microsoft
/
SynapseML
зеркало из https://github.com/microsoft/SynapseML.git
1
0
Форкнуть 0
Код Задачи Пакеты Проекты Релизы Вики Активность

docs: fix broken links (#2027) 

* docs: fix broken links

* Update docs/Explore Algorithms/AI Services/Advanced Usage - Async, Batching, and Multi-Key.ipynb
This commit is contained in:
Mark Hamilton 2023-07-20 17:25:27 +01:00 коммит произвёл GitHub
Родитель 81a36bd245
Коммит df2712afeb
Не найден ключ, соответствующий данной подписи
Идентификатор ключа GPG: 4AEE18F83AFDEB23
1 изменённых файлов: 193 добавлений и 193 удалений
Показать статистику Скачать Patch файл Скачать Diff файл Показать все файлы Свернуть все файлы

386
docs/Explore Algorithms/AI Services/Advanced Usage - Async, Batching, and Multi-Key.ipynb
Просмотреть файл

@ -2,88 +2,99 @@
"cells": [ "cells": [
{ {
"cell_type": "markdown", "cell_type": "markdown",
"source": [
"# Cognitive Services Advanced Guide: Asynchrony, Batching, Multi-Key"
],
"metadata": { "metadata": {
"application/vnd.databricks.v1+cell": { "application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {}, "cellMetadata": {},
"nuid": "1a39046d-a692-44c3-b673-78dfc1f97e08",
"inputWidgets": {}, "inputWidgets": {},
"nuid": "1a39046d-a692-44c3-b673-78dfc1f97e08",
"showTitle": false,
"title": "" "title": ""
} }
} },
"source": [
"# Cognitive Services Advanced Guide: Asynchrony, Batching, Multi-Key"
]
}, },
{ {
"cell_type": "markdown", "cell_type": "markdown",
"source": [
"## Step 1: Imports and Keys"
],
"metadata": { "metadata": {
"application/vnd.databricks.v1+cell": { "application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {}, "cellMetadata": {},
"nuid": "0f8d5274-46d5-4604-be41-1f3f5d481d9a",
"inputWidgets": {}, "inputWidgets": {},
"nuid": "0f8d5274-46d5-4604-be41-1f3f5d481d9a",
"showTitle": false,
"title": "" "title": ""
} }
} },
"source": [
"## Step 1: Imports and Keys"
]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null,
"metadata": {
"application/vnd.databricks.v1+cell": {
"cellMetadata": {},
"inputWidgets": {},
"nuid": "84829dd7-0e7d-4ee3-aa9e-c3aa6ef96c8d",
"showTitle": false,
"title": ""
}
},
"outputs": [],
"source": [ "source": [
"from synapse.ml.core.platform import find_secret\n", "from synapse.ml.core.platform import find_secret\n",
"\n", "\n",
"service_key = find_secret(\"cognitive-api-key\")\n", "service_key = find_secret(\"cognitive-api-key\")\n",
"service_loc = \"eastus\"" "service_loc = \"eastus\""
], ]
},
{
"cell_type": "markdown",
"metadata": { "metadata": {
"application/vnd.databricks.v1+cell": { "application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {}, "cellMetadata": {},
"nuid": "84829dd7-0e7d-4ee3-aa9e-c3aa6ef96c8d",
"inputWidgets": {}, "inputWidgets": {},
"nuid": "8a49dfe2-f00d-4db5-95d5-f119fc09e2ee",
"showTitle": false,
"title": "" "title": ""
} }
}, },
"outputs": [],
"execution_count": 0
},
{
"cell_type": "markdown",
"source": [ "source": [
"## Step 2: Basic Usage" "## Step 2: Basic Usage"
], ]
"metadata": {
"application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {},
"nuid": "8a49dfe2-f00d-4db5-95d5-f119fc09e2ee",
"inputWidgets": {},
"title": ""
}
}
}, },
{ {
"cell_type": "markdown", "cell_type": "markdown",
"metadata": {
"application/vnd.databricks.v1+cell": {
"cellMetadata": {},
"inputWidgets": {},
"nuid": "93d1a1d0-96b5-48a2-9248-0d9facdae679",
"showTitle": false,
"title": ""
}
},
"source": [ "source": [
"Image 1 | Image 2 | Image 3 \n", "Image 1 | Image 2 | Image 3 \n",
