 archived/data_analysis_of_ground_truth_image_classification_output.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Analysis Using a Ground Truth Image Classification Output Manifest\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "\n",
    "This notebook's CI test result for us-west-2 is as follows. CI test results in other regions can be found at the end of the notebook. \n",
    "\n",
    "![This us-west-2 badge failed to load. Check your device's internet connectivity, otherwise the service is currently unavailable](https://prod.us-west-2.tcx-beacon.docs.aws.dev/sagemaker-nb/us-west-2/ground_truth_labeling_jobs|data_analysis_of_ground_truth_image_classification_output|data_analysis_of_ground_truth_image_classification_output.ipynb)\n",
    "\n",
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "1. [Introduction](#Introduction)\n",
    "2. [Postprocess the Output Manifest](#Postprocess-the-output-manifest)\n",
    "3. [Plot class histograms](#Plott-class-histograms)\n",
    "4. [Plot annotated images](#Plot-annotated-images)\n",
    "  1. [Plot a small output sample](#Plot-a-small-output-sample)\n",
    "  2. [Plot the full results](#Plot-the-full-results)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Introduction\n",
    "This notebook demonstrates how to do basic analysis and sanity-checking on the results of a Ground Truth Image Classification job. We assume you have successfully completed a Ground Truth Image Classification annotation job called *MY_JOB_NAME* (of course your job can have a different name).\n",
    "\n",
    "The most important output of Ground Truth is the *output manifest*. To find the manifest, go to your AWS Console, and visit `Amazon SageMaker > Labeling jobs > MY_JOB_NAME`. Click on the URI under `Output  dataset location`. From there, go to `manifests / output`. Your output manifest is called simply `output.manifest`. Note the S3 output manifest path. For example, it could be `s3://MY_DATASET/output/MY_JOB_NAME/manifests/output/output.manifest`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import os\n",
    "import json\n",
    "import random\n",
    "from collections import namedtuple\n",
    "from collections import Counter\n",
    "\n",
    "import imageio\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib.backends.backend_pdf import PdfPages"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Postprocess the output manifest\n",
    "First, we will download the output manifest manfiest (please provide its S3 path in the cell below). Next, we will postprocess it to form four arrays:\n",
    "* `img_uris` contains the S3 URIs of all the images that Ground Truth annotated.\n",
    "* `labels` contains Ground Truth's labels for each image in `img_uris`.\n",
    "* `confidences` contains the confidence of each label in `labels`.\n",
    "* `human` is a flag array that contains 1 at indices corresponding to images annotated by human annotators, and 0 at indices corresponding to images annotated by Ground Truth's automated data labeling."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load the output manifest's annotations.\n",
    "OUTPUT_MANIFEST = (\n",
    "    \"<< your output.manifest S3 path here >>\"  # Replace with the S3 URI for your output manifest.\n",
    ")\n",
    "!aws s3 cp {OUTPUT_MANIFEST} 'output.manifest'\n",
    "\n",
    "with open(\"output.manifest\", \"r\") as f:\n",
    "    output = [json.loads(line.strip()) for line in f.readlines()]\n",
    "\n",
    "# Create data arrays.\n",
    "img_uris = [None] * len(output)\n",
    "confidences = np.zeros(len(output))\n",
    "labels = [None] * len(output)\n",
    "human = np.zeros(len(output))\n",
    "\n",
    "# Find the job name the manifest corresponds to.\n",
    "keys = list(output[0].keys())\n",
    "metakey = keys[np.where([(\"-metadata\" in k) for k in keys])[0][0]]\n",
    "jobname = metakey[:-9]\n",
    "\n",
    "# Extract the data.\n",
    "for datum_id, datum in enumerate(output):\n",
    "    img_uris[datum_id] = datum[\"source-ref\"]\n",
    "    labels[datum_id] = datum[metakey][\"class-name\"]\n",
