{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "toCy3v03Dwx7"
   },
   "source": [
    "##### Copyright 2021 The TensorFlow Hub Authors.\n",
    "\n",
    "Licensed under the Apache License, Version 2.0 (the \"License\");"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:54:37.294201Z",
     "iopub.status.busy": "2022-12-14T21:54:37.293750Z",
     "iopub.status.idle": "2022-12-14T21:54:37.298133Z",
     "shell.execute_reply": "2022-12-14T21:54:37.297520Z"
    },
    "id": "QKe-ubNcDvgv"
   },
   "outputs": [],
   "source": [
    "# Copyright 2021 The TensorFlow Hub Authors. All Rights Reserved.\n",
    "#\n",
    "# Licensed under the Apache License, Version 2.0 (the \"License\");\n",
    "# you may not use this file except in compliance with the License.\n",
    "# You may obtain a copy of the License at\n",
    "#\n",
    "#     http://www.apache.org/licenses/LICENSE-2.0\n",
    "#\n",
    "# Unless required by applicable law or agreed to in writing, software\n",
    "# distributed under the License is distributed on an \"AS IS\" BASIS,\n",
    "# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n",
    "# See the License for the specific language governing permissions and\n",
    "# limitations under the License.\n",
    "# =============================================================================="
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "qFdPvlXBOdUN"
   },
   "source": [
    "# 스트리밍 동작 인식을 위한 MoViNet"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "MfBg1C5NB3X0"
   },
   "source": [
    "<table class=\"tfo-notebook-buttons\" align=\"left\">\n",
    "  <td><a target=\"_blank\" href=\"https://www.tensorflow.org/hub/tutorials/movinet\"><img src=\"https://www.tensorflow.org/images/tf_logo_32px.png\">TensorFlow.org에서 보기</a></td>\n",
    "  <td><a target=\"_blank\" href=\"https://colab.research.google.com/github/tensorflow/docs-l10n/blob/master/site/ko/hub/tutorials/movinet.ipynb\"><img src=\"https://www.tensorflow.org/images/colab_logo_32px.png\">Google Colab에서 실행</a></td>\n",
    "  <td><a target=\"_blank\" href=\"https://github.com/tensorflow/docs-l10n/blob/master/site/ko/hub/tutorials/movinet.ipynb\"><img src=\"https://www.tensorflow.org/images/GitHub-Mark-32px.png\">GitHub에서보기</a></td>\n",
    "  <td><a href=\"https://storage.googleapis.com/tensorflow_docs/docs-l10n/site/ko/hub/tutorials/movinet.ipynb\"><img src=\"https://www.tensorflow.org/images/download_logo_32px.png\">노트북 다운로드</a></td>\n",
    "  <td><a href=\"https://tfhub.dev/google/collections/movinet/1\"><img src=\"https://www.tensorflow.org/images/hub_logo_32px.png\">TF Hub 모델 보기</a></td>\n",
    "</table>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "-vxk2Kbc_KSP"
   },
   "source": [
    "이 튜토리얼은 사전 훈련된 비디오 분류 모델을 실행하여 주어진 비디오에서 활동(댄스, 수영, 자전거 타기 등)을 분류합니다.\n",
    "\n",
    "이 튜토리얼에서 사용된 모델 아키텍처는 [MoViNet](https://arxiv.org/pdf/2103.11511.pdf)(Mobile Video Networks)이라고 합니다. MoVieNets는 거대한 데이터세트([Kinetics 600](https://deepmind.com/research/open-source/kinetics))에서 훈련된 효율적인 비디오 분류 모델 집합입니다.\n",
    "\n",
    "TF Hub에서 사용 가능한 [i3d 모델](https://tfhub.dev/s?q=i3d-kinetics)과 달리 MoViNet은 스트리밍 비디오에서 프레임별 추론도 지원합니다.\n",
    "\n",
    "사전 훈련된 모델은 [TF Hub](https://tfhub.dev/google/collections/movinet/1)에서 사용할 수 있습니다. TF Hub 컬렉션에는 [TFLite](https://tensorflow.org/lite)에 최적화된 양자화된 모델도 포함되어 있습니다.\n",
    "\n",
    "이러한 모델의 소스는 [TensorFlow Model Garden](https://github.com/tensorflow/models/tree/master/official/projects/movinet)에서 얻을 수 있습니다. 여기에는 MoViNet 모델의 구축 및 미세 조정도 포함하는 [이 튜토리얼의 상세 버전](https://colab.sandbox.google.com/github/tensorflow/models/blob/master/official/projects/movinet/movinet_tutorial.ipynb)이 포함됩니다.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "3E96e1UKQ8uR"
   },
   "source": [
    "![jumping jacks plot](https://storage.googleapis.com/tf_model_garden/vision/movinet/artifacts/jumpingjacks_plot.gif)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "8_oLnvJy7kz5"
   },
   "source": [
    "## 설정\n",
    "\n",
    "더 작은 모델(A0-A2)에 대한 추론의 경우, 이 Colab에 대한 CPU가 충분합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:54:37.301953Z",
     "iopub.status.busy": "2022-12-14T21:54:37.301377Z",
     "iopub.status.idle": "2022-12-14T21:54:55.017797Z",
     "shell.execute_reply": "2022-12-14T21:54:55.016663Z"
    },
    "id": "GUgUMGmY1yq-"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "Reading package lists... 0%\r",
      "\r",
      "Reading package lists... 100%\r",
      "\r",
      "Reading package lists... Done\r",
      "\r\n",
      "\r",
      "Building dependency tree... 0%\r",
      "\r",
      "Building dependency tree... 0%\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "Building dependency tree... 50%\r",
      "\r",
      "Building dependency tree... 50%\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "Building dependency tree       \r",
      "\r\n",
      "\r",
      "Reading state information... 0%\r",
      "\r",
      "Reading state information... 0%\r",
      "\r",
      "Reading state information... Done\r",
      "\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The following packages were automatically installed and are no longer required:\r\n",
      "  libatasmart4 libblockdev-fs2 libblockdev-loop2 libblockdev-part-err2\r\n",
      "  libblockdev-part2 libblockdev-swap2 libblockdev-utils2 libblockdev2\r\n",
      "  libparted-fs-resize0\r\n",
      "Use 'sudo apt autoremove' to remove them.\r\n",
      "The following additional packages will be installed:\r\n",
      "  i965-va-driver intel-media-va-driver libaacs0 libaom0 libass9 libavcodec58\r\n",
      "  libavdevice58 libavfilter7 libavformat58 libavresample4 libavutil56\r\n",
      "  libbdplus0 libbluray2 libbs2b0 libcdio-cdda2 libcdio-paranoia2 libcdio18\r\n",
      "  libchromaprint1 libcodec2-0.9 libdc1394-22 libfftw3-double3 libflite1\r\n",
      "  libgme0 libgsm1 libigdgmm11 liblilv-0-0 libmysofa1 libopenal-data libopenal1\r\n",
      "  libopenmpt0 libpostproc55 librubberband2 libsdl2-2.0-0 libserd-0-0 libshine3\r\n",
      "  libsnappy1v5 libsndio7.0 libsord-0-0 libsratom-0-0 libssh-gcrypt-4\r\n",
      "  libswresample3 libswscale5 libva-drm2 libva-x11-2 libva2 libvidstab1.1\r\n",
      "  libwebpmux3 libx264-155 libx265-179 libxvidcore4 libzvbi-common libzvbi0\r\n",
      "  mesa-va-drivers ocl-icd-libopencl1 va-driver-all\r\n",
      "Suggested packages:\r\n",
      "  ffmpeg-doc i965-va-driver-shaders libbluray-bdj libfftw3-bin libfftw3-dev\r\n",
      "  libportaudio2 serdi sndiod sordi\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The following NEW packages will be installed:\r\n",
      "  ffmpeg i965-va-driver intel-media-va-driver libaacs0 libaom0 libass9\r\n",
      "  libavcodec58 libavdevice58 libavfilter7 libavformat58 libavresample4\r\n",
      "  libavutil56 libbdplus0 libbluray2 libbs2b0 libcdio-cdda2 libcdio-paranoia2\r\n",
      "  libcdio18 libchromaprint1 libcodec2-0.9 libdc1394-22 libfftw3-double3\r\n",
      "  libflite1 libgme0 libgsm1 libigdgmm11 liblilv-0-0 libmysofa1 libopenal-data\r\n",
      "  libopenal1 libopenmpt0 libpostproc55 librubberband2 libsdl2-2.0-0\r\n",
      "  libserd-0-0 libshine3 libsnappy1v5 libsndio7.0 libsord-0-0 libsratom-0-0\r\n",
      "  libssh-gcrypt-4 libswresample3 libswscale5 libva-drm2 libva-x11-2 libva2\r\n",
      "  libvidstab1.1 libwebpmux3 libx264-155 libx265-179 libxvidcore4\r\n",
      "  libzvbi-common libzvbi0 mesa-va-drivers ocl-icd-libopencl1 va-driver-all\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 upgraded, 56 newly installed, 0 to remove and 169 not upgraded.\r\n",
      "Need to get 42.5 MB of archives.\r\n",
      "After this operation, 156 MB of additional disk space will be used.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[33m\r",
      "0% [Working]\u001b[0m\r",
      "            \r",
      "Get:1 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libaom0 amd64 1.0.0.errata1-3build1 [1160 kB]\r\n",
      "\u001b[33m\r",
      "0% [1 libaom0 13.7 kB/1160 kB 1%]\u001b[0m\u001b[33m\r",
      "                                 \r",
      "3% [Working]\u001b[0m\r",
      "            \r",
      "Get:2 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libva2 amd64 2.7.0-2 [53.5 kB]\r\n",
      "\u001b[33m\r",
      "3% [2 libva2 13.7 kB/53.5 kB 26%]\u001b[0m\u001b[33m\r",
      "                                 \r",
      "3% [Waiting for headers]\u001b[0m\r",
      "                        \r",
      "Get:3 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libva-drm2 amd64 2.7.0-2 [7044 B]\r\n",
      "\u001b[33m\r",
      "3% [3 libva-drm2 7044 B/7044 B 100%]\u001b[0m\u001b[33m\r",
      "                                    \r",
      "3% [Waiting for headers]\u001b[0m\r",
      "                        \r",
      "Get:4 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libva-x11-2 amd64 2.7.0-2 [11.9 kB]\r\n",
      "\u001b[33m\r",
      "3% [4 libva-x11-2 11.9 kB/11.9 kB 100%]\u001b[0m\u001b[33m\r",
      "                                       \r",
      "4% [Waiting for headers]\u001b[0m\r",
      "                        \r",
      "Get:5 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/main amd64 ocl-icd-libopencl1 amd64 2.2.11-1ubuntu1 [30.3 kB]\r\n",
      "\u001b[33m\r",
      "4% [5 ocl-icd-libopencl1 16.5 kB/30.3 kB 54%]\u001b[0m\u001b[33m\r",
      "                                             \r",
      "4% [Waiting for headers]\u001b[0m\r",
      "                        \r",
      "Get:6 http://us-central1.gce.archive.ubuntu.com/ubuntu focal-updates/universe amd64 libavutil56 amd64 7:4.2.7-0ubuntu0.1 [241 kB]\r\n",
      "\u001b[33m\r",
      "4% [6 libavutil56 13.7 kB/241 kB 6%]\u001b[0m\u001b[33m\r",
      "                                    \r",
      "5% [Waiting for headers]\u001b[0m\r",
      "                        \r",
      "Get:7 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libcodec2-0.9 amd64 0.9.2-2 [7886 kB]\r\n",
      "\u001b[33m\r",
      "5% [7 libcodec2-0.9 6632 B/7886 kB 0%]\u001b[0m"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[33m\r",
      "                                      \r",
      "20% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:8 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libgsm1 amd64 1.0.18-2 [24.4 kB]\r\n",
      "\u001b[33m\r",
      "20% [8 libgsm1 24.4 kB/24.4 kB 100%]\u001b[0m\u001b[33m\r",
      "                                    \r",
      "21% [Working]\u001b[0m\r",
      "             \r",
      "Get:9 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libshine3 amd64 3.1.1-2 [23.2 kB]\r\n",
      "\u001b[33m\r",
      "21% [9 libshine3 18.7 kB/23.2 kB 81%]\u001b[0m\u001b[33m\r",
      "                                     \r",
      "21% [Working]\u001b[0m\r",
      "             \r",
      "Get:10 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/main amd64 libsnappy1v5 amd64 1.1.8-1build1 [16.7 kB]\r\n",
      "\u001b[33m\r",
      "21% [10 libsnappy1v5 8192 B/16.7 kB 49%]\u001b[0m\u001b[33m\r",
      "                                        \r",
      "21% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:11 http://us-central1.gce.archive.ubuntu.com/ubuntu focal-updates/universe amd64 libswresample3 amd64 7:4.2.7-0ubuntu0.1 [57.1 kB]\r\n",
      "\u001b[33m\r",
      "21% [11 libswresample3 43.6 kB/57.1 kB 76%]\u001b[0m\u001b[33m\r",
      "                                           \r",
      "22% [Working]\u001b[0m\r",
      "             \r",
      "Get:12 http://us-central1.gce.archive.ubuntu.com/ubuntu focal-updates/main amd64 libwebpmux3 amd64 0.6.1-2ubuntu0.20.04.1 [19.5 kB]\r\n",
      "\u001b[33m\r",
      "22% [12 libwebpmux3 0 B/19.5 kB 0%]\u001b[0m\u001b[33m\r",
      "                                   \r",
      "22% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:13 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libx264-155 amd64 2:0.155.2917+git0a84d98-2 [521 kB]\r\n",
      "\u001b[33m\r",
      "22% [13 libx264-155 31.8 kB/521 kB 6%]\u001b[0m\u001b[33m\r",
      "                                      \r",
      "24% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:14 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libx265-179 amd64 3.2.1-1build1 [1060 kB]\r\n",
      "\u001b[33m\r",
      "24% [14 libx265-179 14.9 kB/1060 kB 1%]\u001b[0m"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[33m\r",
      "                                       \r",
      "26% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:15 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libxvidcore4 amd64 2:1.3.7-1 [201 kB]\r\n",
      "\u001b[33m\r",
      "26% [15 libxvidcore4 22.9 kB/201 kB 11%]\u001b[0m\u001b[33m\r",
      "                                        \r",
      "27% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:16 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libzvbi-common all 0.2.35-17 [32.5 kB]\r\n",
      "\u001b[33m\r",
      "27% [16 libzvbi-common 17.8 kB/32.5 kB 55%]\u001b[0m\u001b[33m\r",
      "                                           \r",
      "27% [Waiting for headers]\u001b[0m"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "                         \r",
      "Get:17 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libzvbi0 amd64 0.2.35-17 [237 kB]\r\n",
      "\u001b[33m\r",
      "27% [17 libzvbi0 6634 B/237 kB 3%]\u001b[0m\u001b[33m\r",
      "                                  \r",
      "28% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:18 http://us-central1.gce.archive.ubuntu.com/ubuntu focal-updates/universe amd64 libavcodec58 amd64 7:4.2.7-0ubuntu0.1 [4878 kB]\r\n",
      "\u001b[33m\r",
      "28% [18 libavcodec58 9367 B/4878 kB 0%]\u001b[0m"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[33m\r",
      "                                       \r",
      "37% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:19 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libass9 amd64 1:0.14.0-2 [88.0 kB]\r\n",
