compute_target=compute_target,\n",
" entry_script='pytorch_train.py',\n",
" use_gpu=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The `script_params` parameter is a dictionary containing the command-line arguments to your training script `entry_script`. Please note the following:\n",
"- We passed our training data reference `ds_data` to our script's `--data_dir` argument. This will 1) mount our datastore on the remote compute and 2) provide the path to the training data `hymenoptera_data` on our datastore.\n",
"- We specified the output directory as `./outputs`. The `outputs` directory is specially treated by AML in that all the content in this directory gets uploaded to your workspace as part of your run history. The files written to this directory are therefore accessible even once your remote run is over. In this tutorial, we will save our trained model to this output directory.\n",
"\n",
"To leverage the Azure VM's GPU for training, we set `use_gpu=True`."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Submit job\n",
"Run your experiment by submitting your estimator object. Note that this call is asynchronous."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run = experiment.submit(estimator)\n",
"print(run.get_details())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Monitor your run\n",
"You can monitor the progress of the run with a Jupyter widget. Like the run submission, the widget is asynchronous and provides live updates every 10-15 seconds until the job completes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.widgets import RunDetails\n",
"RunDetails(run).show()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.wait_for_completion(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Tune model hyperparameters\n",
"Now that we've seen how to do a simple PyTorch training run using the SDK, let's see if we can further improve the accuracy of our model. We can optimize our model's hyperparameters using Azure Machine Learning's hyperparameter tuning capabilities."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Start a hyperparameter sweep\n",
"First, we will define the hyperparameter space to sweep over. Since our training script uses a learning rate schedule to decay the learning rate every several epochs, let's tune the initial learning rate and the momentum parameters. In this example we will use random sampling to try different configuration sets of hyperparameters to maximize our primary metric, the best validation accuracy (`best_val_acc`).\n",
"\n",
"Then, we specify the early termination policy to use to early terminate poorly performing runs. Here we use the `BanditPolicy`, which will terminate any run that doesn't fall within the slack factor of our primary evaluation metric. In this tutorial, we will apply this policy every epoch (since we report our `best_val_acc` metric every epoch and `evaluation_interval=1`). Notice we will delay the first policy evaluation until after the first `10` epochs (`delay_evaluation=10`).\n",
"Refer [here](https://docs.microsoft.com/azure/machine-learning/service/how-to-tune-hyperparameters#specify-an-early-termination-policy) for more information on the BanditPolicy and other policies available."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.hyperdrive import *\n",
"\n",
"param_sampling = RandomParameterSampling( {\n",
" 'learning_rate': uniform(0.0005, 0.005),\n",
" 'momentum': uniform(0.9, 0.99)\n",
" }\n",
")\n",
"\n",
"early_termination_policy = BanditPolicy(slack_factor=0.15, evaluation_interval=1, delay_evaluation=10)\n",
"\n",
"hyperdrive_run_config = HyperDriveRunConfig(estimator=estimator,\n",
" hyperparameter_sampling=param_sampling, \n",
" policy=early_termination_policy,\n",
" primary_metric_name='best_val_acc',\n",
" primary_metric_goal=PrimaryMetricGoal.MAXIMIZE,\n",
" max_total_runs=20,\n",
" max_concurrent_runs=4)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Finally, lauch the hyperparameter tuning job."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# start the HyperDrive run\n",
"hyperdrive_run = experiment.submit(hyperdrive_run_config)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Monitor HyperDrive runs\n",
"You can monitor the progress of the runs with the following Jupyter widget. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.widgets import RunDetails\n",
"\n",
"RunDetails(hyperdrive_run).show()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"hyperdrive_run.wait_for_completion(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Find and register the best model\n",
"Once all the runs complete, we can find the run that produced the model with the highest accuracy."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run = hyperdrive_run.get_best_run_by_primary_metric()\n",
"best_run_metrics = best_run.get_metrics()\n",
"print(best_run)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print('Best Run is:\\n Validation accuracy: {0:.5f} \\n Learning rate: {1:.5f} \\n Momentum: {2:.5f}'.format(\n",
" best_run_metrics['best_val_acc'][-1],\n",
" best_run_metrics['lr'],\n",
" best_run_metrics['momentum'])\n",
" )"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Finally, register the model from your best-performing run to your workspace. The `model_path` parameter takes in the relative path on the remote VM to the model file in your `outputs` directory. In the next section, we will deploy this registered model as a web service."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"model = best_run.register_model(model_name = 'pytorch-hymenoptera', model_path = 'outputs/model.pt')\n",
"print(model.name, model.id, model.version, sep = '\\t')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Deploy model as web service\n",
"Once you have your trained model, you can deploy the model on Azure. In this tutorial, we will deploy the model as a web service in [Azure Container Instances](https://docs.microsoft.com/en-us/azure/container-instances/) (ACI). For more information on deploying models using Azure ML, refer [here](https://docs.microsoft.com/azure/machine-learning/service/how-to-deploy-and-where)."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create scoring script\n",
"\n",
"First, we will create a scoring script that will be invoked by the web service call. Note that the scoring script must have two required functions:\n",
"* `init()`: In this function, you typically load the model into a `global` object. This function is executed only once when the Docker container is started. \n",
"* `run(input_data)`: In this function, the model is used to predict a value based on the input data. The input and output typically use JSON as serialization and deserialization format, but you are not limited to that.\n",
"\n",
"Refer to the scoring script `pytorch_score.py` for this tutorial. Our web service will use this file to predict whether an image is an ant or a bee. When writing your own scoring script, don't forget to test it locally first before you go and deploy the web service."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create environment file\n",
"Then, we will need to create an environment file (`myenv.yml`) that specifies all of the scoring script's package dependencies. This file is used to ensure that all of those dependencies are installed in the Docker image by AML. In this case, we need to specify `torch`, `torchvision`, `pillow`, and `azureml-sdk`."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile myenv.yml\n",
"name: myenv\n",
"channels:\n",
" - defaults\n",
"dependencies:\n",
" - pip:\n",
" - torch\n",
" - torchvision\n",
" - pillow\n",
" - azureml-core"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Configure the container image\n",
"Now configure the Docker image that you will use to build your ACI container."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.image import ContainerImage\n",
"\n",
"image_config = ContainerImage.image_configuration(execution_script='pytorch_score.py', \n",
" runtime='python', \n",
" conda_file='myenv.yml',\n",
" description='Image with hymenoptera model')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Configure the ACI container\n",
"We are almost ready to deploy. Create a deployment configuration file to specify the number of CPUs and gigabytes of RAM needed for your ACI container. While it depends on your model, the default of `1` core and `1` gigabyte of RAM is usually sufficient for many models."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.webservice import AciWebservice\n",
"\n",
"aciconfig = AciWebservice.deploy_configuration(cpu_cores=1, \n",
" memory_gb=1, \n",
" tags={'data': 'hymenoptera', 'method':'transfer learning', 'framework':'pytorch'},\n",
" description='Classify ants/bees using transfer learning with PyTorch')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Deploy the registered model\n",
"Finally, let's deploy a web service from our registered model. First, retrieve the model from your workspace."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.model import Model\n",
"\n",
"model = Model(ws, name='pytorch-hymenoptera')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Then, deploy the web service using the ACI config and image config files created in the previous steps. We pass the `model` object in a list to the `models` parameter. If you would like to deploy more than one registered model, append the additional models to this list."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%time\n",
"from azureml.core.webservice import Webservice\n",
"\n",
"service_name = 'aci-hymenoptera'\n",
"service = Webservice.deploy_from_model(workspace=ws,\n",
" name=service_name,\n",
" models=[model],\n",
" image_config=image_config,\n",
" deployment_config=aciconfig,)\n",
"\n",
"service.wait_for_deployment(show_output=True)\n",
"print(service.state)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"If your deployment fails for any reason and you need to redeploy, make sure to delete the service before you do so: `service.delete()`"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Tip: If something goes wrong with the deployment, the first thing to look at is the logs from the service by running the following command:**"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"service.get_logs()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Get the web service's HTTP endpoint, which accepts REST client calls. This endpoint can be shared with anyone who wants to test the web service or integrate it into an application."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(service.scoring_uri)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Test the web service\n",
"Finally, let's test our deployed web service. We will send the data as a JSON string to the web service hosted in ACI and use the SDK's `run` API to invoke the service. Here we will take an arbitrary image from our validation data to predict on."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os, json, base64\n",
"from io import BytesIO\n",
"from PIL import Image\n",
"import matplotlib.pyplot as plt\n",
"\n",
"def imgToBase64(img):\n",
" \"\"\"Convert pillow image to base64-encoded image\"\"\"\n",
" imgio = BytesIO()\n",
" img.save(imgio, 'JPEG')\n",
" img_str = base64.b64encode(imgio.getvalue())\n",
" return img_str.decode('utf-8')\n",
"\n",
"test_img = os.path.join('hymenoptera_data', 'val', 'bees', '10870992_eebeeb3a12.jpg') #arbitary image from val dataset\n",
"plt.imshow(Image.open(test_img))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"base64Img = imgToBase64(Image.open(test_img))\n",
"\n",
"result = service.run(input_data=json.dumps({'data': base64Img}))\n",
"print(json.loads(result))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Delete web service\n",
"Once you no longer need the web service, you should delete it."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"service.delete()"
]
}
],
"metadata": {
"authors": [
{
"name": "minxia"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
},
"msauthor": "minxia"
},
"nbformat": 4,
"nbformat_minor": 2
}