    "    confidences[datum_id] = datum[metakey][\"confidence\"]\n",
    "    human[datum_id] = int(datum[metakey][\"human-annotated\"] == \"yes\")\n",
    "labels = np.array(labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plot class histograms\n",
    "Now, let's plot the class histograms. The next cell should produce three subplots:\n",
    "* The Left subplot shows the number of images annotated as belonging to each visual category. The categories will be sorted from the most to the least numerous. Each bar is divided into a 'human' and 'machine' part which shows how many images were annotated as given category by human annotators and by the automated data labeling mechanism.\n",
    "\n",
    "* The Middle subplot is the same as Left, except y-axis is in log-scale. This helps visualize unbalanced datasets where some categories contain orders of magnitude more images than other.\n",
    "\n",
    "* The Right subplot shows the average confidence of images in each category, separately for human and auto-annotated images."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compute the number of annotations in each class.\n",
    "n_classes = len(set(labels))\n",
    "sorted_clnames, class_sizes = zip(*Counter(labels).most_common(n_classes))\n",
    "\n",
    "# Find ids of human-annotated images.\n",
    "human_sizes = [human[labels == clname].sum() for clname in sorted_clnames]\n",
    "class_sizes = np.array(class_sizes)\n",
    "human_sizes = np.array(human_sizes)\n",
    "\n",
    "# Compute the average annotation confidence per class.\n",
    "human_confidences = np.array(\n",
    "    [confidences[np.logical_and(labels == clname, human)] for clname in sorted_clnames]\n",
    ")\n",
    "machine_confidences = [\n",
    "    confidences[np.logical_and(labels == clname, 1 - human)] for clname in sorted_clnames\n",
    "]\n",
    "\n",
    "# If there is no images annotated as a specific class, set the average class confidence to 0.\n",
    "for class_id in range(n_classes):\n",
    "    if human_confidences[class_id].size == 0:\n",
    "        human_confidences[class_id] = np.array([0])\n",
    "    if machine_confidences[class_id].size == 0:\n",
    "        machine_confidences[class_id] = np.array([0])\n",
    "\n",
    "plt.figure(figsize=(9, 3), facecolor=\"white\", dpi=100)\n",
    "plt.subplot(1, 3, 1)\n",
    "plt.title(\"Annotation histogram\")\n",
    "plt.bar(range(n_classes), human_sizes, color=\"gray\", hatch=\"/\", edgecolor=\"k\", label=\"human\")\n",
    "plt.bar(\n",
    "    range(n_classes),\n",
    "    class_sizes - human_sizes,\n",
    "    bottom=human_sizes,\n",
    "    color=\"gray\",\n",
    "    edgecolor=\"k\",\n",
    "    label=\"machine\",\n",
    ")\n",
    "plt.xticks(range(n_classes), sorted_clnames, rotation=90)\n",
    "plt.ylabel(\"Annotation Count\")\n",
    "plt.legend()\n",
    "\n",
    "plt.subplot(1, 3, 2)\n",
    "plt.title(\"Annotation histogram (logscale)\")\n",
    "plt.bar(range(n_classes), human_sizes, color=\"gray\", hatch=\"/\", edgecolor=\"k\", label=\"human\")\n",
    "plt.bar(\n",
    "    range(n_classes),\n",
    "    class_sizes - human_sizes,\n",
    "    bottom=human_sizes,\n",
    "    color=\"gray\",\n",
    "    edgecolor=\"k\",\n",
    "    label=\"machine\",\n",
    ")\n",
    "plt.xticks(range(n_classes), sorted_clnames, rotation=90)\n",
    "plt.yscale(\"log\")\n",
    "\n",
    "plt.subplot(1, 3, 3)\n",
    "plt.title(\"Mean confidences\")\n",
    "plt.bar(\n",
    "    np.arange(n_classes),\n",
    "    [conf.mean() for conf in human_confidences],\n",
    "    color=\"gray\",\n",
    "    hatch=\"/\",\n",
    "    edgecolor=\"k\",\n",
    "    width=0.4,\n",
    ")\n",
    "plt.bar(\n",
    "    np.arange(n_classes) + 0.4,\n",
    "    [conf.mean() for conf in machine_confidences],\n",
    "    color=\"gray\",\n",
    "    edgecolor=\"k\",\n",
    "    width=0.4,\n",
    ")\n",
    "plt.xticks(range(n_classes), sorted_clnames, rotation=90);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plot annotated images\n",
    "In any data science task, it is crucial to plot and inspect the results to check they make sense. In order to do this, we will \n",
