AI-Final-Project.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "4dad7864-b48b-42e6-8e6f-7c9798d4dce5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "hello\n"
     ]
    }
   ],
   "source": [
    "print(\"hello\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "3b78f10a-69b3-4a1f-aaeb-b2cea4f7c660",
   "metadata": {},
   "outputs": [],
   "source": [
    "import cv2\n",
    "import numpy as np\n",
    "import os\n",
    "from matplotlib import pyplot as plt\n",
    "import time\n",
    "import mediapipe as mp # gives us keypoints"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "f940a475-cb09-4e3b-b3df-7d5209e08d6a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting opencv-python\n",
      "  Downloading opencv_python-4.8.1.78-cp37-abi3-macosx_11_0_arm64.whl.metadata (19 kB)\n",
      "Collecting numpy>=1.21.2 (from opencv-python)\n",
      "  Downloading numpy-1.26.2-cp311-cp311-macosx_11_0_arm64.whl.metadata (115 kB)\n",
      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m115.1/115.1 kB\u001b[0m \u001b[31m578.2 kB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0ma \u001b[36m0:00:01\u001b[0m\n",
      "\u001b[?25hDownloading opencv_python-4.8.1.78-cp37-abi3-macosx_11_0_arm64.whl (33.1 MB)\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m33.1/33.1 MB\u001b[0m \u001b[31m13.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m00:01\u001b[0m00:01\u001b[0m\n",
      "\u001b[?25hDownloading numpy-1.26.2-cp311-cp311-macosx_11_0_arm64.whl (14.0 MB)\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m14.0/14.0 MB\u001b[0m \u001b[31m13.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m00:01\u001b[0m00:01\u001b[0m\n",
      "\u001b[?25hInstalling collected packages: numpy, opencv-python\n",
      "Successfully installed numpy-1.26.2 opencv-python-4.8.1.78\n"
     ]
    }
   ],
   "source": [
    "!pip3 install opencv-python"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "1a60cbab-4d47-4423-a487-53ac38eaae97",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting matplotlib\n",
      "  Downloading matplotlib-3.8.2-cp311-cp311-macosx_11_0_arm64.whl.metadata (5.8 kB)\n",
      "Requirement already satisfied: numpy in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (1.26.2)\n",
      "Collecting mediapipe\n",
      "  Downloading mediapipe-0.10.8-cp311-cp311-macosx_11_0_universal2.whl.metadata (9.6 kB)\n",
      "Collecting contourpy>=1.0.1 (from matplotlib)\n",
      "  Downloading contourpy-1.2.0-cp311-cp311-macosx_11_0_arm64.whl.metadata (5.8 kB)\n",
      "Collecting cycler>=0.10 (from matplotlib)\n",
      "  Downloading cycler-0.12.1-py3-none-any.whl.metadata (3.8 kB)\n",
      "Collecting fonttools>=4.22.0 (from matplotlib)\n",
      "  Downloading fonttools-4.45.1-cp311-cp311-macosx_10_9_universal2.whl.metadata (155 kB)\n",
      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m155.2/155.2 kB\u001b[0m \u001b[31m2.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0ma \u001b[36m0:00:01\u001b[0m\n",
      "\u001b[?25hCollecting kiwisolver>=1.3.1 (from matplotlib)\n",
      "  Downloading kiwisolver-1.4.5-cp311-cp311-macosx_11_0_arm64.whl.metadata (6.4 kB)\n",
      "Requirement already satisfied: packaging>=20.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from matplotlib) (23.2)\n",
      "Collecting pillow>=8 (from matplotlib)\n",
      "  Downloading Pillow-10.1.0-cp311-cp311-macosx_11_0_arm64.whl.metadata (9.5 kB)\n",
      "Collecting pyparsing>=2.3.1 (from matplotlib)\n",
      "  Downloading pyparsing-3.1.1-py3-none-any.whl.metadata (5.1 kB)\n",
      "Requirement already satisfied: python-dateutil>=2.7 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from matplotlib) (2.8.2)\n",
      "Collecting absl-py (from mediapipe)\n",
      "  Downloading absl_py-2.0.0-py3-none-any.whl.metadata (2.3 kB)\n",
      "Requirement already satisfied: attrs>=19.1.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from mediapipe) (23.1.0)\n",
      "Collecting flatbuffers>=2.0 (from mediapipe)\n",
      "  Downloading flatbuffers-23.5.26-py2.py3-none-any.whl.metadata (850 bytes)\n",
      "Collecting opencv-contrib-python (from mediapipe)\n",
      "  Downloading opencv_contrib_python-4.8.1.78-cp37-abi3-macosx_11_0_arm64.whl.metadata (19 kB)\n",
      "Collecting protobuf<4,>=3.11 (from mediapipe)\n",
      "  Downloading protobuf-3.20.3-py2.py3-none-any.whl (162 kB)\n",
      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m162.1/162.1 kB\u001b[0m \u001b[31m1.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0ma \u001b[36m0:00:01\u001b[0m\n",
      "\u001b[?25hCollecting sounddevice>=0.4.4 (from mediapipe)\n",
      "  Downloading sounddevice-0.4.6-py3-none-macosx_10_6_x86_64.macosx_10_6_universal2.whl (107 kB)\n",
      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m108.0/108.0 kB\u001b[0m \u001b[31m10.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[?25hRequirement already satisfied: six>=1.5 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from python-dateutil>=2.7->matplotlib) (1.16.0)\n",
      "Requirement already satisfied: CFFI>=1.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from sounddevice>=0.4.4->mediapipe) (1.16.0)\n",
      "Requirement already satisfied: pycparser in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from CFFI>=1.0->sounddevice>=0.4.4->mediapipe) (2.21)\n",
      "Downloading matplotlib-3.8.2-cp311-cp311-macosx_11_0_arm64.whl (7.5 MB)\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m7.5/7.5 MB\u001b[0m \u001b[31m13.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0ma \u001b[36m0:00:01\u001b[0m\n",
      "\u001b[?25hDownloading mediapipe-0.10.8-cp311-cp311-macosx_11_0_universal2.whl (47.0 MB)\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m47.0/47.0 MB\u001b[0m \u001b[31m13.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m00:01\u001b[0m00:01\u001b[0m\n",
      "\u001b[?25hDownloading contourpy-1.2.0-cp311-cp311-macosx_11_0_arm64.whl (243 kB)\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m243.4/243.4 kB\u001b[0m \u001b[31m7.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[?25hDownloading cycler-0.12.1-py3-none-any.whl (8.3 kB)\n",
      "Downloading flatbuffers-23.5.26-py2.py3-none-any.whl (26 kB)\n",
      "Downloading fonttools-4.45.1-cp311-cp311-macosx_10_9_universal2.whl (2.8 MB)\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m2.8/2.8 MB\u001b[0m \u001b[31m13.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m00:01\u001b[0m00:01\u001b[0m\n",
      "\u001b[?25hDownloading kiwisolver-1.4.5-cp311-cp311-macosx_11_0_arm64.whl (66 kB)\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m66.2/66.2 kB\u001b[0m \u001b[31m4.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[?25hDownloading Pillow-10.1.0-cp311-cp311-macosx_11_0_arm64.whl (3.3 MB)\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m3.3/3.3 MB\u001b[0m \u001b[31m13.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m00:01\u001b[0m00:01\u001b[0m\n",
      "\u001b[?25hDownloading pyparsing-3.1.1-py3-none-any.whl (103 kB)\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m103.1/103.1 kB\u001b[0m \u001b[31m9.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[?25hDownloading absl_py-2.0.0-py3-none-any.whl (130 kB)\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m130.2/130.2 kB\u001b[0m \u001b[31m8.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[?25hDownloading opencv_contrib_python-4.8.1.78-cp37-abi3-macosx_11_0_arm64.whl (41.4 MB)\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m41.4/41.4 MB\u001b[0m \u001b[31m13.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m00:01\u001b[0m00:01\u001b[0m\n",
      "\u001b[?25hInstalling collected packages: flatbuffers, pyparsing, protobuf, pillow, opencv-contrib-python, kiwisolver, fonttools, cycler, contourpy, absl-py, sounddevice, matplotlib, mediapipe\n",
      "Successfully installed absl-py-2.0.0 contourpy-1.2.0 cycler-0.12.1 flatbuffers-23.5.26 fonttools-4.45.1 kiwisolver-1.4.5 matplotlib-3.8.2 mediapipe-0.10.8 opencv-contrib-python-4.8.1.78 pillow-10.1.0 protobuf-3.20.3 pyparsing-3.1.1 sounddevice-0.4.6\n"
     ]
    }
   ],
   "source": [