      "\u001b[33m\r",
      "37% [19 libass9 13.7 kB/88.0 kB 16%]\u001b[0m\u001b[33m\r",
      "                                    \r",
      "38% [Waiting for headers]\u001b[0m"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "                         \r",
      "Get:20 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libbluray2 amd64 1:1.2.0-1 [138 kB]\r\n",
      "\u001b[33m\r",
      "38% [20 libbluray2 23.3 kB/138 kB 17%]\u001b[0m\u001b[33m\r",
      "                                      \r",
      "39% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:21 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libchromaprint1 amd64 1.4.3-3build1 [37.6 kB]\r\n",
      "\u001b[33m\r",
      "39% [21 libchromaprint1 10.9 kB/37.6 kB 29%]\u001b[0m\u001b[33m\r",
      "                                            \r",
      "39% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:22 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libgme0 amd64 0.6.2-1build1 [123 kB]\r\n",
      "\u001b[33m\r",
      "39% [22 libgme0 13.7 kB/123 kB 11%]\u001b[0m\u001b[33m\r",
      "                                   \r",
      "40% [Waiting for headers]\u001b[0m"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "                         \r",
      "Get:23 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libopenmpt0 amd64 0.4.11-1build1 [599 kB]\r\n",
      "\u001b[33m\r",
      "40% [23 libopenmpt0 13.7 kB/599 kB 2%]\u001b[0m\u001b[33m\r",
      "                                      \r",
      "41% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:24 http://us-central1.gce.archive.ubuntu.com/ubuntu focal-updates/main amd64 libssh-gcrypt-4 amd64 0.9.3-2ubuntu2.2 [202 kB]\r\n",
      "\u001b[33m\r",
      "41% [24 libssh-gcrypt-4 41.9 kB/202 kB 21%]\u001b[0m\u001b[33m\r",
      "                                           \r",
      "42% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:25 http://us-central1.gce.archive.ubuntu.com/ubuntu focal-updates/universe amd64 libavformat58 amd64 7:4.2.7-0ubuntu0.1 [985 kB]\r\n",
      "\u001b[33m\r",
      "42% [25 libavformat58 13.7 kB/985 kB 1%]\u001b[0m\u001b[33m\r",
      "                                        \r",
      "44% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:26 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libbs2b0 amd64 3.1.0+dfsg-2.2build1 [10.2 kB]\r\n",
      "\u001b[33m\r",
      "44% [26 libbs2b0 10.2 kB/10.2 kB 100%]\u001b[0m\u001b[33m\r",
      "                                      \r",
      "44% [Working]\u001b[0m\r",
      "             \r",
      "Get:27 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libflite1 amd64 2.1-release-3 [12.8 MB]\r\n",
      "\u001b[33m\r",
      "44% [27 libflite1 4096 B/12.8 MB 0%]\u001b[0m"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[33m\r",
      "                                    \r",
      "69% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:28 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libserd-0-0 amd64 0.30.2-1 [46.6 kB]\r\n",
      "\u001b[33m\r",
      "69% [28 libserd-0-0 13.7 kB/46.6 kB 29%]\u001b[0m\u001b[33m\r",
      "                                        \r",
      "69% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:29 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libsord-0-0 amd64 0.16.4-1 [19.5 kB]\r\n",
      "\u001b[33m\r",
      "69% [29 libsord-0-0 13.9 kB/19.5 kB 71%]\u001b[0m\u001b[33m\r",
      "                                        \r",
      "70% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:30 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libsratom-0-0 amd64 0.6.4-1 [16.9 kB]\r\n",
      "\u001b[33m\r",
      "70% [30 libsratom-0-0 16.9 kB/16.9 kB 100%]\u001b[0m\u001b[33m\r",
      "                                           \r",
      "70% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:31 http://us-central1.gce.archive.ubuntu.com/ubuntu focal-updates/universe amd64 liblilv-0-0 amd64 0.24.6-1ubuntu0.1 [40.6 kB]\r\n",
      "\u001b[33m\r",
      "70% [31 liblilv-0-0 6639 B/40.6 kB 16%]\u001b[0m\u001b[33m\r",
      "                                       \r",
      "71% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:32 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libmysofa1 amd64 1.0~dfsg0-1 [39.2 kB]\r\n",
      "\u001b[33m\r",
      "71% [32 libmysofa1 9452 B/39.2 kB 24%]\u001b[0m\u001b[33m\r",
      "                                      \r",
      "71% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:33 http://us-central1.gce.archive.ubuntu.com/ubuntu focal-updates/universe amd64 libpostproc55 amd64 7:4.2.7-0ubuntu0.1 [55.0 kB]\r\n",
      "\u001b[33m\r",
      "71% [33 libpostproc55 13.7 kB/55.0 kB 25%]\u001b[0m\u001b[33m\r",
      "                                          \r",
      "71% [Waiting for headers]\u001b[0m"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "                         \r",
      "Get:34 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/main amd64 libfftw3-double3 amd64 3.3.8-2ubuntu1 [728 kB]\r\n",
      "\u001b[33m\r",
      "71% [34 libfftw3-double3 13.7 kB/728 kB 2%]\u001b[0m\u001b[33m\r",
      "                                           \r",
      "73% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:35 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 librubberband2 amd64 1.8.2-1build1 [89.4 kB]\r\n",
      "\u001b[33m\r",
      "73% [35 librubberband2 2411 B/89.4 kB 3%]\u001b[0m\u001b[33m\r",
      "                                         \r",
      "74% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:36 http://us-central1.gce.archive.ubuntu.com/ubuntu focal-updates/universe amd64 libswscale5 amd64 7:4.2.7-0ubuntu0.1 [157 kB]\r\n",
      "\u001b[33m\r",
      "74% [36 libswscale5 5829 B/157 kB 4%]\u001b[0m\u001b[33m\r",
      "                                     \r",
      "74% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:37 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libvidstab1.1 amd64 1.1.0-2 [35.0 kB]\r\n",
      "\u001b[33m\r",
      "74% [37 libvidstab1.1 10.9 kB/35.0 kB 31%]\u001b[0m\u001b[33m\r",
      "                                          \r",
      "75% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:38 http://us-central1.gce.archive.ubuntu.com/ubuntu focal-updates/universe amd64 libavfilter7 amd64 7:4.2.7-0ubuntu0.1 [1085 kB]\r\n",
      "\u001b[33m\r",
      "75% [38 libavfilter7 9261 B/1085 kB 1%]\u001b[0m\r",
      "                                       \r",
      "Get:39 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/main amd64 libcdio18 amd64 2.0.0-2 [58.6 kB]\r\n",
      "\u001b[33m\r",
      "                                       \r",
      "77% [39 libcdio18 38.4 kB/58.6 kB 66%]\u001b[0m\u001b[33m\r",
      "                                      \r",
      "78% [Working]\u001b[0m\r",
      "             \r",
      "Get:40 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/main amd64 libcdio-cdda2 amd64 10.2+2.0.0-1 [17.6 kB]\r\n",
      "\u001b[33m\r",
      "78% [40 libcdio-cdda2 4096 B/17.6 kB 23%]\u001b[0m\u001b[33m\r",
      "                                         \r",
      "78% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:41 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/main amd64 libcdio-paranoia2 amd64 10.2+2.0.0-1 [16.2 kB]\r\n",
      "\u001b[33m\r",
      "78% [41 libcdio-paranoia2 16.2 kB/16.2 kB 100%]\u001b[0m\u001b[33m\r",
      "                                               \r",
      "78% [Working]\u001b[0m\r",
      "             \r",
      "Get:42 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libdc1394-22 amd64 2.2.5-2.1 [79.6 kB]\r\n",
      "\u001b[33m\r",
      "78% [42 libdc1394-22 4096 B/79.6 kB 5%]\u001b[0m\u001b[33m\r",
      "                                       \r",
      "79% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:43 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libopenal-data all 1:1.19.1-1 [162 kB]\r\n",
      "\u001b[33m\r",
      "79% [43 libopenal-data 13.7 kB/162 kB 8%]\u001b[0m\u001b[33m\r",
      "                                         \r",
      "80% [Waiting for headers]\u001b[0m"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "                         \r",
      "Get:44 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libsndio7.0 amd64 1.5.0-3 [24.5 kB]\r\n",
      "\u001b[33m\r",
      "80% [44 libsndio7.0 2416 B/24.5 kB 10%]\u001b[0m\u001b[33m\r",
      "                                       \r",
      "80% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:45 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libopenal1 amd64 1:1.19.1-1 [492 kB]\r\n",
      "\u001b[33m\r",
      "80% [45 libopenal1 12.3 kB/492 kB 2%]\u001b[0m\u001b[33m\r",
      "                                     \r",
      "81% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:46 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libsdl2-2.0-0 amd64 2.0.10+dfsg1-3 [407 kB]\r\n",
      "\u001b[33m\r",
      "81% [46 libsdl2-2.0-0 13.7 kB/407 kB 3%]\u001b[0m\u001b[33m\r",
      "                                        \r",
      "82% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:47 http://us-central1.gce.archive.ubuntu.com/ubuntu focal-updates/universe amd64 libavdevice58 amd64 7:4.2.7-0ubuntu0.1 [74.3 kB]\r\n",
      "\u001b[33m\r",
      "82% [47 libavdevice58 13.7 kB/74.3 kB 18%]\u001b[0m\u001b[33m\r",
      "                                          \r",
      "83% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:48 http://us-central1.gce.archive.ubuntu.com/ubuntu focal-updates/universe amd64 libavresample4 amd64 7:4.2.7-0ubuntu0.1 [54.2 kB]\r\n",
      "\u001b[33m\r",
      "83% [48 libavresample4 12.3 kB/54.2 kB 23%]\u001b[0m\u001b[33m\r",
      "                                           \r",
      "83% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:49 http://us-central1.gce.archive.ubuntu.com/ubuntu focal-updates/universe amd64 ffmpeg amd64 7:4.2.7-0ubuntu0.1 [1453 kB]\r\n",
      "\u001b[33m\r",
      "83% [49 ffmpeg 2411 B/1453 kB 0%]\u001b[0m"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[33m\r",
      "                                 \r",
      "86% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:50 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libigdgmm11 amd64 20.1.1+ds1-1 [111 kB]\r\n",
      "\u001b[33m\r",
      "86% [50 libigdgmm11 13.7 kB/111 kB 12%]\u001b[0m\u001b[33m\r",
      "                                       \r",
      "87% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:51 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 intel-media-va-driver amd64 20.1.1+dfsg1-1 [1764 kB]\r\n",
      "\u001b[33m\r",
      "87% [51 intel-media-va-driver 24.6 kB/1764 kB 1%]\u001b[0m\u001b[33m\r",
      "                                                 \r",
      "91% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:52 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libaacs0 amd64 0.9.0-2 [50.1 kB]\r\n",
      "\u001b[33m\r",
      "91% [52 libaacs0 2415 B/50.1 kB 5%]\u001b[0m\u001b[33m\r",
      "                                   \r",
      "91% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:53 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 libbdplus0 amd64 0.1.2-3 [47.3 kB]\r\n",
      "\u001b[33m\r",
      "91% [53 libbdplus0 13.7 kB/47.3 kB 29%]\u001b[0m\u001b[33m\r",
      "                                       \r",
      "92% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:54 http://us-central1.gce.archive.ubuntu.com/ubuntu focal-updates/universe amd64 mesa-va-drivers amd64 21.2.6-0ubuntu0.1~20.04.2 [2970 kB]\r\n",
      "\u001b[33m\r",
      "92% [54 mesa-va-drivers 2408 B/2970 kB 0%]\u001b[0m"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[33m\r",
      "                                          \r",
      "98% [Waiting for headers]\u001b[0m\r",
      "                         \r",
      "Get:55 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 i965-va-driver amd64 2.4.0-0ubuntu1 [924 kB]\r\n",
      "\u001b[33m\r",
      "98% [55 i965-va-driver 13.7 kB/924 kB 1%]\u001b[0m\r",
      "                                         \r",
      "Get:56 http://us-central1.gce.archive.ubuntu.com/ubuntu focal/universe amd64 va-driver-all amd64 2.7.0-2 [4020 B]\r\n",
      "\u001b[33m\r",
      "100% [56 va-driver-all 4020 B/4020 B 100%]\u001b[0m\u001b[33m\r",
      "                                          \r",
      "100% [Working]\u001b[0m\r",
      "              \r",
      "Fetched 42.5 MB in 2s (26.4 MB/s)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "Extracting templates from packages: 53%"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "Extracting templates from packages: 100%\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\u001b7\u001b[0;23r\u001b8\u001b[1A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libaom0:amd64.\r\n",
      "(Reading database ... \r",
      "(Reading database ... 5%\r",
      "(Reading database ... 10%\r",
      "(Reading database ... 15%\r",
      "(Reading database ... 20%\r",
      "(Reading database ... 25%\r",
      "(Reading database ... 30%\r",
      "(Reading database ... 35%\r",
      "(Reading database ... 40%\r",
      "(Reading database ... 45%\r",
      "(Reading database ... 50%\r",
      "(Reading database ... 55%\r",
      "(Reading database ... 60%\r",
      "(Reading database ... 65%\r",
      "(Reading database ... 70%\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(Reading database ... 75%\r",
      "(Reading database ... 80%\r",
      "(Reading database ... 85%\r",
      "(Reading database ... 90%\r",
      "(Reading database ... 95%\r",
      "(Reading database ... 100%\r",
      "(Reading database ... 140796 files and directories currently installed.)\r\n",