    "1. Download the input images that Ground Truth annotated.\n",
    "2. Split them by annotated category and whether the annotation was done by human or the auto-labeling mechanism.\n",
    "3. Plot images in each category and human/auto-annoated class.\n",
    "\n",
    "We will download the input images to `LOCAL_IMAGE_DIR` you can choose in the next cell. Note that if this directory already contains images with the same filenames as your Ground Truth input images, we will not re-download the images.\n",
    "\n",
    "If your dataset is large and you do not wish to download and plot **all** the images, simply set `DATASET_SIZE` to a small number. We will pick a random subset of your data for plotting."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "LOCAL_IMG_DIR = '<< choose a local directory name to download the images to >>' # Replace with the name of a local directory to store images.\n",
    "DATASET_SIZE = len(img_uris) # Change this to a reasonable number if your dataset much larger than 10K images.\n",
    "\n",
    "subset_ids = np.random.choice(range(len(img_uris)), DATASET_SIZE, replace=False)\n",
    "img_uris = [img_uris[idx] for idx in subset_ids]\n",
    "labels = labels[subset_ids]\n",
    "confidences = confidences[subset_ids]\n",
    "human = human[subset_ids]\n",
    "\n",
    "img_fnames = [None] * len(output)\n",
    "for img_uri_id, img_uri in enumerate(img_uris):\n",
    "    target_fname = os.path.join(\n",
    "        LOCAL_IMG_DIR, img_uri.split('/')[-1])\n",
    "    if not os.path.isfile(target_fname):\n",
    "        !aws s3 cp {img_uri} {target_fname}\n",
    "    img_fnames[img_uri_id] = target_fname"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plot a small output sample\n",
    "The following cell will create two figures. The first plots `N_SHOW` images in each category, as annotated by humans. The second plots `N_SHOW` images in each category, as annotated by the auto-labeling mechanism. \n",
    "\n",
    "If any category contains less than `N_SHOW` images, that row will not be displayed. By default, `N_SHOW = 10`, but feel free to change this to any other small number."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "N_SHOW = 10\n",
    "\n",
    "plt.figure(figsize=(3 * N_SHOW, 2 + 3 * n_classes), facecolor=\"white\", dpi=60)\n",
    "for class_name_id, class_name in enumerate(sorted_clnames):\n",
    "    class_ids = np.where(np.logical_and(np.array(labels) == class_name, human))[0]\n",
    "    try:\n",
    "        show_ids = np.random.choice(class_ids, N_SHOW, replace=False)\n",
    "    except ValueError:\n",
    "        print(\"Not enough human annotations to show {}\".format(class_name))\n",
    "        continue\n",
    "    for show_id_id, show_id in enumerate(show_ids):\n",
    "        plt.subplot2grid((n_classes, N_SHOW), (class_name_id, show_id_id))\n",
    "        plt.title(\"human: \" + class_name)\n",
    "        plt.imshow(imageio.imread(img_fnames[show_id]))\n",
    "        plt.axis(\"off\")\n",
    "    plt.tight_layout()\n",
    "\n",
    "plt.figure(figsize=(3 * N_SHOW, 2 + 3 * n_classes), facecolor=\"white\", dpi=100)\n",
    "for class_name_id, class_name in enumerate(sorted_clnames):\n",
    "    class_ids = np.where(np.logical_and(np.array(labels) == class_name, 1 - human))[0]\n",
    "    try:\n",
    "        show_ids = np.random.choice(class_ids, N_SHOW, replace=False)\n",
    "    except ValueError:\n",
    "        print(\"Not enough machine annotations to show {}\".format(class_name))\n",
    "        continue\n",
    "    for show_id_id, show_id in enumerate(show_ids):\n",
    "        plt.subplot2grid((n_classes, N_SHOW), (class_name_id, show_id_id))\n",
    "        plt.title(\"machine: \" + class_name)\n",
    "        plt.imshow(imageio.imread(img_fnames[show_id]))\n",
    "        plt.axis(\"off\")\n",
    "    plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plot the full results\n",
