{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "EOgDUDMAG6mn"
   },
   "source": [
    "##### Copyright 2022 The TensorFlow Authors."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "cellView": "form",
    "execution": {
     "iopub.execute_input": "2023-11-07T17:28:45.546703Z",
     "iopub.status.busy": "2023-11-07T17:28:45.546277Z",
     "iopub.status.idle": "2023-11-07T17:28:45.550316Z",
     "shell.execute_reply": "2023-11-07T17:28:45.549694Z"
    },
    "id": "B3PsBDmGG_W8"
   },
   "outputs": [],
   "source": [
    "#@title 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",
    "# https://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."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ifkGYxdCHIof"
   },
   "source": [
    "<table class=\"tfo-notebook-buttons\" align=\"left\">\n",
    "  <td>     <a target=\"_blank\" href=\"https://tensorflow.google.cn/tutorials/video/transfer_learning_with_movinet\"><img src=\"https://tensorflow.google.cn/images/tf_logo_32px.png\">在 TensorFlow.org 上查看</a>\n",
    "</td>\n",
    "  <td>     <a target=\"_blank\" href=\"https://colab.research.google.com/github/tensorflow/docs-l10n/blob/master/site/zh-cn/tutorials/video/transfer_learning_with_movinet.ipynb\"><img src=\"https://tensorflow.google.cn/images/colab_logo_32px.png\">在 Google Colab 中运行</a>\n",
    "</td>\n",
    "  <td>     <a target=\"_blank\" href=\"https://github.com/tensorflow/docs-l10n/blob/master/site/zh-cn/tutorials/video/transfer_learning_with_movinet.ipynb\"><img src=\"https://tensorflow.google.cn/images/GitHub-Mark-32px.png\">在 Github 上查看源代码</a>\n",
    "</td>\n",
    "  <td>     <a href=\"https://storage.googleapis.com/tensorflow_docs/docs-l10n/site/zh-cn/tutorials/video/transfer_learning_with_movinet.ipynb\"><img src=\"https://tensorflow.google.cn/images/download_logo_32px.png\">下载笔记本</a>\n",
    "</td>\n",
    "</table>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "sWxDDkRwLVMC"
   },
   "source": [
    "# 使用 MoViNet 进行视频分类的迁移学习\n",
    "\n",
    "MoViNets（移动视频网络）提供了一系列高效的视频分类模型，支持对流式视频进行推断。在本教程中，您将使用预训练的 MoViNet 模型对来自 [UCF101 数据集](https://www.crcv.ucf.edu/data/UCF101.php)的视频进行分类，特别是针对动作识别任务。预训练模型是一个先前在更大数据集上训练过的已保存网络。可以在 Kondratyuk, D. 等人 2021 年撰写的 [MoViNets: Mobile Video Networks for Efficient Video Recognition](https://arxiv.org/abs/2103.11511) 论文中找到有关 MoViNets 的更多详细信息。在本教程中，您将完成以下任务：\n",
    "\n",
    "- 了解如何下载预训练的 MoViNet 模型\n",
    "- 通过冻结 MoViNet 模型的卷积基，使用带有新分类器的预训练模型创建新模型\n",
    "- 将分类器头替换为新数据集的标签数\n",
    "- 在 [UCF101 数据集](https://www.crcv.ucf.edu/data/UCF101.php)上执行迁移学习\n",
    "\n",
    "本教程下载的模型来自 [official/projects/movinet](https://github.com/tensorflow/models/tree/master/official/projects/movinet)。此仓库包含 TF Hub 在 TensorFlow 2 SavedModel 格式中使用的 MoViNet 模型集合。\n",
    "\n",
    "本视频加载和预处理教程是 TensorFlow 视频教程系列的第一部分。下面是其他三个教程：\n",
    "\n",
    "- [加载视频数据](https://tensorflow.google.cn/tutorials/load_data/video)：本教程解释了本文档中使用的大部分代码；特别是，更详细地解释了如何通过 `FrameGenerator` 类预处理和加载数据。\n",
    "- [构建用于视频分类的 3D CNN 模型](https://tensorflow.google.cn/tutorials/video/video_classification)。请注意，本教程使用分解 3D 数据的空间和时间方面的 (2+1)D CNN；如果使用 MRI 扫描等体数据，请考虑使用 3D CNN 而不是 (2+1)D CNN。\n",
    "- [用于流式动作识别的 MoViNet](https://tensorflow.google.cn/hub/tutorials/movinet)：熟悉 TF Hub 上提供的 MoViNet 模型。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "GidiisyXwK--"
   },
   "source": [
    "## 安装\n",
    "\n",
    "首先，安装并导入一些必要的库，包括：用于检查 ZIP 文件内容的 [remotezip](https://github.com/gtsystem/python-remotezip)，用于使用进度条的 [tqdm](https://github.com/tqdm/tqdm)，用于处理视频文件的 [OpenCV](https://opencv.org/)（确保 `opencv-python` 和 `opencv-python-headless` 是同一版本），以及用于下载预训练 MoViNet 模型的 TensorFlow 模型 ([`tf-models- official`](https://github.com/tensorflow/models/tree/master/official))。TensorFlow 模型软件包是一组使用 TensorFlow 高级 API 的模型。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-11-07T17:28:45.553919Z",
     "iopub.status.busy": "2023-11-07T17:28:45.553413Z",
     "iopub.status.idle": "2023-11-07T17:29:24.815353Z",
     "shell.execute_reply": "2023-11-07T17:29:24.814315Z"
    },
    "id": "nubWhqYdwEXD"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting remotezip\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Downloading remotezip-0.12.1.tar.gz (7.5 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Preparing metadata (setup.py) ... \u001b[?25l-"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b \bdone\r\n",
      "\u001b[?25hRequirement already satisfied: tqdm in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (4.66.1)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting opencv-python==4.5.2.52\r\n",
      "  Downloading opencv_python-4.5.2.52-cp39-cp39-manylinux2014_x86_64.whl (51.0 MB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting opencv-python-headless==4.5.2.52\r\n",
      "  Downloading opencv_python_headless-4.5.2.52-cp39-cp39-manylinux2014_x86_64.whl (38.2 MB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting tf-models-official\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Downloading tf_models_official-2.14.2-py2.py3-none-any.whl.metadata (1.4 kB)\r\n",
      "Requirement already satisfied: numpy>=1.19.3 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from opencv-python==4.5.2.52) (1.26.1)\r\n",
      "Requirement already satisfied: requests in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from remotezip) (2.31.0)\r\n",
      "Collecting tabulate (from remotezip)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Downloading tabulate-0.9.0-py3-none-any.whl (35 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting Cython (from tf-models-official)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Downloading Cython-3.0.5-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (3.2 kB)\r\n",
      "Requirement already satisfied: Pillow in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tf-models-official) (10.0.1)\r\n",
      "Collecting gin-config (from tf-models-official)\r\n",
      "  Downloading gin_config-0.5.0-py3-none-any.whl (61 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting google-api-python-client>=1.6.7 (from tf-models-official)\r\n",
      "  Downloading google_api_python_client-2.107.0-py2.py3-none-any.whl.metadata (6.6 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting immutabledict (from tf-models-official)\r\n",
      "  Downloading immutabledict-3.0.0-py3-none-any.whl.metadata (3.1 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting kaggle>=1.3.9 (from tf-models-official)\r\n",
      "  Downloading kaggle-1.5.16.tar.gz (83 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Preparing metadata (setup.py) ... \u001b[?25l-"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b \bdone\r\n",
      "\u001b[?25hRequirement already satisfied: matplotlib in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tf-models-official) (3.8.1)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting oauth2client (from tf-models-official)\r\n",
      "  Downloading oauth2client-4.1.3-py2.py3-none-any.whl (98 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: pandas>=0.22.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tf-models-official) (2.1.2)\r\n",
      "Requirement already satisfied: psutil>=5.4.3 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tf-models-official) (5.9.6)\r\n",
      "Collecting py-cpuinfo>=3.3.0 (from tf-models-official)\r\n",
      "  Downloading py_cpuinfo-9.0.0-py3-none-any.whl (22 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting pycocotools (from tf-models-official)\r\n",
      "  Downloading pycocotools-2.0.7-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (1.1 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: pyyaml>=6.0.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tf-models-official) (6.0.1)\r\n",
      "Collecting sacrebleu (from tf-models-official)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Downloading sacrebleu-2.3.2-py3-none-any.whl.metadata (57 kB)\r\n",
      "Requirement already satisfied: scipy>=0.19.1 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tf-models-official) (1.11.3)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting sentencepiece (from tf-models-official)\r\n",
      "  Downloading sentencepiece-0.1.99-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (1.3 MB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting seqeval (from tf-models-official)\r\n",
      "  Downloading seqeval-1.2.2.tar.gz (43 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Preparing metadata (setup.py) ... \u001b[?25l-"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b \b\\"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b \bdone\r\n",
      "\u001b[?25hRequirement already satisfied: six in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tf-models-official) (1.16.0)\r\n",
      "Requirement already satisfied: tensorflow-datasets in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tf-models-official) (4.9.3)\r\n",
      "Requirement already satisfied: tensorflow-hub>=0.6.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tf-models-official) (0.15.0)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting tensorflow-model-optimization>=0.4.1 (from tf-models-official)\r\n",
      "  Downloading tensorflow_model_optimization-0.7.5-py2.py3-none-any.whl.metadata (914 bytes)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting tensorflow-text~=2.14.0 (from tf-models-official)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Downloading tensorflow_text-2.14.0-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (1.9 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting tensorflow~=2.14.0 (from tf-models-official)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Downloading tensorflow-2.14.0-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (4.1 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting tf-slim>=1.1.0 (from tf-models-official)\r\n",
      "  Downloading tf_slim-1.1.0-py2.py3-none-any.whl (352 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting httplib2<1.dev0,>=0.15.0 (from google-api-python-client>=1.6.7->tf-models-official)\r\n",
      "  Downloading httplib2-0.22.0-py3-none-any.whl (96 kB)\r\n",
      "Requirement already satisfied: google-auth<3.0.0.dev0,>=1.19.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from google-api-python-client>=1.6.7->tf-models-official) (2.23.4)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting google-auth-httplib2>=0.1.0 (from google-api-python-client>=1.6.7->tf-models-official)\r\n",
      "  Downloading google_auth_httplib2-0.1.1-py2.py3-none-any.whl.metadata (2.1 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting google-api-core!=2.0.*,!=2.1.*,!=2.2.*,!=2.3.0,<3.0.0.dev0,>=1.31.5 (from google-api-python-client>=1.6.7->tf-models-official)\r\n",
      "  Downloading google_api_core-2.12.0-py3-none-any.whl.metadata (2.7 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting uritemplate<5,>=3.0.1 (from google-api-python-client>=1.6.7->tf-models-official)\r\n",
      "  Downloading uritemplate-4.1.1-py2.py3-none-any.whl (10 kB)\r\n",
      "Requirement already satisfied: certifi in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from kaggle>=1.3.9->tf-models-official) (2023.7.22)\r\n",
      "Requirement already satisfied: python-dateutil in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from kaggle>=1.3.9->tf-models-official) (2.8.2)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting python-slugify (from kaggle>=1.3.9->tf-models-official)\r\n",
      "  Downloading python_slugify-8.0.1-py2.py3-none-any.whl (9.7 kB)\r\n",
      "Requirement already satisfied: urllib3 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from kaggle>=1.3.9->tf-models-official) (2.0.7)\r\n",
      "Requirement already satisfied: bleach in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from kaggle>=1.3.9->tf-models-official) (6.1.0)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: pytz>=2020.1 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from pandas>=0.22.0->tf-models-official) (2023.3.post1)\r\n",
      "Requirement already satisfied: tzdata>=2022.1 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from pandas>=0.22.0->tf-models-official) (2023.3)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: absl-py>=1.0.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow~=2.14.0->tf-models-official) (1.4.0)\r\n",
      "Requirement already satisfied: astunparse>=1.6.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow~=2.14.0->tf-models-official) (1.6.3)\r\n",
      "Requirement already satisfied: flatbuffers>=23.5.26 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow~=2.14.0->tf-models-official) (23.5.26)\r\n",
      "Requirement already satisfied: gast!=0.5.0,!=0.5.1,!=0.5.2,>=0.2.1 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow~=2.14.0->tf-models-official) (0.5.4)\r\n",
      "Requirement already satisfied: google-pasta>=0.1.1 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow~=2.14.0->tf-models-official) (0.2.0)\r\n",
      "Requirement already satisfied: h5py>=2.9.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow~=2.14.0->tf-models-official) (3.10.0)\r\n",
      "Requirement already satisfied: libclang>=13.0.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow~=2.14.0->tf-models-official) (16.0.6)\r\n",
      "Requirement already satisfied: ml-dtypes==0.2.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow~=2.14.0->tf-models-official) (0.2.0)\r\n",