":-------------------------:|:-------------------------:|:----------------------:|\n", ":-------------------------:|:-------------------------:|:----------------------:|\n",
"!<img src=\"https://raw.githubusercontent.com/Azure-Samples/cognitive-services-sample-data-files/master/ComputerVision/Images/objects.jpg\" width=\"300\" /> | <img src=\"https://raw.githubusercontent.com/Azure-Samples/cognitive-services-sample-data-files/master/ComputerVision/Images/dog.jpg\" width=\"300\" /> | <img src=\"https://raw.githubusercontent.com/Azure-Samples/cognitive-services-sample-data-files/master/ComputerVision/Images/house.jpg\" width=\"300\" />" "!<img src=\"https://raw.githubusercontent.com/Azure-Samples/cognitive-services-sample-data-files/master/ComputerVision/Images/objects.jpg\" width=\"300\" /> | <img src=\"https://raw.githubusercontent.com/Azure-Samples/cognitive-services-sample-data-files/master/ComputerVision/Images/dog.jpg\" width=\"300\" /> | <img src=\"https://raw.githubusercontent.com/Azure-Samples/cognitive-services-sample-data-files/master/ComputerVision/Images/house.jpg\" width=\"300\" />"
], ]
"metadata": {
"application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {},
"nuid": "93d1a1d0-96b5-48a2-9248-0d9facdae679",
"inputWidgets": {},
"title": ""
}
}
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null,
"metadata": {
"application/vnd.databricks.v1+cell": {
"cellMetadata": {},
"inputWidgets": {},
"nuid": "9e1933f3-06b3-4dfd-a6a2-30d33d7da845",
"showTitle": false,
"title": ""
}
},
"outputs": [],
"source": [ "source": [
"from synapse.ml.cognitive.vision import AnalyzeImage\n", "from synapse.ml.cognitive.vision import AnalyzeImage\n",
"\n", "\n",
@ -108,85 +119,83 @@
")\n", ")\n",
"\n", "\n",
"image_results = analyzer.transform(image_df).cache()" "image_results = analyzer.transform(image_df).cache()"
], ]
"metadata": {
"application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {},
"nuid": "9e1933f3-06b3-4dfd-a6a2-30d33d7da845",
"inputWidgets": {},
"title": ""
}
},
"outputs": [],
"execution_count": 0
}, },
{ {
"cell_type": "markdown", "cell_type": "markdown",
"metadata": {
"application/vnd.databricks.v1+cell": {
"cellMetadata": {},
"inputWidgets": {},
"nuid": "8f759bfa-4b88-4659-a535-d768ddee9e4f",
"showTitle": false,
"title": ""
}
},
"source": [ "source": [
"#### First we'll look at the full response objects:" "#### First we'll look at the full response objects:"
], ]
"metadata": {
"application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {},
"nuid": "8f759bfa-4b88-4659-a535-d768ddee9e4f",
"inputWidgets": {},
"title": ""
}
}
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null,
"metadata": {
"application/vnd.databricks.v1+cell": {
"cellMetadata": {},
"inputWidgets": {},
"nuid": "9fae2ca5-f16a-460b-94ec-e433f24f7fb4",
"showTitle": false,
"title": ""
}
},
"outputs": [],
"source": [ "source": [
"display(image_results)" "display(image_results)"
], ]
"metadata": {
"application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {},
"nuid": "9fae2ca5-f16a-460b-94ec-e433f24f7fb4",
"inputWidgets": {},
"title": ""
}
},
"outputs": [],
"execution_count": 0
}, },
{ {
"cell_type": "markdown", "cell_type": "markdown",
"source": [
"#### We can select out just what we need:"
],
"metadata": { "metadata": {
"application/vnd.databricks.v1+cell": { "application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {}, "cellMetadata": {},
"nuid": "7b08b439-a505-4de3-a71e-63af30453163",
"inputWidgets": {}, "inputWidgets": {},
"nuid": "7b08b439-a505-4de3-a71e-63af30453163",
"showTitle": false,
"title": "" "title": ""
} }
} },
"source": [