57
training/01.train-hyperparameter-tune-deploy-with-pytorch/pytorch_score.py Normal file
View File

@@ -0,0 +1,57 @@
# Copyright (c) Microsoft. All rights reserved.
# Licensed under the MIT license.
import torch
import torch.nn as nn
from torchvision import transforms
import json
import base64
from io import BytesIO
from PIL import Image
from azureml.core.model import Model
def preprocess_image(image_file):
"""Preprocess the input image."""
data_transforms = transforms.Compose([
transforms.Resize(256),
transforms.CenterCrop(224),
transforms.ToTensor(),
transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
])
image = Image.open(image_file)
image = data_transforms(image).float()
image = torch.tensor(image)
image = image.unsqueeze(0)
return image
def base64ToImg(base64ImgString):
base64Img = base64ImgString.encode('utf-8')
decoded_img = base64.b64decode(base64Img)
return BytesIO(decoded_img)
def init():
global model
model_path = Model.get_model_path('pytorch-hymenoptera')
model = torch.load(model_path, map_location=lambda storage, loc: storage)
model.eval()
def run(input_data):
img = base64ToImg(json.loads(input_data)['data'])
img = preprocess_image(img)
# get prediction
output = model(img)
classes = ['ants', 'bees']
softmax = nn.Softmax(dim=1)
pred_probs = softmax(model(img)).detach().numpy()[0]
index = torch.argmax(output, 1)
result = {"label": classes[index], "probability": str(pred_probs[index])}
return result

176
training/01.train-hyperparameter-tune-deploy-with-pytorch/pytorch_train.py Normal file
View File

@@ -0,0 +1,176 @@
# Copyright (c) 2017, PyTorch contributors
# Modifications copyright (C) Microsoft Corporation
# Licensed under the BSD license
# Adapted from https://pytorch.org/tutorials/beginner/transfer_learning_tutorial.html
from __future__ import print_function, division
import torch
import torch.nn as nn
import torch.optim as optim
from torch.optim import lr_scheduler
from torchvision import datasets, models, transforms
import numpy as np
import time
import os
import copy
import argparse
from azureml.core.run import Run
# get the Azure ML run object
run = Run.get_context()
def load_data(data_dir):
"""Load the train/val data."""
# Data augmentation and normalization for training
# Just normalization for validation
data_transforms = {
'train': transforms.Compose([
transforms.RandomResizedCrop(224),
transforms.RandomHorizontalFlip(),
transforms.ToTensor(),
transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
]),
'val': transforms.Compose([
transforms.Resize(256),
transforms.CenterCrop(224),
transforms.ToTensor(),
transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
]),
}
image_datasets = {x: datasets.ImageFolder(os.path.join(data_dir, x),
data_transforms[x])
for x in ['train', 'val']}
dataloaders = {x: torch.utils.data.DataLoader(image_datasets[x], batch_size=4,
shuffle=True, num_workers=0)
for x in ['train', 'val']}
dataset_sizes = {x: len(image_datasets[x]) for x in ['train', 'val']}
class_names = image_datasets['train'].classes
return dataloaders, dataset_sizes, class_names
def train_model(model, criterion, optimizer, scheduler, num_epochs, data_dir):
"""Train the model."""
# load training/validation data
dataloaders, dataset_sizes, class_names = load_data(data_dir)
device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
since = time.time()
best_model_wts = copy.deepcopy(model.state_dict())
best_acc = 0.0
for epoch in range(num_epochs):
print('Epoch {}/{}'.format(epoch, num_epochs - 1))
print('-' * 10)
# Each epoch has a training and validation phase
for phase in ['train', 'val']:
if phase == 'train':
scheduler.step()
model.train() # Set model to training mode
else:
model.eval() # Set model to evaluate mode
running_loss = 0.0
running_corrects = 0
# Iterate over data.
for inputs, labels in dataloaders[phase]:
inputs = inputs.to(device)
labels = labels.to(device)
# zero the parameter gradients
optimizer.zero_grad()
# forward
# track history if only in train
with torch.set_grad_enabled(phase == 'train'):
outputs = model(inputs)
_, preds = torch.max(outputs, 1)
loss = criterion(outputs, labels)
# backward + optimize only if in training phase
if phase == 'train':
loss.backward()
optimizer.step()
# statistics
running_loss += loss.item() * inputs.size(0)
running_corrects += torch.sum(preds == labels.data)
epoch_loss = running_loss / dataset_sizes[phase]
epoch_acc = running_corrects.double() / dataset_sizes[phase]
print('{} Loss: {:.4f} Acc: {:.4f}'.format(
phase, epoch_loss, epoch_acc))
# deep copy the model
if phase == 'val' and epoch_acc > best_acc:
best_acc = epoch_acc
best_model_wts = copy.deepcopy(model.state_dict())
# log the best val accuracy to AML run
run.log('best_val_acc', np.float(best_acc))
print()
time_elapsed = time.time() - since
print('Training complete in {:.0f}m {:.0f}s'.format(
time_elapsed // 60, time_elapsed % 60))
print('Best val Acc: {:4f}'.format(best_acc))
# load best model weights
model.load_state_dict(best_model_wts)
return model
def fine_tune_model(num_epochs, data_dir, learning_rate, momentum):
"""Load a pretrained model and reset the final fully connected layer."""
# log the hyperparameter metrics to the AML run
run.log('lr', np.float(learning_rate))
run.log('momentum', np.float(momentum))
model_ft = models.resnet18(pretrained=True)
num_ftrs = model_ft.fc.in_features
model_ft.fc = nn.Linear(num_ftrs, 2) # only 2 classes to predict
device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
model_ft = model_ft.to(device)
criterion = nn.CrossEntropyLoss()
# Observe that all parameters are being optimized
optimizer_ft = optim.SGD(model_ft.parameters(), lr=learning_rate, momentum=momentum)
# Decay LR by a factor of 0.1 every 7 epochs
exp_lr_scheduler = lr_scheduler.StepLR(optimizer_ft, step_size=7, gamma=0.1)
model = train_model(model_ft, criterion, optimizer_ft, exp_lr_scheduler, num_epochs, data_dir)
return model
def main():
# get command-line arguments
parser = argparse.ArgumentParser()
parser.add_argument('--data_dir', type=str, help='directory of training data')
parser.add_argument('--num_epochs', type=int, default=25, help='number of epochs to train')
parser.add_argument('--output_dir', type=str, help='output directory')
parser.add_argument('--learning_rate', type=float, default=0.001, help='learning rate')
parser.add_argument('--momentum', type=float, default=0.9, help='momentum')
args = parser.parse_args()
print("data directory is: " + args.data_dir)
model = fine_tune_model(args.num_epochs, args.data_dir, args.learning_rate, args.momentum)
os.makedirs(args.output_dir, exist_ok=True)
torch.save(model, os.path.join(args.output_dir, 'model.pt'))
if __name__ == "__main__":
main()

315
training/02.distributed-pytorch-with-horovod/02.distributed-pytorch-with-horovod.ipynb Normal file
View File