      "Requirement already satisfied: opt-einsum>=2.3.2 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow~=2.14.0->tf-models-official) (3.3.0)\r\n",
      "Requirement already satisfied: packaging in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow~=2.14.0->tf-models-official) (23.2)\r\n",
      "Requirement already satisfied: protobuf!=4.21.0,!=4.21.1,!=4.21.2,!=4.21.3,!=4.21.4,!=4.21.5,<5.0.0dev,>=3.20.3 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow~=2.14.0->tf-models-official) (3.20.3)\r\n",
      "Requirement already satisfied: setuptools in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow~=2.14.0->tf-models-official) (68.2.2)\r\n",
      "Requirement already satisfied: termcolor>=1.1.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow~=2.14.0->tf-models-official) (2.3.0)\r\n",
      "Requirement already satisfied: typing-extensions>=3.6.6 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow~=2.14.0->tf-models-official) (4.8.0)\r\n",
      "Requirement already satisfied: wrapt<1.15,>=1.11.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow~=2.14.0->tf-models-official) (1.14.1)\r\n",
      "Requirement already satisfied: tensorflow-io-gcs-filesystem>=0.23.1 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow~=2.14.0->tf-models-official) (0.34.0)\r\n",
      "Requirement already satisfied: grpcio<2.0,>=1.24.3 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow~=2.14.0->tf-models-official) (1.59.2)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting tensorboard<2.15,>=2.14 (from tensorflow~=2.14.0->tf-models-official)\r\n",
      "  Downloading tensorboard-2.14.1-py3-none-any.whl.metadata (1.7 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting tensorflow-estimator<2.15,>=2.14.0 (from tensorflow~=2.14.0->tf-models-official)\r\n",
      "  Downloading tensorflow_estimator-2.14.0-py2.py3-none-any.whl.metadata (1.3 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting keras<2.15,>=2.14.0 (from tensorflow~=2.14.0->tf-models-official)\r\n",
      "  Downloading keras-2.14.0-py3-none-any.whl.metadata (2.4 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: dm-tree~=0.1.1 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow-model-optimization>=0.4.1->tf-models-official) (0.1.8)\r\n",
      "Requirement already satisfied: contourpy>=1.0.1 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from matplotlib->tf-models-official) (1.2.0)\r\n",
      "Requirement already satisfied: cycler>=0.10 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from matplotlib->tf-models-official) (0.12.1)\r\n",
      "Requirement already satisfied: fonttools>=4.22.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from matplotlib->tf-models-official) (4.44.0)\r\n",
      "Requirement already satisfied: kiwisolver>=1.3.1 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from matplotlib->tf-models-official) (1.4.5)\r\n",
      "Requirement already satisfied: pyparsing>=2.3.1 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from matplotlib->tf-models-official) (3.1.1)\r\n",
      "Requirement already satisfied: importlib-resources>=3.2.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from matplotlib->tf-models-official) (6.1.1)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: pyasn1>=0.1.7 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from oauth2client->tf-models-official) (0.5.0)\r\n",
      "Requirement already satisfied: pyasn1-modules>=0.0.5 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from oauth2client->tf-models-official) (0.3.0)\r\n",
      "Requirement already satisfied: rsa>=3.1.4 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from oauth2client->tf-models-official) (4.9)\r\n",
      "Requirement already satisfied: charset-normalizer<4,>=2 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from requests->remotezip) (3.3.2)\r\n",
      "Requirement already satisfied: idna<4,>=2.5 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from requests->remotezip) (3.4)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting portalocker (from sacrebleu->tf-models-official)\r\n",
      "  Downloading portalocker-2.8.2-py3-none-any.whl.metadata (8.5 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting regex (from sacrebleu->tf-models-official)\r\n",
      "  Downloading regex-2023.10.3-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (40 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting colorama (from sacrebleu->tf-models-official)\r\n",
      "  Downloading colorama-0.4.6-py2.py3-none-any.whl (25 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting lxml (from sacrebleu->tf-models-official)\r\n",
      "  Downloading lxml-4.9.3-cp39-cp39-manylinux_2_28_x86_64.whl.metadata (3.8 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: scikit-learn>=0.21.3 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from seqeval->tf-models-official) (1.3.2)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: array-record in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow-datasets->tf-models-official) (0.5.0)\r\n",
      "Requirement already satisfied: click in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow-datasets->tf-models-official) (8.1.7)\r\n",
      "Requirement already satisfied: etils>=0.9.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from etils[enp,epath,etree]>=0.9.0->tensorflow-datasets->tf-models-official) (1.5.2)\r\n",
      "Requirement already satisfied: promise in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow-datasets->tf-models-official) (2.3)\r\n",
      "Requirement already satisfied: tensorflow-metadata in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow-datasets->tf-models-official) (1.14.0)\r\n",
      "Requirement already satisfied: toml in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorflow-datasets->tf-models-official) (0.10.2)\r\n",
      "Requirement already satisfied: wheel<1.0,>=0.23.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from astunparse>=1.6.0->tensorflow~=2.14.0->tf-models-official) (0.41.2)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: fsspec in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from etils[enp,epath,etree]>=0.9.0->tensorflow-datasets->tf-models-official) (2023.10.0)\r\n",
      "Requirement already satisfied: zipp in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from etils[enp,epath,etree]>=0.9.0->tensorflow-datasets->tf-models-official) (3.17.0)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: googleapis-common-protos<2.0.dev0,>=1.56.2 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from google-api-core!=2.0.*,!=2.1.*,!=2.2.*,!=2.3.0,<3.0.0.dev0,>=1.31.5->google-api-python-client>=1.6.7->tf-models-official) (1.61.0)\r\n",
      "Requirement already satisfied: cachetools<6.0,>=2.0.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from google-auth<3.0.0.dev0,>=1.19.0->google-api-python-client>=1.6.7->tf-models-official) (5.3.2)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: joblib>=1.1.1 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from scikit-learn>=0.21.3->seqeval->tf-models-official) (1.3.2)\r\n",
      "Requirement already satisfied: threadpoolctl>=2.0.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from scikit-learn>=0.21.3->seqeval->tf-models-official) (3.2.0)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting google-auth-oauthlib<1.1,>=0.5 (from tensorboard<2.15,>=2.14->tensorflow~=2.14.0->tf-models-official)\r\n",
      "  Downloading google_auth_oauthlib-1.0.0-py2.py3-none-any.whl (18 kB)\r\n",
      "Requirement already satisfied: markdown>=2.6.8 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorboard<2.15,>=2.14->tensorflow~=2.14.0->tf-models-official) (3.5.1)\r\n",
      "Requirement already satisfied: tensorboard-data-server<0.8.0,>=0.7.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorboard<2.15,>=2.14->tensorflow~=2.14.0->tf-models-official) (0.7.2)\r\n",
      "Requirement already satisfied: werkzeug>=1.0.1 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from tensorboard<2.15,>=2.14->tensorflow~=2.14.0->tf-models-official) (3.0.1)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: webencodings in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from bleach->kaggle>=1.3.9->tf-models-official) (0.5.1)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting text-unidecode>=1.3 (from python-slugify->kaggle>=1.3.9->tf-models-official)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Downloading text_unidecode-1.3-py2.py3-none-any.whl (78 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: requests-oauthlib>=0.7.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from google-auth-oauthlib<1.1,>=0.5->tensorboard<2.15,>=2.14->tensorflow~=2.14.0->tf-models-official) (1.3.1)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: importlib-metadata>=4.4 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from markdown>=2.6.8->tensorboard<2.15,>=2.14->tensorflow~=2.14.0->tf-models-official) (6.8.0)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: MarkupSafe>=2.1.1 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from werkzeug>=1.0.1->tensorboard<2.15,>=2.14->tensorflow~=2.14.0->tf-models-official) (2.1.3)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: oauthlib>=3.0.0 in /tmpfs/src/tf_docs_env/lib/python3.9/site-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<1.1,>=0.5->tensorboard<2.15,>=2.14->tensorflow~=2.14.0->tf-models-official) (3.2.2)\r\n",
      "Downloading tf_models_official-2.14.2-py2.py3-none-any.whl (2.7 MB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading google_api_python_client-2.107.0-py2.py3-none-any.whl (12.7 MB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading tensorflow-2.14.0-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (489.8 MB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading tensorflow_model_optimization-0.7.5-py2.py3-none-any.whl (241 kB)\r\n",
      "Downloading tensorflow_text-2.14.0-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (6.5 MB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading Cython-3.0.5-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (3.6 MB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading immutabledict-3.0.0-py3-none-any.whl (4.0 kB)\r\n",
      "Downloading pycocotools-2.0.7-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (435 kB)\r\n",
      "Downloading sacrebleu-2.3.2-py3-none-any.whl (119 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading google_api_core-2.12.0-py3-none-any.whl (121 kB)\r\n",
      "Downloading google_auth_httplib2-0.1.1-py2.py3-none-any.whl (9.3 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading keras-2.14.0-py3-none-any.whl (1.7 MB)\r\n",
      "Downloading tensorboard-2.14.1-py3-none-any.whl (5.5 MB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading tensorflow_estimator-2.14.0-py2.py3-none-any.whl (440 kB)\r\n",
      "Downloading lxml-4.9.3-cp39-cp39-manylinux_2_28_x86_64.whl (8.0 MB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading portalocker-2.8.2-py3-none-any.whl (17 kB)\r\n",
      "Downloading regex-2023.10.3-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (773 kB)\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Building wheels for collected packages: remotezip, kaggle, seqeval\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Building wheel for remotezip (setup.py) ... \u001b[?25l-"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b \bdone\r\n",
      "\u001b[?25h  Created wheel for remotezip: filename=remotezip-0.12.1-py3-none-any.whl size=7934 sha256=71040c701024ca50e6820b29587d9487c50f07b5dabe4c335f9759b36f1543a0\r\n",
      "  Stored in directory: /home/kbuilder/.cache/pip/wheels/60/74/6c/b12b4c8fb4b7ab08f495ce17e88f1e98835268af7a8ad5588f\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Building wheel for kaggle (setup.py) ... \u001b[?25l-"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b \b\\"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b \bdone\r\n",
      "\u001b[?25h  Created wheel for kaggle: filename=kaggle-1.5.16-py3-none-any.whl size=110683 sha256=6d41fc76e591c7d0525ee18ae40a68e35671c21f903befa4f8fd0d0887ff0ca3\r\n",
      "  Stored in directory: /home/kbuilder/.cache/pip/wheels/d2/ed/a5/da3a0cfb13373d1ace41cafa4f2467d858c55c52473ba72799\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Building wheel for seqeval (setup.py) ... \u001b[?25l-"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b \b\\"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\b \bdone\r\n",
      "\u001b[?25h  Created wheel for seqeval: filename=seqeval-1.2.2-py3-none-any.whl size=16162 sha256=6a82fd451c05eee48f90f2958e8725be600286a1c36757efe66d90b8f57d6615\r\n",
      "  Stored in directory: /home/kbuilder/.cache/pip/wheels/e2/a5/92/2c80d1928733611c2747a9820e1324a6835524d9411510c142\r\n",
      "Successfully built remotezip kaggle seqeval\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Installing collected packages: text-unidecode, sentencepiece, py-cpuinfo, gin-config, uritemplate, tf-slim, tensorflow-model-optimization, tensorflow-estimator, tabulate, regex, python-slugify, portalocker, opencv-python-headless, opencv-python, lxml, keras, immutabledict, httplib2, Cython, colorama, sacrebleu, remotezip, oauth2client, kaggle, seqeval, pycocotools, google-auth-oauthlib, google-auth-httplib2, google-api-core, tensorboard, google-api-python-client, tensorflow, tensorflow-text, tf-models-official\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Attempting uninstall: tensorflow-estimator\r\n",