"#### We can select out just what we need:"
]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"source": [ "execution_count": null,
"display(image_results.select(\"analysis_results.description.captions.text\"))"
],
"metadata": { "metadata": {
"application/vnd.databricks.v1+cell": { "application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {}, "cellMetadata": {},
"nuid": "88e738a6-f1bf-4077-8436-984aac858b1b",
"inputWidgets": {}, "inputWidgets": {},
"nuid": "88e738a6-f1bf-4077-8436-984aac858b1b",
"showTitle": false,
"title": "" "title": ""
} }
}, },
"outputs": [], "outputs": [],
"execution_count": 0 "source": [
"display(image_results.select(\"analysis_results.description.captions.text\"))"
]
}, },
{ {
"cell_type": "markdown", "cell_type": "markdown",
"metadata": {
"application/vnd.databricks.v1+cell": {
"cellMetadata": {},
"inputWidgets": {},
"nuid": "c6c2fd12-8c26-4f96-b0a5-0c55c04c182a",
"showTitle": false,
"title": ""
}
},
"source": [ "source": [
"#### What's going on under the hood\n", "#### What's going on under the hood\n",
"\n", "\n",
@ -194,99 +203,101 @@
"\n", "\n",
"When we call the cognitive service transformer, we start cognitive service clients on each of your spark workers.\n", "When we call the cognitive service transformer, we start cognitive service clients on each of your spark workers.\n",
"These clients send requests to the cloud, and turn the JSON responses into Spark Struct Types so that you can access any field that the service returns." "These clients send requests to the cloud, and turn the JSON responses into Spark Struct Types so that you can access any field that the service returns."
], ]
"metadata": {
"application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {},
"nuid": "c6c2fd12-8c26-4f96-b0a5-0c55c04c182a",
"inputWidgets": {},
"title": ""
}
}
}, },
{ {
"cell_type": "markdown", "cell_type": "markdown",
"metadata": {
"application/vnd.databricks.v1+cell": {
"cellMetadata": {},
"inputWidgets": {},
"nuid": "31618622-57db-4973-8ab8-1bab6d7efd2e",
"showTitle": false,
"title": ""
}
},
"source": [ "source": [
"## Step 3: Asynchronous Usage" "## Step 3: Asynchronous Usage"
], ]
"metadata": {
"application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {},
"nuid": "31618622-57db-4973-8ab8-1bab6d7efd2e",
"inputWidgets": {},
"title": ""
}
}
}, },
{ {
"cell_type": "markdown", "cell_type": "markdown",
"metadata": {
"application/vnd.databricks.v1+cell": {
"cellMetadata": {},
"inputWidgets": {},
"nuid": "8e7e5ace-71c2-4170-8b5d-350297b907db",
"showTitle": false,
"title": ""
}
},
"source": [ "source": [
"<img src=\"https://mmlspark.blob.core.windows.net/graphics/Cog%20Service%20NB/async_parallelism.svg\" width=\"700\"/>\n", "<img src=\"https://mmlspark.blob.core.windows.net/graphics/Cog%20Service%20NB/async_parallelism.svg\" width=\"700\"/>\n",
"\n", "\n",
"Apache Spark ordinarily parallelizes a computation to all of it's worker threads. When working with services however this parallelism doesent fully maximize throughput because workers sit idle as requests are processed on the server. The `concurrency` parameter makes sure that each worker can stay busy as they wait for requests to complete." "Apache Spark ordinarily parallelizes a computation to all of it's worker threads. When working with services however this parallelism doesent fully maximize throughput because workers sit idle as requests are processed on the server. The `concurrency` parameter makes sure that each worker can stay busy as they wait for requests to complete."