      "Preparing to unpack .../00-libaom0_1.0.0.errata1-3build1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [  0%]\u001b[49m\u001b[39m [..........................................................] \u001b8Unpacking libaom0:amd64 (1.0.0.errata1-3build1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libva2:amd64.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing to unpack .../01-libva2_2.7.0-2_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [  1%]\u001b[49m\u001b[39m [..........................................................] \u001b8Unpacking libva2:amd64 (2.7.0-2) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libva-drm2:amd64.\r\n",
      "Preparing to unpack .../02-libva-drm2_2.7.0-2_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [  2%]\u001b[49m\u001b[39m [#.........................................................] \u001b8Unpacking libva-drm2:amd64 (2.7.0-2) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libva-x11-2:amd64.\r\n",
      "Preparing to unpack .../03-libva-x11-2_2.7.0-2_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [  3%]\u001b[49m\u001b[39m [#.........................................................] \u001b8Unpacking libva-x11-2:amd64 (2.7.0-2) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package ocl-icd-libopencl1:amd64.\r\n",
      "Preparing to unpack .../04-ocl-icd-libopencl1_2.2.11-1ubuntu1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [  4%]\u001b[49m\u001b[39m [##........................................................] \u001b8Unpacking ocl-icd-libopencl1:amd64 (2.2.11-1ubuntu1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libavutil56:amd64.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing to unpack .../05-libavutil56_7%3a4.2.7-0ubuntu0.1_amd64.deb ...\r\n",
      "Unpacking libavutil56:amd64 (7:4.2.7-0ubuntu0.1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [  5%]\u001b[49m\u001b[39m [###.......................................................] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libcodec2-0.9:amd64.\r\n",
      "Preparing to unpack .../06-libcodec2-0.9_0.9.2-2_amd64.deb ...\r\n",
      "Unpacking libcodec2-0.9:amd64 (0.9.2-2) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [  6%]\u001b[49m\u001b[39m [###.......................................................] \u001b8Selecting previously unselected package libgsm1:amd64.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing to unpack .../07-libgsm1_1.0.18-2_amd64.deb ...\r\n",
      "Unpacking libgsm1:amd64 (1.0.18-2) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [  7%]\u001b[49m\u001b[39m [####......................................................] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libshine3:amd64.\r\n",
      "Preparing to unpack .../08-libshine3_3.1.1-2_amd64.deb ...\r\n",
      "Unpacking libshine3:amd64 (3.1.1-2) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [  8%]\u001b[49m\u001b[39m [####......................................................] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libsnappy1v5:amd64.\r\n",
      "Preparing to unpack .../09-libsnappy1v5_1.1.8-1build1_amd64.deb ...\r\n",
      "Unpacking libsnappy1v5:amd64 (1.1.8-1build1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libswresample3:amd64.\r\n",
      "Preparing to unpack .../10-libswresample3_7%3a4.2.7-0ubuntu0.1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [  9%]\u001b[49m\u001b[39m [#####.....................................................] \u001b8Unpacking libswresample3:amd64 (7:4.2.7-0ubuntu0.1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libwebpmux3:amd64.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing to unpack .../11-libwebpmux3_0.6.1-2ubuntu0.20.04.1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 10%]\u001b[49m\u001b[39m [#####.....................................................] \u001b8Unpacking libwebpmux3:amd64 (0.6.1-2ubuntu0.20.04.1) ...\r\n",
      "Selecting previously unselected package libx264-155:amd64.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing to unpack .../12-libx264-155_2%3a0.155.2917+git0a84d98-2_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 11%]\u001b[49m\u001b[39m [######....................................................] \u001b8Unpacking libx264-155:amd64 (2:0.155.2917+git0a84d98-2) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libx265-179:amd64.\r\n",
      "Preparing to unpack .../13-libx265-179_3.2.1-1build1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 12%]\u001b[49m\u001b[39m [######....................................................] \u001b8Unpacking libx265-179:amd64 (3.2.1-1build1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libxvidcore4:amd64.\r\n",
      "Preparing to unpack .../14-libxvidcore4_2%3a1.3.7-1_amd64.deb ...\r\n",
      "Unpacking libxvidcore4:amd64 (2:1.3.7-1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 13%]\u001b[49m\u001b[39m [#######...................................................] \u001b8Selecting previously unselected package libzvbi-common.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing to unpack .../15-libzvbi-common_0.2.35-17_all.deb ...\r\n",
      "Unpacking libzvbi-common (0.2.35-17) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 14%]\u001b[49m\u001b[39m [########..................................................] \u001b8Selecting previously unselected package libzvbi0:amd64.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing to unpack .../16-libzvbi0_0.2.35-17_amd64.deb ...\r\n",
      "Unpacking libzvbi0:amd64 (0.2.35-17) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 15%]\u001b[49m\u001b[39m [########..................................................] \u001b8Selecting previously unselected package libavcodec58:amd64.\r\n",
      "Preparing to unpack .../17-libavcodec58_7%3a4.2.7-0ubuntu0.1_amd64.deb ...\r\n",
      "Unpacking libavcodec58:amd64 (7:4.2.7-0ubuntu0.1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 16%]\u001b[49m\u001b[39m [#########.................................................] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libass9:amd64.\r\n",
      "Preparing to unpack .../18-libass9_1%3a0.14.0-2_amd64.deb ...\r\n",
      "Unpacking libass9:amd64 (1:0.14.0-2) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libbluray2:amd64.\r\n",
      "Preparing to unpack .../19-libbluray2_1%3a1.2.0-1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 17%]\u001b[49m\u001b[39m [##########................................................] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unpacking libbluray2:amd64 (1:1.2.0-1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libchromaprint1:amd64.\r\n",
      "Preparing to unpack .../20-libchromaprint1_1.4.3-3build1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 18%]\u001b[49m\u001b[39m [##########................................................] \u001b8Unpacking libchromaprint1:amd64 (1.4.3-3build1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libgme0:amd64.\r\n",
      "Preparing to unpack .../21-libgme0_0.6.2-1build1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 19%]\u001b[49m\u001b[39m [###########...............................................] \u001b8Unpacking libgme0:amd64 (0.6.2-1build1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libopenmpt0:amd64.\r\n",
      "Preparing to unpack .../22-libopenmpt0_0.4.11-1build1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 20%]\u001b[49m\u001b[39m [###########...............................................] \u001b8Unpacking libopenmpt0:amd64 (0.4.11-1build1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libssh-gcrypt-4:amd64.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing to unpack .../23-libssh-gcrypt-4_0.9.3-2ubuntu2.2_amd64.deb ...\r\n",
      "Unpacking libssh-gcrypt-4:amd64 (0.9.3-2ubuntu2.2) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 21%]\u001b[49m\u001b[39m [############..............................................] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libavformat58:amd64.\r\n",
      "Preparing to unpack .../24-libavformat58_7%3a4.2.7-0ubuntu0.1_amd64.deb ...\r\n",
      "Unpacking libavformat58:amd64 (7:4.2.7-0ubuntu0.1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 22%]\u001b[49m\u001b[39m [############..............................................] \u001b8Selecting previously unselected package libbs2b0:amd64.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing to unpack .../25-libbs2b0_3.1.0+dfsg-2.2build1_amd64.deb ...\r\n",
      "Unpacking libbs2b0:amd64 (3.1.0+dfsg-2.2build1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 23%]\u001b[49m\u001b[39m [#############.............................................] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libflite1:amd64.\r\n",
      "Preparing to unpack .../26-libflite1_2.1-release-3_amd64.deb ...\r\n",
      "Unpacking libflite1:amd64 (2.1-release-3) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 24%]\u001b[49m\u001b[39m [#############.............................................] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libserd-0-0:amd64.\r\n",
      "Preparing to unpack .../27-libserd-0-0_0.30.2-1_amd64.deb ...\r\n",
      "Unpacking libserd-0-0:amd64 (0.30.2-1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libsord-0-0:amd64.\r\n",
      "Preparing to unpack .../28-libsord-0-0_0.16.4-1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 25%]\u001b[49m\u001b[39m [##############............................................] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unpacking libsord-0-0:amd64 (0.16.4-1) ...\r\n",
      "Selecting previously unselected package libsratom-0-0:amd64.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing to unpack .../29-libsratom-0-0_0.6.4-1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 26%]\u001b[49m\u001b[39m [###############...........................................] \u001b8Unpacking libsratom-0-0:amd64 (0.6.4-1) ...\r\n",
      "Selecting previously unselected package liblilv-0-0:amd64.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing to unpack .../30-liblilv-0-0_0.24.6-1ubuntu0.1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 27%]\u001b[49m\u001b[39m [###############...........................................] \u001b8Unpacking liblilv-0-0:amd64 (0.24.6-1ubuntu0.1) ...\r\n",
      "Selecting previously unselected package libmysofa1:amd64.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing to unpack .../31-libmysofa1_1.0~dfsg0-1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 28%]\u001b[49m\u001b[39m [################..........................................] \u001b8Unpacking libmysofa1:amd64 (1.0~dfsg0-1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libpostproc55:amd64.\r\n",
      "Preparing to unpack .../32-libpostproc55_7%3a4.2.7-0ubuntu0.1_amd64.deb ...\r\n",
      "Unpacking libpostproc55:amd64 (7:4.2.7-0ubuntu0.1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 29%]\u001b[49m\u001b[39m [#################.........................................] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libfftw3-double3:amd64.\r\n",
      "Preparing to unpack .../33-libfftw3-double3_3.3.8-2ubuntu1_amd64.deb ...\r\n",
      "Unpacking libfftw3-double3:amd64 (3.3.8-2ubuntu1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 30%]\u001b[49m\u001b[39m [#################.........................................] \u001b8Selecting previously unselected package librubberband2:amd64.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing to unpack .../34-librubberband2_1.8.2-1build1_amd64.deb ...\r\n",
      "Unpacking librubberband2:amd64 (1.8.2-1build1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 31%]\u001b[49m\u001b[39m [##################........................................] \u001b8Selecting previously unselected package libswscale5:amd64.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing to unpack .../35-libswscale5_7%3a4.2.7-0ubuntu0.1_amd64.deb ...\r\n",
      "Unpacking libswscale5:amd64 (7:4.2.7-0ubuntu0.1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 32%]\u001b[49m\u001b[39m [##################........................................] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libvidstab1.1:amd64.\r\n",
      "Preparing to unpack .../36-libvidstab1.1_1.1.0-2_amd64.deb ...\r\n",
      "Unpacking libvidstab1.1:amd64 (1.1.0-2) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libavfilter7:amd64.\r\n",
      "Preparing to unpack .../37-libavfilter7_7%3a4.2.7-0ubuntu0.1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 33%]\u001b[49m\u001b[39m [###################.......................................] \u001b8Unpacking libavfilter7:amd64 (7:4.2.7-0ubuntu0.1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libcdio18:amd64.\r\n",
      "Preparing to unpack .../38-libcdio18_2.0.0-2_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 34%]\u001b[49m\u001b[39m [###################.......................................] \u001b8Unpacking libcdio18:amd64 (2.0.0-2) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libcdio-cdda2:amd64.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing to unpack .../39-libcdio-cdda2_10.2+2.0.0-1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 35%]\u001b[49m\u001b[39m [####################......................................] \u001b8Unpacking libcdio-cdda2:amd64 (10.2+2.0.0-1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libcdio-paranoia2:amd64.\r\n",
      "Preparing to unpack .../40-libcdio-paranoia2_10.2+2.0.0-1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 36%]\u001b[49m\u001b[39m [####################......................................] \u001b8Unpacking libcdio-paranoia2:amd64 (10.2+2.0.0-1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libdc1394-22:amd64.\r\n",
      "Preparing to unpack .../41-libdc1394-22_2.2.5-2.1_amd64.deb ...\r\n",
      "Unpacking libdc1394-22:amd64 (2.2.5-2.1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 37%]\u001b[49m\u001b[39m [#####################.....................................] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libopenal-data.\r\n",