      "    Found existing installation: tensorflow-estimator 2.15.0\r\n",
      "    Uninstalling tensorflow-estimator-2.15.0:\r\n",
      "      Successfully uninstalled tensorflow-estimator-2.15.0\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Attempting uninstall: keras\r\n",
      "    Found existing installation: keras 2.15.0\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    Uninstalling keras-2.15.0:\r\n",
      "      Successfully uninstalled keras-2.15.0\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Attempting uninstall: google-auth-oauthlib\r\n",
      "    Found existing installation: google-auth-oauthlib 1.1.0\r\n",
      "    Uninstalling google-auth-oauthlib-1.1.0:\r\n",
      "      Successfully uninstalled google-auth-oauthlib-1.1.0\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Attempting uninstall: tensorboard\r\n",
      "    Found existing installation: tensorboard 2.15.1\r\n",
      "    Uninstalling tensorboard-2.15.1:\r\n",
      "      Successfully uninstalled tensorboard-2.15.1\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Attempting uninstall: tensorflow\r\n",
      "    Found existing installation: tensorflow 2.15.0rc1\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    Uninstalling tensorflow-2.15.0rc1:\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "      Successfully uninstalled tensorflow-2.15.0rc1\r\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Successfully installed Cython-3.0.5 colorama-0.4.6 gin-config-0.5.0 google-api-core-2.12.0 google-api-python-client-2.107.0 google-auth-httplib2-0.1.1 google-auth-oauthlib-1.0.0 httplib2-0.22.0 immutabledict-3.0.0 kaggle-1.5.16 keras-2.14.0 lxml-4.9.3 oauth2client-4.1.3 opencv-python-4.5.2.52 opencv-python-headless-4.5.2.52 portalocker-2.8.2 py-cpuinfo-9.0.0 pycocotools-2.0.7 python-slugify-8.0.1 regex-2023.10.3 remotezip-0.12.1 sacrebleu-2.3.2 sentencepiece-0.1.99 seqeval-1.2.2 tabulate-0.9.0 tensorboard-2.14.1 tensorflow-2.14.0 tensorflow-estimator-2.14.0 tensorflow-model-optimization-0.7.5 tensorflow-text-2.14.0 text-unidecode-1.3 tf-models-official-2.14.2 tf-slim-1.1.0 uritemplate-4.1.1\r\n"
     ]
    }
   ],
   "source": [
    "!pip install remotezip tqdm opencv-python==4.5.2.52 opencv-python-headless==4.5.2.52 tf-models-official"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-11-07T17:29:24.819696Z",
     "iopub.status.busy": "2023-11-07T17:29:24.819421Z",
     "iopub.status.idle": "2023-11-07T17:29:28.936954Z",
     "shell.execute_reply": "2023-11-07T17:29:28.936207Z"
    },
    "id": "QImPsudoK9JI"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-11-07 17:29:26.513969: E tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:9342] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered\n",
      "2023-11-07 17:29:26.514019: E tensorflow/compiler/xla/stream_executor/cuda/cuda_fft.cc:609] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered\n",
      "2023-11-07 17:29:26.514050: E tensorflow/compiler/xla/stream_executor/cuda/cuda_blas.cc:1518] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered\n"
     ]
    }
   ],
   "source": [
    "import tqdm\n",
    "import random\n",
    "import pathlib\n",
    "import itertools\n",
    "import collections\n",
    "\n",
    "import cv2\n",
    "import numpy as np\n",
    "import remotezip as rz\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import keras\n",
    "import tensorflow as tf\n",
    "import tensorflow_hub as hub\n",
    "from tensorflow.keras import layers\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "from tensorflow.keras.losses import SparseCategoricalCrossentropy\n",
    "\n",
    "# Import the MoViNet model from TensorFlow Models (tf-models-official) for the MoViNet model\n",
    "from official.projects.movinet.modeling import movinet\n",
    "from official.projects.movinet.modeling import movinet_model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2w3H4dfOPfnm"
   },
   "source": [
    "## 加载数据\n",
    "\n",
    "下面的隐藏单元定义了从 UCF-101 数据集下载数据切片并将其加载到 `tf.data.Dataset` 中的函数。[加载视频数据教程](https://tensorflow.google.cn/tutorials/load_data/video)详细地介绍了此代码。\n",
    "\n",
    "隐藏块末尾的 `FrameGenerator` 类是这里最重要的实用工具。它会创建一个可以将数据馈送到 TensorFlow 数据流水线中的可迭代对象。具体来说，此类包含一个可加载视频帧及其编码标签的 Python 生成器。生成器 (`__call__`) 函数可产生由 `frames_from_video_file` 生成的帧数组以及与帧集关联的标签的独热编码向量。\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "cellView": "form",
    "execution": {
     "iopub.execute_input": "2023-11-07T17:29:28.941750Z",
     "iopub.status.busy": "2023-11-07T17:29:28.941348Z",
     "iopub.status.idle": "2023-11-07T17:29:28.959725Z",
     "shell.execute_reply": "2023-11-07T17:29:28.959155Z"
    },
    "id": "fwEhJ13_PSy6"
   },
   "outputs": [],
   "source": [
    "#@title \n",
    "\n",
    "def list_files_per_class(zip_url):\n",
    "  \"\"\"\n",
    "    List the files in each class of the dataset given the zip URL.\n",
    "\n",
    "    Args:\n",
    "      zip_url: URL from which the files can be unzipped. \n",
    "\n",
    "    Return:\n",
    "      files: List of files in each of the classes.\n",
    "  \"\"\"\n",
    "  files = []\n",
    "  with rz.RemoteZip(URL) as zip:\n",
    "    for zip_info in zip.infolist():\n",
    "      files.append(zip_info.filename)\n",
    "  return files\n",
    "\n",
    "def get_class(fname):\n",
    "  \"\"\"\n",
    "    Retrieve the name of the class given a filename.\n",
    "\n",
    "    Args:\n",
    "      fname: Name of the file in the UCF101 dataset.\n",
    "\n",
    "    Return:\n",
    "      Class that the file belongs to.\n",
    "  \"\"\"\n",
    "  return fname.split('_')[-3]\n",
    "\n",
    "def get_files_per_class(files):\n",
    "  \"\"\"\n",
    "    Retrieve the files that belong to each class. \n",
    "\n",
    "    Args:\n",
    "      files: List of files in the dataset.\n",
    "\n",
    "    Return:\n",
    "      Dictionary of class names (key) and files (values).\n",
    "  \"\"\"\n",
    "  files_for_class = collections.defaultdict(list)\n",
    "  for fname in files:\n",
    "    class_name = get_class(fname)\n",
    "    files_for_class[class_name].append(fname)\n",
    "  return files_for_class\n",
    "\n",
    "def download_from_zip(zip_url, to_dir, file_names):\n",
    "  \"\"\"\n",
    "    Download the contents of the zip file from the zip URL.\n",
    "\n",
    "    Args:\n",
    "      zip_url: Zip URL containing data.\n",
    "      to_dir: Directory to download data to.\n",
    "      file_names: Names of files to download.\n",
    "  \"\"\"\n",
    "  with rz.RemoteZip(zip_url) as zip:\n",
    "    for fn in tqdm.tqdm(file_names):\n",
    "      class_name = get_class(fn)\n",
    "      zip.extract(fn, str(to_dir / class_name))\n",
    "      unzipped_file = to_dir / class_name / fn\n",
    "\n",
    "      fn = pathlib.Path(fn).parts[-1]\n",
    "      output_file = to_dir / class_name / fn\n",
    "      unzipped_file.rename(output_file,)\n",
    "\n",
    "def split_class_lists(files_for_class, count):\n",
    "  \"\"\"\n",
    "    Returns the list of files belonging to a subset of data as well as the remainder of\n",
    "    files that need to be downloaded.\n",
    "\n",
    "    Args:\n",
    "      files_for_class: Files belonging to a particular class of data.\n",
    "      count: Number of files to download.\n",
    "\n",
    "    Return:\n",
    "      split_files: Files belonging to the subset of data.\n",
    "      remainder: Dictionary of the remainder of files that need to be downloaded.\n",
    "  \"\"\"\n",
    "  split_files = []\n",
    "  remainder = {}\n",
    "  for cls in files_for_class:\n",
    "    split_files.extend(files_for_class[cls][:count])\n",
    "    remainder[cls] = files_for_class[cls][count:]\n",
    "  return split_files, remainder\n",
    "\n",
    "def download_ufc_101_subset(zip_url, num_classes, splits, download_dir):\n",
    "  \"\"\"\n",
    "    Download a subset of the UFC101 dataset and split them into various parts, such as\n",
    "    training, validation, and test. \n",
    "\n",
    "    Args:\n",
    "      zip_url: Zip URL containing data.\n",
    "      num_classes: Number of labels.\n",
    "      splits: Dictionary specifying the training, validation, test, etc. (key) division of data \n",
    "              (value is number of files per split).\n",
    "      download_dir: Directory to download data to.\n",
    "\n",
    "    Return:\n",
    "      dir: Posix path of the resulting directories containing the splits of data.\n",
    "  \"\"\"\n",
    "  files = list_files_per_class(zip_url)\n",
    "  for f in files:\n",
    "    tokens = f.split('/')\n",
    "    if len(tokens) <= 2:\n",
    "      files.remove(f) # Remove that item from the list if it does not have a filename\n",
    "\n",
    "  files_for_class = get_files_per_class(files)\n",
    "\n",
    "  classes = list(files_for_class.keys())[:num_classes]\n",
    "\n",
    "  for cls in classes:\n",
    "    new_files_for_class = files_for_class[cls]\n",
    "    random.shuffle(new_files_for_class)\n",
    "    files_for_class[cls] = new_files_for_class\n",
    "\n",
    "  # Only use the number of classes you want in the dictionary\n",
    "  files_for_class = {x: files_for_class[x] for x in list(files_for_class)[:num_classes]}\n",
    "\n",
    "  dirs = {}\n",
    "  for split_name, split_count in splits.items():\n",
    "    print(split_name, \":\")\n",
    "    split_dir = download_dir / split_name\n",
    "    split_files, files_for_class = split_class_lists(files_for_class, split_count)\n",
    "    download_from_zip(zip_url, split_dir, split_files)\n",
    "    dirs[split_name] = split_dir\n",
    "\n",
    "  return dirs\n",
    "\n",
    "def format_frames(frame, output_size):\n",
    "  \"\"\"\n",
    "    Pad and resize an image from a video.\n",
    "\n",
    "    Args:\n",
    "      frame: Image that needs to resized and padded. \n",
    "      output_size: Pixel size of the output frame image.\n",
    "\n",
    "    Return:\n",
    "      Formatted frame with padding of specified output size.\n",
    "  \"\"\"\n",
    "  frame = tf.image.convert_image_dtype(frame, tf.float32)\n",
    "  frame = tf.image.resize_with_pad(frame, *output_size)\n",
    "  return frame\n",
    "\n",
    "def frames_from_video_file(video_path, n_frames, output_size = (224,224), frame_step = 15):\n",
    "  \"\"\"\n",
    "    Creates frames from each video file present for each category.\n",
    "\n",
    "    Args:\n",
    "      video_path: File path to the video.\n",
    "      n_frames: Number of frames to be created per video file.\n",
    "      output_size: Pixel size of the output frame image.\n",
    "\n",
    "    Return:\n",
    "      An NumPy array of frames in the shape of (n_frames, height, width, channels).\n",
    "  \"\"\"\n",
    "  # Read each video frame by frame\n",
    "  result = []\n",
    "  src = cv2.VideoCapture(str(video_path))  \n",
    "\n",
    "  video_length = src.get(cv2.CAP_PROP_FRAME_COUNT)\n",
    "\n",
    "  need_length = 1 + (n_frames - 1) * frame_step\n",
    "\n",
    "  if need_length > video_length:\n",
    "    start = 0\n",
    "  else:\n",
    "    max_start = video_length - need_length\n",
    "    start = random.randint(0, max_start + 1)\n",
    "\n",
    "  src.set(cv2.CAP_PROP_POS_FRAMES, start)\n",
    "  # ret is a boolean indicating whether read was successful, frame is the image itself\n",
    "  ret, frame = src.read()\n",
    "  result.append(format_frames(frame, output_size))\n",
    "\n",
    "  for _ in range(n_frames - 1):\n",
    "    for _ in range(frame_step):\n",
    "      ret, frame = src.read()\n",
    "    if ret:\n",
    "      frame = format_frames(frame, output_size)\n",
    "      result.append(frame)\n",
    "    else:\n",
    "      result.append(np.zeros_like(result[0]))\n",
    "  src.release()\n",
    "  result = np.array(result)[..., [2, 1, 0]]\n",
    "\n",
    "  return result\n",
    "\n",
    "class FrameGenerator:\n",
    "  def __init__(self, path, n_frames, training = False):\n",
    "    \"\"\" Returns a set of frames with their associated label. \n",
    "\n",
    "      Args:\n",
    "        path: Video file paths.\n",
    "        n_frames: Number of frames. \n",
    "        training: Boolean to determine if training dataset is being created.\n",
    "    \"\"\"\n",
    "    self.path = path\n",
    "    self.n_frames = n_frames\n",
    "    self.training = training\n",
    "    self.class_names = sorted(set(p.name for p in self.path.iterdir() if p.is_dir()))\n",
    "    self.class_ids_for_name = dict((name, idx) for idx, name in enumerate(self.class_names))\n",
    "\n",
    "  def get_files_and_class_names(self):\n",
    "    video_paths = list(self.path.glob('*/*.avi'))\n",
    "    classes = [p.parent.name for p in video_paths] \n",
    "    return video_paths, classes\n",
    "\n",
    "  def __call__(self):\n",
    "    video_paths, classes = self.get_files_and_class_names()\n",
    "\n",
    "    pairs = list(zip(video_paths, classes))\n",
    "\n",
    "    if self.training:\n",
    "      random.shuffle(pairs)\n",
    "\n",
    "    for path, name in pairs:\n",
    "      video_frames = frames_from_video_file(path, self.n_frames) \n",
    "      label = self.class_ids_for_name[name] # Encode labels\n",
