Azure
/
azureml-examples
зеркало из https://github.com/Azure/azureml-examples.git
1
0
Форкнуть 0
Код Задачи Пакеты Проекты Релизы Вики Активность

updated minimal image version (#3346)

This commit is contained in:
Satwik Kumar 2024-08-13 11:56:07 -07:00 коммит произвёл GitHub
Родитель a46bb8ef9b
Коммит cabd0a5846
Не найден ключ, соответствующий данной подписи
Идентификатор ключа GPG: B5690EEEBB952194
6 изменённых файлов: 459 добавлений и 459 удалений
Показать статистику Скачать Patch файл Скачать Diff файл Показать все файлы Свернуть все файлы

904
sdk/python/endpoints/online/custom-container/online-endpoints-custom-container-multimodel.ipynb
Просмотреть файл

@ -1,453 +1,453 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Create a multimodel deployment using a custom container"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"In this example we serve two models from the same endpoint and same deployment. \n",
"\n",
"Both model files are registered as a single model asset on Azure and loaded simultaneously in the scoring script. The scoring script parses each request for a \"model\" field and routes the payload accordingly. \n",
"\n",
"A custom container is not necessary for multimodel deployment - the custom container used here simply adds Python requirements via `pip` to an Azure Inference Minimal base image. An equivalent multimodel deployment can be created using a conda file-based environment."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 1. Configure parameters, assets, and clients"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 1.1 Set workspace details"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"subscription_id = \"<SUBSCRIPTION_ID>\"\n",
"resource_group = \"<RESOURCE_GROUP>\"\n",
"workspace_name = \"<AML_WORKSPACE_NAME>\""
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 1.2 Set endpoint details"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import random\n",
"\n",
"endpoint_name = f\"multimod-{random.randint(0,10000)}\""
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 1.3 Set asset paths\n",
"Define the directories containing the two model files as well as a directory which contains the scoring script"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"\n",
"base_path = \"../../../../../cli/endpoints/online/custom-container/minimal/multimodel\"\n",
"models_path = os.path.join(base_path, \"models\")\n",
"code_path = os.path.join(base_path, \"code\")\n",
"test_data_path = os.path.join(base_path, \"test-data\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 1.4 Examine the models folder\n",
"The models folder contains two models which will be loaded simultaneously by the scoring script."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"\n",
"os.listdir(models_path)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 1.5 Examine the scoring script\n",
"\n",
"The scoring script loads both models into a dictionary keyed on their name in the `init` function. In the run function, each request is parsed for a `model` key in the JSON to choose the model. The `data` payload is then passed to the appropriate model.\n",
"\n",
"```python \n",
"import joblib\n",
"import os\n",
"import pandas as pd\n",
"from pathlib import Path\n",
"import json\n",
"\n",
"models = None\n",
"\n",
"\n",
"def init():\n",
" global models\n",
" model_dir = Path(os.getenv(\"AZUREML_MODEL_DIR\")) / \"models\"\n",
" models = {m[:-4]: joblib.load(model_dir / m) for m in os.listdir(model_dir)}\n",
"\n",
"\n",
"def run(data):\n",
" data = json.loads(data)\n",
" model = models[data[\"model\"]]\n",
" payload = pd.DataFrame(data[\"data\"])\n",
" try:\n",
" ret = model.predict(payload)\n",
" return pd.DataFrame(ret).to_json()\n",
" except KeyError:\n",
" raise KeyError(\"No such model\")\n",
"\n",
"``` "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 1.6 Examine the Dockerfile\n",
"The dockerfile is located at `base_path` / `minimal-multimodel.dockerfile`. It uses the AzureML Inference Minimal CPU image as a base and adds relevant dependencies for the scoring script.\n",
"\n",
"```docker\n",
"FROM mcr.microsoft.com/azureml/minimal-ubuntu20.04-py38-cpu-inference:latest\n",
"\n",
"RUN pip install pandas numpy scikit-learn joblib\n",
"```"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 1.7 Create an MLClient instance"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azure.ai.ml import MLClient\n",