      "Preparing to unpack .../42-libopenal-data_1%3a1.19.1-1_all.deb ...\r\n",
      "Unpacking libopenal-data (1:1.19.1-1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 38%]\u001b[49m\u001b[39m [######################....................................] \u001b8Selecting previously unselected package libsndio7.0:amd64.\r\n",
      "Preparing to unpack .../43-libsndio7.0_1.5.0-3_amd64.deb ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unpacking libsndio7.0:amd64 (1.5.0-3) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 39%]\u001b[49m\u001b[39m [######################....................................] \u001b8Selecting previously unselected package libopenal1:amd64.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing to unpack .../44-libopenal1_1%3a1.19.1-1_amd64.deb ...\r\n",
      "Unpacking libopenal1:amd64 (1:1.19.1-1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 40%]\u001b[49m\u001b[39m [#######################...................................] \u001b8Selecting previously unselected package libsdl2-2.0-0:amd64.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing to unpack .../45-libsdl2-2.0-0_2.0.10+dfsg1-3_amd64.deb ...\r\n",
      "Unpacking libsdl2-2.0-0:amd64 (2.0.10+dfsg1-3) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libavdevice58:amd64.\r\n",
      "Preparing to unpack .../46-libavdevice58_7%3a4.2.7-0ubuntu0.1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 41%]\u001b[49m\u001b[39m [#######################...................................] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unpacking libavdevice58:amd64 (7:4.2.7-0ubuntu0.1) ...\r\n",
      "Selecting previously unselected package libavresample4:amd64.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing to unpack .../47-libavresample4_7%3a4.2.7-0ubuntu0.1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 42%]\u001b[49m\u001b[39m [########################..................................] \u001b8Unpacking libavresample4:amd64 (7:4.2.7-0ubuntu0.1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package ffmpeg.\r\n",
      "Preparing to unpack .../48-ffmpeg_7%3a4.2.7-0ubuntu0.1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 43%]\u001b[49m\u001b[39m [#########################.................................] \u001b8Unpacking ffmpeg (7:4.2.7-0ubuntu0.1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package libigdgmm11:amd64.\r\n",
      "Preparing to unpack .../49-libigdgmm11_20.1.1+ds1-1_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 44%]\u001b[49m\u001b[39m [#########################.................................] \u001b8Unpacking libigdgmm11:amd64 (20.1.1+ds1-1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package intel-media-va-driver:amd64.\r\n",
      "Preparing to unpack .../50-intel-media-va-driver_20.1.1+dfsg1-1_amd64.deb ...\r\n",
      "Unpacking intel-media-va-driver:amd64 (20.1.1+dfsg1-1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 45%]\u001b[49m\u001b[39m [##########################................................] \u001b8Selecting previously unselected package libaacs0:amd64.\r\n",
      "Preparing to unpack .../51-libaacs0_0.9.0-2_amd64.deb ...\r\n",
      "Unpacking libaacs0:amd64 (0.9.0-2) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 46%]\u001b[49m\u001b[39m [##########################................................] \u001b8Selecting previously unselected package libbdplus0:amd64.\r\n",
      "Preparing to unpack .../52-libbdplus0_0.1.2-3_amd64.deb ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unpacking libbdplus0:amd64 (0.1.2-3) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 47%]\u001b[49m\u001b[39m [###########################...............................] \u001b8Selecting previously unselected package mesa-va-drivers:amd64.\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Preparing to unpack .../53-mesa-va-drivers_21.2.6-0ubuntu0.1~20.04.2_amd64.deb ...\r\n",
      "Unpacking mesa-va-drivers:amd64 (21.2.6-0ubuntu0.1~20.04.2) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 48%]\u001b[49m\u001b[39m [###########################...............................] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package i965-va-driver:amd64.\r\n",
      "Preparing to unpack .../54-i965-va-driver_2.4.0-0ubuntu1_amd64.deb ...\r\n",
      "Unpacking i965-va-driver:amd64 (2.4.0-0ubuntu1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selecting previously unselected package va-driver-all:amd64.\r\n",
      "Preparing to unpack .../55-va-driver-all_2.7.0-2_amd64.deb ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 49%]\u001b[49m\u001b[39m [############################..............................] \u001b8Unpacking va-driver-all:amd64 (2.7.0-2) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Setting up libgme0:amd64 (0.6.2-1build1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 50%]\u001b[49m\u001b[39m [#############################.............................] \u001b8Setting up libssh-gcrypt-4:amd64 (0.9.3-2ubuntu2.2) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 51%]\u001b[49m\u001b[39m [#############################.............................] \u001b8Setting up libshine3:amd64 (3.1.1-2) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 52%]\u001b[49m\u001b[39m [##############################............................] \u001b8Setting up libgsm1:amd64 (1.0.18-2) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 53%]\u001b[49m\u001b[39m [##############################............................] \u001b8Setting up libx264-155:amd64 (2:0.155.2917+git0a84d98-2) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 54%]\u001b[49m\u001b[39m [###############################...........................] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Setting up libx265-179:amd64 (3.2.1-1build1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 55%]\u001b[49m\u001b[39m [###############################...........................] \u001b8Setting up libaom0:amd64 (1.0.0.errata1-3build1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 56%]\u001b[49m\u001b[39m [################################..........................] \u001b8Setting up libmysofa1:amd64 (1.0~dfsg0-1) ...\r\n",
      "Setting up libdc1394-22:amd64 (2.2.5-2.1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 57%]\u001b[49m\u001b[39m [#################################.........................] \u001b8Setting up libcdio18:amd64 (2.0.0-2) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 58%]\u001b[49m\u001b[39m [#################################.........................] \u001b8Setting up libxvidcore4:amd64 (2:1.3.7-1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 59%]\u001b[49m\u001b[39m [##################################........................] \u001b8Setting up libsnappy1v5:amd64 (1.1.8-1build1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 60%]\u001b[49m\u001b[39m [##################################........................] \u001b8Setting up libass9:amd64 (1:0.14.0-2) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 61%]\u001b[49m\u001b[39m [###################################.......................] \u001b8Setting up libva2:amd64 (2.7.0-2) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 62%]\u001b[49m\u001b[39m [####################################......................] \u001b8Setting up libigdgmm11:amd64 (20.1.1+ds1-1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 63%]\u001b[49m\u001b[39m [####################################......................] \u001b8Setting up libcodec2-0.9:amd64 (0.9.2-2) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 64%]\u001b[49m\u001b[39m [#####################################.....................] \u001b8Setting up intel-media-va-driver:amd64 (20.1.1+dfsg1-1) ...\r\n",
      "Setting up libaacs0:amd64 (0.9.0-2) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 65%]\u001b[49m\u001b[39m [#####################################.....................] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Setting up libfftw3-double3:amd64 (3.3.8-2ubuntu1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 66%]\u001b[49m\u001b[39m [######################################....................] \u001b8Setting up librubberband2:amd64 (1.8.2-1build1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 67%]\u001b[49m\u001b[39m [######################################....................] \u001b8Setting up libsndio7.0:amd64 (1.5.0-3) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 68%]\u001b[49m\u001b[39m [#######################################...................] \u001b8Setting up libbdplus0:amd64 (0.1.2-3) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 69%]\u001b[49m\u001b[39m [########################################..................] \u001b8Setting up libvidstab1.1:amd64 (1.1.0-2) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 70%]\u001b[49m\u001b[39m [########################################..................] \u001b8Setting up libflite1:amd64 (2.1-release-3) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 71%]\u001b[49m\u001b[39m [#########################################.................] \u001b8Setting up libva-drm2:amd64 (2.7.0-2) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 72%]\u001b[49m\u001b[39m [#########################################.................] \u001b8Setting up ocl-icd-libopencl1:amd64 (2.2.11-1ubuntu1) ...\r\n",
      "Setting up libbs2b0:amd64 (3.1.0+dfsg-2.2build1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 73%]\u001b[49m\u001b[39m [##########################################................] \u001b8Setting up libopenal-data (1:1.19.1-1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 74%]\u001b[49m\u001b[39m [###########################################...............] \u001b8Setting up mesa-va-drivers:amd64 (21.2.6-0ubuntu0.1~20.04.2) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 75%]\u001b[49m\u001b[39m [###########################################...............] \u001b8Setting up libbluray2:amd64 (1:1.2.0-1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 76%]\u001b[49m\u001b[39m [############################################..............] \u001b8Setting up libsdl2-2.0-0:amd64 (2.0.10+dfsg1-3) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 77%]\u001b[49m\u001b[39m [############################################..............] \u001b8Setting up libva-x11-2:amd64 (2.7.0-2) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 78%]\u001b[49m\u001b[39m [#############################################.............] \u001b8Setting up libwebpmux3:amd64 (0.6.1-2ubuntu0.20.04.1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 79%]\u001b[49m\u001b[39m [#############################################.............] \u001b8Setting up libopenmpt0:amd64 (0.4.11-1build1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 80%]\u001b[49m\u001b[39m [##############################################............] \u001b8Setting up libzvbi-common (0.2.35-17) ...\r\n",
      "Setting up i965-va-driver:amd64 (2.4.0-0ubuntu1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 81%]\u001b[49m\u001b[39m [###############################################...........] \u001b8Setting up libserd-0-0:amd64 (0.30.2-1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 82%]\u001b[49m\u001b[39m [###############################################...........] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Setting up libzvbi0:amd64 (0.2.35-17) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 83%]\u001b[49m\u001b[39m [################################################..........] \u001b8Setting up libcdio-cdda2:amd64 (10.2+2.0.0-1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 84%]\u001b[49m\u001b[39m [################################################..........] \u001b8Setting up libcdio-paranoia2:amd64 (10.2+2.0.0-1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 85%]\u001b[49m\u001b[39m [#################################################.........] \u001b8Setting up libopenal1:amd64 (1:1.19.1-1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 86%]\u001b[49m\u001b[39m [##################################################........] \u001b8Setting up libavutil56:amd64 (7:4.2.7-0ubuntu0.1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 87%]\u001b[49m\u001b[39m [##################################################........] \u001b8Setting up va-driver-all:amd64 (2.7.0-2) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 88%]\u001b[49m\u001b[39m [###################################################.......] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Setting up libpostproc55:amd64 (7:4.2.7-0ubuntu0.1) ...\r\n",
      "Setting up libsord-0-0:amd64 (0.16.4-1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 89%]\u001b[49m\u001b[39m [###################################################.......] \u001b8Setting up libsratom-0-0:amd64 (0.6.4-1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 90%]\u001b[49m\u001b[39m [####################################################......] \u001b8Setting up libswscale5:amd64 (7:4.2.7-0ubuntu0.1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 91%]\u001b[49m\u001b[39m [####################################################......] \u001b8Setting up liblilv-0-0:amd64 (0.24.6-1ubuntu0.1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 92%]\u001b[49m\u001b[39m [#####################################################.....] \u001b8Setting up libswresample3:amd64 (7:4.2.7-0ubuntu0.1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 93%]\u001b[49m\u001b[39m [######################################################....] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Setting up libavresample4:amd64 (7:4.2.7-0ubuntu0.1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 94%]\u001b[49m\u001b[39m [######################################################....] \u001b8Setting up libavcodec58:amd64 (7:4.2.7-0ubuntu0.1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 95%]\u001b[49m\u001b[39m [#######################################################...] \u001b8Setting up libchromaprint1:amd64 (1.4.3-3build1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 96%]\u001b[49m\u001b[39m [#######################################################...] \u001b8Setting up libavformat58:amd64 (7:4.2.7-0ubuntu0.1) ...\r\n",