    "      yield video_frames, label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-11-07T17:29:28.963072Z",
     "iopub.status.busy": "2023-11-07T17:29:28.962497Z",
     "iopub.status.idle": "2023-11-07T17:30:22.525001Z",
     "shell.execute_reply": "2023-11-07T17:30:22.524225Z"
    },
    "id": "vDHrNLZkPSR9"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train :\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0/300 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 1/300 [00:00<00:46,  6.42it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  1%|          | 2/300 [00:00<00:45,  6.56it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  1%|          | 3/300 [00:00<00:41,  7.11it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  1%|▏         | 4/300 [00:00<00:39,  7.59it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  2%|▏         | 5/300 [00:00<00:40,  7.32it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  2%|▏         | 7/300 [00:00<00:34,  8.57it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  3%|▎         | 8/300 [00:01<00:35,  8.19it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  3%|▎         | 9/300 [00:01<00:36,  7.87it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  3%|▎         | 10/300 [00:01<00:41,  6.93it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  4%|▍         | 12/300 [00:01<00:35,  8.23it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  5%|▍         | 14/300 [00:01<00:31,  9.18it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  5%|▌         | 16/300 [00:01<00:30,  9.46it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  6%|▌         | 17/300 [00:02<00:31,  8.88it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  6%|▋         | 19/300 [00:02<00:27, 10.07it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  7%|▋         | 21/300 [00:02<00:27, 10.32it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  8%|▊         | 23/300 [00:02<00:26, 10.32it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  8%|▊         | 25/300 [00:02<00:25, 10.63it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  9%|▉         | 27/300 [00:02<00:27,  9.96it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 10%|▉         | 29/300 [00:03<00:27,  9.74it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 10%|█         | 31/300 [00:03<00:25, 10.54it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 11%|█         | 33/300 [00:03<00:23, 11.24it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 12%|█▏        | 35/300 [00:03<00:22, 11.57it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 12%|█▏        | 37/300 [00:03<00:21, 12.33it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 13%|█▎        | 39/300 [00:03<00:21, 11.89it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 14%|█▎        | 41/300 [00:04<00:21, 11.78it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 14%|█▍        | 43/300 [00:04<00:21, 11.75it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 15%|█▌        | 45/300 [00:04<00:21, 11.63it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 16%|█▌        | 47/300 [00:04<00:22, 11.49it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 16%|█▋        | 49/300 [00:04<00:21, 11.70it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 17%|█▋        | 51/300 [00:04<00:20, 12.41it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 18%|█▊        | 53/300 [00:05<00:19, 12.58it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 18%|█▊        | 55/300 [00:05<00:19, 12.64it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 19%|█▉        | 57/300 [00:05<00:19, 12.78it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 20%|█▉        | 59/300 [00:05<00:18, 12.95it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 20%|██        | 61/300 [00:05<00:21, 11.24it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 21%|██        | 63/300 [00:06<00:22, 10.49it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 22%|██▏       | 65/300 [00:06<00:27,  8.52it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 22%|██▏       | 67/300 [00:06<00:25,  8.99it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 23%|██▎       | 69/300 [00:06<00:25,  8.92it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 24%|██▎       | 71/300 [00:07<00:24,  9.44it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 24%|██▍       | 72/300 [00:07<00:25,  8.96it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 25%|██▍       | 74/300 [00:07<00:22,  9.99it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 25%|██▌       | 76/300 [00:07<00:26,  8.58it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 26%|██▌       | 77/300 [00:07<00:27,  8.24it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 26%|██▋       | 79/300 [00:07<00:26,  8.26it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 27%|██▋       | 80/300 [00:08<00:27,  7.89it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 27%|██▋       | 82/300 [00:08<00:23,  9.10it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 28%|██▊       | 83/300 [00:08<00:24,  9.04it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 28%|██▊       | 84/300 [00:08<00:26,  8.29it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 29%|██▊       | 86/300 [00:08<00:22,  9.32it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 29%|██▉       | 87/300 [00:08<00:24,  8.66it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 30%|██▉       | 89/300 [00:09<00:24,  8.77it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 30%|███       | 91/300 [00:09<00:21,  9.68it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 31%|███       | 93/300 [00:09<00:21,  9.60it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 31%|███▏      | 94/300 [00:09<00:21,  9.53it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 32%|███▏      | 95/300 [00:09<00:21,  9.54it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 32%|███▏      | 97/300 [00:09<00:20, 10.14it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 33%|███▎      | 99/300 [00:10<00:19, 10.56it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 34%|███▎      | 101/300 [00:10<00:17, 11.25it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 34%|███▍      | 103/300 [00:10<00:16, 11.76it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 35%|███▌      | 105/300 [00:10<00:17, 11.01it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 36%|███▌      | 107/300 [00:10<00:19, 10.02it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 36%|███▋      | 109/300 [00:10<00:18, 10.23it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 37%|███▋      | 111/300 [00:11<00:18, 10.36it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 38%|███▊      | 113/300 [00:11<00:17, 10.86it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 38%|███▊      | 115/300 [00:11<00:15, 11.71it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 39%|███▉      | 117/300 [00:11<00:14, 12.29it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 40%|███▉      | 119/300 [00:11<00:14, 12.55it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 40%|████      | 121/300 [00:11<00:14, 11.94it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 41%|████      | 123/300 [00:12<00:16, 10.73it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 42%|████▏     | 125/300 [00:12<00:16, 10.86it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 42%|████▏     | 127/300 [00:12<00:15, 11.52it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 43%|████▎     | 129/300 [00:12<00:14, 11.47it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 44%|████▎     | 131/300 [00:12<00:14, 11.44it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 44%|████▍     | 133/300 [00:13<00:15, 11.09it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 45%|████▌     | 135/300 [00:13<00:14, 11.51it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 46%|████▌     | 137/300 [00:13<00:13, 11.90it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 46%|████▋     | 139/300 [00:13<00:12, 12.50it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 47%|████▋     | 141/300 [00:13<00:12, 13.03it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 48%|████▊     | 143/300 [00:13<00:12, 12.30it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 48%|████▊     | 145/300 [00:14<00:12, 12.02it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 49%|████▉     | 147/300 [00:14<00:12, 11.83it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 50%|████▉     | 149/300 [00:14<00:12, 11.84it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 50%|█████     | 151/300 [00:14<00:14, 10.47it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 51%|█████     | 153/300 [00:14<00:15,  9.34it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 51%|█████▏    | 154/300 [00:14<00:15,  9.31it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 52%|█████▏    | 156/300 [00:15<00:15,  9.22it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 52%|█████▏    | 157/300 [00:15<00:15,  9.02it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 53%|█████▎    | 158/300 [00:15<00:16,  8.66it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 53%|█████▎    | 159/300 [00:15<00:16,  8.38it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 53%|█████▎    | 160/300 [00:15<00:18,  7.63it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 54%|█████▎    | 161/300 [00:15<00:18,  7.38it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 54%|█████▍    | 162/300 [00:16<00:18,  7.41it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 54%|█████▍    | 163/300 [00:16<00:19,  7.07it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 55%|█████▌    | 165/300 [00:16<00:18,  7.49it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 55%|█████▌    | 166/300 [00:16<00:18,  7.29it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 56%|█████▌    | 167/300 [00:16<00:17,  7.76it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 56%|█████▋    | 169/300 [00:16<00:16,  8.18it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 57%|█████▋    | 171/300 [00:17<00:14,  8.95it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 57%|█████▋    | 172/300 [00:17<00:14,  8.68it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 58%|█████▊    | 174/300 [00:17<00:12, 10.04it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 59%|█████▊    | 176/300 [00:17<00:12,  9.57it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 59%|█████▉    | 177/300 [00:17<00:13,  8.91it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 60%|█████▉    | 179/300 [00:17<00:12,  9.60it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 60%|██████    | 180/300 [00:18<00:13,  8.97it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 61%|██████    | 182/300 [00:18<00:12,  9.32it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 61%|██████    | 183/300 [00:18<00:12,  9.39it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 61%|██████▏   | 184/300 [00:18<00:12,  9.07it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 62%|██████▏   | 186/300 [00:18<00:11,  9.59it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 63%|██████▎   | 188/300 [00:18<00:10, 10.71it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 63%|██████▎   | 190/300 [00:18<00:09, 11.43it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 64%|██████▍   | 192/300 [00:19<00:09, 11.48it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 65%|██████▍   | 194/300 [00:19<00:09, 11.11it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 65%|██████▌   | 196/300 [00:19<00:10, 10.02it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 66%|██████▌   | 198/300 [00:19<00:09, 10.45it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 67%|██████▋   | 200/300 [00:19<00:09, 10.61it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 67%|██████▋   | 202/300 [00:20<00:08, 10.98it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 68%|██████▊   | 204/300 [00:20<00:07, 12.16it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 69%|██████▊   | 206/300 [00:20<00:07, 12.67it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 69%|██████▉   | 208/300 [00:20<00:07, 11.84it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 70%|███████   | 210/300 [00:20<00:07, 11.28it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 71%|███████   | 212/300 [00:20<00:08, 10.70it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 71%|███████▏  | 214/300 [00:21<00:07, 10.82it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 72%|███████▏  | 216/300 [00:21<00:07, 11.42it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 73%|███████▎  | 218/300 [00:21<00:08,  9.44it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 73%|███████▎  | 220/300 [00:21<00:08,  9.33it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 74%|███████▍  | 222/300 [00:22<00:07,  9.91it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 75%|███████▍  | 224/300 [00:22<00:07, 10.05it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 75%|███████▌  | 226/300 [00:22<00:07, 10.46it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 76%|███████▌  | 228/300 [00:22<00:06, 10.68it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 77%|███████▋  | 230/300 [00:22<00:06, 11.41it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 77%|███████▋  | 232/300 [00:22<00:06, 11.16it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 78%|███████▊  | 234/300 [00:23<00:05, 