"from azure.ai.ml.entities import (\n",
" ManagedOnlineEndpoint,\n",
" ManagedOnlineDeployment,\n",
" Model,\n",
" CodeConfiguration,\n",
" Environment,\n",
" BuildContext,\n",
" ProbeSettings,\n",
")\n",
"from azure.identity import DefaultAzureCredential\n",
"\n",
"credential = DefaultAzureCredential()\n",
"ml_client = MLClient(\n",
" credential,\n",
" subscription_id=subscription_id,\n",
" resource_group_name=resource_group,\n",
" workspace_name=workspace_name,\n",
")"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"credential = DefaultAzureCredential()\n",
"ml_client = MLClient(\n",
" credential,\n",
" subscription_id=subscription_id,\n",
" resource_group_name=resource_group,\n",
" workspace_name=workspace_name,\n",
")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 2. Create an endpoint"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 2.1 Define and create the endpoint"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"endpoint = ManagedOnlineEndpoint(name=endpoint_name)\n",
"poller = ml_client.online_endpoints.begin_create_or_update(endpoint)\n",
"poller.wait()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 2.2 Confirm that creation was successful"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azure.ai.ml.exceptions import DeploymentException\n",
"\n",
"status = poller.status()\n",
"if status != \"Succeeded\":\n",
" raise DeploymentException(status)\n",
"else:\n",
" print(\"Endpoint creation succeeded\")\n",
" endpoint = poller.result()\n",
" print(endpoint)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 3. Create the deployment"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 3.1 Define the deployment"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"deployment = ManagedOnlineDeployment(\n",
" name=\"custom-container-multimodel\",\n",
" endpoint_name=endpoint_name,\n",
" model=Model(name=\"minimal-multimodel\", path=models_path),\n",
" code_configuration=CodeConfiguration(\n",
" code=code_path, scoring_script=\"minimal-multimodel-score.py\"\n",
" ),\n",
" environment=Environment(\n",
" name=\"minimal-multimodel\",\n",
" build=BuildContext(\n",
" path=base_path,\n",
" dockerfile_path=\"minimal-multimodel.dockerfile\",\n",
" ),\n",
" inference_config={\n",
" \"liveness_route\": {\"path\": \"/\", \"port\": 5001},\n",
" \"readiness_route\": {\"path\": \"/\", \"port\": 5001},\n",
" \"scoring_route\": {\"path\": \"/score\", \"port\": 5001},\n",
" },\n",
" ),\n",
" instance_type=\"Standard_DS3_v2\",\n",
" instance_count=1,\n",
")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 3.2 Create the deployment"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"poller = ml_client.online_deployments.begin_create_or_update(deployment)\n",
"poller.wait()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 3.3 Confirm that creation was successful"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"status = poller.status()\n",
"if status != \"Succeeded\":\n",
" raise DeploymentException(status)\n",
"else:\n",
" print(\"Deployment creation succeeded\")\n",
" deployment = poller.result()\n",
" print(deployment)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 3.4 Set traffic to 100% "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"endpoint.traffic = {\"custom-container-multimodel\": 100}\n",
"poller = ml_client.begin_create_or_update(endpoint)\n",
"poller.wait()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 4. Test the endpoint\n",
"The `model` JSON field in both JSON payloads indicates which model to score."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 4.1 Test the diabetes model"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import json\n",
"\n",
"res = ml_client.online_endpoints.invoke(\n",
" endpoint_name, request_file=os.path.join(test_data_path, \"diabetes-test-data.json\")\n",
")\n",
"print(json.loads(res))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 4.2 Test the iris model"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"res = ml_client.online_endpoints.invoke(\n",
" endpoint_name, request_file=os.path.join(test_data_path, \"iris-test-data.json\")\n",
")\n",
"print(json.loads(res))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 5. Delete assets"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 5.1 Delete the endpoint"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"poller = ml_client.online_endpoints.begin_delete(name=endpoint_name)"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3.10 - SDK V2",