], ]
"metadata": {
"application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {},
"nuid": "8e7e5ace-71c2-4170-8b5d-350297b907db",
"inputWidgets": {},
"title": ""
}
}
}, },
{ {
"cell_type": "code", "cell_type": "code",
"source": [ "execution_count": null,
"display(analyzer.setConcurrency(3).transform(image_df))"
],
"metadata": { "metadata": {
"application/vnd.databricks.v1+cell": { "application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {}, "cellMetadata": {},
"nuid": "f874a63e-f22e-4c6f-9d54-83f93d140721",
"inputWidgets": {}, "inputWidgets": {},
"nuid": "f874a63e-f22e-4c6f-9d54-83f93d140721",
"showTitle": false,
"title": "" "title": ""
} }
}, },
"outputs": [], "outputs": [],
"execution_count": 0 "source": [
"display(analyzer.setConcurrency(3).transform(image_df))"
]
}, },
{ {
"cell_type": "markdown", "cell_type": "markdown",
"metadata": {
"application/vnd.databricks.v1+cell": {
"cellMetadata": {},
"inputWidgets": {},
"nuid": "f82c9d17-77db-44fa-8d1c-b0b7905c0e31",
"showTitle": false,
"title": ""
}
},
"source": [ "source": [
"#### Faster without extra hardware:\n", "#### Faster without extra hardware:\n",
"<img src=\"https://mmlspark.blob.core.windows.net/graphics/Cog%20Service%20NB/async_relative%20(2).png\" width=\"500\" />" "<img src=\"https://mmlspark.blob.core.windows.net/graphics/Cog%20Service%20NB/async_relative%20(2).png\" width=\"500\" />"
], ]
"metadata": {
"application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {},
"nuid": "f82c9d17-77db-44fa-8d1c-b0b7905c0e31",
"inputWidgets": {},
"title": ""
}
}
}, },
{ {
"cell_type": "markdown", "cell_type": "markdown",
"source": [
"## Step 4: Batching"
],
"metadata": { "metadata": {
"application/vnd.databricks.v1+cell": { "application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {}, "cellMetadata": {},
"nuid": "d54b3f5e-8d44-486f-97a3-0b8528934e73",
"inputWidgets": {}, "inputWidgets": {},
"nuid": "d54b3f5e-8d44-486f-97a3-0b8528934e73",
"showTitle": false,
"title": "" "title": ""
} }
} },
"source": [
"## Step 4: Batching"
]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null,
"metadata": {
"application/vnd.databricks.v1+cell": {
"cellMetadata": {},
"inputWidgets": {},
"nuid": "c3092f7b-105b-4171-9649-f04b189d76a0",
"showTitle": false,
"title": ""
}
},
"outputs": [],
"source": [ "source": [
"from synapse.ml.cognitive.text import TextSentiment\n", "from synapse.ml.cognitive.text import TextSentiment\n",
"\n", "\n",
@ -312,36 +323,36 @@
"\n", "\n",
"# Show the results of your text query\n", "# Show the results of your text query\n",
"display(sentiment.transform(text_df).select(\"text\", \"sentiment.document.sentiment\"))" "display(sentiment.transform(text_df).select(\"text\", \"sentiment.document.sentiment\"))"
], ]
},
{
"cell_type": "markdown",
"metadata": { "metadata": {
"application/vnd.databricks.v1+cell": { "application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {}, "cellMetadata": {},
"nuid": "c3092f7b-105b-4171-9649-f04b189d76a0",
"inputWidgets": {}, "inputWidgets": {},
"nuid": "ee4a9f18-d845-4059-9edd-9bd625a75a1a",
"showTitle": false,
"title": ""
}
},
"source": [
"## Step 5: Multi-Key"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"application/vnd.databricks.v1+cell": {
"cellMetadata": {},
"inputWidgets": {},
"nuid": "a6f89d8b-7cd1-42be-8310-62989c80deb2",
"showTitle": false,
"title": "" "title": ""
} }
}, },
"outputs": [], "outputs": [],