11.54it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 79%|███████▊  | 236/300 [00:23<00:05, 11.57it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 79%|███████▉  | 238/300 [00:23<00:05, 11.99it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 80%|████████  | 240/300 [00:23<00:04, 12.87it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 81%|████████  | 242/300 [00:23<00:05, 10.86it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 81%|████████▏ | 244/300 [00:23<00:05, 11.11it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 82%|████████▏ | 246/300 [00:24<00:04, 11.29it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 83%|████████▎ | 248/300 [00:24<00:04, 11.62it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 83%|████████▎ | 250/300 [00:24<00:04, 10.34it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 84%|████████▍ | 252/300 [00:24<00:04, 10.74it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 85%|████████▍ | 254/300 [00:24<00:04, 10.13it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 85%|████████▌ | 256/300 [00:25<00:04, 10.74it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 86%|████████▌ | 258/300 [00:25<00:04, 10.46it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 87%|████████▋ | 260/300 [00:25<00:03, 10.80it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 87%|████████▋ | 262/300 [00:25<00:03, 11.57it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 88%|████████▊ | 264/300 [00:25<00:03, 10.70it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 89%|████████▊ | 266/300 [00:26<00:04,  8.45it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 89%|████████▉ | 268/300 [00:26<00:03,  9.56it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 90%|█████████ | 270/300 [00:26<00:03,  9.72it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 91%|█████████ | 272/300 [00:26<00:02,  9.54it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 91%|█████████▏| 274/300 [00:26<00:02,  9.35it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 92%|█████████▏| 276/300 [00:27<00:02,  9.92it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 93%|█████████▎| 278/300 [00:27<00:02,  9.60it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 93%|█████████▎| 279/300 [00:27<00:02,  8.72it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 94%|█████████▎| 281/300 [00:27<00:01,  9.60it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 94%|█████████▍| 283/300 [00:27<00:01, 10.85it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 95%|█████████▌| 285/300 [00:28<00:01, 10.02it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 96%|█████████▌| 287/300 [00:28<00:01, 10.02it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 96%|█████████▋| 289/300 [00:28<00:01, 10.61it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 97%|█████████▋| 291/300 [00:28<00:00, 11.32it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 98%|█████████▊| 293/300 [00:28<00:00, 12.27it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 98%|█████████▊| 295/300 [00:28<00:00, 10.50it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 99%|█████████▉| 297/300 [00:29<00:00, 11.04it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|█████████▉| 299/300 [00:29<00:00,  9.83it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|██████████| 300/300 [00:29<00:00, 10.20it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test :\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 0/200 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  0%|          | 1/200 [00:00<00:29,  6.66it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  1%|          | 2/200 [00:00<00:25,  7.79it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  2%|▏         | 3/200 [00:00<00:26,  7.38it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  2%|▏         | 4/200 [00:00<00:27,  7.12it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  2%|▎         | 5/200 [00:00<00:28,  6.74it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  3%|▎         | 6/200 [00:00<00:37,  5.23it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  4%|▎         | 7/200 [00:01<00:36,  5.34it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  4%|▍         | 8/200 [00:01<00:35,  5.38it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  4%|▍         | 9/200 [00:01<00:34,  5.58it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  5%|▌         | 10/200 [00:01<00:33,  5.67it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  6%|▌         | 11/200 [00:01<00:33,  5.64it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  6%|▌         | 12/200 [00:01<00:29,  6.43it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  7%|▋         | 14/200 [00:02<00:23,  7.78it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  8%|▊         | 16/200 [00:02<00:21,  8.55it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  8%|▊         | 17/200 [00:02<00:21,  8.45it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "  9%|▉         | 18/200 [00:02<00:29,  6.24it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 10%|█         | 20/200 [00:03<00:25,  7.15it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 11%|█         | 22/200 [00:03<00:20,  8.66it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 12%|█▏        | 24/200 [00:03<00:18,  9.51it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 13%|█▎        | 26/200 [00:03<00:16, 10.65it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 14%|█▍        | 28/200 [00:03<00:18,  9.36it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 15%|█▌        | 30/200 [00:03<00:19,  8.77it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 16%|█▌        | 32/200 [00:04<00:17,  9.88it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 17%|█▋        | 34/200 [00:04<00:15, 10.52it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 18%|█▊        | 36/200 [00:04<00:17,  9.13it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 18%|█▊        | 37/200 [00:04<00:19,  8.51it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 20%|█▉        | 39/200 [00:04<00:16,  9.83it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 20%|██        | 41/200 [00:05<00:17,  9.29it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 21%|██        | 42/200 [00:05<00:18,  8.73it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 22%|██▏       | 43/200 [00:05<00:18,  8.52it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 22%|██▎       | 45/200 [00:05<00:16,  9.43it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 23%|██▎       | 46/200 [00:05<00:17,  8.58it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 24%|██▍       | 48/200 [00:05<00:15, 10.02it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 25%|██▌       | 50/200 [00:06<00:15,  9.54it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 26%|██▌       | 52/200 [00:06<00:14,  9.99it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 27%|██▋       | 54/200 [00:06<00:16,  8.73it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 28%|██▊       | 55/200 [00:06<00:17,  8.11it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 28%|██▊       | 57/200 [00:06<00:16,  8.70it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 29%|██▉       | 58/200 [00:07<00:15,  8.92it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 30%|██▉       | 59/200 [00:07<00:16,  8.52it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 30%|███       | 60/200 [00:07<00:19,  7.03it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 30%|███       | 61/200 [00:07<00:18,  7.51it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 32%|███▏      | 63/200 [00:07<00:14,  9.14it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 32%|███▎      | 65/200 [00:07<00:14,  9.11it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 34%|███▎      | 67/200 [00:08<00:13, 10.16it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 34%|███▍      | 69/200 [00:08<00:11, 11.01it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 36%|███▌      | 71/200 [00:08<00:12, 10.44it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 36%|███▋      | 73/200 [00:08<00:11, 11.42it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 38%|███▊      | 75/200 [00:08<00:10, 11.38it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 38%|███▊      | 77/200 [00:08<00:12,  9.49it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 40%|███▉      | 79/200 [00:09<00:13,  9.05it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 40%|████      | 81/200 [00:09<00:12,  9.71it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 42%|████▏     | 83/200 [00:09<00:12,  9.18it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 42%|████▎     | 85/200 [00:09<00:11, 10.03it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 44%|████▎     | 87/200 [00:10<00:13,  8.69it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 44%|████▍     | 89/200 [00:10<00:11,  9.92it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 46%|████▌     | 91/200 [00:10<00:10, 10.31it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 46%|████▋     | 93/200 [00:10<00:10, 10.14it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 48%|████▊     | 95/200 [00:10<00:10,  9.75it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 48%|████▊     | 97/200 [00:10<00:09, 10.63it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 50%|████▉     | 99/200 [00:11<00:09, 11.00it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 50%|█████     | 101/200 [00:11<00:08, 11.08it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 52%|█████▏    | 103/200 [00:11<00:09,  9.98it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 52%|█████▎    | 105/200 [00:11<00:09,  9.67it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 53%|█████▎    | 106/200 [00:11<00:10,  8.96it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 54%|█████▎    | 107/200 [00:12<00:10,  8.85it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 54%|█████▍    | 108/200 [00:12<00:10,  8.73it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 55%|█████▌    | 110/200 [00:12<00:09,  9.27it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 56%|█████▌    | 111/200 [00:12<00:10,  8.44it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 56%|█████▌    | 112/200 [00:12<00:10,  8.27it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 57%|█████▋    | 114/200 [00:12<00:09,  9.18it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 57%|█████▊    | 115/200 [00:12<00:09,  9.23it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 58%|█████▊    | 117/200 [00:13<00:08, 10.02it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 60%|█████▉    | 119/200 [00:13<00:08,  9.10it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 60%|██████    | 121/200 [00:13<00:08,  9.30it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 61%|██████    | 122/200 [00:13<00:08,  8.98it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 62%|██████▏   | 123/200 [00:13<00:09,  8.20it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 62%|██████▏   | 124/200 [00:14<00:09,  7.97it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 62%|██████▎   | 125/200 [00:14<00:09,  7.53it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 63%|██████▎   | 126/200 [00:14<00:11,  6.72it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 64%|██████▎   | 127/200 [00:14<00:10,  7.00it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 64%|██████▍   | 128/200 [00:14<00:09,  7.58it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 64%|██████▍   | 129/200 [00:14<00:10,  6.85it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 66%|██████▌   | 131/200 [00:15<00:09,  7.36it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 66%|██████▌   | 132/200 [00:15<00:08,  7.62it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 67%|██████▋   | 134/200 [00:15<00:07,  8.81it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 68%|██████▊   | 136/200 [00:15<00:07,  8.70it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 69%|██████▉   | 138/200 [00:15<00:06,  9.69it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 70%|███████   | 140/200 [00:15<00:05, 10.43it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 71%|███████   | 142/200 [00:16<00:05,  9.88it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 72%|███████▏  | 144/200 [00:16<00:06,  8.71it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 72%|███████▎  | 145/200 [00:16<00:06,  8.82it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 73%|███████▎  | 146/200 [00:16<00:06,  8.82it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 74%|███████▎  | 147/200 [00:16<00:06,  8.17it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 74%|███████▍  | 149/200 [00:16<00:05,  8.85it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 75%|███████▌  | 150/200 [00:17<00:06,  7.95it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 76%|███████▌  | 151/200 [00:17<00:06,  7.71it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 76%|███████▋  | 153/200 [00:17<00:05,  8.76it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 77%|███████▋  | 154/200 [00:17<00:05,  8.91it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 78%|███████▊  | 155/200 [00:17<00:05,  8.37it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 78%|███████▊  | 157/200 [00:17<00:04,  9.74it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 80%|███████▉  | 159/200 [00:17<00:03, 11.38it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 