    "!pip3 install matplotlib numpy mediapipe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "6b59547f-a3db-4276-aa1f-afa675ffdda1",
   "metadata": {},
   "outputs": [],
   "source": [
    "mp_holistic = mp.solutions.holistic # brings the holistic model\n",
    "mp_drawing = mp.solutions.drawing_utils # getting the drawing utilities"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "116daa57-66b3-4566-b164-51ea615cefe2",
   "metadata": {},
   "outputs": [],
   "source": [
    "def mp_detection(image, model):\n",
    "    # when we receive the feed from open cv it is in the BGR format and for mediapipe to be able to make a detection we need to convert it into RGB\n",
    "    image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB) # converts an input image from one color space to another\n",
    "    image.flags.writeable = False # making the writeable status of our image to false\n",
    "    results = model.process(image) # makes the prediction\n",
    "    image.flags.writeable = True \n",
    "    image = cv2.cvtColor(image, cv2.COLOR_RGB2BGR) # Converting the color back to BGR\n",
    "    return image, results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "839ea1e4-4e43-4c2e-8687-cc6a03c319d1",
   "metadata": {},
   "outputs": [],
   "source": [
    "# this function actually just draws the keypoints on the image based on the results that we get from the mp_detection function that we get\n",
    "def draw_landmarks(image, results):\n",
    "    # each draw landmark function requires a frame, landmarks list and the connection map to be able to plot correct keypoints \n",
    "    mp_drawing.draw_landmarks(image, results.face_landmarks, mp_holistic.FACEMESH_TESSELATION)\n",
    "    mp_drawing.draw_landmarks(image, results.pose_landmarks, mp_holistic.POSE_CONNECTIONS)\n",
    "    mp_drawing.draw_landmarks(image, results.left_hand_landmarks, mp_holistic.HAND_CONNECTIONS)\n",
    "    mp_drawing.draw_landmarks(image, results.right_hand_landmarks, mp_holistic.HAND_CONNECTIONS)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "647ab4f4-e000-4e36-8a3e-c79bb05c8782",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "I0000 00:00:1701559149.478533       1 gl_context.cc:344] GL version: 2.1 (2.1 Metal - 86), renderer: Apple M2\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n"
     ]
    }
   ],
   "source": [
    "capture = cv2.VideoCapture(1) # this line effectively accesses our webcam\n",
    "# set mediapipe model\n",
    "with mp_holistic.Holistic(min_detection_confidence=0.5, min_tracking_confidence=0.5) as holistic:\n",
    "    # loops through the video capture as long as we are still accessing our webcam\n",
    "    while capture.isOpened():\n",
    "      # reads the feed\n",
    "      ret, frame = capture.read()\n",
    "\n",
    "      # make detections  \n",
    "      image, results = mp_detection(frame, holistic)\n",
    "      print(results)\n",
    "\n",
    "      # draw landmarks\n",
    "      draw_landmarks(image, results)\n",
    "\n",
    "      # shows to screen\n",
    "      cv2.imshow('OpenCV Feed', image)\n",
    "    \n",
    "      # it breaks out of our loop if we press 'q' on our keyboard\n",
    "      if cv2.waitKey(10) & 0xFF == ord('q'):\n",
    "        break\n",
    "    capture.release()\n",
    "    cv2.destroyAllWindows()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "c816b55b-12bb-4073-b5ac-29f2eaea7d87",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "468"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(results.face_landmarks.landmark)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "f655f94c-066d-473c-a393-bfdfb5c2baf0",
   "metadata": {},
   "outputs": [],
   "source": [
    "draw_landmarks(frame, results)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "2aea5a99-7b3d-4a54-b7e2-d78126273185",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x2d4ce8150>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(cv2.cvtColor(frame, cv2.COLOR_BGR2RGB))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "90425f54-06c7-4dde-b409-b4691d4d0a09",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "33"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(results.pose_landmarks.landmark)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "98bb7cae-4210-4949-8660-22fefd24859f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.])"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# stores all keypoints in a flattened way which will be passed to the LSTM model\n",
    "# since the result will only have a particular landmark if it was present in the feed at the time camera was capturing our video but if not\n",
    "# then that particular landmark when accessed would give us an error, so to avoid that we are creating 4 different variables for different \n",
    "# landmarks and storing them in a flattened way if they are present else we just make an array of zeros\n",
    "pose_values = np.array([[res.x, res.y, res.z, res.visibility] for res in results.pose_landmarks.landmark]).flatten() if results.pose_landmarks else np.zeros(132)\n",
    "face = np.array([[res.x, res.y, res.z] for res in results.face_landmarks.landmark]).flatten() if results.face_landmarks else np.zeros(1404)\n",
    "lh_values = np.array([[res.x, res.y, res.z] for res in results.left_hand_landmarks.landmark]).flatten() if results.left_hand_landmarks else np.zeros(21*3)\n",
    "rh_values = np.array([[res.x, res.y, res.z] for res in results.right_hand_landmarks.landmark]).flatten() if results.right_hand_landmarks else np.zeros(21*3)\n",
    "rh_values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "c34a03f0-8fa5-43f9-9e51-71046a45a09b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# here we have just made it into a function and return the concatenated results\n",
    "def extract_values(results):\n",
    "    pose_values = np.array([[res.x, res.y, res.z, res.visibility] for res in results.pose_landmarks.landmark]).flatten() if results.pose_landmarks else np.zeros(132)\n",
    "    face_values = np.array([[res.x, res.y, res.z] for res in results.face_landmarks.landmark]).flatten() if results.face_landmarks else np.zeros(1404)\n",
    "    lh_values = np.array([[res.x, res.y, res.z] for res in results.left_hand_landmarks.landmark]).flatten() if results.left_hand_landmarks else np.zeros(21*3)\n",
    "    rh_values = np.array([[res.x, res.y, res.z] for res in results.right_hand_landmarks.landmark]).flatten() if results.right_hand_landmarks else np.zeros(21*3)\n",
    "    return np.concatenate([pose_values, face_values, lh_values, rh_values])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "2f99e3cc-79f3-44c8-a8f1-93728745139c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.50911266,  0.6620363 , -1.17543077, ...,  0.        ,\n",
       "        0.        ,  0.        ])"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result_test = extract_values(results)\n",
    "result_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "d702ef6e-6085-4c81-a5bf-280dc62948b9",
   "metadata": {},
   "outputs": [],
   "source": [
    "np.save('0', result_test) # saving each frame as a numpy array"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "837c120d-b29c-4e55-b496-8a4d13b27c5d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.50911266,  0.6620363 , -1.17543077, ...,  0.        ,\n",
       "        0.        ,  0.        ])"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.load('0.npy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "d08bce5c-7227-4d72-8b17-9b73f1e47187",
   "metadata": {},
   "outputs": [],
   "source": [
    "DATA_PATH = os.path.join('MP_Data') # path for exported data\n",
    "\n",
    "# actions to detect\n",
    "actions = np.array(['hello', 'thanks', 'iloveyou'])\n",
    "\n",
    "# 30 vidoes worth of data\n",
    "no_sequences = 30\n",
    "\n",
    "# videos are going to be 30 frames in length so basically 30*1662 keypoints\n",
    "sequence_length = 30"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "e4771003-3695-45b7-9ced-1f5b0cceceec",
   "metadata": {},
   "outputs": [],
   "source": [
    "# effectively here we create a different folder for each action that is \"hello, \"thankyou\" and \"Iloveyou\"\n",