    "Finally, we plot all the results to a large pdf file. The pdf (called `ground_truth.pdf`) will display 100 images per page. Each page will contain images belonging to the same category, and annotated either by human annotators or by the auto-labeling mechanism. You can use this pdf to investigate exactly which images were annotated as which class at a glance.\n",
    "\n",
    "This might take a while, and the resulting pdf might be very large. For a dataset of 1K images, the process takes only a minute and creates a 10MB-large pdf. You can set `N_SHOW_PER_CLASS` to a small number if you want to limit the max number of examples shown in each category."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "N_SHOW_PER_CLASS = np.inf\n",
    "plt.figure(figsize=(10, 10), facecolor=\"white\", dpi=100)\n",
    "\n",
    "with PdfPages(\"ground_truth.pdf\") as pdf:\n",
    "    for class_name in sorted_clnames:\n",
    "        # Plot images annotated as class_name by humans.\n",
    "        plt.clf()\n",
    "        plt.text(0.1, 0.5, s=\"Images annotated as {} by humans\".format(class_name), fontsize=20)\n",
    "        plt.axis(\"off\")\n",
    "\n",
    "        class_ids = np.where(np.logical_and(np.array(labels) == class_name, human))[0]\n",
    "        for img_id_id, img_id in enumerate(class_ids):\n",
    "            if img_id_id == N_SHOW_PER_CLASS:\n",
    "                break\n",
    "            if img_id_id % 100 == 0:\n",
    "                pdf.savefig()\n",
    "                plt.clf()\n",
    "                print(\n",
    "                    \"Plotting human annotations of {}, {}/{}...\".format(\n",
    "                        class_name, img_id_id, min(len(class_ids), N_SHOW_PER_CLASS)\n",
    "                    )\n",
    "                )\n",
    "            plt.subplot(10, 10, (img_id_id % 100) + 1)\n",
    "            plt.imshow(imageio.imread(img_fnames[img_id]), aspect=\"auto\")\n",
    "            plt.axis(\"off\")\n",
    "        pdf.savefig()\n",
    "\n",
    "        # Plot images annotated as class_name by machines.\n",
    "        plt.clf()\n",
    "        plt.text(0.1, 0.5, s=\"Images annotated as {} by machines\".format(class_name), fontsize=20)\n",
    "        plt.axis(\"off\")\n",
    "\n",
    "        class_ids = np.where(np.logical_and(np.array(labels) == class_name, 1 - human))[0]\n",
    "        for img_id_id, img_id in enumerate(class_ids):\n",
    "            if img_id_id == N_SHOW_PER_CLASS:\n",
    "                break\n",
    "            if img_id_id % 100 == 0:\n",
    "                pdf.savefig()\n",
    "                plt.clf()\n",
    "                print(\n",
    "                    \"Plotting machine annotations of {}, {}/{}...\".format(\n",
    "                        class_name, img_id_id, min(len(class_ids), N_SHOW_PER_CLASS)\n",
    "                    )\n",
    "                )\n",
    "            plt.subplot(10, 10, (img_id_id % 100) + 1)\n",
    "            plt.imshow(imageio.imread(img_fnames[img_id]), aspect=\"auto\")\n",
    "            plt.axis(\"off\")\n",
    "        pdf.savefig()\n",
    "plt.clf()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Notebook CI Test Results\n",
    "\n",
    "This notebook was tested in multiple regions. The test results are as follows, except for us-west-2 which is shown at the top of the notebook.\n",
    "\n",
    "![This us-east-1 badge failed to load. Check your device's internet connectivity, otherwise the service is currently unavailable](https://prod.us-west-2.tcx-beacon.docs.aws.dev/sagemaker-nb/us-east-1/ground_truth_labeling_jobs|data_analysis_of_ground_truth_image_classification_output|data_analysis_of_ground_truth_image_classification_output.ipynb)\n",
    "\n",
    "![This us-east-2 badge failed to load. Check your device's internet connectivity, otherwise the service is currently unavailable](https://prod.us-west-2.tcx-beacon.docs.aws.dev/sagemaker-nb/us-east-2/ground_truth_labeling_jobs|data_analysis_of_ground_truth_image_classification_output|data_analysis_of_ground_truth_image_classification_output.ipynb)\n",
    "\n",