"language": "python",
"name": "python310-sdkv2"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.10.6"
},
"vscode": {
"interpreter": {
"hash": "e530ce6154f972640d3e5b626ff5929e0848c7598c5ca98c96181f27d47882a4"
}
}
},
"nbformat": 4,
"nbformat_minor": 2
}
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Create a multimodel deployment using a custom container"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"In this example we serve two models from the same endpoint and same deployment. \n",
"\n",
"Both model files are registered as a single model asset on Azure and loaded simultaneously in the scoring script. The scoring script parses each request for a \"model\" field and routes the payload accordingly. \n",
"\n",
"A custom container is not necessary for multimodel deployment - the custom container used here simply adds Python requirements via `pip` to an Azure Inference Minimal base image. An equivalent multimodel deployment can be created using a conda file-based environment."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 1. Configure parameters, assets, and clients"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 1.1 Set workspace details"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"subscription_id = \"<SUBSCRIPTION_ID>\"\n",
"resource_group = \"<RESOURCE_GROUP>\"\n",
"workspace_name = \"<AML_WORKSPACE_NAME>\""
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 1.2 Set endpoint details"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import random\n",
"\n",
"endpoint_name = f\"multimod-{random.randint(0,10000)}\""
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 1.3 Set asset paths\n",
"Define the directories containing the two model files as well as a directory which contains the scoring script"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"\n",
"base_path = \"../../../../../cli/endpoints/online/custom-container/minimal/multimodel\"\n",
"models_path = os.path.join(base_path, \"models\")\n",
"code_path = os.path.join(base_path, \"code\")\n",
"test_data_path = os.path.join(base_path, \"test-data\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 1.4 Examine the models folder\n",
"The models folder contains two models which will be loaded simultaneously by the scoring script."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"\n",
"os.listdir(models_path)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 1.5 Examine the scoring script\n",
"\n",
"The scoring script loads both models into a dictionary keyed on their name in the `init` function. In the run function, each request is parsed for a `model` key in the JSON to choose the model. The `data` payload is then passed to the appropriate model.\n",
"\n",
"```python \n",
"import joblib\n",
"import os\n",
"import pandas as pd\n",
"from pathlib import Path\n",
"import json\n",
"\n",
"models = None\n",
"\n",
"\n",
"def init():\n",
" global models\n",
" model_dir = Path(os.getenv(\"AZUREML_MODEL_DIR\")) / \"models\"\n",
" models = {m[:-4]: joblib.load(model_dir / m) for m in os.listdir(model_dir)}\n",
"\n",
"\n",
"def run(data):\n",
" data = json.loads(data)\n",
" model = models[data[\"model\"]]\n",
" payload = pd.DataFrame(data[\"data\"])\n",
" try:\n",
" ret = model.predict(payload)\n",
" return pd.DataFrame(ret).to_json()\n",
" except KeyError:\n",
" raise KeyError(\"No such model\")\n",
"\n",
"``` "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 1.6 Examine the Dockerfile\n",
"The dockerfile is located at `base_path` / `minimal-multimodel.dockerfile`. It uses the AzureML Inference Minimal CPU image as a base and adds relevant dependencies for the scoring script.\n",
"\n",
"```docker\n",
"FROM mcr.microsoft.com/azureml/minimal-ubuntu22.04-py39-cpu-inference:latest\n",
"\n",
"RUN pip install pandas numpy scikit-learn joblib\n",
"```"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 1.7 Create an MLClient instance"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azure.ai.ml import MLClient\n",
"from azure.ai.ml.entities import (\n",
" ManagedOnlineEndpoint,\n",
" ManagedOnlineDeployment,\n",
" Model,\n",
" CodeConfiguration,\n",
" Environment,\n",
" BuildContext,\n",
" ProbeSettings,\n",
")\n",
"from azure.identity import DefaultAzureCredential\n",
"\n",
"credential = DefaultAzureCredential()\n",
"ml_client = MLClient(\n",