      "Setting up libavfilter7:amd64 (7:4.2.7-0ubuntu0.1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 97%]\u001b[49m\u001b[39m [########################################################..] \u001b8Setting up libavdevice58:amd64 (7:4.2.7-0ubuntu0.1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 98%]\u001b[49m\u001b[39m [########################################################..] \u001b8Setting up ffmpeg (7:4.2.7-0ubuntu0.1) ...\r\n",
      "\u001b7\u001b[24;0f\u001b[42m\u001b[30mProgress: [ 99%]\u001b[49m\u001b[39m [#########################################################.] \u001b8"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Processing triggers for man-db (2.9.1-1) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Processing triggers for libc-bin (2.31-0ubuntu9.9) ...\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r\n",
      "\u001b7\u001b[0;24r\u001b8\u001b[1A\u001b[J"
     ]
    }
   ],
   "source": [
    "!sudo apt install -y ffmpeg\n",
    "!pip install -q mediapy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:54:55.022075Z",
     "iopub.status.busy": "2022-12-14T21:54:55.021818Z",
     "iopub.status.idle": "2022-12-14T21:54:59.047445Z",
     "shell.execute_reply": "2022-12-14T21:54:59.046350Z"
    },
    "id": "s3khsunT7kWa"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[33mWARNING: Skipping opencv-python-headless as it is not installed.\u001b[0m\u001b[33m\r\n",
      "\u001b[0m"
     ]
    }
   ],
   "source": [
    "!pip uninstall -q -y opencv-python-headless\n",
    "!pip install -q \"opencv-python-headless<4.3\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:54:59.052076Z",
     "iopub.status.busy": "2022-12-14T21:54:59.051294Z",
     "iopub.status.idle": "2022-12-14T21:55:01.498882Z",
     "shell.execute_reply": "2022-12-14T21:55:01.498168Z"
    },
    "id": "dI_1csl6Q-gH"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-12-14 21:55:00.393078: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory\n",
      "2022-12-14 21:55:00.393186: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory\n",
      "2022-12-14 21:55:00.393197: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.\n"
     ]
    }
   ],
   "source": [
    "# Import libraries\n",
    "import pathlib\n",
    "\n",
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "import mediapy as media\n",
    "import numpy as np\n",
    "import PIL\n",
    "\n",
    "import tensorflow as tf\n",
    "import tensorflow_hub as hub\n",
    "import tqdm\n",
    "\n",
    "mpl.rcParams.update({\n",
    "    'font.size': 10,\n",
    "})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Pn8K9oWbmREi"
   },
   "source": [
    "kinetics 600 레이블 목록을 가져오고 처음 몇 개의 레이블을 인쇄합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:55:01.503301Z",
     "iopub.status.busy": "2022-12-14T21:55:01.502639Z",
     "iopub.status.idle": "2022-12-14T21:55:01.511867Z",
     "shell.execute_reply": "2022-12-14T21:55:01.511205Z"
    },
    "id": "2VJUAcjhkfb3"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['background', 'tench', 'goldfish', 'great white shark',\n",
       "       'tiger shark', 'hammerhead', 'electric ray', 'stingray', 'cock',\n",
       "       'hen', 'ostrich', 'brambling', 'goldfinch', 'house finch', 'junco',\n",
       "       'indigo bunting', 'robin', 'bulbul', 'jay', 'magpie'], dtype='<U30')"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labels_path = tf.keras.utils.get_file(\n",
    "    fname='labels.txt',\n",
    "    origin='https://raw.githubusercontent.com/tensorflow/models/f8af2291cced43fc9f1d9b41ddbf772ae7b0d7d2/official/projects/movinet/files/kinetics_600_labels.txt'\n",
    ")\n",
    "labels_path = pathlib.Path(labels_path)\n",
    "\n",
    "lines = labels_path.read_text().splitlines()\n",
    "KINETICS_600_LABELS = np.array([line.strip() for line in lines])\n",
    "KINETICS_600_LABELS[:20]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "G9BU5XsOmaq3"
   },
   "source": [
    "분류를 위한 간단한 예제 비디오를 제공하기 위해 팔 벌려 뛰기를 수행 중인 짧은 gif를 로드할 수 있습니다.\n",
    "\n",
    "![jumping jacks](https://github.com/tensorflow/models/raw/f8af2291cced43fc9f1d9b41ddbf772ae7b0d7d2/official/projects/movinet/files/jumpingjack.gif)\n",
    "\n",
    "저작자 표시: CC-BY 라이선스에 따라 [Coach Bobby Bluford](https://www.youtube.com/watch?v=-AxHpj-EuPg)가 YouTube에서 공유한 영상입니다."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "8aFKMbr4mfSg"
   },
   "source": [
    "gif를 다운로드합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:55:01.515304Z",
     "iopub.status.busy": "2022-12-14T21:55:01.514856Z",
     "iopub.status.idle": "2022-12-14T21:55:02.075544Z",
     "shell.execute_reply": "2022-12-14T21:55:02.074836Z"
    },
    "id": "w62jqXhaSb15"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading data from https://github.com/tensorflow/models/raw/f8af2291cced43fc9f1d9b41ddbf772ae7b0d7d2/official/projects/movinet/files/jumpingjack.gif\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "  8192/783318 [..............................] - ETA: 0s"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r",
      "783318/783318 [==============================] - 0s 0us/step\n"
     ]
    }
   ],
   "source": [
    "jumpingjack_url = 'https://github.com/tensorflow/models/raw/f8af2291cced43fc9f1d9b41ddbf772ae7b0d7d2/official/projects/movinet/files/jumpingjack.gif'\n",
    "jumpingjack_path = tf.keras.utils.get_file(\n",
    "    fname='jumpingjack.gif',\n",
    "    origin=jumpingjack_url,\n",
    "    cache_dir='.', cache_subdir='.',\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "hdRS_22PebfB"
   },
   "source": [
    "gif 파일을 `tf.Tensor`로 읽는 함수를 정의합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:55:02.079652Z",
     "iopub.status.busy": "2022-12-14T21:55:02.078889Z",
     "iopub.status.idle": "2022-12-14T21:55:02.084026Z",
     "shell.execute_reply": "2022-12-14T21:55:02.083380Z"
    },
    "id": "mPhmCu6oSi5f"
   },
   "outputs": [],
   "source": [
    "#@title\n",
    "# Read and process a video\n",
    "def load_gif(file_path, image_size=(224, 224)):\n",
    "  \"\"\"Loads a gif file into a TF tensor.\n",
    "\n",
    "  Use images resized to match what's expected by your model.\n",
    "  The model pages say the \"A2\" models expect 224 x 224 images at 5 fps\n",
    "\n",
    "  Args:\n",
    "    file_path: path to the location of a gif file.\n",
    "    image_size: a tuple of target size.\n",
    "\n",
    "  Returns:\n",
    "    a video of the gif file\n",
    "  \"\"\"\n",
    "  # Load a gif file, convert it to a TF tensor\n",
    "  raw = tf.io.read_file(file_path)\n",
    "  video = tf.io.decode_gif(raw)\n",
    "  # Resize the video\n",
    "  video = tf.image.resize(video, image_size)\n",
    "  # change dtype to a float32\n",
    "  # Hub models always want images normalized to [0,1]\n",
    "  # ref: https://www.tensorflow.org/hub/common_signatures/images#input\n",
    "  video = tf.cast(video, tf.float32) / 255.\n",
    "  return video"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Xx7cZm8vpDJm"
   },
   "source": [
    "비디오의 형상은 `(frames, height, width, colors)`입니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:55:02.087244Z",
     "iopub.status.busy": "2022-12-14T21:55:02.086756Z",
     "iopub.status.idle": "2022-12-14T21:55:05.675034Z",
     "shell.execute_reply": "2022-12-14T21:55:05.674343Z"
    },
    "id": "E7k_PmbFSkHv"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "TensorShape([13, 224, 224, 3])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "jumpingjack=load_gif(jumpingjack_path)\n",
    "jumpingjack.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "LcKFy3oedBvF"
   },
   "source": [
    "## 모델을 사용하는 방법\n",
    "\n",
    "이 섹션에는 [TensorFlow Hub의 모델](https://tfhub.dev/google/collections/movinet/1)을 사용하는 방법을 보여주는 연습이 포함되어 있습니다. 작동 중인 모델만 보고 싶다면 다음 섹션으로 건너뛰세요.\n",
    "\n",
    "각 모델에는 `base` 및 `streaming`의 두 가지 버전이 있습니다.\n",
    "\n",
    "- `base` 버전은 비디오를 입력으로 사용하고 프레임에 대한 평균 확률을 반환합니다.\n",
    "- `streaming` 버전은 비디오 프레임과 RNN 상태를 입력으로 사용하고 해당 프레임과 새 RNN 상태에 대한 예측을 반환합니다. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "WQO6Zb8Hm-9q"
   },
   "source": [
    "### 기본 모델"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "RfnYU20JnPqp"
   },
   "source": [
    "[TensorFlow Hub에서 사전 훈련된 모델](https://tfhub.dev/tensorflow/movinet/a2/base/kinetics-600/classification/3)을 다운로드합니다. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:55:05.678898Z",
     "iopub.status.busy": "2022-12-14T21:55:05.678632Z",
     "iopub.status.idle": "2022-12-14T21:55:21.547403Z",
     "shell.execute_reply": "2022-12-14T21:55:21.546643Z"
    },
    "id": "FnpPo6HSR7qv"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 14.9 s, sys: 760 ms, total: 15.7 s\n",
      "Wall time: 15.9 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "id = 'a2'\n",
    "mode = 'base'\n",
    "version = '3'\n",
    "hub_url = f'https://tfhub.dev/tensorflow/movinet/{id}/{mode}/kinetics-600/classification/{version}'\n",
    "model = hub.load(hub_url)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "jvaFwKhxndmb"
   },
   "source": [
    "이 버전의 모델에는 하나의 `signature`가 있습니다. 형상이 `(batch, frames, height, width, colors)`인 `tf.float32`의 `image` 인수를 취합니다. 그리고 하나의 출력, 즉 형상이 `(batch, classes)`인 로짓의 `tf.float32` 텐서를 포함하는 사전을 반환합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:55:21.551345Z",
     "iopub.status.busy": "2022-12-14T21:55:21.550691Z",
     "iopub.status.idle": "2022-12-14T21:55:21.555157Z",
     "shell.execute_reply": "2022-12-14T21:55:21.554513Z"
    },
    "id": "7GzZ4Y03T_gH"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "signature_wrapper(*, image)\n",
      "  Args:\n",
      "    image: float32 Tensor, shape=(None, None, None, None, 3)\n",
      "  Returns:\n",
      "    {'classifier_head': <1>}\n",
      "      <1>: float32 Tensor, shape=(None, 600)\n"
     ]
    }
   ],
   "source": [
    "sig = model.signatures['serving_default']\n",
    "print(sig.pretty_printed_signature())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "M4Xny1ANomi4"
   },
   "source": [
    "비디오에서 이 서명을 실행하려면 먼저 비디오에 외부 `batch` 차원을 추가해야 합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:55:21.558378Z",
     "iopub.status.busy": "2022-12-14T21:55:21.557867Z",
     "iopub.status.idle": "2022-12-14T21:55:32.526171Z",
     "shell.execute_reply": "2022-12-14T21:55:32.525440Z"
    },
    "id": "LBOFEDG1XvZE"
   },
   "outputs": [],
   "source": [
    "#warmup\n",
    "sig(image = jumpingjack[tf.newaxis, :1]);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:55:32.530375Z",
     "iopub.status.busy": "2022-12-14T21:55:32.530091Z",
     "iopub.status.idle": "2022-12-14T21:55:39.061046Z",
     "shell.execute_reply": "2022-12-14T21:55:39.060273Z"
    },
    "id": "jCeW3KycVbGn"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(600,)\n",
      "\n",
      "CPU times: user 4.13 s, sys: 50.3 ms, total: 4.18 s\n",
      "Wall time: 6.53 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "logits = sig(image = jumpingjack[tf.newaxis, ...])\n",
    "logits = logits['classifier_head'][0]\n",
    "\n",
    "print(logits.shape)\n",
    "print()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "AE8doqkPpxED"
   },
   "source": [
    "위의 출력 처리를 나중에 패키징하는 `get_top_k` 함수를 정의합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:55:39.064960Z",
     "iopub.status.busy": "2022-12-14T21:55:39.064264Z",
     "iopub.status.idle": "2022-12-14T21:55:39.069363Z",
     "shell.execute_reply": "2022-12-14T21:55:39.068783Z"
    },
    "id": "OozPNO6LvZ00"
   },
   "outputs": [],
   "source": [
    "#@title\n",
    "# Get top_k labels and probabilities\n",
    "def get_top_k(probs, k=5, label_map=KINETICS_600_LABELS):\n",
    "  \"\"\"Outputs the top k model labels and probabilities on the given video.\n",
    "\n",
    "  Args:\n",
    "    probs: probability tensor of shape (num_frames, num_classes) that represents\n",
    "      the probability of each class on each frame.\n",
    "    k: the number of top predictions to select.\n",
    "    label_map: a list of labels to map logit indices to label strings.\n",
    "\n",
    "  Returns:\n",
    "    a tuple of the top-k labels and probabilities.\n",
    "  \"\"\"\n",
    "  # Sort predictions to find top_k\n",
    "  top_predictions = tf.argsort(probs, axis=-1, direction='DESCENDING')[:k]\n",
    "  # collect the labels of top_k predictions\n",
    "  top_labels = tf.gather(label_map, top_predictions, axis=-1)\n",
    "  # decode lablels\n",
    "  top_labels = [label.decode('utf8') for label in top_labels.numpy()]\n",
    "  # top_k probabilities of the predictions\n",
    "  top_probs = tf.gather(probs, top_predictions, axis=-1).numpy()\n",
    "  return tuple(zip(top_labels, top_probs))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "kTfKMT29pP_Z"
   },
   "source": [
    "`logits`을 확률로 변환하고 비디오의 상위 5개 클래스를 찾습니다. 모델은 비디오가 `jumping jacks`일 가능성이 있음을 확인합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:55:39.072893Z",
     "iopub.status.busy": "2022-12-14T21:55:39.072462Z",
     "iopub.status.idle": "2022-12-14T21:55:39.087757Z",
     "shell.execute_reply": "2022-12-14T21:55:39.087073Z"
    },
    "id": "Z-SrNGsGV5Mt"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "EntleBucher         : 0.834\n",
      "home theater        : 0.008\n",
      "miniature poodle    : 0.003\n",
      "redshank            : 0.003\n",
      "spider monkey       : 0.002\n"
     ]
    }
   ],
   "source": [