    "# and inside these folders we will have a folder for each sequence of action i.e since there are 30 sequences we will have subfolders with name \n",
    "# 0-29 in each of these action folders\n",
    "for action in actions:\n",
    "    for sequence in range(no_sequences):\n",
    "        try:\n",
    "            os.makedirs(os.path.join(DATA_PATH, action, str(sequence)))\n",
    "        except:\n",
    "            pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "6f236f5a-b93f-4cb1-bf4b-fd13de7f47c4",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "I0000 00:00:1701559213.601917       1 gl_context.cc:344] GL version: 2.1 (2.1 Metal - 86), renderer: Apple M2\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n"
     ]
    }
   ],
   "source": [
    "capture = cv2.VideoCapture(1) # this line effectively accesses our webcam\n",
    "# set mediapipe model\n",
    "with mp_holistic.Holistic(min_detection_confidence=0.5, min_tracking_confidence=0.5) as holistic:\n",
    "    # loops through the video capture as long as we are still accessing our webcam\n",
    "    for action in actions:\n",
    "        # looping through the sequences/videos\n",
    "        for sequence in range(no_sequences):\n",
    "            # looping through video length/sequence length\n",
    "            for frame_num in range(sequence_length):\n",
    "                \n",
    "                ret, frame = capture.read()\n",
    "\n",
    "                # make detections  \n",
    "                image, results = mp_detection(frame, holistic)\n",
    "                print(results)\n",
    "            \n",
    "                # draw landmarks\n",
    "                draw_landmarks(image, results)\n",
    "\n",
    "                # apply collection logic\n",
    "                if frame_num == 0:\n",
    "                    cv2.putText(image, 'STARTING COLECTION', (120, 200), cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 255, 0), 4, cv2.LINE_AA)\n",
    "                    cv2.putText(image, 'Collecting frames for {} video number {}'.format(action, sequence), (15, 12), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 255), 1, cv2.LINE_AA)\n",
    "                    cv2.waitKey(2000)\n",
    "                else:\n",
    "                    cv2.putText(image, 'Collecting frames for {} video number {}'.format(action, sequence), (15, 12), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 255), 1, cv2.LINE_AA)\n",
    "\n",
    "                # we extract the keypoints using the function we created above and save each frame as a numpy array in the respective folder\n",
    "                # so for eg. for action \"hello\" in sequence \"0\" i.e folder 0 we will have 30 numpy arrays as .npy files from 0-29\n",
    "                # SO ESSENTIALLY 30 SEQUENCES (VIDEOS) AND 30 FRAMES PER SEQUENCE (VIDEO)\n",
    "                keypoints = extract_values(results)\n",
    "                npy_path = os.path.join(DATA_PATH, action, str(sequence), str(frame_num))\n",
    "                np.save(npy_path, keypoints)\n",
    "                # shows to screen\n",
    "                cv2.imshow('OpenCV Feed', image)\n",
    "      \n",
    "    \n",
    "                # it breaks out of our loop if we press 'q' on our keyboard\n",
    "                if cv2.waitKey(10) & 0xFF == ord('q'):\n",
    "                    break\n",
    "    capture.release()\n",
    "    cv2.destroyAllWindows()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "62eef9e2-cac9-47c5-aed8-b56e10b35b85",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: scikit-learn in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (1.3.2)\n",
      "Requirement already satisfied: numpy<2.0,>=1.17.3 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from scikit-learn) (1.26.2)\n",
      "Requirement already satisfied: scipy>=1.5.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from scikit-learn) (1.11.4)\n",
      "Requirement already satisfied: joblib>=1.1.1 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from scikit-learn) (1.3.2)\n",
      "Requirement already satisfied: threadpoolctl>=2.0.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from scikit-learn) (3.2.0)\n",
      "Requirement already satisfied: tensorflow in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (2.15.0)\n",
      "Requirement already satisfied: tensorflow-macos==2.15.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow) (2.15.0)\n",
      "Requirement already satisfied: absl-py>=1.0.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (2.0.0)\n",
      "Requirement already satisfied: astunparse>=1.6.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (1.6.3)\n",
      "Requirement already satisfied: flatbuffers>=23.5.26 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (23.5.26)\n",
      "Requirement already satisfied: gast!=0.5.0,!=0.5.1,!=0.5.2,>=0.2.1 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (0.5.4)\n",
      "Requirement already satisfied: google-pasta>=0.1.1 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (0.2.0)\n",
      "Requirement already satisfied: h5py>=2.9.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (3.10.0)\n",
      "Requirement already satisfied: libclang>=13.0.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (16.0.6)\n",
      "Requirement already satisfied: ml-dtypes~=0.2.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (0.2.0)\n",
      "Requirement already satisfied: numpy<2.0.0,>=1.23.5 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (1.26.2)\n",
      "Requirement already satisfied: opt-einsum>=2.3.2 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (3.3.0)\n",
      "Requirement already satisfied: packaging in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (23.2)\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 /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (3.20.3)\n",
      "Requirement already satisfied: setuptools in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (65.5.0)\n",
      "Requirement already satisfied: six>=1.12.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (1.16.0)\n",
      "Requirement already satisfied: termcolor>=1.1.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (2.3.0)\n",
      "Requirement already satisfied: typing-extensions>=3.6.6 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (4.8.0)\n",
      "Requirement already satisfied: wrapt<1.15,>=1.11.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (1.14.1)\n",
      "Requirement already satisfied: tensorflow-io-gcs-filesystem>=0.23.1 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (0.34.0)\n",
      "Requirement already satisfied: grpcio<2.0,>=1.24.3 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (1.59.3)\n",
      "Requirement already satisfied: tensorboard<2.16,>=2.15 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (2.15.1)\n",
      "Requirement already satisfied: tensorflow-estimator<2.16,>=2.15.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (2.15.0)\n",
      "Requirement already satisfied: keras<2.16,>=2.15.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorflow-macos==2.15.0->tensorflow) (2.15.0)\n",
      "Requirement already satisfied: wheel<1.0,>=0.23.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from astunparse>=1.6.0->tensorflow-macos==2.15.0->tensorflow) (0.42.0)\n",
      "Requirement already satisfied: google-auth<3,>=1.6.3 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorboard<2.16,>=2.15->tensorflow-macos==2.15.0->tensorflow) (2.23.4)\n",
      "Requirement already satisfied: google-auth-oauthlib<2,>=0.5 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorboard<2.16,>=2.15->tensorflow-macos==2.15.0->tensorflow) (1.1.0)\n",
      "Requirement already satisfied: markdown>=2.6.8 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorboard<2.16,>=2.15->tensorflow-macos==2.15.0->tensorflow) (3.5.1)\n",
      "Requirement already satisfied: requests<3,>=2.21.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorboard<2.16,>=2.15->tensorflow-macos==2.15.0->tensorflow) (2.31.0)\n",