    "![This us-west-1 badge failed to load. Check your device's internet connectivity, otherwise the service is currently unavailable](https://prod.us-west-2.tcx-beacon.docs.aws.dev/sagemaker-nb/us-west-1/ground_truth_labeling_jobs|data_analysis_of_ground_truth_image_classification_output|data_analysis_of_ground_truth_image_classification_output.ipynb)\n",
    "\n",
    "![This ca-central-1 badge failed to load. Check your device's internet connectivity, otherwise the service is currently unavailable](https://prod.us-west-2.tcx-beacon.docs.aws.dev/sagemaker-nb/ca-central-1/ground_truth_labeling_jobs|data_analysis_of_ground_truth_image_classification_output|data_analysis_of_ground_truth_image_classification_output.ipynb)\n",
    "\n",
    "![This sa-east-1 badge failed to load. Check your device's internet connectivity, otherwise the service is currently unavailable](https://prod.us-west-2.tcx-beacon.docs.aws.dev/sagemaker-nb/sa-east-1/ground_truth_labeling_jobs|data_analysis_of_ground_truth_image_classification_output|data_analysis_of_ground_truth_image_classification_output.ipynb)\n",
    "\n",
    "![This eu-west-1 badge failed to load. Check your device's internet connectivity, otherwise the service is currently unavailable](https://prod.us-west-2.tcx-beacon.docs.aws.dev/sagemaker-nb/eu-west-1/ground_truth_labeling_jobs|data_analysis_of_ground_truth_image_classification_output|data_analysis_of_ground_truth_image_classification_output.ipynb)\n",
    "\n",
    "![This eu-west-2 badge failed to load. Check your device's internet connectivity, otherwise the service is currently unavailable](https://prod.us-west-2.tcx-beacon.docs.aws.dev/sagemaker-nb/eu-west-2/ground_truth_labeling_jobs|data_analysis_of_ground_truth_image_classification_output|data_analysis_of_ground_truth_image_classification_output.ipynb)\n",
    "\n",
    "![This eu-west-3 badge failed to load. Check your device's internet connectivity, otherwise the service is currently unavailable](https://prod.us-west-2.tcx-beacon.docs.aws.dev/sagemaker-nb/eu-west-3/ground_truth_labeling_jobs|data_analysis_of_ground_truth_image_classification_output|data_analysis_of_ground_truth_image_classification_output.ipynb)\n",
    "\n",
    "![This eu-central-1 badge failed to load. Check your device's internet connectivity, otherwise the service is currently unavailable](https://prod.us-west-2.tcx-beacon.docs.aws.dev/sagemaker-nb/eu-central-1/ground_truth_labeling_jobs|data_analysis_of_ground_truth_image_classification_output|data_analysis_of_ground_truth_image_classification_output.ipynb)\n",
    "\n",
    "![This eu-north-1 badge failed to load. Check your device's internet connectivity, otherwise the service is currently unavailable](https://prod.us-west-2.tcx-beacon.docs.aws.dev/sagemaker-nb/eu-north-1/ground_truth_labeling_jobs|data_analysis_of_ground_truth_image_classification_output|data_analysis_of_ground_truth_image_classification_output.ipynb)\n",
    "\n",
    "![This ap-southeast-1 badge failed to load. Check your device's internet connectivity, otherwise the service is currently unavailable](https://prod.us-west-2.tcx-beacon.docs.aws.dev/sagemaker-nb/ap-southeast-1/ground_truth_labeling_jobs|data_analysis_of_ground_truth_image_classification_output|data_analysis_of_ground_truth_image_classification_output.ipynb)\n",
    "\n",
    "![This ap-southeast-2 badge failed to load. Check your device's internet connectivity, otherwise the service is currently unavailable](https://prod.us-west-2.tcx-beacon.docs.aws.dev/sagemaker-nb/ap-southeast-2/ground_truth_labeling_jobs|data_analysis_of_ground_truth_image_classification_output|data_analysis_of_ground_truth_image_classification_output.ipynb)\n",
    "\n",
    "![This ap-northeast-1 badge failed to load. Check your device's internet connectivity, otherwise the service is currently unavailable](https://prod.us-west-2.tcx-beacon.docs.aws.dev/sagemaker-nb/ap-northeast-1/ground_truth_labeling_jobs|data_analysis_of_ground_truth_image_classification_output|data_analysis_of_ground_truth_image_classification_output.ipynb)\n",
    "\n",