80%|████████  | 161/200 [00:18<00:03, 10.70it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 82%|████████▏ | 163/200 [00:18<00:03,  9.36it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 82%|████████▏ | 164/200 [00:18<00:03,  9.24it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 82%|████████▎ | 165/200 [00:18<00:04,  8.36it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 84%|████████▎ | 167/200 [00:18<00:03,  9.78it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 84%|████████▍ | 169/200 [00:19<00:03,  9.23it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 86%|████████▌ | 171/200 [00:19<00:03,  9.30it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 86%|████████▌ | 172/200 [00:19<00:02,  9.39it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 86%|████████▋ | 173/200 [00:19<00:03,  8.32it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 87%|████████▋ | 174/200 [00:19<00:03,  8.50it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 88%|████████▊ | 175/200 [00:19<00:03,  8.03it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 88%|████████▊ | 176/200 [00:20<00:03,  7.69it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 89%|████████▉ | 178/200 [00:20<00:02,  9.02it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 90%|█████████ | 180/200 [00:20<00:01, 10.34it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 91%|█████████ | 182/200 [00:20<00:02,  8.42it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 92%|█████████▏| 183/200 [00:20<00:02,  8.10it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 92%|█████████▏| 184/200 [00:20<00:02,  7.97it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 92%|█████████▎| 185/200 [00:21<00:01,  8.27it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 94%|█████████▎| 187/200 [00:21<00:01,  8.90it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 94%|█████████▍| 188/200 [00:21<00:01,  7.71it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 94%|█████████▍| 189/200 [00:21<00:01,  7.74it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 95%|█████████▌| 190/200 [00:21<00:01,  7.26it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 96%|█████████▌| 191/200 [00:21<00:01,  6.59it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 96%|█████████▌| 192/200 [00:22<00:01,  6.83it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 96%|█████████▋| 193/200 [00:22<00:01,  6.40it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 98%|█████████▊| 195/200 [00:22<00:00,  7.61it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 98%|█████████▊| 196/200 [00:22<00:00,  7.89it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      " 98%|█████████▊| 197/200 [00:22<00:00,  7.54it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|█████████▉| 199/200 [00:22<00:00,  8.13it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|██████████| 200/200 [00:23<00:00,  8.49it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r",
      "100%|██████████| 200/200 [00:23<00:00,  8.69it/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "URL = 'https://storage.googleapis.com/thumos14_files/UCF101_videos.zip'\n",
    "download_dir = pathlib.Path('./UCF101_subset/')\n",
    "subset_paths = download_ufc_101_subset(URL, \n",
    "                        num_classes = 10, \n",
    "                        splits = {\"train\": 30, \"test\": 20}, \n",
    "                        download_dir = download_dir)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "aYYShfhMx9DW"
   },
   "source": [
    "创建训练并测试数据集："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-11-07T17:30:22.528494Z",
     "iopub.status.busy": "2023-11-07T17:30:22.528083Z",
     "iopub.status.idle": "2023-11-07T17:30:23.137215Z",
     "shell.execute_reply": "2023-11-07T17:30:23.136530Z"
    },
    "id": "-twTu3_Bx-iJ"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-11-07 17:30:23.068909: W tensorflow/core/common_runtime/gpu/gpu_device.cc:2211] Cannot dlopen some GPU libraries. Please make sure the missing libraries mentioned above are installed properly if you would like to use GPU. Follow the guide at https://www.tensorflow.org/install/gpu for how to download and setup the required libraries for your platform.\n",
      "Skipping registering GPU devices...\n"
     ]
    }
   ],
   "source": [
    "batch_size = 8\n",
    "num_frames = 8\n",
    "\n",
    "output_signature = (tf.TensorSpec(shape = (None, None, None, 3), dtype = tf.float32),\n",
    "                    tf.TensorSpec(shape = (), dtype = tf.int16))\n",
    "\n",
    "train_ds = tf.data.Dataset.from_generator(FrameGenerator(subset_paths['train'], num_frames, training = True),\n",
    "                                          output_signature = output_signature)\n",
    "train_ds = train_ds.batch(batch_size)\n",
    "\n",
    "test_ds = tf.data.Dataset.from_generator(FrameGenerator(subset_paths['test'], num_frames),\n",
    "                                         output_signature = output_signature)\n",
    "test_ds = test_ds.batch(batch_size)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "d7stgmuBCGQT"
   },
   "source": [
    "此处生成的标签表示类的编码。例如，“ApplyEyeMakeup”被映射到整数。查看训练数据的标签，确保数据集已被充分重排。 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-11-07T17:30:23.141396Z",
     "iopub.status.busy": "2023-11-07T17:30:23.140900Z",
     "iopub.status.idle": "2023-11-07T17:30:29.396180Z",
     "shell.execute_reply": "2023-11-07T17:30:29.395316Z"
    },
    "id": "k9L2-toXCOQq"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor([2 9 2 5 4 9 7 9], shape=(8,), dtype=int16)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor([8 8 2 4 9 2 9 2], shape=(8,), dtype=int16)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor([7 1 8 1 2 6 0 6], shape=(8,), dtype=int16)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor([0 5 2 1 4 7 1 2], shape=(8,), dtype=int16)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor([8 3 1 4 7 1 4 0], shape=(8,), dtype=int16)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor([2 4 5 8 6 2 0 4], shape=(8,), dtype=int16)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor([2 2 1 5 6 1 1 8], shape=(8,), dtype=int16)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor([1 1 8 0 5 8 0 2], shape=(8,), dtype=int16)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor([0 4 6 6 2 9 6 4], shape=(8,), dtype=int16)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor([7 3 5 3 6 7 9 0], shape=(8,), dtype=int16)\n"
     ]
    }
   ],
   "source": [
    "for frames, labels in train_ds.take(10):\n",
    "  print(labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "YZ3qwZnpfy9c"
   },
   "source": [
    "查看数据的形状。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-11-07T17:30:29.400080Z",
     "iopub.status.busy": "2023-11-07T17:30:29.399461Z",
     "iopub.status.idle": "2023-11-07T17:30:29.403579Z",
     "shell.execute_reply": "2023-11-07T17:30:29.402931Z"
    },
    "id": "b6MqP4m2fyQT"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape: (8, 8, 224, 224, 3)\n",
      "Label: (8,)\n"
     ]
    }
   ],
   "source": [
    "print(f\"Shape: {frames.shape}\")\n",
    "print(f\"Label: {labels.shape}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "lxbhPqXGvc_F"
   },
   "source": [
    "## 什么是 MoViNets？\n",
    "\n",
    "如前所述，[MoViNets](https://arxiv.org/abs/2103.11511) 是用于流式传输视频或动作识别等任务中的在线推断的视频分类模型。考虑使用 MoViNets 对您的视频数据进行分类以进行动作识别。\n",
    "\n",
    "基于 2D 帧的分类器高效且可简单地运行整个视频，或者一次流式传输一帧。由于它们不能考虑时间上下文，它们的准确率有限，并且可能会在帧与帧之间给出不一致的输出。\n",
    "\n",
    "一个简单的 3D CNN 使用双向时间上下文，可以提高准确率和时间一致性。这些网络可能需要更多资源，并且由于它们着眼于未来，不能用于流式传输数据。\n",
    "\n",
    "![Standard convolution](https://tensorflow.google.cn/images/tutorials/video/standard_convolution.png)\n",
    "\n",
    "MoViNet 架构使用沿时间轴“因果”的 3D 卷积（如 `padding=\"causal\"` 的 `layers.Conv1D`）。这提供了两种方式的一些优点，主要是它允许高效流式传输。\n",
    "\n",
    "![Causal convolution](https://tensorflow.google.cn/images/tutorials/video/causal_convolution.png)\n",
    "\n",
    "因果卷积确保仅使用直到时间 *t* 的输入来计算时间 *t* 的输出。为了演示这如何提高流式传输的效率，请从您可能熟悉的一个更简单示例开始：RNN。RNN 通过时间向前传递状态：\n",
    "\n",
    "![RNN model](https://tensorflow.google.cn/images/tutorials/video/rnn_comparison.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-11-07T17:30:29.406994Z",
     "iopub.status.busy": "2023-11-07T17:30:29.406430Z",
     "iopub.status.idle": "2023-11-07T17:30:29.469791Z",
     "shell.execute_reply": "2023-11-07T17:30:29.469118Z"
    },
    "id": "dMvDkgfFZC6a"
   },
   "outputs": [],
   "source": [
    "gru = layers.GRU(units=4, return_sequences=True, return_state=True)\n",
    "\n",
    "inputs = tf.random.normal(shape=[1, 10, 8]) # (batch, sequence, channels)\n",
    "\n",
    "result, state = gru(inputs) # Run it all at once"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "T7xyb5C4bTs7"
   },
   "source": [
    "通过设置 RNN 的 `return_sequences=True` 参数，可以要求它在计算结束时返回状态。这样，您就可以暂停，随后从上次中断的地方继续，以获得完全相同的结果：\n",
    "\n",
    "![States passing in RNNs](https://tensorflow.google.cn/images/tutorials/video/rnn_state_passing.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-11-07T17:30:29.473445Z",
     "iopub.status.busy": "2023-11-07T17:30:29.472863Z",
     "iopub.status.idle": "2023-11-07T17:30:29.501580Z",
     "shell.execute_reply": "2023-11-07T17:30:29.500941Z"
    },
    "id": "bI8FOPRRXXPa"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True\n",
      "True\n"
     ]
    }
   ],
   "source": [
    "first_half, state = gru(inputs[:, :5, :])   # run the first half, and capture the state\n",
    "second_half, _ = gru(inputs[:,5:, :], initial_state=state)  # Use the state to continue where you left off.\n",
    "\n",
    "print(np.allclose(result[:, :5,:], first_half))\n",
    "print(np.allclose(result[:, 5:,:], second_half))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "KM3MArumY_Qk"
   },
   "source": [
    "如果小心处理，因果卷积能以相同的方式使用。Le Paine 等人在 [Fast Wavenet Generation Algorithm](https://arxiv.org/abs/1611.09482) 中使用了这种技术。在 [MoVinet 论文](https://arxiv.org/abs/2103.11511)中，`state` 被称为“流缓冲区”。\n",
    "\n",
    "![States passed in causal convolution](https://tensorflow.google.cn/images/tutorials/video/causal_conv_states.png)\n",
    "\n",
    "通过向前传递这一点状态，可以避免重新计算上面显示的整个感受域。 "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "1UsxiPs8yA2e"
   },
   "source": [
    "## 下载预训练的 MoViNet 模型\n",
    "\n",
    "在本部分中，您将：\n",
    "\n",
    "1. 可以使用 TensorFlow 模型的 [`official/projects/movinet`](https://github.com/tensorflow/models/tree/master/official/projects/movinet) 中提供的开源代码创建 MoViNet 模型。\n",
    "2. 加载预训练的权重。\n",
    "3. 冻结卷积基或除最终分类器头之外的所有其他层，以加快微调速度。\n",
    "\n",
    "要构建模型，您可以从 `a0` 配置开始，因为在针对其他模型进行基准分析时，它的训练速度最快。查看 [TensorFlow Model Garden 上的可用 MoViNet 模型](https://github.com/tensorflow/models/blob/master/official/projects/movinet/configs/movinet.py)，了解哪些模型可能适用于您的用例。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-11-07T17:30:29.505473Z",
     "iopub.status.busy": "2023-11-07T17:30:29.504901Z",
     "iopub.status.idle": "2023-11-07T17:30:36.245622Z",
     "shell.execute_reply": "2023-11-07T17:30:36.244773Z"
    },
    "id": "rhSCM6cee05F"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "movinet_a0_base/\r\n",
      "movinet_a0_base/checkpoint\r\n",
      "movinet_a0_base/ckpt-1.data-00000-of-00001\r\n",
      "movinet_a0_base/ckpt-1.index\r\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.checkpoint.checkpoint.CheckpointLoadStatus at 0x7f120079dfd0>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_id = 'a0'\n",
    "resolution = 224\n",
    "\n",
    "tf.keras.backend.clear_session()\n",
    "\n",
    "backbone = movinet.Movinet(model_id=model_id)\n",
    "backbone.trainable = False\n",
    "\n",
    "# Set num_classes=600 to load the pre-trained weights from the original model\n",
    "model = movinet_model.MovinetClassifier(backbone=backbone, num_classes=600)\n",
    "model.build([None, None, None, None, 3])\n",
    "\n",
    "# Load pre-trained weights\n",
    "!wget https://storage.googleapis.com/tf_model_garden/vision/movinet/movinet_a0_base.tar.gz -O movinet_a0_base.tar.gz -q\n",
    "!tar -xvf movinet_a0_base.tar.gz\n",
    "\n",
    "checkpoint_dir = f'movinet_{model_id}_base'\n",
    "checkpoint_path = tf.train.latest_checkpoint(checkpoint_dir)\n",
    "checkpoint = tf.train.Checkpoint(model=model)\n",
    "status = checkpoint.restore(checkpoint_path)\n",
    "status.assert_existing_objects_matched()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "BW23HVNtCXff"
   },
   "source": [
    "要构建分类器，请创建一个采用主干和数据集中的类数的函数。`build_classifier` 函数将采用主干和数据集中的类数来构建分类器。在这种情况下，新分类器将采用 `num_classes` 个输出（UCF101 的此子集有 10 个类）。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-11-07T17:30:36.249610Z",
     "iopub.status.busy": "2023-11-07T17:30:36.249312Z",
     "iopub.status.idle": "2023-11-07T17:30:36.253653Z",
     "shell.execute_reply": "2023-11-07T17:30:36.252982Z"
    },
    "id": "6cfAelbU5Gi3"
   },
   "outputs": [],
   "source": [
    "def build_classifier(batch_size, num_frames, resolution, backbone, num_classes):\n",
    "  \"\"\"Builds a classifier on top of a backbone model.\"\"\"\n",
    "  model = movinet_model.MovinetClassifier(\n",
    "      backbone=backbone,\n",
    "      num_classes=num_classes)\n",
    "  model.build([batch_size, num_frames, resolution, resolution, 3])\n",
    "\n",
    "  return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-11-07T17:30:36.256915Z",
     "iopub.status.busy": "2023-11-07T17:30:36.256452Z",