      "Requirement already satisfied: tensorboard-data-server<0.8.0,>=0.7.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorboard<2.16,>=2.15->tensorflow-macos==2.15.0->tensorflow) (0.7.2)\n",
      "Requirement already satisfied: werkzeug>=1.0.1 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from tensorboard<2.16,>=2.15->tensorflow-macos==2.15.0->tensorflow) (3.0.1)\n",
      "Requirement already satisfied: cachetools<6.0,>=2.0.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from google-auth<3,>=1.6.3->tensorboard<2.16,>=2.15->tensorflow-macos==2.15.0->tensorflow) (5.3.2)\n",
      "Requirement already satisfied: pyasn1-modules>=0.2.1 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from google-auth<3,>=1.6.3->tensorboard<2.16,>=2.15->tensorflow-macos==2.15.0->tensorflow) (0.3.0)\n",
      "Requirement already satisfied: rsa<5,>=3.1.4 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from google-auth<3,>=1.6.3->tensorboard<2.16,>=2.15->tensorflow-macos==2.15.0->tensorflow) (4.9)\n",
      "Requirement already satisfied: requests-oauthlib>=0.7.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from google-auth-oauthlib<2,>=0.5->tensorboard<2.16,>=2.15->tensorflow-macos==2.15.0->tensorflow) (1.3.1)\n",
      "Requirement already satisfied: charset-normalizer<4,>=2 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from requests<3,>=2.21.0->tensorboard<2.16,>=2.15->tensorflow-macos==2.15.0->tensorflow) (3.3.2)\n",
      "Requirement already satisfied: idna<4,>=2.5 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from requests<3,>=2.21.0->tensorboard<2.16,>=2.15->tensorflow-macos==2.15.0->tensorflow) (3.6)\n",
      "Requirement already satisfied: urllib3<3,>=1.21.1 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from requests<3,>=2.21.0->tensorboard<2.16,>=2.15->tensorflow-macos==2.15.0->tensorflow) (2.1.0)\n",
      "Requirement already satisfied: certifi>=2017.4.17 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from requests<3,>=2.21.0->tensorboard<2.16,>=2.15->tensorflow-macos==2.15.0->tensorflow) (2023.11.17)\n",
      "Requirement already satisfied: MarkupSafe>=2.1.1 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from werkzeug>=1.0.1->tensorboard<2.16,>=2.15->tensorflow-macos==2.15.0->tensorflow) (2.1.3)\n",
      "Requirement already satisfied: pyasn1<0.6.0,>=0.4.6 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from pyasn1-modules>=0.2.1->google-auth<3,>=1.6.3->tensorboard<2.16,>=2.15->tensorflow-macos==2.15.0->tensorflow) (0.5.1)\n",
      "Requirement already satisfied: oauthlib>=3.0.0 in /Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<2,>=0.5->tensorboard<2.16,>=2.15->tensorflow-macos==2.15.0->tensorflow) (3.2.2)\n"
     ]
    }
   ],
   "source": [
    "!pip3 install scikit-learn\n",
    "!pip3 install tensorflow\n",
    "from sklearn.model_selection import train_test_split\n",
    "from tensorflow.keras.utils import to_categorical"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "57f2fc06-eeda-4a23-a4d5-26cbd295444a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'hello': 0, 'thanks': 1, 'iloveyou': 2}"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "label_map = {label:num for num, label in enumerate(actions)}\n",
    "label_map"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "bcd4d877-43d5-4d3e-9e16-1e1d326d5022",
   "metadata": {},
   "outputs": [],
   "source": [
    "# here we create two new lists to be able to work on our data better\n",
    "# sequences will contain data for 90 videos that are 30 frames each where each one has 1662 keypoints\n",
    "# labels will have the label for the corresponding action\n",
    "sequences, labels = [], []\n",
    "for action in actions:\n",
    "    for sequence in range(no_sequences):\n",
    "        window = []\n",
    "        for frame_num in range(sequence_length):\n",
    "            res = np.load(os.path.join(DATA_PATH, action, str(sequence), \"{}.npy\".format(frame_num)))\n",
    "            window.append(res)\n",
    "        sequences.append(window)\n",
    "        labels.append(label_map[action])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "404ce6de-2c04-465b-8156-4203c7abf800",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(90, 30, 1662)"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.array(sequences).shape # output : 90 videos, 30 frames each and each one has 1662 keypoints"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "48c7ea4d-caa0-4a28-9d07-5242fc257c01",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(90,)"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.array(labels).shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "ef19dd64-046f-42e1-9a95-e85ac2511fb4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(90, 30, 1662)"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = np.array(sequences)\n",
    "X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "69ad0730-362b-4b77-bdca-9cf560f46c35",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [1, 0, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 1, 0],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1],\n",
       "       [0, 0, 1]])"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# here we one hot encode our labels which is basically a list of numbers to a suitable format\n",
    "y = to_categorical(labels).astype(int)\n",
    "y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "c8e1eb45-ddaf-48ea-9771-fd6fb34ff749",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.05)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "0e4f444b-c85e-4247-8c4f-b43b64eaa836",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(85, 30, 1662)"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "fcafe1b3-d884-4cd6-b02a-c5bc028c1767",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import Sequential\n",
    "# lSTM gives us the temporal component and allows us to build our neural network which essentially allows us to perform action detection\n",
    "# also importing dense which is a normal fully connected layer\n",
    "from tensorflow.keras.layers import LSTM, Dense\n",
    "from tensorflow.keras.callbacks import TensorBoard"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "4d073c9c-d181-439a-9803-5955196d1a88",
   "metadata": {},
   "outputs": [],
   "source": [
    "# this is just part of the development in which we made use of the tensor board web app to monitor our neural network\n",
    "log_dir = os.path.join('Logs')\n",
    "tb_callback = TensorBoard(log_dir=log_dir)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "c7760a1a-cd98-4a36-8d69-7fb277d5d7ab",
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "# 3 LSTM layers and 3 Dense layers\n",
    "# when tensorflow with an LSTM layer, if we stack them together we need to return the sequences as the next layer will need them\n",
    "# and the shape signifies 30 frames and 1662 keypoints\n",
    "model.add(LSTM(64, return_sequences=True, activation='relu', input_shape=(30, 1662)))\n",
    "model.add(LSTM(128, return_sequences=True, activation='relu'))\n",
    "model.add(LSTM(64, return_sequences=False, activation='relu'))\n",
    "model.add(Dense(64, activation='relu'))\n",
    "model.add(Dense(32, activation='relu'))\n",
    "model.add(Dense(actions.shape[0], activation='softmax')) # softmax - returns values with probablity between 0-1 and their sum adds up to one\n",
    "# the highest probability amongst them will signify which action it most likely is\n",
    "\n",
    "# Reason behind using Mediapipe with LSTM\n",
    "# During research we found that the current state of the art models use CNN layers followed by LSTM layers\n",
    "# But on going down this road we found that the model gave extremely poor accuracy and wont be useful as we were trying to predict actions in real time\n",
    "# So main reasons for doing this was 1. we needed less data to produce a highly accurate model 2. this one would be faster to train given that\n",
    "# we are doing this project on our computers 3. it would thereby give faster detections "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "6e8844d6-c2ca-4a51-8f79-0de8459c18db",
   "metadata": {},
   "outputs": [],
   "source": [