    "probs = tf.nn.softmax(logits, axis=-1)\n",
    "for label, p in get_top_k(probs):\n",
    "  print(f'{label:20s}: {p:.3f}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ltdijoQpqjxZ"
   },
   "source": [
    "### 스트리밍 모델"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "9dqdUPQXq45b"
   },
   "source": [
    "이전 섹션에서는 전체 비디오를 실행하는 모델을 사용했습니다. 종종 비디오를 처리할 때 마지막에 하나의 예측을 원하지 않고 프레임별로 예측을 업데이트하고 싶을 수 있습니다. 모델의 `stream` 버전을 사용하면 가능합니다.\n",
    "\n",
    "모델의 `stream` 버전을 로드합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:55:39.091220Z",
     "iopub.status.busy": "2022-12-14T21:55:39.090624Z",
     "iopub.status.idle": "2022-12-14T21:56:25.037394Z",
     "shell.execute_reply": "2022-12-14T21:56:25.036612Z"
    },
    "id": "mxt0hRXFZkAM"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 43.8 s, sys: 2.19 s, total: 46 s\n",
      "Wall time: 45.9 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "id = 'a2'\n",
    "mode = 'stream'\n",
    "version = '3'\n",
    "hub_url = f'https://tfhub.dev/tensorflow/movinet/{id}/{mode}/kinetics-600/classification/{version}'\n",
    "model = hub.load(hub_url)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "pDswtsGgsYGS"
   },
   "source": [
    "이 모델을 사용하는 것은 `base` 모델보다 약간 더 복잡합니다. 모델의 RNN에 대한 내부 상태를 추적해야 합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:56:25.041343Z",
     "iopub.status.busy": "2022-12-14T21:56:25.040799Z",
     "iopub.status.idle": "2022-12-14T21:56:25.045633Z",
     "shell.execute_reply": "2022-12-14T21:56:25.045003Z"
    },
    "id": "0fM_Vb1VsbDm"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['call', 'init_states']"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list(model.signatures.keys())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ojr1_iYCtPvp"
   },
   "source": [
    "`init_states` 서명은 비디오의 **형상** `(batch, frames, height, width, colors)`를 입력으로 사용하고 초기 RNN 상태를 포함하는 텐서의 방대한 사전을 반환합니다. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:56:25.049240Z",
     "iopub.status.busy": "2022-12-14T21:56:25.048704Z",
     "iopub.status.idle": "2022-12-14T21:56:25.053641Z",
     "shell.execute_reply": "2022-12-14T21:56:25.053014Z"
    },
    "id": "67loYFGpo_RP"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "signature_wrapper(*, input_shape).\n",
      "  Args:.\n",
      "    input_shape: int32 Tensor, shape=(5,).\n",
      "  Returns:.\n",
      "    {'state/b0/l0/pool_buffer': <1>, 'state/b0/l0/pool_frame_count': <2>, 'state/b0/l1/pool_buffer': <3>, 'state/b0/l1/pool_frame_count': <4>, 'state/b0/l1/stream_buffer': <5>, 'state/b0/l2/pool_buffer': <6>, 'state/b0/l2/pool_frame_count': <7>, 'state/b0/l2/stream_buffer': <8>, 'state/b1/l0/pool_buffer': <9>, 'state/b1/l0/pool_frame_count': <10>, 'state/b1/l0/stream_buffer': <11>, 'state/b1/l1/pool_buffer': <12>, 'state/b1/l1/pool_frame_count': <13>, 'state/b1/l1/stream_buffer': <14>, 'state/b1/l2/pool_buffer': <15>, 'state/b1/l2/pool_frame_count': <16>, 'state/b1/l2/stream_buffer': <17>, 'state/b1/l3/pool_buffer': <18>, 'state/b1/l3/pool_frame_count': <19>, 'state/b1/l3/stream_buffer': <20>, 'state/b1/l4/pool_buffer': <21>, 'state/b1/l4/pool_frame_count': <22>, 'state/b1/l4/stream_buffer': <23>, 'state/b2/l0/pool_buffer': <24>, 'state/b2/l0/pool_frame_count': <25>, 'state/b2/l0/stream_buffer': <26>, 'state/b2/l1/pool_buffer': <27>, 'state/b2/l1/pool_frame_count': <28>, 'state/b2/l1/stream_buffer': <29>, 'state/b2/l2/pool_buffer': <30>, 'state/b2/l2/pool_frame_count': <31>, 'state/b2/l2/stream_buffer': <32>, 'state/b2/l3/pool_buffer': <33>, 'state/b2/l3/pool_frame_count': <34>, 'state/b2/l3/stream_buffer': <35>, 'state/b2/l4/pool_buffer': <36>, 'state/b2/l4/pool_frame_count': <37>, 'state/b2/l4/stream_buffer': <38>, 'state/b3/l0/pool_buffer': <39>, 'state/b3/l0/pool_frame_count': <40>, 'state/b3/l0/stream_buffer': <41>, 'state/b3/l1/pool_buffer': <42>, 'state/b3/l1/pool_frame_count': <43>, 'state/b3/l1/stream_buffer': <44>, 'state/b3/l2/pool_buffer': <45>, 'state/b3/l2/pool_frame_count': <46>, 'state/b3/l2/stream_buffer': <47>, 'state/b3/l3/pool_buffer': <48>, 'state/b3/l3/pool_frame_count': <49>, 'state/b3/l3/stream_buffer': <50>, 'state/b3/l4/pool_buffer': <51>, 'state/b3/l4/pool_frame_count': <52>, 'state/b3/l5/pool_buffer': <53>, 'state/b3/l5/pool_frame_count': <54>, 'state/b3/l5/stream_buffer': <55>, 'state/b4/l0/pool_buffer': <56>, 'state/b4/l0/pool_frame_count': <57>, 'state/b4/l0/stream_buffer': <58>, 'state/b4/l1/pool_buffer': <59>, 'state/b4/l1/pool_frame_count': <60>, 'state/b4/l2/pool_buffer': <61>, 'state/b4/l2/pool_frame_count': <62>, 'state/b4/l3/pool_buffer': <63>, 'state/b4/l3/pool_frame_count': <64>, 'state/b4/l4/pool_buffer': <65>, 'state/b4/l4/pool_frame_count': <66>, 'state/b4/l5/pool_buffer': <67>, 'state/b4/l5/pool_frame_count': <68>, 'state/b4/l5/stream_buffer': <69>, 'state/b4/l6/pool_buffer': <70>, 'state/b4/l6/pool_frame_count': <71>, 'state/head/pool_buffer': <72>, 'state/head/pool_frame_count': <73>}.\n",
      "      <1>: float32 Tensor, shape=(None, 1, 1, 1, 40).\n",
      "      <2>: int32 Tensor, shape=(1,).\n",
      "      <3>: float32 Tensor, shape=(None, 1, 1, 1, 40).\n",
      "      <4>: int32 Tensor, shape=(1,).\n",
      "      <5>: float32 Tensor, shape=(None, 2, None, None, 40).\n",
      "      ...\n"
     ]
    }
   ],
   "source": [
    "lines = model.signatures['init_states'].pretty_printed_signature().splitlines()\n",
    "lines = lines[:10]\n",
    "lines.append('      ...')\n",
    "print('.\\n'.join(lines))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:56:25.057146Z",
     "iopub.status.busy": "2022-12-14T21:56:25.056553Z",
     "iopub.status.idle": "2022-12-14T21:56:25.290533Z",
     "shell.execute_reply": "2022-12-14T21:56:25.289787Z"
    },
    "id": "v5lG3vejn5df"
   },
   "outputs": [],
   "source": [
    "initial_state = model.init_states(jumpingjack[tf.newaxis, ...].shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:56:25.294575Z",
     "iopub.status.busy": "2022-12-14T21:56:25.294300Z",
     "iopub.status.idle": "2022-12-14T21:56:25.298810Z",
     "shell.execute_reply": "2022-12-14T21:56:25.298222Z"
    },
    "id": "J3DwmyHnuhH_"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type(initial_state)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:56:25.302115Z",
     "iopub.status.busy": "2022-12-14T21:56:25.301497Z",
     "iopub.status.idle": "2022-12-14T21:56:25.306011Z",
     "shell.execute_reply": "2022-12-14T21:56:25.305464Z"
    },
    "id": "K8SyiEU6tB-e"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['state/b0/l0/pool_buffer',\n",
       " 'state/b0/l0/pool_frame_count',\n",
       " 'state/b0/l1/pool_buffer',\n",
       " 'state/b0/l1/pool_frame_count',\n",
       " 'state/b0/l1/stream_buffer']"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list(sorted(initial_state.keys()))[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "xeMCzJMBvwRF"
   },
   "source": [
    "RNN의 초기 상태가 확보되면 상태와 비디오 프레임을 입력으로 전달할 수 있습니다(비디오 프레임의 `(batch, frames, height, width, colors)` 형상 유지). 모델은 `(logits, state)` 쌍을 반환합니다.\n",
    "\n",
    "첫 번째 프레임을 본 후 모델은 비디오가 \"팔 벌려 뛰기\"인지 확신하지 못합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:56:25.309109Z",
     "iopub.status.busy": "2022-12-14T21:56:25.308877Z",
     "iopub.status.idle": "2022-12-14T21:56:25.314347Z",
     "shell.execute_reply": "2022-12-14T21:56:25.313689Z"
    },
    "id": "McSLdIgtsI3d"
   },
   "outputs": [],
   "source": [
    "inputs = initial_state.copy()\n",
    "\n",
    "# Add the batch axis, take the first frme, but keep the frame-axis.\n",
    "inputs['image'] = jumpingjack[tf.newaxis, 0:1, ...] "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:56:25.317715Z",
     "iopub.status.busy": "2022-12-14T21:56:25.317193Z",
     "iopub.status.idle": "2022-12-14T21:56:32.462447Z",
     "shell.execute_reply": "2022-12-14T21:56:32.461704Z"
    },
    "id": "WlH7PqLPX664"
   },
   "outputs": [],
   "source": [
    "# warmup\n",
    "model(inputs);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:56:32.466758Z",
     "iopub.status.busy": "2022-12-14T21:56:32.466149Z",
     "iopub.status.idle": "2022-12-14T21:56:32.512728Z",
     "shell.execute_reply": "2022-12-14T21:56:32.512068Z"
    },
    "id": "7uzNXtu7X5sr"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "miniature schnauzer : 0.427\n",
      "corkscrew           : 0.134\n",
      "miniature poodle    : 0.056\n",
      "church              : 0.053\n",
      "fly                 : 0.039\n",
      "\n"
     ]
    }
   ],
   "source": [
    "logits, new_state = model(inputs)\n",
    "logits = logits[0]\n",
    "probs = tf.nn.softmax(logits, axis=-1)\n",
    "\n",
    "for label, p in get_top_k(probs):\n",
    "  print(f'{label:20s}: {p:.3f}')\n",
    "\n",
    "print()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "oLU644FQwXSb"
   },
   "source": [
    "루프에서 모델을 실행하여 각 프레임과 함께 업데이트된 상태를 전달하면 모델이 올바른 결과로 빠르게 수렴됩니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:56:32.516190Z",
     "iopub.status.busy": "2022-12-14T21:56:32.515550Z",
     "iopub.status.idle": "2022-12-14T21:56:33.029414Z",
     "shell.execute_reply": "2022-12-14T21:56:33.028651Z"
    },
    "id": "Fzm7T4ImmIEg"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 552 ms, sys: 17.7 ms, total: 570 ms\n",
      "Wall time: 509 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "state = initial_state.copy()\n",
    "all_logits = []\n",
    "\n",
    "for n in range(len(jumpingjack)):\n",
    "  inputs = state\n",
    "  inputs['image'] = jumpingjack[tf.newaxis, n:n+1, ...]\n",
    "  result, state = model(inputs)\n",
    "  all_logits.append(logits)\n",
    "\n",
    "probabilities = tf.nn.softmax(all_logits, axis=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:56:33.033116Z",
     "iopub.status.busy": "2022-12-14T21:56:33.032478Z",
     "iopub.status.idle": "2022-12-14T21:56:33.039179Z",
     "shell.execute_reply": "2022-12-14T21:56:33.038502Z"
    },
    "id": "B7UtHoSWcOT2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "miniature schnauzer : 0.427\n",
      "corkscrew           : 0.134\n",
      "miniature poodle    : 0.056\n",
      "church              : 0.053\n",
      "fly                 : 0.039\n"
     ]
    }
   ],
   "source": [
    "for label, p in get_top_k(probabilities[-1]):\n",
    "  print(f'{label:20s}: {p:.3f}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:56:33.042393Z",
     "iopub.status.busy": "2022-12-14T21:56:33.041863Z",
     "iopub.status.idle": "2022-12-14T21:56:33.198798Z",
     "shell.execute_reply": "2022-12-14T21:56:33.198060Z"
    },
    "id": "6ffV3NhZcsrv"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkAAAAGwCAYAAABB4NqyAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8o6BhiAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAwXUlEQVR4nO3df1RUdcLH8c8AAooKCgGiKP6oyB8pSpKapkcKrUez3cpcQzNr233UVLLSfTZNd4tsn3XNZHUlM2st7YdZa62bkpqWpUGU1vobxDRANAFREWfu80fHeXYCbW7OMEP3/TpnznG+c++dz9xTzOfc+517bYZhGAIAALCQAF8HAAAAqG8UIAAAYDkUIAAAYDkUIAAAYDkUIAAAYDkUIAAAYDkUIAAAYDlBvg7gjxwOh44ePapmzZrJZrP5Og4AAHCDYRiqrKxUXFycAgIufYyHAlSHo0ePKj4+3tcxAADAT3D48GG1adPmkstQgOrQrFkzSd/vwObNm/s4DQAAcEdFRYXi4+Od3+OXQgGqw4XTXs2bN6cAAQDQwLgzfYVJ0AAAwHL8ogBlZWUpISFBoaGhSklJ0fbt291ab+XKlbLZbBoxYoTL+L333iubzebyGDJkiBeSAwCAhsjnBWjVqlXKyMjQrFmzlJeXp+7duystLU2lpaWXXK+wsFDTpk1T//7963x9yJAh+vbbb52PV1991RvxAQBAA+TzAjRv3jw98MADGjdunDp37qzFixerSZMmeuGFFy66jt1u1+jRozV79mx16NChzmVCQkIUGxvrfLRo0cJbHwEAADQwPi1A586dU25urlJTU51jAQEBSk1N1bZt2y663pw5cxQdHa3x48dfdJlNmzYpOjpaV199tX7729/q+PHjF122urpaFRUVLg8AAPDz5dMCVFZWJrvdrpiYGJfxmJgYFRcX17nO1q1btXTpUmVnZ190u0OGDNFLL72knJwczZ07V5s3b9bQoUNlt9vrXD4zM1Ph4eHOB9cAAgDg561B/Qy+srJS6enpys7OVlRU1EWXu/vuu53/7tatm6699lp17NhRmzZt0uDBg2stP2PGDGVkZDifX7iOAAAA+HnyaQGKiopSYGCgSkpKXMZLSkoUGxtba/kDBw6osLBQw4YNc445HA5JUlBQkPbs2aOOHTvWWq9Dhw6KiorS/v376yxAISEhCgkJudyPAwAAGgifngILDg5Wr169lJOT4xxzOBzKyclRnz59ai2fmJionTt3Kj8/3/kYPny4Bg0apPz8/Isetfnmm290/PhxtWrVymufBQAANBw+PwWWkZGhsWPHKjk5Wb1799b8+fNVVVWlcePGSZLGjBmj1q1bKzMzU6GhoeratavL+hEREZLkHD916pRmz56tX/7yl4qNjdWBAwf06KOPqlOnTkpLS6vXzwYAAPyTzwvQyJEjdezYMc2cOVPFxcXq0aOH1q1b55wYXVRU9KN3dP1PgYGB+vLLL7V8+XKdPHlScXFxuvnmm/WHP/yB01wAAECSZDMMw/B1CH9TUVGh8PBwlZeXcy8wAAAaCDPf3z6/ECIAAEB9owABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADLoQABAADL8YsClJWVpYSEBIWGhiolJUXbt293a72VK1fKZrNpxIgRF13mN7/5jWw2m+bPn++ZsAAAoMHzeQFatWqVMjIyNGvWLOXl5al79+5KS0tTaWnpJdcrLCzUtGnT1L9//4su89Zbb+mTTz5RXFycp2MDAIAGzOcFaN68eXrggQc0btw4de7cWYsXL1aTJk30wgsvXHQdu92u0aNHa/bs2erQoUOdyxw5ckSTJk3SihUr1KhRI2/FBwAADZBPC9C5c+eUm5ur1NRU51hAQIBSU1O1bdu2i643Z84cRUdHa/z48XW+7nA4lJ6erkceeURdunT50RzV1dWqqKhweQAAgJ8vnxagsrIy2e12xcTEuIzHxMSouLi4znW2bt2qpUuXKjs7+6LbnTt3roKCgvTQQw+5lSMzM1Ph4eHOR3x8vPsfAgAANDg+PwVmRmVlpdLT05Wdna2oqKg6l8nNzdWzzz6rF198UTabza3tzpgxQ+Xl5c7H4cOHPRkbAAD4mSBfvnlUVJQCAwNVUlLiMl5SUqLY2Nhayx84cECFhYUaNmyYc8zhcEiSgoKCtGfPHm3ZskWlpaVq27atcxm73a6HH35Y8+fPV2FhYa3thoSEKCQkxEOfCgAA+DufFqDg4GD16tVLOTk5zp+yOxwO5eTkaOLEibWWT0xM1M6dO13Gfv/736uyslLPPvus4uPjlZ6e7jKnSJLS0tKUnp6ucePGee2zAACAhsOnBUiSMjIyNHbsWCUnJ6t3796aP3++qqqqnGVlzJgxat26tTIzMxUaGqquXbu6rB8RESFJzvHIyEhFRka6LNOoUSPFxsbq6quv9v4HAgAAfs/nBWjkyJE6duyYZs6cqeLiYvXo0UPr1q1zTowuKipSQECDmqoEAAD8nM0wDMPXIfxNRUWFwsPDVV5erubNm/s6DgAAcIOZ728OrQAAAMuhAAEAAMuhAAEAAMuhAAEAAMuhAAEAAMuhAAEAAMuhAAEAAMuhAAEAAMuhAAEAAMsxdSuMkydP6q233tKWLVt06NAhnT59WldccYWSkpKUlpamvn37eisnAACAx7h1BOjo0aO6//771apVK/3xj3/UmTNn1KNHDw0ePFht2rTRxo0bddNNN6lz585atWqVtzMDAABcFreOACUlJWns2LHKzc1V586d61zmzJkzWrNmjebPn6/Dhw9r2rRpHg0KAADgKW7dDPX48eOKjIx0e6Nml/c33AwVAICGx+M3QzVbZhpy+QEAAD9/bk+C/vDDD12eDxgwwONhAAAA6oPbBWjs2LHOf9tsNh08eNArgQAAALzN7QJUUFDgzRwAAAD1xtSFEGtqajR48GDt27fPW3kAAAC8zlQBatSokb788ktvZQEAAKgXpm+Fcc8992jp0qXeyAIAAFAvTN0KQ5LOnz+vF154QRs2bFCvXr0UFhbm8vq8efM8Fg4AAMAbTBegXbt2qWfPnpKkvXv3urxms9k8kwoAAMCLTBegjRs3eiMHAABAvTE9B+iC/fv361//+pfOnDkjSXLjjhoAAAB+wXQBOn78uAYPHqyrrrpKt9xyi7799ltJ0vjx4/Xwww97PCAAAICnmS5AU6dOVaNGjVRUVKQmTZo4x0eOHKl169Z5NBwAAIA3mJ4D9P777+tf//qX2rRp4zJ+5ZVX6tChQx4LBgAA4C2mjwBVVVW5HPm54MSJEwoJCfFIKAAAAG8yXYD69++vl156yfncZrPJ4XDomWee0aBBgzwaDgAAwBtMnwJ75plnNHjwYH322Wc6d+6cHn30UX311Vc6ceKEPvroI29kBAAA8CjTR4C6du2qvXv36oYbbtBtt92mqqoq/eIXv9Dnn3+ujh07eiMjAACAR9kMLuBTS0VFhcLDw1VeXq7mzZv7Og4AAHCDme9v00eAOnTooHHjxqm6utplvKysTB06dDC7OQAAgHpnugAVFhbqo48+Uv/+/VVcXOwct9vt/AweAAA0CKYLkM1m07p169SmTRv16tVLO3bs8EYuAAAArzFdgAzDUNOmTbV69WqNGTNGN954o/7+9797IxsAAIBXmP4ZvM1mc/47MzNTXbp00QMPPKBRo0Z5NBgAAIC3mC5AP/zR2D333KOOHTvq9ttv91goAAAAbzJdgBwOR62xPn366IsvvtDu3bs9EgoAAMCbTBegi4mJiVFMTIynNgcAAOA1pgtQ+/btXeYB/dDBgwcvKxAAAIC3mS5AU6ZMcXleU1Ojzz//XOvWrdMjjzziqVwAAABeY7oATZ48uc7xrKwsffbZZ5cdCAAAwNtMXwfoYoYOHao333zTU5sDAADwGo8VoDfeeEMtW7b01OYAAAC8xvQpsKSkJJdJ0IZhqLi4WMeOHdNf//pXj4YDAADwBtMFaMSIES7PAwICdMUVV2jgwIFKTEz0VC4AAACvsRk/vLQzVFFRofDwcJWXl6t58+a+jgMAANxg5vv7si6EePbsWZ07d85ljMIAAAD8nelJ0FVVVZo4caKio6MVFhamFi1auDwAAAD8nekC9Oijj+qDDz7QokWLFBISoueff16zZ89WXFycXnrpJW9kBAAA8CjTp8D+8Y9/6KWXXtLAgQM1btw49e/fX506dVK7du20YsUKjR492hs5AQAAPMb0EaATJ06oQ4cOkr6f73PixAlJ0g033KAPP/zQs+kAAAC8wHQB6tChgwoKCiRJiYmJeu211yR9f2QoIiLCo+EAAAC8wXQBGjdunL744gtJ0vTp05WVlaXQ0FBNnTqVm6ECAIAG4bKvA3To0CHl5uaqU6dOuvbaaz2Vy6e4DhAAAA1PvV0HSJLatWundu3aXe5mAAAA6s1PKkA5OTnKyclRaWmpHA6Hy2svvPCCR4IBAAB4i+kCNHv2bM2ZM0fJyclq1aqVy41RAQAAGgLTBWjx4sV68cUXlZ6e7o08AAAAXmf6V2Dnzp1T3759vZEFAACgXpguQPfff79eeeUVb2QBAACoF6ZPgZ09e1ZLlizRhg0bdO2116pRo0Yur8+bN89j4QAAALzBdAH68ssv1aNHD0nSrl27XF5jQjQAAGgITBegjRs3eiMHAABAvTE9BwgAAKChM30EqKqqSk8//fRFL4R48OBBj4UDAADwBtMF6P7779fmzZuVnp7OhRABAECDZLoA/fOf/9S7776rfv36eSxEVlaW/vSnP6m4uFjdu3fXc889p969e//oeitXrtSoUaN02223ac2aNc7xJ554QitXrtThw4cVHBysXr166cknn1RKSorHMgMAgIbL9BygFi1aqGXLlh4LsGrVKmVkZGjWrFnKy8tT9+7dlZaWptLS0kuuV1hYqGnTpql///61Xrvqqqu0cOFC7dy5U1u3blVCQoJuvvlmHTt2zGO5AQBAw2UzDMMws8Lf//53vf3221q+fLmaNGly2QFSUlJ03XXXaeHChZIkh8Oh+Ph4TZo0SdOnT69zHbvdrgEDBui+++7Tli1bdPLkSZcjQD9UUVGh8PBwbdiwQYMHD671enV1taqrq12Wj4+PV3l5uZo3b355HxAAANSLC9/37nx/u3UKLCkpyWWuz/79+xUTE6OEhIRaF0LMy8tzO+i5c+eUm5urGTNmOMcCAgKUmpqqbdu2XXS9OXPmKDo6WuPHj9eWLVt+9D2WLFmi8PBwde/evc5lMjMzNXv2bLdzAwCAhs2tAjRixAivvHlZWZnsdrtiYmJcxmNiYrR79+4619m6dauWLl2q/Pz8S2577dq1uvvuu3X69Gm1atVK69evV1RUVJ3LzpgxQxkZGc7nF44AAQCAnye3CtCsWbO8ncMtlZWVSk9PV3Z29kXLzAWDBg1Sfn6+ysrKlJ2drbvuukuffvqpoqOjay0bEhKikJAQb8UGAAB+xvSvwHbs2CGHw1HrF1WffvqpAgMDlZyc7Pa2oqKiFBgYqJKSEpfxkpISxcbG1lr+wIEDKiws1LBhw5xjF65DFBQUpD179qhjx46SpLCwMHXq1EmdOnXS9ddfryuvvFJLly51Od0GAACsyfSvwCZMmKDDhw/XGj9y5IgmTJhgalsXfqKek5PjHHM4HMrJyVGfPn1qLZ+YmKidO3cqPz/f+Rg+fLjzaM+lTls5HA6Xic4AAMC6TB8B+vrrr9WzZ89a40lJSfr6669NB8jIyNDYsWOVnJys3r17a/78+aqqqtK4ceMkSWPGjFHr1q2VmZmp0NBQde3a1WX9iIgISXKOV1VV6cknn9Tw4cPVqlUrlZWVKSsrS0eOHNGdd95pOh8AAPj5MV2AQkJCVFJSog4dOriMf/vttwoKMr05jRw5UseOHdPMmTNVXFysHj16aN26dc6J0UVFRQoIcP9AVWBgoHbv3q3ly5errKxMkZGRuu6667RlyxZ16dLFdD4AAPDzY/o6QKNGjdK3336rt99+W+Hh4ZKkkydPasSIEYqOjtZrr73mlaD1ycx1BAAAgH/w+HWA/tP//u//asCAAWrXrp2SkpIkSfn5+YqJidHLL7/80xIDAADUI9MFqHXr1vryyy+1YsUKffHFF2rcuLHGjRunUaNG1booIgAAgD8yP2lH3//E/Ne//rWnswAAANQL0z+DX758ud59913n80cffVQRERHq27evDh065NFwAAAA3mC6AD311FNq3LixJGnbtm1auHChnnnmGUVFRWnq1KkeDwgAAOBppk+BHT58WJ06dZIkrVmzRnfccYd+/etfq1+/fho4cKCn8wEAAHic6SNATZs21fHjxyVJ77//vm666SZJUmhoqM6cOePZdAAAAF5g+gjQTTfdpPvvv19JSUnau3evbrnlFknSV199pYSEBE/nAwAA8DjTR4CysrLUp08fHTt2TG+++aYiIyMlSbm5uRo1apTHAwIAAHia6StBWwFXggYAoOHx6pWg8dMZhqEzNXZfxwAAwOcaNwqUzWbz2ftTgOrRmRq7Os/8l69jAADgc1/PSVOTYN/VENNzgAAAABo6jgDVo8aNAvX1nDRfxwAAwOcaNwr06fv/pAJ0/vx5bdq0SQcOHNCvfvUrNWvWTEePHlXz5s3VtGlTT2f82bDZbD493AcAAL5n+tv40KFDGjJkiIqKilRdXa2bbrpJzZo109y5c1VdXa3Fixd7IycAAIDHmJ4DNHnyZCUnJ+u7775z3hNMkm6//Xbl5OR4NBwAAIA3mD4CtGXLFn388ccKDg52GU9ISNCRI0c8FgwAAMBbTB8BcjgcsttrX8vmm2++UbNmzTwSCgAAwJtMF6Cbb75Z8+fPdz632Ww6deqUZs2a5bwvGAAAgD8zfSuMw4cPa8iQITIMQ/v27VNycrL27dunqKgoffjhh4qOjvZW1nrDrTAAAGh4zHx//6R7gZ0/f16rVq3SF198oVOnTqlnz54aPXq0y6TohowCBABAw+O1AlRTU6PExEStXbtW11xzzWUH9VcUIAAAGh4z39+m5gA1atRIZ8+evaxwAAAAvmZ6EvSECRM0d+5cnT9/3ht5AAAAvM70dYB27NihnJwcvf/+++rWrZvCwsJcXl+9erXHwgEAAHiD6QIUERGhX/7yl97IAgAAUC9MF6Bly5Z5IwcAAEC9MT0HCAAAoKEzfQSoffv2stlsF3394MGDlxUIAADA20wXoClTprg8r6mp0eeff65169bpkUce8VQuAAAArzFdgCZPnlzneFZWlj777LPLDgQAAOBtHpsDNHToUL355pue2hwAAIDXeKwAvfHGG2rZsqWnNgcAAOA1pk+BJSUluUyCNgxDxcXFOnbsmP761796NBwAAIA3mC5At912m0sBCggI0BVXXKGBAwcqMTHRo+EAAAC8wdTd4K2Cu8EDANDweO1u8JIUGBio0tLSWuPHjx9XYGCg2c0BAADUO9MF6GIHjKqrqxUcHHzZgQAAALzN7TlACxYskCTZbDY9//zzatq0qfM1u92uDz/8kDlAAACgQXC7AP3lL3+R9P0RoMWLF7uc7goODlZCQoIWL17s+YQAAAAe5nYBKigokCQNGjRIq1evVosWLbwWCgAAwJtM/wx+48aN3sgBAABQb0wXIEn65ptv9M4776ioqEjnzp1zeW3evHkeCQYAAOAtpgtQTk6Ohg8frg4dOmj37t3q2rWrCgsLZRiGevbs6Y2MAAAAHmX6Z/AzZszQtGnTtHPnToWGhurNN9/U4cOHdeONN+rOO+/0RkYAAACPMl2A/v3vf2vMmDGSpKCgIJ05c0ZNmzbVnDlzNHfuXI8HBAAA8DTTBSgsLMw576dVq1Y6cOCA87WysjLPJQMAAPAS03OArr/+em3dulXXXHONbrnlFj388MPauXOnVq9ereuvv94bGQEAADzKdAGaN2+eTp06JUmaPXu2Tp06pVWrVunKK6/kF2AAAKBB4G7wdeBu8AAANDxevRs8AABAQ+fWKbCWLVtq7969ioqKUosWLWSz2S667IkTJzwWDgAAwBvcKkB/+ctf1KxZM0nS/PnzvZkHAADA65gDVAfmAAEA0PCY+f7+SfcCczgc2r9/v0pLS+VwOFxeGzBgwE/ZJAAAQL0xXYA++eQT/epXv9KhQ4f0w4NHNptNdrvdY+EAAAC8wXQB+s1vfqPk5GS9++67atWq1SUnRAMAAPgj0wVo3759euONN9SpUydv5AEAAPA609cBSklJ0f79+72RBQAAoF6YPgI0adIkPfzwwyouLla3bt3UqFEjl9evvfZaj4UDAADwBtM/gw8IqH3QyGazyTCMn80kaH4GDwBAw+PVn8EXFBT85GAAAAD+wHQBateunTdyAAAA1Bu3CtA777yjoUOHqlGjRnrnnXcuuezw4cM9EgwAAMBb3JoDFBAQoOLiYkVHR9c5B8i5MeYAAQAAH/H4HKD/vN3FD299AQAA0NCYvg6QN2RlZSkhIUGhoaFKSUnR9u3b3Vpv5cqVstlsGjFihHOspqZGjz32mLp166awsDDFxcVpzJgxOnr0qJfSAwCAhuYn3Qx1x44d2rhxY503Q503b56pba1atUoZGRlavHixUlJSNH/+fKWlpWnPnj2Kjo6+6HqFhYWaNm2a+vfv7zJ++vRp5eXl6fHHH1f37t313XffafLkyRo+fLg+++wzU9kAAMDPk+nrAD311FP6/e9/r6uvvloxMTEu9wKz2Wz64IMPTAVISUnRddddp4ULF0r6/hRbfHy8Jk2apOnTp9e5jt1u14ABA3Tfffdpy5YtOnnypNasWXPR99ixY4d69+6tQ4cOqW3btj+aiTlAAAA0PF69DtCzzz6rF154Qffee+9Pzed07tw55ebmasaMGc6xgIAApaamatu2bRddb86cOYqOjtb48eO1ZcuWH32f8vJy2Ww2RURE1Pl6dXW1qqurnc8rKirc/xAAAKDBMT0HKCAgQP369fPIm5eVlclutysmJsZlPCYmRsXFxXWus3XrVi1dulTZ2dluvcfZs2f12GOPadSoURdtg5mZmQoPD3c+4uPjzX0QAADQoJguQFOnTlVWVpY3svyoyspKpaenKzs7W1FRUT+6fE1Nje666y4ZhqFFixZddLkZM2aovLzc+Th8+LAnYwMAAD9j+hTYtGnTdOutt6pjx47q3LlzrZuhrl692u1tRUVFKTAwUCUlJS7jJSUlio2NrbX8gQMHVFhYqGHDhjnHLkzCDgoK0p49e9SxY0dJ/19+Dh06pA8++OCS5wJDQkIUEhLidm4AANCwmT4C9NBDD2njxo266qqrFBkZ6XLqKDw83NS2goOD1atXL+Xk5DjHHA6HcnJy1KdPn1rLJyYmaufOncrPz3c+hg8frkGDBik/P9956upC+dm3b582bNigyMhIsx8TAAD8jJk+ArR8+XK9+eabuvXWWz0SICMjQ2PHjlVycrJ69+6t+fPnq6qqSuPGjZMkjRkzRq1bt1ZmZqZCQ0PVtWtXl/UvTGy+MF5TU6M77rhDeXl5Wrt2rex2u3M+UcuWLRUcHOyR3AAAoOEyXYBatmzpPM3kCSNHjtSxY8c0c+ZMFRcXq0ePHlq3bp1zYnRRUdElb7/xQ0eOHHHer6xHjx4ur23cuFEDBw70VHQAANBAmb4O0LJly7Ru3TotW7ZMTZo08VYun+I6QAAANDxevQ7QggULdODAAcXExCghIaHWJOi8vDyzmwQAAKhXpgvQf953CwAAoCEyfQrMCjgFBgBAw2Pm+9sv7gYPAABQn9w+Bda+fXuXG58ePHjQK4EAAAC8ze0C9OKLL3oxBgAAQP1xuwDdeOON3swBAABQb9yaA1RVVWVqo2aXBwAAqE9uFaBOnTrp6aef1rfffnvRZQzD0Pr16zV06FAtWLDAYwEBAAA8za1TYJs2bdLvfvc7PfHEE+revbuSk5MVFxen0NBQfffdd/r666+1bds2BQUFacaMGXrwwQe9nRsAAOAnM3UdoKKiIr3++uvasmWLDh06pDNnzigqKkpJSUlKS0vT0KFDFRgY6M289YLrAAEA0PCY+f7mQoh1oAABANDw1NuFEA3DEP0JAAA0ND+pAC1dulRdu3ZVaGioQkND1bVrVz3//POezgYAAOAVpm+GOnPmTM2bN0+TJk1Snz59JEnbtm3T1KlTVVRUpDlz5ng8JAAAgCeZngN0xRVXaMGCBRo1apTL+KuvvqpJkyaprKzMowF9gTlAAAA0PF6dA1RTU6Pk5ORa47169dL58+fNbg4AAKDemS5A6enpWrRoUa3xJUuWaPTo0R4JBQAA4E2m5wBJ30+Cfv/993X99ddLkj799FMVFRVpzJgxysjIcC43b948z6QEAADwINMFaNeuXerZs6ck6cCBA5KkqKgoRUVFadeuXc7lbDabhyICAAB4lukCtHHjRm/kAAAAqDeXdSFEAACAhogCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALIcCBAAALMfnBSgrK0sJCQkKDQ1VSkqKtm/f7tZ6K1eulM1m04gRI1zGV69erZtvvlmRkZGy2WzKz8/3fGgAANCg+bQArVq1ShkZGZo1a5by8vLUvXt3paWlqbS09JLrFRYWatq0aerfv3+t16qqqnTDDTdo7ty53ooNAAAaOJthGIav3jwlJUXXXXedFi5cKElyOByKj4/XpEmTNH369DrXsdvtGjBggO677z5t2bJFJ0+e1Jo1a2otV1hYqPbt2+vzzz9Xjx49Lpmjurpa1dXVzucVFRWKj49XeXm5mjdv/pM/HwAAqD8VFRUKDw936/vbZ0eAzp07p9zcXKWmpv5/mIAApaamatu2bRddb86cOYqOjtb48eM9liUzM1Ph4eHOR3x8vMe2DQAA/I/PClBZWZnsdrtiYmJcxmNiYlRcXFznOlu3btXSpUuVnZ3t0SwzZsxQeXm583H48GGPbh8AAPiXIF8HcFdlZaXS09OVnZ2tqKgoj247JCREISEhHt0mAADwXz4rQFFRUQoMDFRJSYnLeElJiWJjY2stf+DAARUWFmrYsGHOMYfDIUkKCgrSnj171LFjR++GBgAAPws+OwUWHBysXr16KScnxznmcDiUk5OjPn361Fo+MTFRO3fuVH5+vvMxfPhwDRo0SPn5+czbAQAAbvPpKbCMjAyNHTtWycnJ6t27t+bPn6+qqiqNGzdOkjRmzBi1bt1amZmZCg0NVdeuXV3Wj4iIkCSX8RMnTqioqEhHjx6VJO3Zs0eSFBsbW+eRJQAAYD0+LUAjR47UsWPHNHPmTBUXF6tHjx5at26dc2J0UVGRAgLMHaR65513nAVKku6++25J0qxZs/TEE094LDsAAGi4fHodIH9l5joCAADAPzSI6wABAAD4CgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYDgUIAABYTpCvA/gjwzAkSRUVFT5OAgAA3HXhe/vC9/ilUIDqUFlZKUmKj4/3cRIAAGBWZWWlwsPDL7mMzXCnJlmMw+HQ0aNH1axZM9lsNo9uu6KiQvHx8Tp8+LCaN2/u0W3/3LCv3Me+ch/7yn3sK/exr9znzX1lGIYqKysVFxengIBLz/LhCFAdAgIC1KZNG6++R/PmzfmfxE3sK/exr9zHvnIf+8p97Cv3eWtf/diRnwuYBA0AACyHAgQAACyHAlTPQkJCNGvWLIWEhPg6it9jX7mPfeU+9pX72FfuY1+5z1/2FZOgAQCA5XAECAAAWA4FCAAAWA4FCAAAWA4FCAAAWA4FqB5lZWUpISFBoaGhSklJ0fbt230dye9kZmbquuuuU7NmzRQdHa0RI0Zoz549vo7VIDz99NOy2WyaMmWKr6P4rSNHjuiee+5RZGSkGjdurG7duumzzz7zdSy/Y7fb9fjjj6t9+/Zq3LixOnbsqD/84Q9u3V/p5+7DDz/UsGHDFBcXJ5vNpjVr1ri8bhiGZs6cqVatWqlx48ZKTU3Vvn37fBPWxy61r2pqavTYY4+pW7duCgsLU1xcnMaMGaOjR4/WWz4KUD1ZtWqVMjIyNGvWLOXl5al79+5KS0tTaWmpr6P5lc2bN2vChAn65JNPtH79etXU1Ojmm29WVVWVr6P5tR07duhvf/ubrr32Wl9H8Vvfffed+vXrp0aNGumf//ynvv76a/35z39WixYtfB3N78ydO1eLFi3SwoUL9e9//1tz587VM888o+eee87X0XyuqqpK3bt3V1ZWVp2vP/PMM1qwYIEWL16sTz/9VGFhYUpLS9PZs2frOanvXWpfnT59Wnl5eXr88ceVl5en1atXa8+ePRo+fHj9BTRQL3r37m1MmDDB+dxutxtxcXFGZmamD1P5v9LSUkOSsXnzZl9H8VuVlZXGlVdeaaxfv9648cYbjcmTJ/s6kl967LHHjBtuuMHXMRqEW2+91bjvvvtcxn7xi18Yo0eP9lEi/yTJeOutt5zPHQ6HERsba/zpT39yjp08edIICQkxXn31VR8k9B8/3Fd12b59uyHJOHToUL1k4ghQPTh37pxyc3OVmprqHAsICFBqaqq2bdvmw2T+r7y8XJLUsmVLHyfxXxMmTNCtt97q8t8XanvnnXeUnJysO++8U9HR0UpKSlJ2dravY/mlvn37KicnR3v37pUkffHFF9q6dauGDh3q42T+raCgQMXFxS7/L4aHhyslJYW/9W4oLy+XzWZTREREvbwfN0OtB2VlZbLb7YqJiXEZj4mJ0e7du32Uyv85HA5NmTJF/fr1U9euXX0dxy+tXLlSeXl52rFjh6+j+L2DBw9q0aJFysjI0O9+9zvt2LFDDz30kIKDgzV27Fhfx/Mr06dPV0VFhRITExUYGCi73a4nn3xSo0eP9nU0v1ZcXCxJdf6tv/Aa6nb27Fk99thjGjVqVL3dTJYCBL81YcIE7dq1S1u3bvV1FL90+PBhTZ48WevXr1doaKiv4/g9h8Oh5ORkPfXUU5KkpKQk7dq1S4sXL6YA/cBrr72mFStW6JVXXlGXLl2Un5+vKVOmKC4ujn0Fj6upqdFdd90lwzC0aNGientfToHVg6ioKAUGBqqkpMRlvKSkRLGxsT5K5d8mTpyotWvXauPGjWrTpo2v4/il3NxclZaWqmfPngoKClJQUJA2b96sBQsWKCgoSHa73dcR/UqrVq3UuXNnl7FrrrlGRUVFPkrkvx555BFNnz5dd999t7p166b09HRNnTpVmZmZvo7m1y78PedvvfsulJ9Dhw5p/fr19Xb0R6IA1Yvg4GD16tVLOTk5zjGHw6GcnBz16dPHh8n8j2EYmjhxot566y198MEHat++va8j+a3Bgwdr586dys/Pdz6Sk5M1evRo5efnKzAw0NcR/Uq/fv1qXVJh7969ateunY8S+a/Tp08rIMD16yEwMFAOh8NHiRqG9u3bKzY21uVvfUVFhT799FP+1tfhQvnZt2+fNmzYoMjIyHp9f06B1ZOMjAyNHTtWycnJ6t27t+bPn6+qqiqNGzfO19H8yoQJE/TKK6/o7bffVrNmzZznzcPDw9W4cWMfp/MvzZo1qzU3KiwsTJGRkcyZqsPUqVPVt29fPfXUU7rrrru0fft2LVmyREuWLPF1NL8zbNgwPfnkk2rbtq26dOmizz//XPPmzdN9993n62g+d+rUKe3fv9/5vKCgQPn5+WrZsqXatm2rKVOm6I9//KOuvPJKtW/fXo8//rji4uI0YsQI34X2kUvtq1atWumOO+5QXl6e1q5dK7vd7vx737JlSwUHB3s/YL381gyGYRjGc889Z7Rt29YIDg42evfubXzyySe+juR3JNX5WLZsma+jNQj8DP7S/vGPfxhdu3Y1QkJCjMTERGPJkiW+juSXKioqjMmTJxtt27Y1QkNDjQ4dOhj/8z//Y1RXV/s6ms9t3Lixzr9RY8eONQzj+5/CP/7440ZMTIwREhJiDB482NizZ49vQ/vIpfZVQUHBRf/eb9y4sV7y2QyDS3sCAABrYQ4QAACwHAoQAACwHAoQAACwHAoQAACwHAoQAACwHAoQAACwHAoQAACwHAoQAACwHAoQAACwHAoQAJ+69957ZbPZaj3+8x5CDcXrr7+uvn37SpI+/vhjdejQwceJAFwMN0MF4HNDhgzRsmXLXMauuOKKWsudO3eufm6S+BNt27ZN/fr1kyRt2bLF+W8A/ocjQAB8LiQkRLGxsS6PwMBADRw4UBMnTtSUKVMUFRWltLQ0SdK8efPUrVs3hYWFKT4+Xv/93/+tU6dOObf34osvKiIiQmvXrtXVV1+tJk2a6I477tDp06e1fPlyJSQkqEWLFnrooYdkt9ud61VXV2vatGlq3bq1wsLClJKSok2bNrn9OT7++GNn6dm6dSsFCPBjFCAAfm358uUKDg7WRx99pMWLF0uSAgICtGDBAn311Vdavny5PvjgAz366KMu650+fVoLFizQypUrtW7dOm3atEm333673nvvPb333nt6+eWX9be//U1vvPGGc52JEydq27ZtWrlypb788kvdeeedGjJkiPbt23fRfK+88ooiIiIUERGh7du3Kz09XREREXrvvfc0bdo0RURE6JVXXvHOzgHw09XLPecB4CLGjh1rBAYGGmFhYc7HHXfcYRiGYdx4441GUlLSj27j9ddfNyIjI53Ply1bZkgy9u/f7xx78MEHjSZNmhiVlZXOsbS0NOPBBx80DMMwDh06ZAQGBhpHjhxx2fbgwYONGTNmXPS9KysrjYKCAiM7O9vo0qWLUVBQYLz99ttGq1atjIKCAqOgoMDlPQH4B+YAAfC5QYMGadGiRc7nYWFhzn/36tWr1vIbNmxQZmamdu/erYqKCp0/f15nz57V6dOn1aRJE0lSkyZN1LFjR+c6MTExSkhIUNOmTV3GSktLJUk7d+6U3W7XVVdd5fJe1dXVioyMvGj2pk2bqmnTpsrLy9Ntt92mhIQErVixQrfccosSEhLM7QgA9YYCBMDnwsLC1KlTp4u+9p8KCwv1X//1X/rtb3+rJ598Ui1bttTWrVs1fvx4nTt3zlmAGjVq5LKezWarc8zhcEiSTp06pcDAQOXm5iowMNBluf8sTf+pqKhInTt3liSdPXtWQUFBevbZZ1VdXa2AgACtXLlS99xzj/PUHQD/QQEC0KDk5ubK4XDoz3/+swICvp/G+Nprr132dpOSkmS321VaWqr+/fu7tU5cXJzy8/NVXFys1NRU5efny263q0ePHtqyZYtatmyp5s2bX3Y2AJ7HJGgADUqnTp1UU1Oj5557TgcPHtTLL7/skSMsV111lUaPHq0xY8Zo9erVKigo0Pbt25WZmal33323znWCgoLUqVMnffPNN0pJSVFiYqKOHz+uDh06qHfv3urUqZOio6MvOxsAz6MAAWhQunfvrnnz5mnu3Lnq2rWrVqxYoczMTI9se9myZRozZowefvhhXX311RoxYoR27Nihtm3bXnK9TZs2acCAAZKkzZs3O/8NwH/ZDMMwfB0CAACgPnEECAAAWA4FCAAAWA4FCAAAWA4FCAAAWA4FCAAAWA4FCAAAWA4FCAAAWA4FCAAAWA4FCAAAWA4FCAAAWA4FCAAAWM7/AT4A56gKJrWCAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "id = tf.argmax(probabilities[-1])\n",
    "plt.plot(probabilities[:, id])\n",
    "plt.xlabel('Frame #')\n",
    "plt.ylabel(f\"p('{KINETICS_600_LABELS[id]}')\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "d7MZ_AfRW845"
   },
   "source": [
    "`base` 모델을 실행한 이전 섹션보다 최종 확률이 훨씬 더 확실하다는 것을 알 수 있습니다. `base` 모델은 전체 프레임에 대한 예측의 평균을 반환합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:56:33.202770Z",
     "iopub.status.busy": "2022-12-14T21:56:33.202131Z",
     "iopub.status.idle": "2022-12-14T21:56:33.210450Z",
     "shell.execute_reply": "2022-12-14T21:56:33.209733Z"
    },
    "id": "0Wij4tsyW8dR"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "miniature schnauzer : 0.427\n",
      "corkscrew           : 0.134\n",
      "miniature poodle    : 0.056\n",
      "church              : 0.053\n",
      "fly                 : 0.039\n"
     ]
    }
   ],
   "source": [
    "for label, p in get_top_k(tf.reduce_mean(probabilities, axis=0)):\n",
    "  print(f'{label:20s}: {p:.3f}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "qLUoC9ejggGo"
   },
   "source": [
    "## 시간 경과에 따른 예측 애니메이션\n",
    "\n",
    "이전 섹션에서는 이러한 모델을 사용하는 방법에 대해 자세히 설명했습니다. 이 섹션은 이를 바탕으로 멋진 추론 애니메이션을 생성합니다. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "OnFqOXazoWgy"
   },
   "source": [
    "아래의 숨겨진 셀은 이 섹션에서 사용되는 도우미 함수를 정의합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "cellView": "form",
    "execution": {
     "iopub.execute_input": "2022-12-14T21:56:33.214428Z",
     "iopub.status.busy": "2022-12-14T21:56:33.213970Z",