     "iopub.status.idle": "2023-11-07T17:30:37.999139Z",
     "shell.execute_reply": "2023-11-07T17:30:37.998329Z"
    },
    "id": "9HWSk-u7oPUZ"
   },
   "outputs": [],
   "source": [
    "model = build_classifier(batch_size, num_frames, resolution, backbone, 10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "JhbX7qdTN8lc"
   },
   "source": [
    "对于本教程，选择 `tf.keras.optimizers.Adam` 优化器和 `tf.keras.losses.SparseCategoricalCrossentropy` 损失函数。使用指标参数查看每个步骤中模型性能的准确率。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-11-07T17:30:38.003340Z",
     "iopub.status.busy": "2023-11-07T17:30:38.002834Z",
     "iopub.status.idle": "2023-11-07T17:30:38.022158Z",
     "shell.execute_reply": "2023-11-07T17:30:38.021570Z"
    },
    "id": "dVqBLrn1tBsd"
   },
   "outputs": [],
   "source": [
    "num_epochs = 2\n",
    "\n",
    "loss_obj = tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True)\n",
    "\n",
    "optimizer = tf.keras.optimizers.Adam(learning_rate = 0.001)\n",
    "\n",
    "model.compile(loss=loss_obj, optimizer=optimizer, metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "VflEr_t6CuQu"
   },
   "source": [
    "训练模型。两个周期后，观察训练集和测试集的低损失和高准确率。 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-11-07T17:30:38.025944Z",
     "iopub.status.busy": "2023-11-07T17:30:38.025264Z",
     "iopub.status.idle": "2023-11-07T17:56:26.980044Z",
     "shell.execute_reply": "2023-11-07T17:56:26.979216Z"
    },
    "id": "9ZeiYzI0tqQG"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/2\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-11-07 17:30:49.251128: E ./tensorflow/compiler/xla/stream_executor/stream_executor_internal.h:124] SetPriority unimplemented for this stream.\n",
      "2023-11-07 17:30:49.387365: E ./tensorflow/compiler/xla/stream_executor/stream_executor_internal.h:124] SetPriority unimplemented for this stream.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "      1/Unknown - 24s 24s/step - loss: 2.2871 - accuracy: 0.1250"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\r",
      "      2/Unknown - 37s 12s/step - loss: 2.2236 - accuracy: 0.3125"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\r",
      "      3/Unknown - 49s 12s/step - loss: 2.1787 - accuracy: 0.3750"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\r",
      "      4/Unknown - 62s 12s/step - loss: 2.1110 - accuracy: 0.4062"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\r",
      "      5/Unknown - 74s 12s/step - loss: 2.0345 - accuracy: 0.4000"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\r",
      "      6/Unknown - 86s 12s/step - loss: 1.9710 - accuracy: 0.4375"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\r",
      "      7/Unknown - 99s 12s/step - loss: 1.9378 - accuracy: 0.4643"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\r",
      "      8/Unknown - 111s 12s/step - loss: 1.9194 - accuracy: 0.4688"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "      9/Unknown - 123s 12s/step - loss: 1.8709 - accuracy: 0.5000"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     10/Unknown - 136s 12s/step - loss: 1.8010 - accuracy: 0.5375"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     11/Unknown - 148s 12s/step - loss: 1.7230 - accuracy: 0.5568"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     12/Unknown - 160s 12s/step - loss: 1.6488 - accuracy: 0.5833"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     13/Unknown - 173s 12s/step - loss: 1.5995 - accuracy: 0.6154"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     14/Unknown - 185s 12s/step - loss: 1.5441 - accuracy: 0.6429"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     15/Unknown - 197s 12s/step - loss: 1.5073 - accuracy: 0.6583"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     16/Unknown - 210s 12s/step - loss: 1.4792 - accuracy: 0.6719"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     17/Unknown - 222s 12s/step - loss: 1.4502 - accuracy: 0.6838"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     18/Unknown - 234s 12s/step - loss: 1.4147 - accuracy: 0.6944"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     19/Unknown - 247s 12s/step - loss: 1.3872 - accuracy: 0.6974"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     20/Unknown - 259s 12s/step - loss: 1.3603 - accuracy: 0.6938"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     21/Unknown - 271s 12s/step - loss: 1.3240 - accuracy: 0.7024"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     22/Unknown - 284s 12s/step - loss: 1.2887 - accuracy: 0.7159"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     23/Unknown - 296s 12s/step - loss: 1.2520 - accuracy: 0.7283"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     24/Unknown - 309s 12s/step - loss: 1.2208 - accuracy: 0.7344"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     25/Unknown - 321s 12s/step - loss: 1.1851 - accuracy: 0.7400"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     26/Unknown - 334s 12s/step - loss: 1.1598 - accuracy: 0.7404"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     27/Unknown - 346s 12s/step - loss: 1.1294 - accuracy: 0.7500"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     28/Unknown - 359s 12s/step - loss: 1.1045 - accuracy: 0.7545"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     29/Unknown - 371s 12s/step - loss: 1.0767 - accuracy: 0.7629"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     30/Unknown - 383s 12s/step - loss: 1.0498 - accuracy: 0.7708"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     31/Unknown - 396s 12s/step - loss: 1.0223 - accuracy: 0.7782"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     32/Unknown - 408s 12s/step - loss: 0.9927 - accuracy: 0.7852"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     33/Unknown - 420s 12s/step - loss: 0.9854 - accuracy: 0.7803"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     34/Unknown - 433s 12s/step - loss: 0.9604 - accuracy: 0.7868"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     35/Unknown - 445s 12s/step - loss: 0.9376 - accuracy: 0.7929"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     36/Unknown - 457s 12s/step - loss: 0.9195 - accuracy: 0.7951"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     37/Unknown - 470s 12s/step - loss: 0.8998 - accuracy: 0.8007"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     38/Unknown - 479s 12s/step - loss: 0.8940 - accuracy: 0.8000"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "38/38 [==============================] - 785s 21s/step - loss: 0.8940 - accuracy: 0.8000 - val_loss: 0.1508 - val_accuracy: 0.9900\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 2/2\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      " 1/38 [..............................] - ETA: 8:07 - loss: 0.0802 - accuracy: 1.0000"
     ]
    },
    {
     "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\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",
      " 2/38 [>.............................] - ETA: 7:24 - loss: 0.0697 - accuracy: 1.0000"
     ]
    },
    {
     "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\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",
      " 3/38 [=>............................] - ETA: 7:11 - loss: 0.0536 - accuracy: 1.0000"
     ]
    },
    {
     "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\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",
      " 4/38 [==>...........................] - ETA: 6:59 - loss: 0.1014 - accuracy: 0.9688"
     ]
    },
    {
     "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\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",
      " 5/38 [==>...........................] - ETA: 6:45 - loss: 0.1108 - accuracy: 0.9750"
     ]
    },
    {
     "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\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",
      " 6/38 [===>..........................] - ETA: 6:33 - loss: 0.0981 - accuracy: 0.9792"
     ]
    },
    {
     "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\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",
      " 7/38 [====>.........................] - ETA: 6:20 - loss: 0.1141 - accuracy: 0.9643"
     ]
    },
    {
     "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\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",
      " 8/38 [=====>........................] - ETA: 6:08 - loss: 0.1098 - accuracy: 0.9688"
     ]
    },
    {
     "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\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",
      " 9/38 [======>.......................] - ETA: 5:55 - loss: 0.1111 - accuracy: 0.9722"
     ]
    },
    {
     "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\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",
      "10/38 [======>.......................] - ETA: 5:43 - loss: 0.1072 - accuracy: 0.9750"
     ]
    },
    {
     "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\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",
      "11/38 [=======>......................] - ETA: 5:31 - loss: 0.1037 - accuracy: 0.9773"
     ]
    },
    {
     "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\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",
      "12/38 [========>.....................] - ETA: 5:19 - loss: 0.1087 - accuracy: 0.9792"
     ]
    },
    {
     "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\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",
      "13/38 [=========>....................] - ETA: 5:07 - loss: 0.1055 - accuracy: 0.9808"
     ]
    },
    {
     "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\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",
      "14/38 [==========>...................] - ETA: 4:54 - loss: 0.1035 - accuracy: 0.9821"
     ]
    },
    {
     "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\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",
      "15/38 [==========>...................] - ETA: 4:42 - loss: 0.1061 - accuracy: 0.9833"
     ]
    },
    {
     "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\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",
      "16/38 [===========>..................] - ETA: 4:30 - loss: 0.1110 - accuracy: 0.9766"
     ]
    },
    {
     "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\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",
      "17/38 [============>.................] - ETA: 4:17 - loss: 0.1326 - accuracy: 0.9706"
     ]
    },
    {
     "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\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",
      "18/38 [=============>................] - ETA: 4:05 - loss: 0.1337 - accuracy: 0.9653"
     ]
    },
    {
     "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\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",
      "19/38 [==============>...............] - ETA: 3:53 - loss: 0.1407 - accuracy: 0.9605"
     ]
    },
    {
     "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\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",
      "20/38 [==============>...............] - ETA: 3:40 - loss: 0.1366 - accuracy: 0.9625"
     ]
    },
    {
     "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\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",
      "21/38 [===============>..............] - ETA: 3:28 - loss: 0.1308 - accuracy: 0.9643"
     ]
    },
    {
     "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\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",
      "22/38 [================>.............] - ETA: 3:16 - loss: 0.1312 - accuracy: 0.9602"
     ]
    },
    {
     "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\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",
      "23/38 [=================>............] - ETA: 3:04 - loss: 0.1285 - accuracy: 0.9620"
     ]
    },
    {
     "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\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",
      "24/38 [=================>............] - ETA: 2:51 - loss: 0.1241 - accuracy: 0.9635"
     ]
    },
    {
     "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\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",
      "25/38 [==================>...........] - ETA: 2:39 - loss: 0.1220 - accuracy: 0.9650"
     ]
    },
    {
     "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\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",
      "26/38 [===================>..........] - ETA: 2:27 - loss: 0.1228 - accuracy: 0.9615"
     ]
    },
    {
     "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\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",
      "27/38 [====================>.........] - ETA: 2:14 - loss: 0.1196 - accuracy: 0.9630"
     ]
    },
    {
     "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\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",
      "28/38 [=====================>........] - ETA: 2:02 - loss: 0.1165 - accuracy: 0.9643"
     ]
    },
    {
     "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\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",
      "29/38 [=====================>........] - ETA: 1:50 - loss: 0.1161 - accuracy: 0.9655"
     ]
    },
    {
     "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\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",
      "30/38 [======================>.......] - ETA: 1:38 - loss: 0.1141 - accuracy: 0.9667"
     ]
    },
    {