" credential,\n",
" subscription_id=subscription_id,\n",
" resource_group_name=resource_group,\n",
" workspace_name=workspace_name,\n",
")"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"credential = DefaultAzureCredential()\n",
"ml_client = MLClient(\n",
" credential,\n",
" subscription_id=subscription_id,\n",
" resource_group_name=resource_group,\n",
" workspace_name=workspace_name,\n",
")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 2. Create an endpoint"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 2.1 Define and create the endpoint"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"endpoint = ManagedOnlineEndpoint(name=endpoint_name)\n",
"poller = ml_client.online_endpoints.begin_create_or_update(endpoint)\n",
"poller.wait()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 2.2 Confirm that creation was successful"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azure.ai.ml.exceptions import DeploymentException\n",
"\n",
"status = poller.status()\n",
"if status != \"Succeeded\":\n",
" raise DeploymentException(status)\n",
"else:\n",
" print(\"Endpoint creation succeeded\")\n",
" endpoint = poller.result()\n",
" print(endpoint)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 3. Create the deployment"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 3.1 Define the deployment"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"deployment = ManagedOnlineDeployment(\n",
" name=\"custom-container-multimodel\",\n",
" endpoint_name=endpoint_name,\n",
" model=Model(name=\"minimal-multimodel\", path=models_path),\n",
" code_configuration=CodeConfiguration(\n",
" code=code_path, scoring_script=\"minimal-multimodel-score.py\"\n",
" ),\n",
" environment=Environment(\n",
" name=\"minimal-multimodel\",\n",
" build=BuildContext(\n",
" path=base_path,\n",
" dockerfile_path=\"minimal-multimodel.dockerfile\",\n",
" ),\n",
" inference_config={\n",
" \"liveness_route\": {\"path\": \"/\", \"port\": 5001},\n",
" \"readiness_route\": {\"path\": \"/\", \"port\": 5001},\n",
" \"scoring_route\": {\"path\": \"/score\", \"port\": 5001},\n",
" },\n",
" ),\n",
" instance_type=\"Standard_DS3_v2\",\n",
" instance_count=1,\n",
")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 3.2 Create the deployment"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"poller = ml_client.online_deployments.begin_create_or_update(deployment)\n",
"poller.wait()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 3.3 Confirm that creation was successful"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"status = poller.status()\n",
"if status != \"Succeeded\":\n",
" raise DeploymentException(status)\n",
"else:\n",
" print(\"Deployment creation succeeded\")\n",
" deployment = poller.result()\n",
" print(deployment)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 3.4 Set traffic to 100% "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"endpoint.traffic = {\"custom-container-multimodel\": 100}\n",
"poller = ml_client.begin_create_or_update(endpoint)\n",
"poller.wait()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 4. Test the endpoint\n",
"The `model` JSON field in both JSON payloads indicates which model to score."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 4.1 Test the diabetes model"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import json\n",
"\n",
"res = ml_client.online_endpoints.invoke(\n",
" endpoint_name, request_file=os.path.join(test_data_path, \"diabetes-test-data.json\")\n",
")\n",
"print(json.loads(res))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 4.2 Test the iris model"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"res = ml_client.online_endpoints.invoke(\n",
" endpoint_name, request_file=os.path.join(test_data_path, \"iris-test-data.json\")\n",
")\n",
"print(json.loads(res))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 5. Delete assets"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 5.1 Delete the endpoint"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"poller = ml_client.online_endpoints.begin_delete(name=endpoint_name)"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3.10 - SDK V2",
"language": "python",
"name": "python310-sdkv2"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.10.6"
},
"vscode": {
"interpreter": {
"hash": "e530ce6154f972640d3e5b626ff5929e0848c7598c5ca98c96181f27d47882a4"
}
}
},
"nbformat": 4,
"nbformat_minor": 2
}