    "# we now go ahead and compile our model\n",
    "# we use a loss function called categorical_crossentropy as this is a multiclass classification problem\n",
    "model.compile(optimizer='Adam', loss='categorical_crossentropy', metrics=['categorical_accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "ba6ce63e-4dcc-4463-b6dd-2bfa0e603c0e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/2000\n",
      "3/3 [==============================] - 2s 26ms/step - loss: 7.0560 - categorical_accuracy: 0.3412\n",
      "Epoch 2/2000\n",
      "3/3 [==============================] - 0s 29ms/step - loss: 18.3076 - categorical_accuracy: 0.3765\n",
      "Epoch 3/2000\n",
      "3/3 [==============================] - 0s 29ms/step - loss: 42.9183 - categorical_accuracy: 0.2588\n",
      "Epoch 4/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 103.6959 - categorical_accuracy: 0.4118\n",
      "Epoch 5/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 115.9782 - categorical_accuracy: 0.2941\n",
      "Epoch 6/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 174.9014 - categorical_accuracy: 0.3176\n",
      "Epoch 7/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 87.6973 - categorical_accuracy: 0.3647\n",
      "Epoch 8/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 58.5829 - categorical_accuracy: 0.3882\n",
      "Epoch 9/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 144.9060 - categorical_accuracy: 0.3059\n",
      "Epoch 10/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 149.4675 - categorical_accuracy: 0.2353\n",
      "Epoch 11/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 70.1748 - categorical_accuracy: 0.4118\n",
      "Epoch 12/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 41.1014 - categorical_accuracy: 0.4118\n",
      "Epoch 13/2000\n",
      "3/3 [==============================] - 0s 31ms/step - loss: 103.2512 - categorical_accuracy: 0.4235\n",
      "Epoch 14/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 108.7631 - categorical_accuracy: 0.3647\n",
      "Epoch 15/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 158.9485 - categorical_accuracy: 0.3647\n",
      "Epoch 16/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 234.6702 - categorical_accuracy: 0.2706\n",
      "Epoch 17/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 273.4725 - categorical_accuracy: 0.3412\n",
      "Epoch 18/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 426.4129 - categorical_accuracy: 0.3176\n",
      "Epoch 19/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 278.3083 - categorical_accuracy: 0.3176\n",
      "Epoch 20/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 616.3901 - categorical_accuracy: 0.4000\n",
      "Epoch 21/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 258.1302 - categorical_accuracy: 0.3059\n",
      "Epoch 22/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 292.9443 - categorical_accuracy: 0.3412\n",
      "Epoch 23/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 165.4491 - categorical_accuracy: 0.3647\n",
      "Epoch 24/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 153.4972 - categorical_accuracy: 0.3176\n",
      "Epoch 25/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 516.6609 - categorical_accuracy: 0.3059\n",
      "Epoch 26/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 301.0374 - categorical_accuracy: 0.3176\n",
      "Epoch 27/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 144.8918 - categorical_accuracy: 0.2353\n",
      "Epoch 28/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 162.5491 - categorical_accuracy: 0.4235\n",
      "Epoch 29/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 95.3326 - categorical_accuracy: 0.3059\n",
      "Epoch 30/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 102.7709 - categorical_accuracy: 0.3176\n",
      "Epoch 31/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 112.6616 - categorical_accuracy: 0.2824\n",
      "Epoch 32/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 164.3832 - categorical_accuracy: 0.3059\n",
      "Epoch 33/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 134.0790 - categorical_accuracy: 0.3647\n",
      "Epoch 34/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 23.4930 - categorical_accuracy: 0.2706\n",
      "Epoch 35/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 16.0163 - categorical_accuracy: 0.3529\n",
      "Epoch 36/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 61.2990 - categorical_accuracy: 0.2941\n",
      "Epoch 37/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 60.1881 - categorical_accuracy: 0.3412\n",
      "Epoch 38/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 93.4828 - categorical_accuracy: 0.2941\n",
      "Epoch 39/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 83.2673 - categorical_accuracy: 0.3529\n",
      "Epoch 40/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 51.7259 - categorical_accuracy: 0.3647\n",
      "Epoch 41/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 96.2778 - categorical_accuracy: 0.3529\n",
      "Epoch 42/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 213.4282 - categorical_accuracy: 0.3647\n",
      "Epoch 43/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 216.3444 - categorical_accuracy: 0.2706\n",
      "Epoch 44/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 52.9689 - categorical_accuracy: 0.2941\n",
      "Epoch 45/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 105.4374 - categorical_accuracy: 0.3882\n",
      "Epoch 46/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 153.1924 - categorical_accuracy: 0.2941\n",
      "Epoch 47/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 273.1635 - categorical_accuracy: 0.3412\n",
      "Epoch 48/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 329.6787 - categorical_accuracy: 0.2588\n",
      "Epoch 49/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 492.3907 - categorical_accuracy: 0.3765\n",
      "Epoch 50/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 393.6571 - categorical_accuracy: 0.3647\n",
      "Epoch 51/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 424.7514 - categorical_accuracy: 0.4118\n",
      "Epoch 52/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 500.4011 - categorical_accuracy: 0.2824\n",
      "Epoch 53/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 902.2084 - categorical_accuracy: 0.3412\n",
      "Epoch 54/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 2186.5381 - categorical_accuracy: 0.4000\n",
      "Epoch 55/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1149.9041 - categorical_accuracy: 0.4235\n",
      "Epoch 56/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1634.1129 - categorical_accuracy: 0.3059\n",
      "Epoch 57/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 3130.8794 - categorical_accuracy: 0.3294\n",
      "Epoch 58/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 5145.6987 - categorical_accuracy: 0.3882\n",
      "Epoch 59/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 12431.2783 - categorical_accuracy: 0.3412\n",
      "Epoch 60/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 8967.0137 - categorical_accuracy: 0.3176\n",
      "Epoch 61/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 9273.4600 - categorical_accuracy: 0.2471\n",
      "Epoch 62/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 7359.9097 - categorical_accuracy: 0.3412\n",
      "Epoch 63/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 2564.2236 - categorical_accuracy: 0.3294\n",
      "Epoch 64/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 3359.9155 - categorical_accuracy: 0.3176\n",
      "Epoch 65/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1257.7609 - categorical_accuracy: 0.2824\n",
      "Epoch 66/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1413.0988 - categorical_accuracy: 0.3412\n",
      "Epoch 67/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1137.9507 - categorical_accuracy: 0.3412\n",
      "Epoch 68/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 692.8610 - categorical_accuracy: 0.3412\n",