    "![This ap-northeast-2 badge failed to load. Check your device's internet connectivity, otherwise the service is currently unavailable](https://prod.us-west-2.tcx-beacon.docs.aws.dev/sagemaker-nb/ap-northeast-2/ground_truth_labeling_jobs|data_analysis_of_ground_truth_image_classification_output|data_analysis_of_ground_truth_image_classification_output.ipynb)\n",
    "\n",
    "![This ap-south-1 badge failed to load. Check your device's internet connectivity, otherwise the service is currently unavailable](https://prod.us-west-2.tcx-beacon.docs.aws.dev/sagemaker-nb/ap-south-1/ground_truth_labeling_jobs|data_analysis_of_ground_truth_image_classification_output|data_analysis_of_ground_truth_image_classification_output.ipynb)\n"
   ]
  }
 ],
 "metadata": {
  "availableInstances": [
   {
    "_defaultOrder": 0,
    "_isFastLaunch": true,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 4,
    "name": "ml.t3.medium",
    "vcpuNum": 2
   },
   {
    "_defaultOrder": 1,
    "_isFastLaunch": false,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 8,
    "name": "ml.t3.large",
    "vcpuNum": 2
   },
   {
    "_defaultOrder": 2,
    "_isFastLaunch": false,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 16,
    "name": "ml.t3.xlarge",
    "vcpuNum": 4
   },
   {
    "_defaultOrder": 3,
    "_isFastLaunch": false,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 32,
    "name": "ml.t3.2xlarge",
    "vcpuNum": 8
   },
   {
    "_defaultOrder": 4,
    "_isFastLaunch": true,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 8,
    "name": "ml.m5.large",
    "vcpuNum": 2
   },
   {
    "_defaultOrder": 5,
    "_isFastLaunch": false,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 16,
    "name": "ml.m5.xlarge",
    "vcpuNum": 4
   },
   {
    "_defaultOrder": 6,
    "_isFastLaunch": false,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 32,
    "name": "ml.m5.2xlarge",
    "vcpuNum": 8
   },
   {
    "_defaultOrder": 7,
    "_isFastLaunch": false,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 64,
    "name": "ml.m5.4xlarge",
    "vcpuNum": 16
   },
   {
    "_defaultOrder": 8,
    "_isFastLaunch": false,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 128,
    "name": "ml.m5.8xlarge",
    "vcpuNum": 32
   },
   {
    "_defaultOrder": 9,
    "_isFastLaunch": false,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 192,
    "name": "ml.m5.12xlarge",
    "vcpuNum": 48
   },
   {
    "_defaultOrder": 10,
    "_isFastLaunch": false,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 256,
    "name": "ml.m5.16xlarge",
    "vcpuNum": 64
   },
   {
    "_defaultOrder": 11,
    "_isFastLaunch": false,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 384,
    "name": "ml.m5.24xlarge",
    "vcpuNum": 96
   },
   {
    "_defaultOrder": 12,
    "_isFastLaunch": false,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 8,
    "name": "ml.m5d.large",
    "vcpuNum": 2
   },
   {
    "_defaultOrder": 13,
    "_isFastLaunch": false,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 16,
    "name": "ml.m5d.xlarge",
    "vcpuNum": 4
   },
   {
    "_defaultOrder": 14,
    "_isFastLaunch": false,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 32,
    "name": "ml.m5d.2xlarge",
    "vcpuNum": 8
   },
   {
    "_defaultOrder": 15,
    "_isFastLaunch": false,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 64,
    "name": "ml.m5d.4xlarge",
    "vcpuNum": 16
   },
   {
    "_defaultOrder": 16,
    "_isFastLaunch": false,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 128,
    "name": "ml.m5d.8xlarge",
    "vcpuNum": 32
   },
   {
    "_defaultOrder": 17,
    "_isFastLaunch": false,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 192,
    "name": "ml.m5d.12xlarge",
    "vcpuNum": 48
   },
   {
    "_defaultOrder": 18,
    "_isFastLaunch": false,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 256,
    "name": "ml.m5d.16xlarge",
    "vcpuNum": 64
   },
   {
    "_defaultOrder": 19,
    "_isFastLaunch": false,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 384,
    "name": "ml.m5d.24xlarge",
    "vcpuNum": 96
   },
   {
    "_defaultOrder": 20,
    "_isFastLaunch": false,
    "category": "General purpose",
    "gpuNum": 0,
    "hideHardwareSpecs": true,
