{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "87f87f5b-05e4-475f-b4d7-d0c4672923d1",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sagemaker.tensorflow import TensorFlowProcessor\n",
    "from sagemaker.processing import  ProcessingOutput\n",
    "from sagemaker import get_execution_role\n",
    "from sagemaker.local import LocalSession\n",
    "import sagemaker"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9dd9927b-4081-4f1e-9cda-5b68dc5baf8d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create a local session for local testing. \n",
    "session = LocalSession()\n",
    "\n",
    "# A Role that SageMaker processing job can assume to perform feature engineering task for the labeled dataset.\n",
    "role = sagemaker.get_execution_role() # or overwrite with you own.\n",
    "\n",
    "# Name of the S3 bucket for input data and output dataset.\n",
    "default_bucket = session.default_bucket()\n",
    "\n",
    "# Name of the location on S3 where the SageMaker Ground Truth job outputs the annotation. \n",
    "# e.g. this could be found in the SageMaker Ground Truth labeling job (via console of API)\n",
    "# For instance, if using the SageMaker Console, navigate to Ground Truth -> Labeling Jobs -> Name of the labeling job -> Output dataset location\n",
    "sm_groundtruth_dataset_location=\"s3://<bucket_location>\"\n",
    "\n",
    "# S3 location for the output produced by the Sagemaker processing job.\n",
    "s3_output_location=\"s3://<s3 processing output location>\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "142316c5-1642-4007-94fe-1261107fa08e",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Initialize the TensorFlowProcessor\n",
    "tp = TensorFlowProcessor(\n",
    "    framework_version='2.5.1',\n",
    "    role=role,\n",
    "    instance_type='local',\n",
    "    instance_count=1,\n",
    "    base_job_name='processor-tracknet',\n",
    "    py_version='py37',\n",
    "    sagemaker_session=session\n",
    ")\n",
    "\n",
    "\n",
    "tp.run(\n",
    "    code='preprocess.py',\n",
    "    source_dir='../src/processing',\n",
    "    arguments=[\"--ground_truth_output_dataset_location\", sm_groundtruth_dataset_location],\n",
    "    outputs=[\n",
    "        ProcessingOutput(\n",
    "            output_name='processing_output',\n",
    "            source='/opt/ml/processing/output',\n",
    "            destination=s3_output_location\n",
    "        )\n",
    "    ]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f708f679-893b-4482-8683-8026e831cff5",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "conda_python3",
   "language": "python",
   "name": "conda_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": "3.10.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}