      "Epoch 69/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 593.0701 - categorical_accuracy: 0.3176\n",
      "Epoch 70/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 508.5493 - categorical_accuracy: 0.3412\n",
      "Epoch 71/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 303.2937 - categorical_accuracy: 0.2824\n",
      "Epoch 72/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 139.1952 - categorical_accuracy: 0.3765\n",
      "Epoch 73/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 134.7635 - categorical_accuracy: 0.3294\n",
      "Epoch 74/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 148.9759 - categorical_accuracy: 0.3765\n",
      "Epoch 75/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 104.3925 - categorical_accuracy: 0.4118\n",
      "Epoch 76/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 284.0808 - categorical_accuracy: 0.3294\n",
      "Epoch 77/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 296.9342 - categorical_accuracy: 0.3059\n",
      "Epoch 78/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 161.2019 - categorical_accuracy: 0.3059\n",
      "Epoch 79/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 990.2282 - categorical_accuracy: 0.2941\n",
      "Epoch 80/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 476.4444 - categorical_accuracy: 0.3412\n",
      "Epoch 81/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 152.5647 - categorical_accuracy: 0.3647\n",
      "Epoch 82/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 62.6641 - categorical_accuracy: 0.2824\n",
      "Epoch 83/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 56.7778 - categorical_accuracy: 0.3529\n",
      "Epoch 84/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 29.9703 - categorical_accuracy: 0.3176\n",
      "Epoch 85/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 186.3420 - categorical_accuracy: 0.3647\n",
      "Epoch 86/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 155.0587 - categorical_accuracy: 0.3176\n",
      "Epoch 87/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 45.1309 - categorical_accuracy: 0.3765\n",
      "Epoch 88/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 76.4569 - categorical_accuracy: 0.3882\n",
      "Epoch 89/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 118.5712 - categorical_accuracy: 0.3412\n",
      "Epoch 90/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 361.9853 - categorical_accuracy: 0.3176\n",
      "Epoch 91/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 116.9058 - categorical_accuracy: 0.3412\n",
      "Epoch 92/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 143.9207 - categorical_accuracy: 0.3294\n",
      "Epoch 93/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 77.6913 - categorical_accuracy: 0.2941\n",
      "Epoch 94/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 243.8011 - categorical_accuracy: 0.3412\n",
      "Epoch 95/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 41.9528 - categorical_accuracy: 0.2941\n",
      "Epoch 96/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 81.3274 - categorical_accuracy: 0.2588\n",
      "Epoch 97/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 69.0206 - categorical_accuracy: 0.2471\n",
      "Epoch 98/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 132.9934 - categorical_accuracy: 0.2824\n",
      "Epoch 99/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 89.1113 - categorical_accuracy: 0.3882\n",
      "Epoch 100/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 83.7921 - categorical_accuracy: 0.4118\n",
      "Epoch 101/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 47.2715 - categorical_accuracy: 0.3765\n",
      "Epoch 102/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 107.7991 - categorical_accuracy: 0.3412\n",
      "Epoch 103/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 98.1364 - categorical_accuracy: 0.3412\n",
      "Epoch 104/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 51.6771 - categorical_accuracy: 0.2941\n",
      "Epoch 105/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 90.0554 - categorical_accuracy: 0.3765\n",
      "Epoch 106/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 71.8619 - categorical_accuracy: 0.2588\n",
      "Epoch 107/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 33.0223 - categorical_accuracy: 0.3412\n",
      "Epoch 108/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 30.2408 - categorical_accuracy: 0.4706\n",
      "Epoch 109/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 60.6913 - categorical_accuracy: 0.2000\n",
      "Epoch 110/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 28.2937 - categorical_accuracy: 0.3176\n",
      "Epoch 111/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 34.4506 - categorical_accuracy: 0.2471\n",
      "Epoch 112/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 37.8304 - categorical_accuracy: 0.3059\n",
      "Epoch 113/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 52.0024 - categorical_accuracy: 0.2824\n",
      "Epoch 114/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 30.4440 - categorical_accuracy: 0.3882\n",
      "Epoch 115/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 22.5877 - categorical_accuracy: 0.3529\n",
      "Epoch 116/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 61.9297 - categorical_accuracy: 0.3412\n",
      "Epoch 117/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 55.3389 - categorical_accuracy: 0.2824\n",
      "Epoch 118/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 22.6703 - categorical_accuracy: 0.3059\n",
      "Epoch 119/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 17.1709 - categorical_accuracy: 0.3882\n",
      "Epoch 120/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 29.3451 - categorical_accuracy: 0.2588\n",
      "Epoch 121/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 18.5217 - categorical_accuracy: 0.4118\n",
      "Epoch 122/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 37.7661 - categorical_accuracy: 0.3412\n",
      "Epoch 123/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 32.2148 - categorical_accuracy: 0.3647\n",
      "Epoch 124/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 27.1224 - categorical_accuracy: 0.3647\n",
      "Epoch 125/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 49.0206 - categorical_accuracy: 0.3176\n",
      "Epoch 126/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 38.2867 - categorical_accuracy: 0.2941\n",
      "Epoch 127/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 34.2654 - categorical_accuracy: 0.3176\n",
      "Epoch 128/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 42.0596 - categorical_accuracy: 0.3412\n",
      "Epoch 129/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 22.7873 - categorical_accuracy: 0.3176\n",
      "Epoch 130/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 30.2603 - categorical_accuracy: 0.3294\n",
      "Epoch 131/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 18.5120 - categorical_accuracy: 0.4000\n",
      "Epoch 132/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 44.3866 - categorical_accuracy: 0.3529\n",
      "Epoch 133/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 54.0364 - categorical_accuracy: 0.2824\n",
      "Epoch 134/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 27.3491 - categorical_accuracy: 0.2941\n",
      "Epoch 135/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 22.9925 - categorical_accuracy: 0.3294\n",
      "Epoch 136/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 42.7826 - categorical_accuracy: 0.3412\n",
      "Epoch 137/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 65.7174 - categorical_accuracy: 0.2471\n",
      "Epoch 138/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 32.7514 - categorical_accuracy: 0.3529\n",
      "Epoch 139/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 47.2920 - categorical_accuracy: 0.3529\n",
      "Epoch 140/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 39.3452 - categorical_accuracy: 0.3765\n",