@@ -0,0 +1,315 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# 02. Distributed PyTorch with Horovod\n",
"In this tutorial, you will train a PyTorch model on the [MNIST](http://yann.lecun.com/exdb/mnist/) dataset using distributed training via [Horovod](https://github.com/uber/horovod)."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"* Understand the [architecture and terms](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture) introduced by Azure Machine Learning (AML)\n",
"* Go through the [00.configuration.ipynb](https://github.com/Azure/MachineLearningNotebooks/blob/master/00.configuration.ipynb) notebook to:\n",
" * install the AML SDK\n",
" * create a workspace and its configuration file (`config.json`)\n",
"* Review the [tutorial](https://aka.ms/aml-notebook-pytorch) on single-node PyTorch training using the SDK"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"Diagnostics"
]
},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize workspace\n",
"\n",
"Initialize a [Workspace](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture#workspace) object from the existing workspace you created in the Prerequisites step. `Workspace.from_config()` creates a workspace object from the details stored in `config.json`."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.workspace import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print('Workspace name: ' + ws.name, \n",
" 'Azure region: ' + ws.location, \n",
" 'Subscription id: ' + ws.subscription_id, \n",
" 'Resource group: ' + ws.resource_group, sep = '\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create a remote compute target\n",
"You will need to create a [compute target](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture#compute-target) to execute your training script on. In this tutorial, you create an [Azure Batch AI](https://docs.microsoft.com/azure/batch-ai/overview) cluster as your training compute resource. This code creates a cluster for you if it does not already exist in your workspace.\n",
"\n",
"**Creation of the cluster takes approximately 5 minutes.** If the cluster is already in your workspace this code will skip the cluster creation process."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import ComputeTarget, BatchAiCompute\n",
"from azureml.core.compute_target import ComputeTargetException\n",
"\n",
"# choose a name for your cluster\n",
"cluster_name = \"gpucluster\"\n",
"\n",
"try:\n",
" compute_target = ComputeTarget(workspace=ws, name=cluster_name)\n",
" print('Found existing compute target.')\n",
"except ComputeTargetException:\n",
" print('Creating a new compute target...')\n",
" compute_config = BatchAiCompute.provisioning_configuration(vm_size='STANDARD_NC6', \n",
" autoscale_enabled=True,\n",
" cluster_min_nodes=0, \n",
" cluster_max_nodes=4)\n",
"\n",
" # create the cluster\n",
" compute_target = ComputeTarget.create(ws, cluster_name, compute_config)\n",
"\n",
" compute_target.wait_for_completion(show_output=True)\n",
"\n",
" # Use the 'status' property to get a detailed status for the current cluster. \n",
" print(compute_target.status.serialize())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The above code creates a GPU cluster. If you instead want to create a CPU cluster, provide a different VM size to the `vm_size` parameter, such as `STANDARD_D2_V2`."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train model on the remote compute\n",
"Now that we have the cluster ready to go, let's run our distributed training job."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create a project directory\n",
"Create a directory that will contain all the necessary code from your local machine that you will need access to on the remote resource. This includes the training script and any additional files your training script depends on."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"\n",
"project_folder = './pytorch-distr-hvd'\n",
"os.makedirs(project_folder, exist_ok=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copy the training script `pytorch_horovod_mnist.py` into this project directory."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import shutil\n",
"shutil.copy('pytorch_horovod_mnist.py', project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create an experiment\n",
"Create an [Experiment](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture#experiment) to track all the runs in your workspace for this distributed PyTorch tutorial. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Experiment\n",
"\n",
"experiment_name = 'pytorch-distr-hvd'\n",
"experiment = Experiment(ws, name=experiment_name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create a PyTorch estimator\n",
"The AML SDK's PyTorch estimator enables you to easily submit PyTorch training jobs for both single-node and distributed runs. For more information on the PyTorch estimator, refer [here](https://docs.microsoft.com/azure/machine-learning/service/how-to-train-pytorch)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.dnn import PyTorch\n",
"\n",
"estimator = PyTorch(source_directory=project_folder,\n",
" compute_target=compute_target,\n",
" entry_script='pytorch_horovod_mnist.py',\n",
" node_count=2,\n",
" process_count_per_node=1,\n",
" distributed_backend='mpi',\n",
" use_gpu=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The above code specifies that we will run our training script on `2` nodes, with one worker per node. In order to execute a distributed run using MPI/Horovod, you must provide the argument `distributed_backend='mpi'`. Using this estimator with these settings, PyTorch, Horovod and their dependencies will be installed for you. However, if your script also uses other packages, make sure to install them via the `PyTorch` constructor's `pip_packages` or `conda_packages` parameters."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Submit job\n",
"Run your experiment by submitting your estimator object. Note that this call is asynchronous."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run = experiment.submit(estimator)\n",
"print(run.get_details())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Monitor your run\n",
"You can monitor the progress of the run with a Jupyter widget. Like the run submission, the widget is asynchronous and provides live updates every 10-15 seconds until the job completes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.widgets import RunDetails\n",
"RunDetails(run).show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Alternatively, you can block until the script has completed training before running more code."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.wait_for_completion(show_output=True) # this provides a verbose log"
]
}
],
"metadata": {
"authors": [
{
"name": "minxia"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
},
"msauthor": "minxia"
},
"nbformat": 4,
"nbformat_minor": 2
}

0
training/distributed-pytorch-with-horovod/pytorch_horovod_mnist.py → training/02.distributed-pytorch-with-horovod/pytorch_horovod_mnist.py
View File

1
training/03.train-hyperparameter-tune-deploy-with-tensorflow/.gitignore vendored Normal file
View File

@@ -0,0 +1 @@
/data/

1624
training/03.train-hyperparameter-tune-deploy-with-tensorflow/.ipynb_checkpoints/03.train-hyperparameter-tune-deploy-with-tensorflow-checkpoint.ipynb Normal file
View File

File diff suppressed because it is too large Load Diff

1674
training/03.train-hyperparameter-tune-deploy-with-tensorflow/03.train-hyperparameter-tune-deploy-with-tensorflow.ipynb Normal file
View File

File diff suppressed because it is too large Load Diff

BIN
training/03.train-hyperparameter-tune-deploy-with-tensorflow/nn.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 119 KiB

106
training/03.train-hyperparameter-tune-deploy-with-tensorflow/tf_mnist.py Normal file
View File

@@ -0,0 +1,106 @@
# Copyright (c) Microsoft Corporation. All rights reserved.
# Licensed under the MIT License.
import numpy as np
import argparse
import os
import tensorflow as tf
from azureml.core import Run
from utils import load_data
print("TensorFlow version:", tf.VERSION)
parser = argparse.ArgumentParser()
parser.add_argument('--data-folder', type=str, dest='data_folder', help='data folder mounting point')
parser.add_argument('--batch-size', type=int, dest='batch_size', default=50, help='mini batch size for training')
parser.add_argument('--first-layer-neurons', type=int, dest='n_hidden_1', default=100,
help='# of neurons in the first layer')
parser.add_argument('--second-layer-neurons', type=int, dest='n_hidden_2', default=100,
help='# of neurons in the second layer')
parser.add_argument('--learning-rate', type=float, dest='learning_rate', default=0.01, help='learning rate')
args = parser.parse_args()
data_folder = os.path.join(args.data_folder, 'mnist')
print('training dataset is stored here:', data_folder)
X_train = load_data(os.path.join(data_folder, 'train-images.gz'), False) / 255.0
X_test = load_data(os.path.join(data_folder, 'test-images.gz'), False) / 255.0
y_train = load_data(os.path.join(data_folder, 'train-labels.gz'), True).reshape(-1)
y_test = load_data(os.path.join(data_folder, 'test-labels.gz'), True).reshape(-1)
print(X_train.shape, y_train.shape, X_test.shape, y_test.shape, sep='\n')
training_set_size = X_train.shape[0]
n_inputs = 28 * 28
n_h1 = args.n_hidden_1
n_h2 = args.n_hidden_2
n_outputs = 10
learning_rate = args.learning_rate
n_epochs = 50
batch_size = args.batch_size
with tf.name_scope('network'):
# construct the DNN
X = tf.placeholder(tf.float32, shape=(None, n_inputs), name='X')
y = tf.placeholder(tf.int64, shape=(None), name='y')
h1 = tf.layers.dense(X, n_h1, activation=tf.nn.relu, name='h1')
h2 = tf.layers.dense(h1, n_h2, activation=tf.nn.relu, name='h2')
output = tf.layers.dense(h2, n_outputs, name='output')
with tf.name_scope('train'):
cross_entropy = tf.nn.sparse_softmax_cross_entropy_with_logits(labels=y, logits=output)
loss = tf.reduce_mean(cross_entropy, name='loss')
optimizer = tf.train.GradientDescentOptimizer(learning_rate)
train_op = optimizer.minimize(loss)
with tf.name_scope('eval'):
correct = tf.nn.in_top_k(output, y, 1)
acc_op = tf.reduce_mean(tf.cast(correct, tf.float32))
init = tf.global_variables_initializer()
saver = tf.train.Saver()
# start an Azure ML run
run = Run.get_context()
with tf.Session() as sess:
init.run()
for epoch in range(n_epochs):
# randomly shuffle training set
indices = np.random.permutation(training_set_size)
X_train = X_train[indices]
y_train = y_train[indices]
# batch index
b_start = 0
b_end = b_start + batch_size
for _ in range(training_set_size // batch_size):
# get a batch
X_batch, y_batch = X_train[b_start: b_end], y_train[b_start: b_end]
# update batch index for the next batch
b_start = b_start + batch_size
b_end = min(b_start + batch_size, training_set_size)
# train
sess.run(train_op, feed_dict={X: X_batch, y: y_batch})
# evaluate training set
acc_train = acc_op.eval(feed_dict={X: X_batch, y: y_batch})
# evaluate validation set
acc_val = acc_op.eval(feed_dict={X: X_test, y: y_test})
# log accuracies
run.log('training_acc', np.float(acc_train))
run.log('validation_acc', np.float(acc_val))
print(epoch, '-- Training accuracy:', acc_train, '\b Validation accuracy:', acc_val)
y_hat = np.argmax(output.eval(feed_dict={X: X_test}), axis=1)
run.log('final_acc', np.float(acc_val))
os.makedirs('./outputs/model', exist_ok=True)
# files saved in the "./outputs" folder are automatically uploaded into run history
saver.save(sess, './outputs/model/mnist-tf.model')

