{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "1540e9a4-128d-4a8b-b56c-450d2f7f44eb",
   "metadata": {},
   "source": [
    "# Notebook - Invoke SageMaker Inference Endpoint"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d889b717-a81a-4068-9c94-b0c18092ab84",
   "metadata": {},
   "source": [
    "**NOTICE: Before running this notebook, please track and wait for the run of the pipeline created by the accompanied notebook `kfp_with_sagemaker_training_s3.ipynb` to complete. Once it completes, use\n",
    "this notebook to involke a SageMaker Endpoint and submit a sample image for inference** \n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0e4c35a6-c776-46c9-9bae-e4eb484d473b",
   "metadata": {},
   "source": [
    "Before running this notebok, \n",
    "- Make sure that a sample image is uploaded to the notebook instance (below `1000_dog.png` is put in the folder where this notebook is located).\n",
    "- Update SageMaker Endpoint name below (`Endpoint-cnn-dist-job-2023-05-04-05-47-05-124` for example). There are two ways to retrieve the Endpoint name: \n",
    "    1. In the left pane of SageMaker Console, go to `Inference > Endpoint`\n",
    "    2. Go to the graph visualization of the run of the pipeline, click the component for `SageMaker - Deploy Model`, and go to `Input/Output > endpoint_name`"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c19c19ae-285f-4566-8552-c5151e944197",
   "metadata": {},
   "source": [
    "## Preparation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9e8ed44e-6959-4b6b-8f2a-557ac143bdf0",
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "import boto3\n",
    "import numpy as np\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a6ec76cf-4329-4009-9fa5-acc8456632b7",
   "metadata": {},
   "outputs": [],
   "source": [
    "# create a boto3 client for SageMaker\n",
    "client = boto3.client('runtime.sagemaker')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fbd42f21-8c71-4156-b41e-f8ab8d187d79",
   "metadata": {},
   "outputs": [],
   "source": [
    "# load a sample image\n",
    "file_name = '1000_dog.png'\n",
    "with open(file_name, 'rb') as f:\n",
    "    payload = f.read()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c17d6599-64ab-4319-9c99-9febf9c9991d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# map between label and index for CIFAR-10 (index 1 corresponds to 'automobile')\n",
    "labels = [\n",
    "    'airplane',\n",
    "    'automobile',\n",
    "    'bird',\n",
    "    'cat',\n",
    "    'deer',\n",
    "    'dog',\n",
    "    'frog',\n",
    "    'horse',\n",
    "    'ship',\n",
    "    'truck'\n",
    "]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0e801a31-5487-4098-a9f5-d8774ae9525d",
   "metadata": {},
   "source": [
    "## Invole SageMaker Endpoint for inference"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "297da926-cd3f-4519-85be-5b4ae09e35f0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# NOTE: ensure you update the endpoint name\n",
    "endpoint_name = 'Endpoint-cnn-dist-job-2023-05-04-05-47-05-124'\n",
    "\n",
    "response = client.invoke_endpoint(\n",
    "    EndpointName=endpoint_name, \n",
    "    ContentType='application/x-image', \n",
    "    Body=payload\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "847c1535-23f7-402e-b70a-61fe66697b5a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# show the prediction result\n",
    "\n",
    "pred_raw = json.loads(response['Body'].read().decode())\n",
    "pred_scores = pred_raw['score'][0]\n",
    "pred_index = np.argmax(pred_scores)\n",
    "pred_label = labels[pred_index]\n",
    "\n",
    "print(f\"prediction result: index {pred_index}, label {pred_label}\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a0a0da04-6d66-44d0-90b1-4c2df38be438",
   "metadata": {},
   "source": [
    "### Delete Endpoint"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "75520f74-52e6-4cfe-ab19-d00adc3f8269",
   "metadata": {},
   "outputs": [],
   "source": [
    "# # delete endpoint\n",
    "# sagemaker_client = boto3.client('sagemaker')\n",
    "# sagemaker_client.delete_endpoint(EndpointName=endpoint_name)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6f31fd76-6a6f-4417-8b1c-b2db656927e0",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}