      "Epoch 141/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 34.3625 - categorical_accuracy: 0.3176\n",
      "Epoch 142/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 19.0845 - categorical_accuracy: 0.3412\n",
      "Epoch 143/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 25.2754 - categorical_accuracy: 0.3294\n",
      "Epoch 144/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 30.9778 - categorical_accuracy: 0.2824\n",
      "Epoch 145/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 35.3865 - categorical_accuracy: 0.3176\n",
      "Epoch 146/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 48.1447 - categorical_accuracy: 0.3176\n",
      "Epoch 147/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 41.6610 - categorical_accuracy: 0.3294\n",
      "Epoch 148/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 60.5712 - categorical_accuracy: 0.3176\n",
      "Epoch 149/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 23.2945 - categorical_accuracy: 0.3882\n",
      "Epoch 150/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 45.5088 - categorical_accuracy: 0.4000\n",
      "Epoch 151/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 30.4175 - categorical_accuracy: 0.4471\n",
      "Epoch 152/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 20.9255 - categorical_accuracy: 0.4118\n",
      "Epoch 153/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 41.5725 - categorical_accuracy: 0.3647\n",
      "Epoch 154/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 20.7798 - categorical_accuracy: 0.4235\n",
      "Epoch 155/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 44.0436 - categorical_accuracy: 0.3412\n",
      "Epoch 156/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 21.0784 - categorical_accuracy: 0.4588\n",
      "Epoch 157/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 28.5118 - categorical_accuracy: 0.4118\n",
      "Epoch 158/2000\n",
      "3/3 [==============================] - 0s 104ms/step - loss: 12.7817 - categorical_accuracy: 0.3176\n",
      "Epoch 159/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 8.4490 - categorical_accuracy: 0.4353\n",
      "Epoch 160/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 10.3906 - categorical_accuracy: 0.4471\n",
      "Epoch 161/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 9.9034 - categorical_accuracy: 0.5765\n",
      "Epoch 162/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 12.9864 - categorical_accuracy: 0.5176\n",
      "Epoch 163/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 9.1659 - categorical_accuracy: 0.4353\n",
      "Epoch 164/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 16.3320 - categorical_accuracy: 0.6118\n",
      "Epoch 165/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 8.6959 - categorical_accuracy: 0.5647\n",
      "Epoch 166/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 14.6832 - categorical_accuracy: 0.3765\n",
      "Epoch 167/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 10.4539 - categorical_accuracy: 0.4118\n",
      "Epoch 168/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 6.9715 - categorical_accuracy: 0.5294\n",
      "Epoch 169/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 4.8324 - categorical_accuracy: 0.4588\n",
      "Epoch 170/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 5.9126 - categorical_accuracy: 0.5059\n",
      "Epoch 171/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 11.4599 - categorical_accuracy: 0.5765\n",
      "Epoch 172/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 5.8332 - categorical_accuracy: 0.4706\n",
      "Epoch 173/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 11.4158 - categorical_accuracy: 0.5647\n",
      "Epoch 174/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 18.0153 - categorical_accuracy: 0.2824\n",
      "Epoch 175/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 16.6163 - categorical_accuracy: 0.4588\n",
      "Epoch 176/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 19.3762 - categorical_accuracy: 0.4588\n",
      "Epoch 177/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 10.1981 - categorical_accuracy: 0.5647\n",
      "Epoch 178/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 11.4766 - categorical_accuracy: 0.6353\n",
      "Epoch 179/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 5.3876 - categorical_accuracy: 0.6706\n",
      "Epoch 180/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 9.0696 - categorical_accuracy: 0.5765\n",
      "Epoch 181/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 8.6040 - categorical_accuracy: 0.4235\n",
      "Epoch 182/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 11.5815 - categorical_accuracy: 0.5294\n",
      "Epoch 183/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 10.9004 - categorical_accuracy: 0.2706\n",
      "Epoch 184/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 6.9249 - categorical_accuracy: 0.4235\n",
      "Epoch 185/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2494 - categorical_accuracy: 0.6353\n",
      "Epoch 186/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 2.6908 - categorical_accuracy: 0.5294\n",
      "Epoch 187/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 2.8712 - categorical_accuracy: 0.6588\n",
      "Epoch 188/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 4.1217 - categorical_accuracy: 0.5529\n",
      "Epoch 189/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 14.4116 - categorical_accuracy: 0.4235\n",
      "Epoch 190/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 25.7566 - categorical_accuracy: 0.3059\n",
      "Epoch 191/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 22.7785 - categorical_accuracy: 0.3765\n",
      "Epoch 192/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 18.9611 - categorical_accuracy: 0.5294\n",
      "Epoch 193/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 5.4589 - categorical_accuracy: 0.6706\n",
      "Epoch 194/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 13.9476 - categorical_accuracy: 0.4471\n",
      "Epoch 195/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 10.2282 - categorical_accuracy: 0.6353\n",
      "Epoch 196/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 9.9231 - categorical_accuracy: 0.5412\n",
      "Epoch 197/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 5.3846 - categorical_accuracy: 0.5647\n",
      "Epoch 198/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 4.8233 - categorical_accuracy: 0.6941\n",
      "Epoch 199/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 7.0785 - categorical_accuracy: 0.6706\n",
      "Epoch 200/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 9.6681 - categorical_accuracy: 0.6353\n",
      "Epoch 201/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 5.2732 - categorical_accuracy: 0.5412\n",
      "Epoch 202/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 8.3229 - categorical_accuracy: 0.4235\n",
      "Epoch 203/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 4.2628 - categorical_accuracy: 0.6353\n",
      "Epoch 204/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 5.3976 - categorical_accuracy: 0.6000\n",
      "Epoch 205/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 5.0024 - categorical_accuracy: 0.6706\n",
      "Epoch 206/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 9.1907 - categorical_accuracy: 0.6353\n",
      "Epoch 207/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 9.2147 - categorical_accuracy: 0.6235\n",
      "Epoch 208/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 9.8820 - categorical_accuracy: 0.5882\n",
      "Epoch 209/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 3.2657 - categorical_accuracy: 0.6353\n",
      "Epoch 210/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 2.1862 - categorical_accuracy: 0.5529\n",
      "Epoch 211/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.4373 - categorical_accuracy: 0.6353\n",
      "Epoch 212/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 2.5246 - categorical_accuracy: 0.6353\n",