    "memoryGiB": 0,
    "name": "ml.geospatial.interactive",
    "supportedImageNames": [
     "sagemaker-geospatial-v1-0"
    ],
    "vcpuNum": 0
   },
   {
    "_defaultOrder": 21,
    "_isFastLaunch": true,
    "category": "Compute optimized",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 4,
    "name": "ml.c5.large",
    "vcpuNum": 2
   },
   {
    "_defaultOrder": 22,
    "_isFastLaunch": false,
    "category": "Compute optimized",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 8,
    "name": "ml.c5.xlarge",
    "vcpuNum": 4
   },
   {
    "_defaultOrder": 23,
    "_isFastLaunch": false,
    "category": "Compute optimized",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 16,
    "name": "ml.c5.2xlarge",
    "vcpuNum": 8
   },
   {
    "_defaultOrder": 24,
    "_isFastLaunch": false,
    "category": "Compute optimized",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 32,
    "name": "ml.c5.4xlarge",
    "vcpuNum": 16
   },
   {
    "_defaultOrder": 25,
    "_isFastLaunch": false,
    "category": "Compute optimized",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 72,
    "name": "ml.c5.9xlarge",
    "vcpuNum": 36
   },
   {
    "_defaultOrder": 26,
    "_isFastLaunch": false,
    "category": "Compute optimized",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 96,
    "name": "ml.c5.12xlarge",
    "vcpuNum": 48
   },
   {
    "_defaultOrder": 27,
    "_isFastLaunch": false,
    "category": "Compute optimized",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 144,
    "name": "ml.c5.18xlarge",
    "vcpuNum": 72
   },
   {
    "_defaultOrder": 28,
    "_isFastLaunch": false,
    "category": "Compute optimized",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 192,
    "name": "ml.c5.24xlarge",
    "vcpuNum": 96
   },
   {
    "_defaultOrder": 29,
    "_isFastLaunch": true,
    "category": "Accelerated computing",
    "gpuNum": 1,
    "hideHardwareSpecs": false,
    "memoryGiB": 16,
    "name": "ml.g4dn.xlarge",
    "vcpuNum": 4
   },
   {
    "_defaultOrder": 30,
    "_isFastLaunch": false,
    "category": "Accelerated computing",
    "gpuNum": 1,
    "hideHardwareSpecs": false,
    "memoryGiB": 32,
    "name": "ml.g4dn.2xlarge",
    "vcpuNum": 8
   },
   {
    "_defaultOrder": 31,
    "_isFastLaunch": false,
    "category": "Accelerated computing",
    "gpuNum": 1,
    "hideHardwareSpecs": false,
    "memoryGiB": 64,
    "name": "ml.g4dn.4xlarge",
    "vcpuNum": 16
   },
   {
    "_defaultOrder": 32,
    "_isFastLaunch": false,
    "category": "Accelerated computing",
    "gpuNum": 1,
    "hideHardwareSpecs": false,
    "memoryGiB": 128,
    "name": "ml.g4dn.8xlarge",
    "vcpuNum": 32
   },
   {
    "_defaultOrder": 33,
    "_isFastLaunch": false,
    "category": "Accelerated computing",
    "gpuNum": 4,
    "hideHardwareSpecs": false,
    "memoryGiB": 192,
    "name": "ml.g4dn.12xlarge",
    "vcpuNum": 48
   },
   {
    "_defaultOrder": 34,
    "_isFastLaunch": false,
    "category": "Accelerated computing",
    "gpuNum": 1,
    "hideHardwareSpecs": false,
    "memoryGiB": 256,
    "name": "ml.g4dn.16xlarge",
    "vcpuNum": 64
   },
   {
    "_defaultOrder": 35,
    "_isFastLaunch": false,
    "category": "Accelerated computing",
    "gpuNum": 1,
    "hideHardwareSpecs": false,
    "memoryGiB": 61,
    "name": "ml.p3.2xlarge",
    "vcpuNum": 8
   },
   {
    "_defaultOrder": 36,
    "_isFastLaunch": false,
    "category": "Accelerated computing",
    "gpuNum": 4,
    "hideHardwareSpecs": false,
    "memoryGiB": 244,
    "name": "ml.p3.8xlarge",
    "vcpuNum": 32
   },
   {
    "_defaultOrder": 37,
    "_isFastLaunch": false,
    "category": "Accelerated computing",
    "gpuNum": 8,
    "hideHardwareSpecs": false,
    "memoryGiB": 488,
    "name": "ml.p3.16xlarge",
    "vcpuNum": 64
   },
   {
    "_defaultOrder": 38,
    "_isFastLaunch": false,
    "category": "Accelerated computing",
    "gpuNum": 8,
    "hideHardwareSpecs": false,
    "memoryGiB": 768,
    "name": "ml.p3dn.24xlarge",
    "vcpuNum": 96
   },
   {
    "_defaultOrder": 39,
    "_isFastLaunch": false,
    "category": "Memory Optimized",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 16,