27
training/03.train-hyperparameter-tune-deploy-with-tensorflow/utils.py Normal file
View File

@@ -0,0 +1,27 @@
# Copyright (c) Microsoft Corporation. All rights reserved.
# Licensed under the MIT License.
import gzip
import numpy as np
import struct
# load compressed MNIST gz files and return numpy arrays
def load_data(filename, label=False):
with gzip.open(filename) as gz:
struct.unpack('I', gz.read(4))
n_items = struct.unpack('>I', gz.read(4))
if not label:
n_rows = struct.unpack('>I', gz.read(4))[0]
n_cols = struct.unpack('>I', gz.read(4))[0]
res = np.frombuffer(gz.read(n_items[0] * n_rows * n_cols), dtype=np.uint8)
res = res.reshape(n_items[0], n_rows * n_cols)
else:
res = np.frombuffer(gz.read(n_items[0]), dtype=np.uint8)
res = res.reshape(n_items[0], 1)
return res
# one-hot encode a 1-D array
def one_hot_encode(array, num_of_classes):
return np.eye(num_of_classes)[array.reshape(-1)]

2
training/04.distributed-tensorflow-with-horovod/.gitignore vendored Normal file
View File

@@ -0,0 +1,2 @@
/data/
/tf-distr-hvd/

386
training/04.distributed-tensorflow-with-horovod/04.distributed-tensorflow-with-horovod.ipynb Normal file
View File

@@ -0,0 +1,386 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# 04. Distributed Tensorflow with Horovod\n",
"In this tutorial, you will train a word2vec model in TensorFlow using distributed training via [Horovod](https://github.com/uber/horovod)."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"* Understand the [architecture and terms](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture) introduced by Azure Machine Learning (AML)\n",
"* Go through the [00.configuration.ipynb](https://github.com/Azure/MachineLearningNotebooks/blob/master/00.configuration.ipynb) notebook to:\n",
" * install the AML SDK\n",
" * create a workspace and its configuration file (`config.json`)\n",
"* Review the [tutorial](https://aka.ms/aml-notebook-hyperdrive) on single-node TensorFlow training using the SDK"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"Diagnostics"
]
},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize workspace\n",
"Initialize a [Workspace](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture#workspace) object from the existing workspace you created in the Prerequisites step. `Workspace.from_config()` creates a workspace object from the details stored in `config.json`."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.workspace import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print('Workspace name: ' + ws.name, \n",
" 'Azure region: ' + ws.location, \n",
" 'Subscription id: ' + ws.subscription_id, \n",
" 'Resource group: ' + ws.resource_group, sep = '\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create a remote compute target\n",
"You will need to create a [compute target](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture#compute-target) to execute your training script on. In this tutorial, you create an [Azure Batch AI](https://docs.microsoft.com/azure/batch-ai/overview) cluster as your training compute resource. This code creates a cluster for you if it does not already exist in your workspace.\n",
"\n",
"**Creation of the cluster takes approximately 5 minutes.** If the cluster is already in your workspace this code will skip the cluster creation process."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import ComputeTarget, BatchAiCompute\n",
"from azureml.core.compute_target import ComputeTargetException\n",
"\n",
"# choose a name for your cluster\n",
"cluster_name = \"gpucluster\"\n",
"\n",
"try:\n",
" compute_target = ComputeTarget(workspace=ws, name=cluster_name)\n",
" print('Found existing compute target')\n",
"except ComputeTargetException:\n",
" print('Creating a new compute target...')\n",
" compute_config = BatchAiCompute.provisioning_configuration(vm_size='STANDARD_NC6', \n",
" autoscale_enabled=True,\n",
" cluster_min_nodes=0, \n",
" cluster_max_nodes=4)\n",
"\n",
" # create the cluster\n",
" compute_target = ComputeTarget.create(ws, cluster_name, compute_config)\n",
"\n",
" compute_target.wait_for_completion(show_output=True)\n",
"\n",
" # Use the 'status' property to get a detailed status for the current cluster. \n",
" print(compute_target.status.serialize())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The above code creates a GPU cluster. If you instead want to create a CPU cluster, provide a different VM size to the `vm_size` parameter, such as `STANDARD_D2_V2`."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Upload data to datastore\n",
"To make data accessible for remote training, AML provides a convenient way to do so via a [Datastore](https://docs.microsoft.com/azure/machine-learning/service/how-to-access-data). The datastore provides a mechanism for you to upload/download data to Azure Storage, and interact with it from your remote compute targets. \n",
"\n",
"If your data is already stored in Azure, or you download the data as part of your training script, you will not need to do this step. For this tutorial, although you can download the data in your training script, we will demonstrate how to upload the training data to a datastore and access it during training to illustrate the datastore functionality."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"First, download the training data from [here](http://mattmahoney.net/dc/text8.zip) to your local machine:"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"import urllib\n",
"\n",
"os.makedirs('./data', exist_ok=True)\n",
"download_url = 'http://mattmahoney.net/dc/text8.zip'\n",
"urllib.request.urlretrieve(download_url, filename='./data/text8.zip')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Each workspace is associated with a default datastore. In this tutorial, we will upload the training data to this default datastore. The below code will upload the contents of the data directory to the path `./data` on the default datastore."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ds = ws.get_default_datastore()\n",
"print(ds.datastore_type, ds.account_name, ds.container_name)\n",
"\n",
"ds.upload(src_dir='data', target_path='data', overwrite=True, show_progress=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"For convenience, let's get a reference to the path on the datastore with the zip file of training data. We can do so using the `path` method. In the next section, we can then pass this reference to our training script's `--input_data` argument. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"path_on_datastore = 'data/text8.zip'\n",
"ds_data = ds.path(path_on_datastore)\n",
"print(ds_data)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train model on the remote compute"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create a project directory\n",
"Create a directory that will contain all the necessary code from your local machine that you will need access to on the remote resource. This includes the training script, and any additional files your training script depends on."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"project_folder = './tf-distr-hvd'\n",
"os.makedirs(project_folder, exist_ok=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copy the training script `tf_horovod_word2vec.py` into this project directory."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import shutil\n",
"shutil.copy('tf_horovod_word2vec.py', project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create an experiment\n",
"Create an [Experiment](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture#experiment) to track all the runs in your workspace for this distributed TensorFlow tutorial. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Experiment\n",
"\n",
"experiment_name = 'tf-distr-hvd'\n",
"experiment = Experiment(ws, name=experiment_name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create a TensorFlow estimator\n",
"The AML SDK's TensorFlow estimator enables you to easily submit TensorFlow training jobs for both single-node and distributed runs. For more information on the TensorFlow estimator, refer [here](https://docs.microsoft.com/azure/machine-learning/service/how-to-train-tensorflow)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.dnn import TensorFlow\n",
"\n",
"script_params={\n",
" '--input_data': ds_data\n",
"}\n",
"\n",
"estimator= TensorFlow(source_directory=project_folder,\n",
" compute_target=compute_target,\n",
" script_params=script_params,\n",
" entry_script='tf_horovod_word2vec.py',\n",
" node_count=2,\n",
" process_count_per_node=1,\n",
" distributed_backend='mpi',\n",
" use_gpu=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The above code specifies that we will run our training script on `2` nodes, with one worker per node. In order to execute a distributed run using MPI/Horovod, you must provide the argument `distributed_backend='mpi'`. Using this estimator with these settings, TensorFlow, Horovod and their dependencies will be installed for you. However, if your script also uses other packages, make sure to install them via the `TensorFlow` constructor's `pip_packages` or `conda_packages` parameters.\n",
"\n",
"Note that we passed our training data reference `ds_data` to our script's `--input_data` argument. This will 1) mount our datastore on the remote compute and 2) provide the path to the data zip file on our datastore."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Submit job\n",
"Run your experiment by submitting your estimator object. Note that this call is asynchronous."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run = experiment.submit(estimator)\n",
"print(run)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Monitor your run\n",
"You can monitor the progress of the run with a Jupyter widget. Like the run submission, the widget is asynchronous and provides live updates every 10-15 seconds until the job completes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.widgets import RunDetails\n",
"RunDetails(run).show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Alternatively, you can block until the script has completed training before running more code."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.wait_for_completion(show_output=True)"
]
}
],
"metadata": {
"authors": [
{
"name": "roastala"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
},
"msauthor": "minxia"
},
"nbformat": 4,
"nbformat_minor": 2
}