     "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\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",
      "31/38 [=======================>......] - ETA: 1:25 - loss: 0.1117 - accuracy: 0.9677"
     ]
    },
    {
     "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\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",
      "32/38 [========================>.....] - ETA: 1:13 - loss: 0.1122 - accuracy: 0.9688"
     ]
    },
    {
     "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\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",
      "33/38 [=========================>....] - ETA: 1:01 - loss: 0.1102 - accuracy: 0.9697"
     ]
    },
    {
     "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\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",
      "34/38 [=========================>....] - ETA: 49s - loss: 0.1079 - accuracy: 0.9706 "
     ]
    },
    {
     "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\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",
      "35/38 [==========================>...] - ETA: 36s - loss: 0.1059 - accuracy: 0.9714"
     ]
    },
    {
     "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\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",
      "36/38 [===========================>..] - ETA: 24s - loss: 0.1036 - accuracy: 0.9722"
     ]
    },
    {
     "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\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",
      "37/38 [============================>.] - ETA: 12s - loss: 0.1024 - accuracy: 0.9730"
     ]
    },
    {
     "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\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",
      "38/38 [==============================] - ETA: 0s - loss: 0.1018 - accuracy: 0.9733 "
     ]
    },
    {
     "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\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",
      "38/38 [==============================] - 764s 20s/step - loss: 0.1018 - accuracy: 0.9733 - val_loss: 0.0713 - val_accuracy: 0.9850\n"
     ]
    }
   ],
   "source": [
    "results = model.fit(train_ds,\n",
    "                    validation_data=test_ds,\n",
    "                    epochs=num_epochs,\n",
    "                    validation_freq=1,\n",
    "                    verbose=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "KkLl2zF8G9W0"
   },
   "source": [
    "## 评估模型\n",
    "\n",
    "该模型在训练数据集上取得了很高的准确率。接下来，使用 Keras `Model.evaluate` 在测试集上对其进行评估。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-11-07T17:56:26.983712Z",
     "iopub.status.busy": "2023-11-07T17:56:26.983454Z",
     "iopub.status.idle": "2023-11-07T18:01:28.033252Z",
     "shell.execute_reply": "2023-11-07T18:01:28.032504Z"
    },
    "id": "NqgbzOiKuxxT"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "      1/Unknown - 13s 13s/step - loss: 0.0174 - accuracy: 1.0000"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\r",
      "      2/Unknown - 25s 12s/step - loss: 0.0258 - accuracy: 1.0000"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\r",
      "      3/Unknown - 37s 12s/step - loss: 0.0198 - accuracy: 1.0000"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\r",
      "      4/Unknown - 49s 12s/step - loss: 0.0155 - accuracy: 1.0000"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\r",
      "      5/Unknown - 61s 12s/step - loss: 0.0136 - accuracy: 1.0000"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\r",
      "      6/Unknown - 73s 12s/step - loss: 0.0118 - accuracy: 1.0000"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\r",
      "      7/Unknown - 85s 12s/step - loss: 0.0111 - accuracy: 1.0000"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\r",
      "      8/Unknown - 97s 12s/step - loss: 0.0211 - accuracy: 1.0000"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\r",
      "      9/Unknown - 109s 12s/step - loss: 0.0578 - accuracy: 0.9861"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     10/Unknown - 121s 12s/step - loss: 0.0711 - accuracy: 0.9875"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     11/Unknown - 133s 12s/step - loss: 0.0697 - accuracy: 0.9886"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     12/Unknown - 145s 12s/step - loss: 0.0887 - accuracy: 0.9792"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     13/Unknown - 157s 12s/step - loss: 0.0961 - accuracy: 0.9808"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     14/Unknown - 169s 12s/step - loss: 0.0960 - accuracy: 0.9821"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     15/Unknown - 181s 12s/step - loss: 0.0946 - accuracy: 0.9833"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     16/Unknown - 193s 12s/step - loss: 0.0909 - accuracy: 0.9844"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     17/Unknown - 205s 12s/step - loss: 0.0877 - accuracy: 0.9853"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     18/Unknown - 217s 12s/step - loss: 0.0843 - accuracy: 0.9861"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     19/Unknown - 229s 12s/step - loss: 0.0889 - accuracy: 0.9803"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     20/Unknown - 241s 12s/step - loss: 0.0861 - accuracy: 0.9812"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     21/Unknown - 253s 12s/step - loss: 0.0865 - accuracy: 0.9821"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     22/Unknown - 265s 12s/step - loss: 0.0834 - accuracy: 0.9830"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     23/Unknown - 277s 12s/step - loss: 0.0863 - accuracy: 0.9783"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     24/Unknown - 289s 12s/step - loss: 0.0831 - accuracy: 0.9792"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "     25/Unknown - 301s 12s/step - loss: 0.0820 - accuracy: 0.9800"
     ]
    },
    {
     "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\b\b\b\b\b\b\b\b\b\r",
      "25/25 [==============================] - 301s 12s/step - loss: 0.0820 - accuracy: 0.9800\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'loss': 0.08196067810058594, 'accuracy': 0.9800000190734863}"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(test_ds, return_dict=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "OkFst2gsHBwD"
   },
   "source": [
    "要进一步呈现模型性能，请使用[混淆矩阵](https://tensorflow.google.cn/api_docs/python/tf/math/confusion_matrix)。混淆矩阵允许评估分类模型的性能，而不仅仅是准确率。为了构建此多类分类问题的混淆矩阵，需要获得测试集中的实际值和预测值。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-11-07T18:01:28.037079Z",
     "iopub.status.busy": "2023-11-07T18:01:28.036459Z",
     "iopub.status.idle": "2023-11-07T18:01:28.041530Z",
     "shell.execute_reply": "2023-11-07T18:01:28.040792Z"
    },
    "id": "hssSdW9XHF_j"
   },
   "outputs": [],
   "source": [
    "def get_actual_predicted_labels(dataset):\n",
    "  \"\"\"\n",
    "    Create a list of actual ground truth values and the predictions from the model.\n",
    "\n",
    "    Args:\n",
    "      dataset: An iterable data structure, such as a TensorFlow Dataset, with features and labels.\n",
    "\n",
    "    Return:\n",
    "      Ground truth and predicted values for a particular dataset.\n",
    "  \"\"\"\n",
    "  actual = [labels for _, labels in dataset.unbatch()]\n",
    "  predicted = model.predict(dataset)\n",
    "\n",
    "  actual = tf.stack(actual, axis=0)\n",
    "  predicted = tf.concat(predicted, axis=0)\n",
    "  predicted = tf.argmax(predicted, axis=1)\n",
    "\n",
    "  return actual, predicted"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-11-07T18:01:28.044979Z",
     "iopub.status.busy": "2023-11-07T18:01:28.044402Z",
     "iopub.status.idle": "2023-11-07T18:01:28.049296Z",
     "shell.execute_reply": "2023-11-07T18:01:28.048691Z"
    },
    "id": "2TmTue6THGWO"
   },
   "outputs": [],
   "source": [
    "def plot_confusion_matrix(actual, predicted, labels, ds_type):\n",
    "  cm = tf.math.confusion_matrix(actual, predicted)\n",
    "  ax = sns.heatmap(cm, annot=True, fmt='g')\n",
    "  sns.set(rc={'figure.figsize':(12, 12)})\n",
    "  sns.set(font_scale=1.4)\n",
    "  ax.set_title('Confusion matrix of action recognition for ' + ds_type)\n",
    "  ax.set_xlabel('Predicted Action')\n",
    "  ax.set_ylabel('Actual Action')\n",
    "  plt.xticks(rotation=90)\n",
    "  plt.yticks(rotation=0)\n",
    "  ax.xaxis.set_ticklabels(labels)\n",
    "  ax.yaxis.set_ticklabels(labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-11-07T18:01:28.052527Z",
     "iopub.status.busy": "2023-11-07T18:01:28.051929Z",
     "iopub.status.idle": "2023-11-07T18:01:28.055688Z",
     "shell.execute_reply": "2023-11-07T18:01:28.055069Z"
    },
    "id": "4RK1A1C1HH6V"
   },
   "outputs": [],
   "source": [
    "fg = FrameGenerator(subset_paths['train'], num_frames, training = True)\n",
    "label_names = list(fg.class_ids_for_name.keys())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-11-07T18:01:28.058905Z",
     "iopub.status.busy": "2023-11-07T18:01:28.058311Z",
     "iopub.status.idle": "2023-11-07T18:06:48.880105Z",
     "shell.execute_reply": "2023-11-07T18:06:48.879332Z"
    },
    "id": "r4AFi2e5HKEO"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "      1/Unknown - 16s 16s/step"
     ]
    },
    {
     "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\r",
      "      2/Unknown - 28s 12s/step"
     ]
    },
    {
     "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\r",
      "      3/Unknown - 40s 12s/step"
     ]
    },
    {
     "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\r",
      "      4/Unknown - 52s 12s/step"
     ]
    },
    {
     "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\r",
      "      5/Unknown - 64s 12s/step"
     ]
    },
    {
     "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\r",
      "      6/Unknown - 76s 12s/step"
     ]
    },
    {
     "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\r",
      "      7/Unknown - 88s 12s/step"
     ]
    },
    {
     "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\r",
      "      8/Unknown - 100s 12s/step"
     ]
    },
    {
     "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\r",
      "      9/Unknown - 112s 12s/step"
     ]
    },
    {
     "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\r",
      "     10/Unknown - 124s 12s/step"
     ]
    },
    {
     "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\r",
      "     11/Unknown - 136s 12s/step"
     ]
    },
    {
     "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\r",
      "     12/Unknown - 148s 12s/step"
     ]
    },
    {
     "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\r",
      "     13/Unknown - 160s 12s/step"
     ]
    },
    {
     "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\r",
      "     14/Unknown - 172s 12s/step"
     ]
    },
    {
     "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\r",
      "     15/Unknown - 184s 12s/step"
     ]
    },
    {
     "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\r",
      "     16/Unknown - 196s 12s/step"
     ]
    },
    {
     "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\r",
      "     17/Unknown - 208s 12s/step"
     ]
    },
    {
     "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\r",
      "     18/Unknown - 220s 12s/step"
     ]
    },
    {
     "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\r",
      "     19/Unknown - 233s 12s/step"
     ]
    },
    {
     "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\r",
      "     20/Unknown - 245s 12s/step"
     ]
    },
    {
     "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\r",
      "     21/Unknown - 257s 12s/step"
     ]
    },
    {
     "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\r",
      "     22/Unknown - 269s 12s/step"
     ]
    },
    {
     "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\r",
      "     23/Unknown - 280s 12s/step"
     ]
    },
    {
     "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\r",
      "     24/Unknown - 293s 12s/step"
     ]
    },
    {
     "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\r",
      "     25/Unknown - 305s 12s/step"
     ]
    },
    {
     "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\r",
      "25/25 [==============================] - 305s 12s/step\n"
     ]
    }
   ],
   "source": [
    "actual, predicted = get_actual_predicted_labels(test_ds)\n",
    "plot_confusion_matrix(actual, predicted, label_names, 'test')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ddQG9sYxa1Ib"
   },
   "source": [
    "## 后续步骤\n",
    "\n",
    "现在，您已经对 MoViNet 模型以及如何利用各种 TensorFlow API（例如，用于迁移学习）有了一定的了解，请尝试将本教程中的代码用于您自己的数据集。数据不必限于视频数据。MRI 扫描等体数据也可与 3D CNN 一起使用。[用于精神分裂症和控制分类的基于脑 MRI 的 3D 卷积神经网络](https://arxiv.org/pdf/2003.08818.pdf)中提到的 NUSDAT 和 IMH 数据集可能是 MRI 数据的两个此类来源。\n",
    "\n",
    "特别是，使用本教程和其他视频数据与分类教程中使用的 `FrameGenerator` 类可以帮助您将数据加载到模型中。\n",
    "\n",
    "要详细了解如何在 TensorFlow 中处理视频数据，请查看以下教程：\n",
    "\n",
    "- [加载视频数据](https://tensorflow.google.cn/tutorials/load_data/video)\n",
    "- [构建用于视频分类的 3D CNN 模型](https://tensorflow.google.cn/tutorials/video/video_classification)\n",
    "- [用于流式动作识别的 MoViNet](https://tensorflow.google.cn/hub/tutorials/movinet)"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "name": "transfer_learning_with_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": "3.9.18"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}