    "name": "ml.r5.large",
    "vcpuNum": 2
   },
   {
    "_defaultOrder": 40,
    "_isFastLaunch": false,
    "category": "Memory Optimized",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 32,
    "name": "ml.r5.xlarge",
    "vcpuNum": 4
   },
   {
    "_defaultOrder": 41,
    "_isFastLaunch": false,
    "category": "Memory Optimized",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 64,
    "name": "ml.r5.2xlarge",
    "vcpuNum": 8
   },
   {
    "_defaultOrder": 42,
    "_isFastLaunch": false,
    "category": "Memory Optimized",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 128,
    "name": "ml.r5.4xlarge",
    "vcpuNum": 16
   },
   {
    "_defaultOrder": 43,
    "_isFastLaunch": false,
    "category": "Memory Optimized",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 256,
    "name": "ml.r5.8xlarge",
    "vcpuNum": 32
   },
   {
    "_defaultOrder": 44,
    "_isFastLaunch": false,
    "category": "Memory Optimized",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 384,
    "name": "ml.r5.12xlarge",
    "vcpuNum": 48
   },
   {
    "_defaultOrder": 45,
    "_isFastLaunch": false,
    "category": "Memory Optimized",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 512,
    "name": "ml.r5.16xlarge",
    "vcpuNum": 64
   },
   {
    "_defaultOrder": 46,
    "_isFastLaunch": false,
    "category": "Memory Optimized",
    "gpuNum": 0,
    "hideHardwareSpecs": false,
    "memoryGiB": 768,
    "name": "ml.r5.24xlarge",
    "vcpuNum": 96
   },
   {
    "_defaultOrder": 47,
    "_isFastLaunch": false,
    "category": "Accelerated computing",
    "gpuNum": 1,
    "hideHardwareSpecs": false,
    "memoryGiB": 16,
    "name": "ml.g5.xlarge",
    "vcpuNum": 4
   },
   {
    "_defaultOrder": 48,
    "_isFastLaunch": false,
    "category": "Accelerated computing",
    "gpuNum": 1,
    "hideHardwareSpecs": false,
    "memoryGiB": 32,
    "name": "ml.g5.2xlarge",
    "vcpuNum": 8
   },
   {
    "_defaultOrder": 49,
    "_isFastLaunch": false,
    "category": "Accelerated computing",
    "gpuNum": 1,
    "hideHardwareSpecs": false,
    "memoryGiB": 64,
    "name": "ml.g5.4xlarge",
    "vcpuNum": 16
   },
   {
    "_defaultOrder": 50,
    "_isFastLaunch": false,
    "category": "Accelerated computing",
    "gpuNum": 1,
    "hideHardwareSpecs": false,
    "memoryGiB": 128,
    "name": "ml.g5.8xlarge",
    "vcpuNum": 32
   },
   {
    "_defaultOrder": 51,
    "_isFastLaunch": false,
    "category": "Accelerated computing",
    "gpuNum": 1,
    "hideHardwareSpecs": false,
    "memoryGiB": 256,
    "name": "ml.g5.16xlarge",
    "vcpuNum": 64
   },
   {
    "_defaultOrder": 52,
    "_isFastLaunch": false,
    "category": "Accelerated computing",
    "gpuNum": 4,
    "hideHardwareSpecs": false,
    "memoryGiB": 192,
    "name": "ml.g5.12xlarge",
    "vcpuNum": 48
   },
   {
    "_defaultOrder": 53,
    "_isFastLaunch": false,
    "category": "Accelerated computing",
    "gpuNum": 4,
    "hideHardwareSpecs": false,
    "memoryGiB": 384,
    "name": "ml.g5.24xlarge",
    "vcpuNum": 96
   },
   {
    "_defaultOrder": 54,
    "_isFastLaunch": false,
    "category": "Accelerated computing",
    "gpuNum": 8,
    "hideHardwareSpecs": false,
    "memoryGiB": 768,
    "name": "ml.g5.48xlarge",
    "vcpuNum": 192
   },
   {
    "_defaultOrder": 55,
    "_isFastLaunch": false,
    "category": "Accelerated computing",
    "gpuNum": 8,
    "hideHardwareSpecs": false,
    "memoryGiB": 1152,
    "name": "ml.p4d.24xlarge",
    "vcpuNum": 96
   },
   {
    "_defaultOrder": 56,
    "_isFastLaunch": false,
    "category": "Accelerated computing",
    "gpuNum": 8,
    "hideHardwareSpecs": false,
    "memoryGiB": 1152,
    "name": "ml.p4de.24xlarge",
    "vcpuNum": 96
   }
  ],
  "instance_type": "ml.t3.medium",
  "kernelspec": {
   "display_name": "Python 3 (Data Science 3.0)",
   "language": "python",
   "name": "python3__SAGEMAKER_INTERNAL__arn:aws:sagemaker:us-east-1:081325390199:image/sagemaker-data-science-310-v1"
  },
  "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"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}