     "iopub.status.idle": "2022-12-14T21:56:33.231787Z",
     "shell.execute_reply": "2022-12-14T21:56:33.231114Z"
    },
    "id": "dx55NK3ZoZeh"
   },
   "outputs": [],
   "source": [
    "#@title\n",
    "# Get top_k labels and probabilities predicted using MoViNets streaming model\n",
    "def get_top_k_streaming_labels(probs, k=5, label_map=KINETICS_600_LABELS):\n",
    "  \"\"\"Returns the top-k labels over an entire video sequence.\n",
    "\n",
    "  Args:\n",
    "    probs: probability tensor of shape (num_frames, num_classes) that represents\n",
    "      the probability of each class on each frame.\n",
    "    k: the number of top predictions to select.\n",
    "    label_map: a list of labels to map logit indices to label strings.\n",
    "\n",
    "  Returns:\n",
    "    a tuple of the top-k probabilities, labels, and logit indices\n",
    "  \"\"\"\n",
    "  top_categories_last = tf.argsort(probs, -1, 'DESCENDING')[-1, :1]\n",
    "  # Sort predictions to find top_k\n",
    "  categories = tf.argsort(probs, -1, 'DESCENDING')[:, :k]\n",
    "  categories = tf.reshape(categories, [-1])\n",
    "\n",
    "  counts = sorted([\n",
    "      (i.numpy(), tf.reduce_sum(tf.cast(categories == i, tf.int32)).numpy())\n",
    "      for i in tf.unique(categories)[0]\n",
    "  ], key=lambda x: x[1], reverse=True)\n",
    "\n",
    "  top_probs_idx = tf.constant([i for i, _ in counts[:k]])\n",
    "  top_probs_idx = tf.concat([top_categories_last, top_probs_idx], 0)\n",
    "  # find unique indices of categories\n",
    "  top_probs_idx = tf.unique(top_probs_idx)[0][:k+1]\n",
    "  # top_k probabilities of the predictions\n",
    "  top_probs = tf.gather(probs, top_probs_idx, axis=-1)\n",
    "  top_probs = tf.transpose(top_probs, perm=(1, 0))\n",
    "  # collect the labels of top_k predictions\n",
    "  top_labels = tf.gather(label_map, top_probs_idx, axis=0)\n",
    "  # decode the top_k labels\n",
    "  top_labels = [label.decode('utf8') for label in top_labels.numpy()]\n",
    "\n",
    "  return top_probs, top_labels, top_probs_idx\n",
    "\n",
    "# Plot top_k predictions at a given time step\n",
    "def plot_streaming_top_preds_at_step(\n",
    "    top_probs,\n",
    "    top_labels,\n",
    "    step=None,\n",
    "    image=None,\n",
    "    legend_loc='lower left',\n",
    "    duration_seconds=10,\n",
    "    figure_height=500,\n",
    "    playhead_scale=0.8,\n",
    "    grid_alpha=0.3):\n",
    "  \"\"\"Generates a plot of the top video model predictions at a given time step.\n",
    "\n",
    "  Args:\n",
    "    top_probs: a tensor of shape (k, num_frames) representing the top-k\n",
    "      probabilities over all frames.\n",
    "    top_labels: a list of length k that represents the top-k label strings.\n",
    "    step: the current time step in the range [0, num_frames].\n",
    "    image: the image frame to display at the current time step.\n",
    "    legend_loc: the placement location of the legend.\n",
    "    duration_seconds: the total duration of the video.\n",
    "    figure_height: the output figure height.\n",
    "    playhead_scale: scale value for the playhead.\n",
    "    grid_alpha: alpha value for the gridlines.\n",
    "\n",
    "  Returns:\n",
    "    A tuple of the output numpy image, figure, and axes.\n",
    "  \"\"\"\n",
    "  # find number of top_k labels and frames in the video\n",
    "  num_labels, num_frames = top_probs.shape\n",
    "  if step is None:\n",
    "    step = num_frames\n",
    "  # Visualize frames and top_k probabilities of streaming video\n",
    "  fig = plt.figure(figsize=(6.5, 7), dpi=300)\n",
    "  gs = mpl.gridspec.GridSpec(8, 1)\n",
    "  ax2 = plt.subplot(gs[:-3, :])\n",
    "  ax = plt.subplot(gs[-3:, :])\n",
    "  # display the frame\n",
    "  if image is not None:\n",
    "    ax2.imshow(image, interpolation='nearest')\n",
    "    ax2.axis('off')\n",
    "  # x-axis (frame number)\n",
    "  preview_line_x = tf.linspace(0., duration_seconds, num_frames)\n",
    "  # y-axis (top_k probabilities)\n",
    "  preview_line_y = top_probs\n",
    "\n",
    "  line_x = preview_line_x[:step+1]\n",
    "  line_y = preview_line_y[:, :step+1]\n",
    "\n",
    "  for i in range(num_labels):\n",
    "    ax.plot(preview_line_x, preview_line_y[i], label=None, linewidth='1.5',\n",
    "            linestyle=':', color='gray')\n",
    "    ax.plot(line_x, line_y[i], label=top_labels[i], linewidth='2.0')\n",
    "\n",
    "\n",
    "  ax.grid(which='major', linestyle=':', linewidth='1.0', alpha=grid_alpha)\n",
    "  ax.grid(which='minor', linestyle=':', linewidth='0.5', alpha=grid_alpha)\n",
    "\n",
    "  min_height = tf.reduce_min(top_probs) * playhead_scale\n",
    "  max_height = tf.reduce_max(top_probs)\n",
    "  ax.vlines(preview_line_x[step], min_height, max_height, colors='red')\n",
    "  ax.scatter(preview_line_x[step], max_height, color='red')\n",
    "\n",
    "  ax.legend(loc=legend_loc)\n",
    "\n",
    "  plt.xlim(0, duration_seconds)\n",
    "  plt.ylabel('Probability')\n",
    "  plt.xlabel('Time (s)')\n",
    "  plt.yscale('log')\n",
    "\n",
    "  fig.tight_layout()\n",
    "  fig.canvas.draw()\n",
    "\n",
    "  data = np.frombuffer(fig.canvas.tostring_rgb(), dtype=np.uint8)\n",
    "  data = data.reshape(fig.canvas.get_width_height()[::-1] + (3,))\n",
    "  plt.close()\n",
    "\n",
    "  figure_width = int(figure_height * data.shape[1] / data.shape[0])\n",
    "  image = PIL.Image.fromarray(data).resize([figure_width, figure_height])\n",
    "  image = np.array(image)\n",
    "\n",
    "  return image\n",
    "\n",
    "# Plotting top_k predictions from MoViNets streaming model\n",
    "def plot_streaming_top_preds(\n",
    "    probs,\n",
    "    video,\n",
    "    top_k=5,\n",
    "    video_fps=25.,\n",
    "    figure_height=500,\n",
    "    use_progbar=True):\n",
    "  \"\"\"Generates a video plot of the top video model predictions.\n",
    "\n",
    "  Args:\n",
    "    probs: probability tensor of shape (num_frames, num_classes) that represents\n",
    "      the probability of each class on each frame.\n",
    "    video: the video to display in the plot.\n",
    "    top_k: the number of top predictions to select.\n",
    "    video_fps: the input video fps.\n",
    "    figure_fps: the output video fps.\n",
    "    figure_height: the height of the output video.\n",
    "    use_progbar: display a progress bar.\n",
    "\n",
    "  Returns:\n",
    "    A numpy array representing the output video.\n",
    "  \"\"\"\n",
    "  # select number of frames per second\n",
    "  video_fps = 8.\n",
    "  # select height of the image\n",
    "  figure_height = 500\n",
    "  # number of time steps of the given video\n",
    "  steps = video.shape[0]\n",
    "  # estimate duration of the video (in seconds)\n",
    "  duration = steps / video_fps\n",
    "  # estiamte top_k probabilities and corresponding labels\n",
    "  top_probs, top_labels, _ = get_top_k_streaming_labels(probs, k=top_k)\n",
    "\n",
    "  images = []\n",
    "  step_generator = tqdm.trange(steps) if use_progbar else range(steps)\n",
    "  for i in step_generator:\n",
    "    image = plot_streaming_top_preds_at_step(\n",
    "        top_probs=top_probs,\n",
    "        top_labels=top_labels,\n",
    "        step=i,\n",
    "        image=video[i],\n",
    "        duration_seconds=duration,\n",
    "        figure_height=figure_height,\n",
    "    )\n",
    "    images.append(image)\n",
    "\n",
    "  return np.array(images)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "eLgFBslcZOQO"
   },
   "source": [
    "비디오 프레임에서 스트리밍 모델을 실행하고 로짓을 수집하여 시작합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:56:33.235232Z",
     "iopub.status.busy": "2022-12-14T21:56:33.234772Z",
     "iopub.status.idle": "2022-12-14T21:56:33.491851Z",
     "shell.execute_reply": "2022-12-14T21:56:33.490928Z"
    },
    "id": "tXWR13wthnK5"
   },
   "outputs": [],
   "source": [
    "init_states = model.init_states(jumpingjack[tf.newaxis].shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:56:33.495901Z",
     "iopub.status.busy": "2022-12-14T21:56:33.495617Z",
     "iopub.status.idle": "2022-12-14T21:56:34.014083Z",
     "shell.execute_reply": "2022-12-14T21:56:34.013379Z"
    },
    "id": "YqSkt7l8ltwt"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0/13 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 23%|██▎       | 3/13 [00:00<00:00, 25.00it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 46%|████▌     | 6/13 [00:00<00:00, 25.61it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 69%|██████▉   | 9/13 [00:00<00:00, 26.31it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 92%|█████████▏| 12/13 [00:00<00:00, 25.98it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|██████████| 13/13 [00:00<00:00, 25.79it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "# Insert your video clip here\n",
    "video = jumpingjack\n",
    "images = tf.split(video[tf.newaxis], video.shape[0], axis=1)\n",
    "\n",
    "all_logits = []\n",
    "\n",
    "# To run on a video, pass in one frame at a time\n",
    "states = init_states\n",
    "for image in tqdm.tqdm(images):\n",
    "  # predictions for each frame\n",
    "  logits, states = model({**states, 'image': image})\n",
    "  all_logits.append(logits)\n",
    "\n",
    "# concatinating all the logits\n",
    "logits = tf.concat(all_logits, 0)\n",
    "# estimating probabilities\n",
    "probs = tf.nn.softmax(logits, axis=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:56:34.017872Z",
     "iopub.status.busy": "2022-12-14T21:56:34.017297Z",
     "iopub.status.idle": "2022-12-14T21:56:34.024446Z",
     "shell.execute_reply": "2022-12-14T21:56:34.023803Z"
    },
    "id": "OOGcCMMJyuPl"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top_k predictions and their probablities\n",
      "\n",
      "EntleBucher         : 0.999\n",
      "home theater        : 0.000\n",
      "redshank            : 0.000\n",
      "spoonbill           : 0.000\n",
      "candle              : 0.000\n"
     ]
    }
   ],
   "source": [
    "final_probs = probs[-1]\n",
    "print('Top_k predictions and their probablities\\n')\n",
    "for label, p in get_top_k(final_probs):\n",
    "  print(f'{label:20s}: {p:.3f}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "GaybT0rbZct-"
   },
   "source": [
    "확률 시퀀스를 비디오로 변환합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:56:34.028207Z",
     "iopub.status.busy": "2022-12-14T21:56:34.027585Z",
     "iopub.status.idle": "2022-12-14T21:56:41.999709Z",
     "shell.execute_reply": "2022-12-14T21:56:41.998869Z"
    },
    "id": "Xdox556CtMRb"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0/13 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  8%|▊         | 1/13 [00:00<00:06,  1.73it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 15%|█▌        | 2/13 [00:01<00:05,  1.92it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 23%|██▎       | 3/13 [00:01<00:05,  1.95it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 31%|███       | 4/13 [00:02<00:04,  1.94it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 38%|███▊      | 5/13 [00:02<00:04,  1.93it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 46%|████▌     | 6/13 [00:03<00:03,  1.92it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 54%|█████▍    | 7/13 [00:03<00:03,  1.90it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 62%|██████▏   | 8/13 [00:05<00:04,  1.12it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 69%|██████▉   | 9/13 [00:05<00:03,  1.28it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 77%|███████▋  | 10/13 [00:06<00:02,  1.43it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 85%|████████▍ | 11/13 [00:06<00:01,  1.56it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 92%|█████████▏| 12/13 [00:07<00:00,  1.66it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|██████████| 13/13 [00:07<00:00,  1.74it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|██████████| 13/13 [00:07<00:00,  1.64it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "# Generate a plot and output to a video tensor\n",
    "plot_video = plot_streaming_top_preds(probs, video, video_fps=8.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-14T21:56:42.003596Z",
     "iopub.status.busy": "2022-12-14T21:56:42.002980Z",
     "iopub.status.idle": "2022-12-14T21:56:42.404721Z",
     "shell.execute_reply": "2022-12-14T21:56:42.403818Z"
    },
    "id": "NSStKE9klCs3"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"show_videos\" style=\"border-spacing:0px;\"><tr><td style=\"padding:1px;\"><video controls width=\"464\" height=\"500\" style=\"object-fit:cover;\" loop autoplay>\n",
       "      <source src=\"data:video/mp4;base64,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\" type=\"video/mp4\"/>\n",
       "      This browser does not support the video tag.\n",
       "      </video></td></tr></table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# For gif format, set codec='gif'\n",
    "media.show_video(plot_video, fps=3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "LCImgZ3OdJw7"
   },
   "source": [
    "## 리소스\n",
    "\n",
    "사전 훈련된 모델은 [TF Hub](https://tfhub.dev/google/collections/movinet/1)에서 사용할 수 있습니다. TF Hub 컬렉션에는 [TFLite](https://tensorflow.org/lite)에 최적화된 양자화된 모델도 포함되어 있습니다.\n",
    "\n",
    "이러한 모델의 소스는 [TensorFlow Model Garden](https://github.com/tensorflow/models/tree/master/official/projects/movinet)에서 제공합니다. 여기에는 MoViNet 모델 구축 및 미세 조정에 대한 내용도 담긴 [이 튜토리얼의 상세 버전](https://colab.sandbox.google.com/github/tensorflow/models/blob/master/official/projects/movinet/movinet_tutorial.ipynb)이 포함됩니다. "
   ]
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "movinet.ipynb",
   "toc_visible": true
  },
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.9.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}