"execution_count": 0
},
{
"cell_type": "markdown",
"source": [
"## Step 5: Multi-Key"
],
"metadata": {
"application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {},
"nuid": "ee4a9f18-d845-4059-9edd-9bd625a75a1a",
"inputWidgets": {},
"title": ""
}
}
},
{
"cell_type": "code",
"source": [ "source": [
"from synapse.ml.cognitive.text import TextSentiment\n", "from synapse.ml.cognitive.text import TextSentiment\n",
"from pyspark.sql.functions import udf\n", "from pyspark.sql.functions import udf\n",
@ -359,71 +370,60 @@
"image_df2 = image_df.withColumn(\"key\", random_key())\n", "image_df2 = image_df.withColumn(\"key\", random_key())\n",
"\n", "\n",
"results = analyzer.setSubscriptionKeyCol(\"key\").transform(image_df2)" "results = analyzer.setSubscriptionKeyCol(\"key\").transform(image_df2)"
], ]
"metadata": {
"application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {},
"nuid": "a6f89d8b-7cd1-42be-8310-62989c80deb2",
"inputWidgets": {},
"title": ""
}
},
"outputs": [],
"execution_count": 0
}, },
{ {
"cell_type": "code", "cell_type": "code",
"source": [ "execution_count": null,
"display(results.select(\"key\", \"analysis_results.description.captions.text\"))"
],
"metadata": { "metadata": {
"application/vnd.databricks.v1+cell": { "application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {}, "cellMetadata": {},
"nuid": "c2f0ff6f-688e-4ca0-88eb-9eb8bda66786",
"inputWidgets": {}, "inputWidgets": {},
"nuid": "c2f0ff6f-688e-4ca0-88eb-9eb8bda66786",
"showTitle": false,
"title": "" "title": ""
} }
}, },
"outputs": [], "outputs": [],
"execution_count": 0 "source": [
"display(results.select(\"key\", \"analysis_results.description.captions.text\"))"
]
}, },
{ {
"cell_type": "markdown", "cell_type": "markdown",
"source": [
"## Learn More\n",
"- [Explore other cogntive services](https://microsoft.github.io/SynapseML/docs/Explore%20Algorithms/CognitiveServices%20-%20Overview/)\n",
"- [Read our paper \"Large-Scale Intelligent Microservices\"](https://arxiv.org/abs/2009.08044)"
],
"metadata": { "metadata": {
"application/vnd.databricks.v1+cell": { "application/vnd.databricks.v1+cell": {
"showTitle": false,
"cellMetadata": {}, "cellMetadata": {},
"nuid": "1ed7401d-28f7-4133-93e3-08e145772502",
"inputWidgets": {}, "inputWidgets": {},
"nuid": "1ed7401d-28f7-4133-93e3-08e145772502",
"showTitle": false,
"title": "" "title": ""
} }
} },
"source": [
"## Learn More\n",
"- [Explore other cogntive services](./Overview)\n",
"- [Read our paper \"Large-Scale Intelligent Microservices\"](https://arxiv.org/abs/2009.08044)"
]
} }
], ],
"metadata": { "metadata": {
"application/vnd.databricks.v1+notebook": { "application/vnd.databricks.v1+notebook": {
"notebookName": "CognitiveServices - Advanced Usage: Async, Batching, and Multi-Key",
"dashboards": [], "dashboards": [],
"language": "python",
"notebookMetadata": { "notebookMetadata": {
"pythonIndentUnit": 2 "pythonIndentUnit": 2
}, },
"language": "python", "notebookName": "CognitiveServices - Advanced Usage: Async, Batching, and Multi-Key",
"widgets": {}, "notebookOrigID": 3743502060540796,
"notebookOrigID": 3743502060540796 "widgets": {}
}, },
"kernelspec": { "kernelspec": {
"name": "python3", "display_name": "Python 3 (ipykernel)",
"language": "python", "language": "python",
"display_name": "Python 3 (ipykernel)" "name": "python3"
} }
}, },
"nbformat": 4, "nbformat": 4,
"nbformat_minor": 0 "nbformat_minor": 0
} }