2
sdk/python/endpoints/online/llm/semantic-kernel/environment/serving/Dockerfile
Просмотреть файл

@ -1,4 +1,4 @@
FROM mcr.microsoft.com/azureml/minimal-ubuntu20.04-py38-cpu-inference:latest
FROM mcr.microsoft.com/azureml/minimal-ubuntu22.04-py39-cpu-inference:latest
COPY . .

6
sdk/python/endpoints/online/managed/online-endpoints-inference-schema.ipynb
Просмотреть файл

@ -238,7 +238,7 @@
" code=\"inference-schema/code\", scoring_script=\"score-numpy.py\"\n",
" ),\n",
" environment=Environment(\n",
" image=\"mcr.microsoft.com/azureml/minimal-ubuntu20.04-py38-cpu-inference\",\n",
" image=\"mcr.microsoft.com/azureml/minimal-ubuntu22.04-py39-cpu-inference\",\n",
" conda_file=\"inference-schema/env.yml\",\n",
" ),\n",
" instance_type=\"Standard_DS3_v2\",\n",
@ -381,7 +381,7 @@
" code=\"inference-schema/code/\", scoring_script=\"score-standard.py\"\n",
" ),\n",
" environment=Environment(\n",
" image=\"mcr.microsoft.com/azureml/minimal-ubuntu20.04-py38-cpu-inference\",\n",
" image=\"mcr.microsoft.com/azureml/minimal-ubuntu22.04-py39-cpu-inference\",\n",
" conda_file=\"inference-schema/env.yml\",\n",
" ),\n",
" instance_type=\"Standard_DS3_v2\",\n",
@ -528,7 +528,7 @@
" code=\"inference-schema/code\", scoring_script=\"score-pandas.py\"\n",
" ),\n",
" environment=Environment(\n",
" image=\"mcr.microsoft.com/azureml/minimal-ubuntu20.04-py38-cpu-inference\",\n",
" image=\"mcr.microsoft.com/azureml/minimal-ubuntu22.04-py39-cpu-inference\",\n",
" conda_file=\"inference-schema/env.yml\",\n",
" ),\n",
" instance_type=\"Standard_DS3_v2\",\n",

2
sdk/python/endpoints/online/managed/online-endpoints-keyvault.ipynb
Просмотреть файл

@ -338,7 +338,7 @@
" ),\n",
" environment=Environment(\n",
" conda_file=\"keyvault/env.yml\",\n",
" image=\"mcr.microsoft.com/azureml/minimal-ubuntu20.04-py38-cpu-inference:latest\",\n",
" image=\"mcr.microsoft.com/azureml/minimal-ubuntu22.04-py39-cpu-inference:latest\",\n",
" ),\n",
" environment_variables={\n",
" \"KV_SECRET_MULTIPLIER\": f\"multiplier@https://{keyvault_name}.vault.azure.net\"\n",

2
sdk/python/endpoints/online/managed/online-endpoints-multimodel.ipynb
Просмотреть файл

@ -282,7 +282,7 @@
"source": [
"environment = Environment(\n",
" name=\"minimal-multimodel-conda\",\n",
" image=\"mcr.microsoft.com/azureml/minimal-ubuntu20.04-py38-cpu-inference\",\n",
" image=\"mcr.microsoft.com/azureml/minimal-ubuntu22.04-py39-cpu-inference\",\n",
" conda_file=conda_file_path,\n",
")\n",
"environment = ml_client.environments.create_or_update(environment)"

2
sdk/python/endpoints/online/managed/online-endpoints-openapi.ipynb
Просмотреть файл

@ -219,7 +219,7 @@
" code=\"openapi/code-decorated\", scoring_script=\"score.py\"\n",
" ),\n",
" environment=Environment(\n",
" image=\"mcr.microsoft.com/azureml/minimal-ubuntu20.04-py38-cpu-inference\",\n",
" image=\"mcr.microsoft.com/azureml/minimal-ubuntu22.04-py39-cpu-inference\",\n",
" conda_file=\"openapi/env.yml\",\n",
" ),\n",
" instance_type=\"Standard_DS3_v2\",\n",