0
training/distributed-tensorflow-with-horovod/tf_horovod_word2vec.py → training/04.distributed-tensorflow-with-horovod/tf_horovod_word2vec.py
View File

1
training/05.distributed-tensorflow-with-parameter-server/.gitignore vendored Normal file
View File

@@ -0,0 +1 @@
/tf-distr-ps/

313
training/05.distributed-tensorflow-with-parameter-server/05.distributed-tensorflow-with-parameter-server.ipynb Normal file
View File

@@ -0,0 +1,313 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# 05. Distributed TensorFlow with parameter server\n",
"In this tutorial, you will train a TensorFlow model on the [MNIST](http://yann.lecun.com/exdb/mnist/) dataset using native [distributed TensorFlow](https://www.tensorflow.org/deploy/distributed)."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"* Understand the [architecture and terms](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture) introduced by Azure Machine Learning (AML)\n",
"* Go through the [00.configuration.ipynb](https://github.com/Azure/MachineLearningNotebooks/blob/master/00.configuration.ipynb) notebook to:\n",
" * install the AML SDK\n",
" * create a workspace and its configuration file (`config.json`)\n",
"* Review the [tutorial](https://aka.ms/aml-notebook-hyperdrive) on single-node TensorFlow training using the SDK"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"Diagnostics"
]
},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize workspace\n",
"Initialize a [Workspace](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture#workspace) object from the existing workspace you created in the Prerequisites step. `Workspace.from_config()` creates a workspace object from the details stored in `config.json`."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.workspace import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print('Workspace name: ' + ws.name, \n",
" 'Azure region: ' + ws.location, \n",
" 'Subscription id: ' + ws.subscription_id, \n",
" 'Resource group: ' + ws.resource_group, sep = '\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create a remote compute target\n",
"You will need to create a [compute target](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture#compute-target) to execute your training script on. In this tutorial, you create an [Azure Batch AI](https://docs.microsoft.com/azure/batch-ai/overview) cluster as your training compute resource. This code creates a cluster for you if it does not already exist in your workspace.\n",
"\n",
"**Creation of the cluster takes approximately 5 minutes.** If the cluster is already in your workspace this code will skip the cluster creation process."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import ComputeTarget, BatchAiCompute\n",
"from azureml.core.compute_target import ComputeTargetException\n",
"\n",
"# choose a name for your cluster\n",
"cluster_name = \"gpucluster\"\n",
"\n",
"try:\n",
" compute_target = ComputeTarget(workspace=ws, name=cluster_name)\n",
" print('Found existing compute target.')\n",
"except ComputeTargetException:\n",
" print('Creating a new compute target...')\n",
" compute_config = BatchAiCompute.provisioning_configuration(vm_size='STANDARD_NC6', \n",
" autoscale_enabled=True,\n",
" cluster_min_nodes=0, \n",
" cluster_max_nodes=4)\n",
"\n",
" # create the cluster\n",
" compute_target = ComputeTarget.create(ws, cluster_name, compute_config)\n",
"\n",
" compute_target.wait_for_completion(show_output=True)\n",
"\n",
" # Use the 'status' property to get a detailed status for the current cluster. \n",
" print(compute_target.status.serialize())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train model on the remote compute\n",
"Now that we have the cluster ready to go, let's run our distributed training job."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create a project directory\n",
"Create a directory that will contain all the necessary code from your local machine that you will need access to on the remote resource. This includes the training script, and any additional files your training script depends on."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"\n",
"project_folder = './tf-distr-ps'\n",
"os.makedirs(project_folder, exist_ok=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copy the training script `tf_mnist_replica.py` into this project directory."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import shutil\n",
"shutil.copy('tf_mnist_replica.py', project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create an experiment\n",
"Create an [Experiment](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture#experiment) to track all the runs in your workspace for this distributed TensorFlow tutorial. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Experiment\n",
"\n",
"experiment_name = 'tf-distr-ps'\n",
"experiment = Experiment(ws, name=experiment_name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create a TensorFlow estimator\n",
"The AML SDK's TensorFlow estimator enables you to easily submit TensorFlow training jobs for both single-node and distributed runs. For more information on the TensorFlow estimator, refer [here](https://docs.microsoft.com/azure/machine-learning/service/how-to-train-tensorflow)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.dnn import TensorFlow\n",
"\n",
"script_params={\n",
" '--num_gpus': 1\n",
"}\n",
"\n",
"estimator = TensorFlow(source_directory=project_folder,\n",
" compute_target=compute_target,\n",
" script_params=script_params,\n",
" entry_script='tf_mnist_replica.py',\n",
" node_count=2,\n",
" worker_count=2,\n",
" parameter_server_count=1, \n",
" distributed_backend='ps',\n",
" use_gpu=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The above code specifies that we will run our training script on `2` nodes, with two workers and one parameter server. In order to execute a native distributed TensorFlow run, you must provide the argument `distributed_backend='ps'`. Using this estimator with these settings, TensorFlow and its dependencies will be installed for you. However, if your script also uses other packages, make sure to install them via the `TensorFlow` constructor's `pip_packages` or `conda_packages` parameters."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Submit job\n",
"Run your experiment by submitting your estimator object. Note that this call is asynchronous."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run = experiment.submit(estimator)\n",
"print(run.get_details())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Monitor your run\n",
"You can monitor the progress of the run with a Jupyter widget. Like the run submission, the widget is asynchronous and provides live updates every 10-15 seconds until the job completes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.widgets import RunDetails\n",
"RunDetails(run).show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Alternatively, you can block until the script has completed training before running more code."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.wait_for_completion(show_output=True) # this provides a verbose log"
]
}
],
"metadata": {
"authors": [
{
"name": "minxia"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
},
"msauthor": "minxia"
},
"nbformat": 4,
"nbformat_minor": 2
}

0
training/distributed-tensorflow-with-parameter-server/tf_mnist_replica.py → training/05.distributed-tensorflow-with-parameter-server/tf_mnist_replica.py
View File

391
training/06.distributed-cntk-with-custom-docker/06.distributed-cntk-with-custom-docker.ipynb Normal file
View File

@@ -0,0 +1,391 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# 06. Distributed CNTK using custom docker images\n",
"In this tutorial, you will train a CNTK model on the [MNIST](http://yann.lecun.com/exdb/mnist/) dataset using a custom docker image and distributed training."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"* Understand the [architecture and terms](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture) introduced by Azure Machine Learning\n",
"* Go through the [00.configuration.ipynb]() notebook to:\n",
" * install the AML SDK\n",
" * create a workspace and its configuration file (`config.json`)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"Diagnostics"
]
},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize workspace\n",
"\n",
"Initialize a [Workspace](https://review.docs.microsoft.com/en-us/azure/machine-learning/service/concept-azure-machine-learning-architecture?branch=release-ignite-aml#workspace) object from the existing workspace you created in the Prerequisites step. `Workspace.from_config()` creates a workspace object from the details stored in `config.json`."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.workspace import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print('Workspace name: ' + ws.name, \n",
" 'Azure region: ' + ws.location, \n",
" 'Subscription id: ' + ws.subscription_id, \n",
" 'Resource group: ' + ws.resource_group, sep = '\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create a remote compute target\n",
"You will need to create a [compute target](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture#compute-target) to execute your training script on. In this tutorial, you create an [Azure Batch AI](https://docs.microsoft.com/azure/batch-ai/overview) cluster as your training compute resource. This code creates a cluster for you if it does not already exist in your workspace.\n",
"\n",
"**Creation of the cluster takes approximately 5 minutes.** If the cluster is already in your workspace this code will skip the cluster creation process."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import ComputeTarget, BatchAiCompute\n",
"from azureml.core.compute_target import ComputeTargetException\n",
"\n",
"# choose a name for your cluster\n",
"cluster_name = \"gpucluster\"\n",
"\n",
"try:\n",
" compute_target = ComputeTarget(workspace=ws, name=cluster_name)\n",
" print('Found existing compute target.')\n",
"except ComputeTargetException:\n",
" print('Creating a new compute target...')\n",
" compute_config = BatchAiCompute.provisioning_configuration(vm_size='STANDARD_NC6', \n",
" autoscale_enabled=True,\n",
" cluster_min_nodes=0, \n",
" cluster_max_nodes=4)\n",
"\n",
" # create the cluster\n",
" compute_target = ComputeTarget.create(ws, cluster_name, compute_config)\n",
"\n",
" compute_target.wait_for_completion(show_output=True)\n",
"\n",
" # Use the 'status' property to get a detailed status for the current cluster. \n",
" print(compute_target.status.serialize())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Upload training data\n",
"For this tutorial, we will be using the MNIST dataset.\n",
"\n",
"First, let's download the dataset. We've included the `install_mnist.py` script to download the data and convert it to a CNTK-supported format. Our data files will get written to a directory named `'mnist'`."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import install_mnist\n",
"\n",
"install_mnist.main('mnist')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"To make the data accessible for remote training, you will need to upload the data from your local machine to the cloud. AML provides a convenient way to do so via a [Datastore](https://docs.microsoft.com/azure/machine-learning/service/how-to-access-data). The datastore provides a mechanism for you to upload/download data, and interact with it from your remote compute targets. \n",
"\n",
"Each workspace is associated with a default datastore. In this tutorial, we will upload the training data to this default datastore, which we will then mount on the remote compute for training in the next section."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ds = ws.get_default_datastore()\n",
"print(ds.datastore_type, ds.account_name, ds.container_name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The following code will upload the training data to the path `./mnist` on the default datastore."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ds.upload(src_dir='./mnist', target_path='./mnist')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now let's get a reference to the path on the datastore with the training data. We can do so using the `path` method. In the next section, we can then pass this reference to our training script's `--data_dir` argument. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"path_on_datastore = 'mnist'\n",
"ds_data = ds.path(path_on_datastore)\n",
"print(ds_data)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train model on the remote compute\n",
"Now that we have the cluster ready to go, let's run our distributed training job."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create a project directory\n",
"Create a directory that will contain all the necessary code from your local machine that you will need access to on the remote resource. This includes the training script, and any additional files your training script depends on."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"\n",
"project_folder = './cntk-distr'\n",
"os.makedirs(project_folder, exist_ok=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copy the training script `cntk_distr_mnist.py` into this project directory."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import shutil\n",
"shutil.copy('cntk_distr_mnist.py', project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create an experiment\n",
"Create an [experiment](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture#experiment) to track all the runs in your workspace for this distributed CNTK tutorial. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Experiment\n",
"\n",
"experiment_name = 'cntk-distr'\n",
"experiment = Experiment(ws, name=experiment_name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create an Estimator\n",
"The AML SDK's base Estimator enables you to easily submit custom scripts for both single-node and distributed runs. You should this generic estimator for training code using frameworks such as sklearn or CNTK that don't have corresponding custom estimators. For more information on using the generic estimator, refer [here](https://docs.microsoft.com/azure/machine-learning/service/how-to-train-ml-models)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.estimator import *\n",
"\n",
"script_params = {\n",
" '--num_epochs': 50,\n",
" '--data_dir': ds_data.as_mount(),\n",
" '--output_dir': './outputs'\n",
"}\n",
"\n",
"estimator = Estimator(source_directory=project_folder,\n",
" compute_target=compute_target,\n",
" entry_script='cntk_distr_mnist.py',\n",
" script_params=script_params,\n",
" node_count=2,\n",
" process_count_per_node=1,\n",
" distributed_backend='mpi', \n",
" pip_packages=['cntk-gpu==2.6'],\n",
" custom_docker_base_image='microsoft/mmlspark:gpu-0.12',\n",
" use_gpu=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We would like to train our model using a [pre-built Docker container](https://hub.docker.com/r/microsoft/mmlspark/). To do so, specify the name of the docker image to the argument `custom_docker_base_image`. You can only provide images available in public docker repositories such as Docker Hub using this argument. To use an image from a private docker repository, use the constructor's `environment_definition` parameter instead. Finally, we provide the `cntk` package to `pip_packages` to install CNTK 2.6 on our custom image.\n",
"\n",
"The above code specifies that we will run our training script on `2` nodes, with one worker per node. In order to run distributed CNTK, which uses MPI, you must provide the argument `distributed_backend='mpi'`."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Submit job\n",
"Run your experiment by submitting your estimator object. Note that this call is asynchronous."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run = experiment.submit(estimator)\n",
"print(run.get_details())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Monitor your run\n",
"You can monitor the progress of the run with a Jupyter widget. Like the run submission, the widget is asynchronous and provides live updates every 10-15 seconds until the job completes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.widgets import RunDetails\n",
"RunDetails(run).show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Alternatively, you can block until the script has completed training before running more code."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"run.wait_for_completion(show_output=True)"
]
}
],
"metadata": {
"authors": [
{
"name": "minxia"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

117
training/06.distributed-cntk-with-custom-docker/cntk_distr_mnist.py Normal file
View File

@@ -0,0 +1,117 @@
# Copyright (c) Microsoft. All rights reserved.
# Licensed under the MIT license.
# Adapted from:
# https://github.com/Microsoft/CNTK/blob/master/Examples/Image/Classification/ConvNet/Python/ConvNet_MNIST.py
# ====================================================================
"""Train a CNN model on the MNIST dataset via distributed training."""
from __future__ import print_function
import numpy as np
import os
import cntk as C
import argparse
from cntk.train.training_session import CheckpointConfig, TestConfig
def create_reader(path, is_training, input_dim, label_dim, total_number_of_samples):
"""Define the reader for both training and evaluation action."""
return C.io.MinibatchSource(C.io.CTFDeserializer(path, C.io.StreamDefs(
features=C.io.StreamDef(field='features', shape=input_dim),
labels=C.io.StreamDef(field='labels', shape=label_dim)
)), randomize=is_training, max_samples=total_number_of_samples)
def convnet_mnist(max_epochs, output_dir, data_dir, debug_output=False, epoch_size=60000, minibatch_size=64):
"""Creates and trains a feedforward classification model for MNIST images."""
image_height = 28
image_width = 28
num_channels = 1
input_dim = image_height * image_width * num_channels
num_output_classes = 10
# Input variables denoting the features and label data
input_var = C.ops.input_variable((num_channels, image_height, image_width), np.float32)
label_var = C.ops.input_variable(num_output_classes, np.float32)
# Instantiate the feedforward classification model
scaled_input = C.ops.element_times(C.ops.constant(0.00390625), input_var)
with C.layers.default_options(activation=C.ops.relu, pad=False):
conv1 = C.layers.Convolution2D((5, 5), 32, pad=True)(scaled_input)
pool1 = C.layers.MaxPooling((3, 3), (2, 2))(conv1)
conv2 = C.layers.Convolution2D((3, 3), 48)(pool1)
pool2 = C.layers.MaxPooling((3, 3), (2, 2))(conv2)
conv3 = C.layers.Convolution2D((3, 3), 64)(pool2)
f4 = C.layers.Dense(96)(conv3)
drop4 = C.layers.Dropout(0.5)(f4)
z = C.layers.Dense(num_output_classes, activation=None)(drop4)
ce = C.losses.cross_entropy_with_softmax(z, label_var)
pe = C.metrics.classification_error(z, label_var)
# Load train data
reader_train = create_reader(os.path.join(data_dir, 'Train-28x28_cntk_text.txt'), True,
input_dim, num_output_classes, max_epochs * epoch_size)
# Load test data
reader_test = create_reader(os.path.join(data_dir, 'Test-28x28_cntk_text.txt'), False,
input_dim, num_output_classes, C.io.FULL_DATA_SWEEP)
# Set learning parameters
lr_per_sample = [0.001] * 10 + [0.0005] * 10 + [0.0001]
lr_schedule = C.learning_parameter_schedule_per_sample(lr_per_sample, epoch_size=epoch_size)
mms = [0] * 5 + [0.9990239141819757]
mm_schedule = C.learners.momentum_schedule_per_sample(mms, epoch_size=epoch_size)
# Instantiate the trainer object to drive the model training
local_learner = C.learners.momentum_sgd(z.parameters, lr_schedule, mm_schedule)
progress_printer = C.logging.ProgressPrinter(
tag='Training',
rank=C.train.distributed.Communicator.rank(),
num_epochs=max_epochs,
)
learner = C.train.distributed.data_parallel_distributed_learner(local_learner)
trainer = C.Trainer(z, (ce, pe), learner, progress_printer)
# define mapping from reader streams to network inputs
input_map_train = {
input_var: reader_train.streams.features,
label_var: reader_train.streams.labels
}
input_map_test = {
input_var: reader_test.streams.features,
label_var: reader_test.streams.labels
}
C.logging.log_number_of_parameters(z)
print()
C.train.training_session(
trainer=trainer,
mb_source=reader_train,
model_inputs_to_streams=input_map_train,
mb_size=minibatch_size,
progress_frequency=epoch_size,
checkpoint_config=CheckpointConfig(frequency=epoch_size,
filename=os.path.join(output_dir, "ConvNet_MNIST")),
test_config=TestConfig(reader_test, minibatch_size=minibatch_size,
model_inputs_to_streams=input_map_test)
).train()
return
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('--num_epochs', help='Total number of epochs to train', type=int, default='40')
parser.add_argument('--output_dir', help='Output directory', required=False, default='outputs')
parser.add_argument('--data_dir', help='Directory with training data')
args = parser.parse_args()
os.makedirs(args.output_dir, exist_ok=True)
convnet_mnist(args.num_epochs, args.output_dir, args.data_dir)
# Must call MPI finalize when process exit without exceptions
C.train.distributed.Communicator.finalize()

96
training/06.distributed-cntk-with-custom-docker/install_mnist.py Normal file
View File

@@ -0,0 +1,96 @@
# Copyright (c) Microsoft. All rights reserved.
# Licensed under the MIT license.
# Script from:
# https://github.com/Microsoft/CNTK/blob/master/Examples/Image/DataSets/MNIST/install_mnist.py
from __future__ import print_function
try:
from urllib.request import urlretrieve
except ImportError:
from urllib import urlretrieve
import gzip
import os
import struct
import numpy as np
def loadData(src, cimg):
print('Downloading ' + src)
gzfname, h = urlretrieve(src, './delete.me')
print('Done.')
try:
with gzip.open(gzfname) as gz:
n = struct.unpack('I', gz.read(4))
# Read magic number.
if n[0] != 0x3080000:
raise Exception('Invalid file: unexpected magic number.')
# Read number of entries.
n = struct.unpack('>I', gz.read(4))[0]
if n != cimg:
raise Exception('Invalid file: expected {0} entries.'.format(cimg))
crow = struct.unpack('>I', gz.read(4))[0]
ccol = struct.unpack('>I', gz.read(4))[0]
if crow != 28 or ccol != 28:
raise Exception('Invalid file: expected 28 rows/cols per image.')
# Read data.
res = np.fromstring(gz.read(cimg * crow * ccol), dtype=np.uint8)
finally:
os.remove(gzfname)
return res.reshape((cimg, crow * ccol))
def loadLabels(src, cimg):
print('Downloading ' + src)
gzfname, h = urlretrieve(src, './delete.me')
print('Done.')
try:
with gzip.open(gzfname) as gz:
n = struct.unpack('I', gz.read(4))
# Read magic number.
if n[0] != 0x1080000:
raise Exception('Invalid file: unexpected magic number.')
# Read number of entries.
n = struct.unpack('>I', gz.read(4))
if n[0] != cimg:
raise Exception('Invalid file: expected {0} rows.'.format(cimg))
# Read labels.
res = np.fromstring(gz.read(cimg), dtype=np.uint8)
finally:
os.remove(gzfname)
return res.reshape((cimg, 1))
def load(dataSrc, labelsSrc, cimg):
data = loadData(dataSrc, cimg)
labels = loadLabels(labelsSrc, cimg)
return np.hstack((data, labels))
def savetxt(filename, ndarray):
with open(filename, 'w') as f:
labels = list(map(' '.join, np.eye(10, dtype=np.uint).astype(str)))
for row in ndarray:
row_str = row.astype(str)
label_str = labels[row[-1]]
feature_str = ' '.join(row_str[:-1])
f.write('|labels {} |features {}\n'.format(label_str, feature_str))
def main(data_dir):
os.makedirs(data_dir, exist_ok=True)
train = load('http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz',
'http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz', 60000)
print('Writing train text file...')
train_txt = os.path.join(data_dir, 'Train-28x28_cntk_text.txt')
savetxt(train_txt, train)
print('Done.')
test = load('http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz',
'http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz', 10000)
print('Writing test text file...')
test_txt = os.path.join(data_dir, 'Test-28x28_cntk_text.txt')
savetxt(test_txt, test)
print('Done.')
if __name__ == "__main__":
main('mnist')

531
training/07.tensorboard/07.tensorboard.ipynb Normal file
View File

@@ -0,0 +1,531 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# 40. Tensorboard Integration with Run History\n",
"\n",
"1. Run a Tensorflow job locally and view its TB output live.\n",
"2. The same, for a DSVM.\n",
"3. And once more, with Batch AI.\n",
"4. Finally, we'll collect all of these historical runs together into a single Tensorboard graph."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prerequisites\n",
"Make sure you go through the [00. Installation and Configuration](00.configuration.ipynb) Notebook first if you haven't."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Diagnostics\n",
"Opt-in diagnostics for better experience, quality, and security of future releases."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"Diagnostics"
]
},
"outputs": [],
"source": [
"from azureml.telemetry import set_diagnostics_collection\n",
"set_diagnostics_collection(send_diagnostics = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize Workspace\n",
"\n",
"Initialize a workspace object from persisted configuration."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print('Workspace name: ' + ws.name, \n",
" 'Azure region: ' + ws.location, \n",
" 'Subscription id: ' + ws.subscription_id, \n",
" 'Resource group: ' + ws.resource_group, sep = '\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Set experiment name and create project\n",
"Choose a name for your run history container in the workspace, and create a folder for the project."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from os import path, makedirs\n",
"experiment_name = 'tensorboard-demo'\n",
"\n",
"# experiment folder\n",
"exp_dir = './sample_projects/' + experiment_name\n",
"\n",
"if not path.exists(exp_dir):\n",
" makedirs(exp_dir)\n",
"\n",
"# runs we started in this session, for the finale\n",
"runs = []"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Download Tensorflow Tensorboard demo code\n",
"\n",
"Tensorflow's repository has an MNIST demo with extensive Tensorboard instrumentation. We'll use it here for our purposes.\n",
"\n",
"Note that we don't need to make any code changes at all - the code works without modification from the Tensorflow repository."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import requests\n",
"import os\n",
"import tempfile\n",
"tf_code = requests.get(\"https://raw.githubusercontent.com/tensorflow/tensorflow/r1.8/tensorflow/examples/tutorials/mnist/mnist_with_summaries.py\")\n",
"with open(os.path.join(exp_dir, \"mnist_with_summaries.py\"), \"w\") as file:\n",
" file.write(tf_code.text)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configure and run locally\n",
"\n",
"We'll start by running this locally. While it might not initially seem that useful to use this for a local run - why not just run TB against the files generated locally? - even in this case there is some value to using this feature. Your local run will be registered in the run history, and your Tensorboard logs will be uploaded to the artifact store associated with this run. Later, you'll be able to restore the logs from any run, regardless of where it happened.\n",
"\n",
"Note that for this run, you will need to install Tensorflow on your local machine by yourself. Further, the Tensorboard module (that is, the one included with Tensorflow) must be accessible to this notebook's kernel, as the local machine is what runs Tensorboard."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.runconfig import RunConfiguration\n",
"\n",
"# Create a run configuration.\n",
"run_config = RunConfiguration()\n",
"run_config.environment.python.user_managed_dependencies = True\n",
"\n",
"# You can choose a specific Python environment by pointing to a Python path \n",
"#run_config.environment.python.interpreter_path = '/home/ninghai/miniconda3/envs/sdk2/bin/python'"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Experiment, Run\n",
"from azureml.core.script_run_config import ScriptRunConfig\n",
"import tensorflow as tf\n",
"\n",
"logs_dir = os.path.join(os.curdir, \"logs\")\n",
"data_dir = os.path.abspath(os.path.join(os.curdir, \"mnist_data\"))\n",
"\n",
"if not path.exists(data_dir):\n",
" makedirs(data_dir)\n",
"\n",
"os.environ[\"TEST_TMPDIR\"] = data_dir\n",
"\n",
"# Writing logs to ./logs results in their being uploaded to Artifact Service,\n",
"# and thus, made accessible to our Tensorboard instance.\n",
"arguments_list = [\"--log_dir\", logs_dir]\n",
"\n",
"# Create an experiment\n",
"exp = Experiment(ws, experiment_name)\n",
"\n",
"# If you would like the run to go for longer, add --max_steps 5000 to the arguments list:\n",
"# arguments_list += [\"--max_steps\", \"5000\"]\n",
"\n",
"script = ScriptRunConfig(exp_dir,\n",
" script=\"mnist_with_summaries.py\",\n",
" run_config=run_config,\n",
" arguments=arguments_list)\n",
"\n",
"run = exp.submit(script)\n",
"# You can also wait for the run to complete\n",
"# run.wait_for_completion(show_output=True)\n",
"runs.append(run)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Start Tensorboard\n",
"\n",
"Now, while the run is in progress, we just need to start Tensorboard with the run as its target, and it will begin streaming logs."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.contrib.tensorboard import Tensorboard\n",
"\n",
"# The Tensorboard constructor takes an array of runs, so be sure and pass it in as a single-element array here\n",
"tb = Tensorboard([run])\n",
"\n",
"# If successful, start() returns a string with the URI of the instance.\n",
"tb.start()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Stop Tensorboard\n",
"\n",
"When you're done, make sure to call the `stop()` method of the Tensorboard object, or it will stay running even after your job completes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"tb.stop()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Now, with a DSVM\n",
"\n",
"Tensorboard uploading works with all compute targets. Here we demonstrate it from a DSVM.\n",
"Note that the Tensorboard instance itself will be run by the notebook kernel. Again, this means this notebook's kernel must have access to the Tensorboard module.\n",
"\n",
"If you are unfamiliar with DSVM configuration, check [04. Train in a remote VM (Ubuntu DSVM)](04.train-on-remote-vm.ipynb) for a more detailed breakdown."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import DsvmCompute\n",
"from azureml.core.compute_target import ComputeTargetException\n",
"\n",
"compute_target_name = 'cpu-dsvm'\n",
"\n",
"try:\n",
" compute_target = DsvmCompute(workspace = ws, name = compute_target_name)\n",
" print('found existing:', compute_target.name)\n",
"except ComputeTargetException:\n",
" print('creating new.')\n",
" dsvm_config = DsvmCompute.provisioning_configuration(vm_size = \"Standard_D2_v2\")\n",
" compute_target = DsvmCompute.create(ws, name = compute_target_name, provisioning_configuration = dsvm_config)\n",
" compute_target.wait_for_completion(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Submit run using TensorFlow estimator\n",
"\n",
"Instead of manually configuring the DSVM environment, we can use the TensorFlow estimator and everything is set up automatically."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.train.dnn import TensorFlow\n",
"\n",
"script_params = {\"--log_dir\": \"./logs\"}\n",
"\n",
"# If you want the run to go longer, set --max-steps to a higher number.\n",
"# script_params[\"--max_steps\"] = \"5000\"\n",
"\n",
"tf_estimator = TensorFlow(source_directory=exp_dir,\n",
" compute_target=compute_target,\n",
" entry_script='mnist_with_summaries.py',\n",
" script_params=script_params)\n",
"\n",
"run = exp.submit(tf_estimator)\n",
"\n",
"runs.append(run)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Start Tensorboard with this run\n",
"\n",
"Just like before."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# The Tensorboard constructor takes an array of runs, so be sure and pass it in as a single-element array here\n",
"tb = Tensorboard([run])\n",
"\n",
"# If successful, start() returns a string with the URI of the instance.\n",
"tb.start()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Stop Tensorboard\n",
"\n",
"When you're done, make sure to call the `stop()` method of the Tensorboard object, or it will stay running even after your job completes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"tb.stop()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Once more, with a Batch AI cluster\n",
"\n",
"Just to prove we can, let's create a Batch AI cluster using MLC, and run our demo there, as well."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import BatchAiCompute\n",
"\n",
"clust_name = ws.name + \"cpu\"\n",
"\n",
"try:\n",
" # If you already have a cluster named this, we don't need to make a new one.\n",
" cts = ws.compute_targets \n",
" compute_target = cts[clust_name]\n",
" assert compute_target.type == 'BatchAI'\n",
"except:\n",
" # Let's make a new one here.\n",
" provisioning_config = BatchAiCompute.provisioning_configuration(cluster_max_nodes=2, \n",
" autoscale_enabled=True, \n",
" cluster_min_nodes=1,\n",
" vm_size='Standard_D11_V2')\n",
" \n",
" compute_target = BatchAiCompute.create(ws, clust_name, provisioning_config)\n",
" compute_target.wait_for_completion(show_output=True, min_node_count=1, timeout_in_minutes=20)\n",
"print(compute_target.name)\n",
" # For a more detailed view of current BatchAI cluster status, use the 'status' property \n",
" # print(compute_target.status.serialize())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Submit run using TensorFlow estimator\n",
"\n",
"Again, we can use the TensorFlow estimator and everything is set up automatically."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"script_params = {\"--log_dir\": \"./logs\"}\n",
"\n",
"# If you want the run to go longer, set --max-steps to a higher number.\n",
"# script_params[\"--max_steps\"] = \"5000\"\n",
"\n",
"tf_estimator = TensorFlow(source_directory=exp_dir,\n",
" compute_target=compute_target,\n",
" entry_script='mnist_with_summaries.py',\n",
" script_params=script_params)\n",
"\n",
"run = exp.submit(tf_estimator)\n",
"\n",
"runs.append(run)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Start Tensorboard with this run\n",
"\n",
"Once more..."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# The Tensorboard constructor takes an array of runs, so be sure and pass it in as a single-element array here\n",
"tb = Tensorboard([run])\n",
"\n",
"# If successful, start() returns a string with the URI of the instance.\n",
"tb.start()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Stop Tensorboard\n",
"\n",
"When you're done, make sure to call the `stop()` method of the Tensorboard object, or it will stay running even after your job completes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"tb.stop()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Finale\n",
"\n",
"If you've paid close attention, you'll have noticed that we've been saving the run objects in an array as we went along. We can start a Tensorboard instance that combines all of these run objects into a single process. This way, you can compare historical runs. You can even do this with live runs; if you made some of those previous runs longer via the `--max_steps` parameter, they might still be running, and you'll see them live in this instance as well."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# The Tensorboard constructor takes an array of runs...\n",
"# and it turns out that we have been building one of those all along.\n",
"tb = Tensorboard(runs)\n",
"\n",
"# If successful, start() returns a string with the URI of the instance.\n",
"tb.start()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Stop Tensorboard\n",
"\n",
"As you might already know, make sure to call the `stop()` method of the Tensorboard object, or it will stay running (until you kill the kernel associated with this notebook, at least)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"tb.stop()"
]
}
],
"metadata": {
"authors": [
{
"name": "roastala"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

Some files were not shown because too many files have changed in this diff Show More