      "Epoch 213/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.9367 - categorical_accuracy: 0.5647\n",
      "Epoch 214/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 2.8138 - categorical_accuracy: 0.6353\n",
      "Epoch 215/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 6.4118 - categorical_accuracy: 0.5059\n",
      "Epoch 216/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 2.7376 - categorical_accuracy: 0.6471\n",
      "Epoch 217/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 3.5278 - categorical_accuracy: 0.6118\n",
      "Epoch 218/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 4.6371 - categorical_accuracy: 0.6471\n",
      "Epoch 219/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 4.4658 - categorical_accuracy: 0.6471\n",
      "Epoch 220/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 2.7163 - categorical_accuracy: 0.6000\n",
      "Epoch 221/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 4.2760 - categorical_accuracy: 0.6471\n",
      "Epoch 222/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 3.1327 - categorical_accuracy: 0.6588\n",
      "Epoch 223/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 3.7825 - categorical_accuracy: 0.6824\n",
      "Epoch 224/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 3.3872 - categorical_accuracy: 0.6235\n",
      "Epoch 225/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 4.8778 - categorical_accuracy: 0.6471\n",
      "Epoch 226/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 6.8246 - categorical_accuracy: 0.6471\n",
      "Epoch 227/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 4.4274 - categorical_accuracy: 0.7059\n",
      "Epoch 228/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 15.7353 - categorical_accuracy: 0.6706\n",
      "Epoch 229/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 6.6588 - categorical_accuracy: 0.6353\n",
      "Epoch 230/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 6.6422 - categorical_accuracy: 0.6118\n",
      "Epoch 231/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 8.1738 - categorical_accuracy: 0.5529\n",
      "Epoch 232/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 6.6378 - categorical_accuracy: 0.6471\n",
      "Epoch 233/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 8.9886 - categorical_accuracy: 0.6706\n",
      "Epoch 234/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 7.2731 - categorical_accuracy: 0.6353\n",
      "Epoch 235/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 7.6407 - categorical_accuracy: 0.5765\n",
      "Epoch 236/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 8.8284 - categorical_accuracy: 0.5529\n",
      "Epoch 237/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 7.0482 - categorical_accuracy: 0.6706\n",
      "Epoch 238/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 6.9461 - categorical_accuracy: 0.6941\n",
      "Epoch 239/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 6.5671 - categorical_accuracy: 0.6235\n",
      "Epoch 240/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 3.9758 - categorical_accuracy: 0.6706\n",
      "Epoch 241/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 3.3479 - categorical_accuracy: 0.6941\n",
      "Epoch 242/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 2.8870 - categorical_accuracy: 0.5059\n",
      "Epoch 243/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 2.6055 - categorical_accuracy: 0.5882\n",
      "Epoch 244/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 4.1440 - categorical_accuracy: 0.6353\n",
      "Epoch 245/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 6.2510 - categorical_accuracy: 0.5412\n",
      "Epoch 246/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 6.1697 - categorical_accuracy: 0.5176\n",
      "Epoch 247/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 6.3092 - categorical_accuracy: 0.6588\n",
      "Epoch 248/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 4.0035 - categorical_accuracy: 0.6471\n",
      "Epoch 249/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 4.5315 - categorical_accuracy: 0.5647\n",
      "Epoch 250/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 5.7474 - categorical_accuracy: 0.5059\n",
      "Epoch 251/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 2.7599 - categorical_accuracy: 0.6824\n",
      "Epoch 252/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 3.1743 - categorical_accuracy: 0.6353\n",
      "Epoch 253/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 6.5424 - categorical_accuracy: 0.6118\n",
      "Epoch 254/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 3.0390 - categorical_accuracy: 0.6235\n",
      "Epoch 255/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 7.9502 - categorical_accuracy: 0.4588\n",
      "Epoch 256/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 2.6942 - categorical_accuracy: 0.7059\n",
      "Epoch 257/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 5.1041 - categorical_accuracy: 0.6118\n",
      "Epoch 258/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 5.1566 - categorical_accuracy: 0.5294\n",
      "Epoch 259/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 6.7355 - categorical_accuracy: 0.6706\n",
      "Epoch 260/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 5.0208 - categorical_accuracy: 0.6353\n",
      "Epoch 261/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 4.8394 - categorical_accuracy: 0.6235\n",
      "Epoch 262/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 4.0415 - categorical_accuracy: 0.5882\n",
      "Epoch 263/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 4.6824 - categorical_accuracy: 0.6706\n",
      "Epoch 264/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 9.5237 - categorical_accuracy: 0.6353\n",
      "Epoch 265/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 9.1060 - categorical_accuracy: 0.6706\n",
      "Epoch 266/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 5.2248 - categorical_accuracy: 0.7294\n",
      "Epoch 267/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 15.2188 - categorical_accuracy: 0.6353\n",
      "Epoch 268/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 11.3541 - categorical_accuracy: 0.5882\n",
      "Epoch 269/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 12.7877 - categorical_accuracy: 0.6588\n",
      "Epoch 270/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 7.9077 - categorical_accuracy: 0.6118\n",
      "Epoch 271/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 6.3384 - categorical_accuracy: 0.5882\n",
      "Epoch 272/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 6.1812 - categorical_accuracy: 0.6471\n",
      "Epoch 273/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 8.8297 - categorical_accuracy: 0.6235\n",
      "Epoch 274/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 3.5785 - categorical_accuracy: 0.6118\n",
      "Epoch 275/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 3.4370 - categorical_accuracy: 0.6235\n",
      "Epoch 276/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 8.0732 - categorical_accuracy: 0.6588\n",
      "Epoch 277/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 6.0609 - categorical_accuracy: 0.6235\n",
      "Epoch 278/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 5.6388 - categorical_accuracy: 0.5412\n",
      "Epoch 279/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 2.4978 - categorical_accuracy: 0.5882\n",
      "Epoch 280/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 2.4766 - categorical_accuracy: 0.5765\n",
      "Epoch 281/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 2.2398 - categorical_accuracy: 0.6588\n",
      "Epoch 282/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 3.6426 - categorical_accuracy: 0.6000\n",
      "Epoch 283/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 2.4135 - categorical_accuracy: 0.6941\n",
      "Epoch 284/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 3.3874 - categorical_accuracy: 0.6706\n",
      "Epoch 285/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 2.8948 - categorical_accuracy: 0.5647\n",
      "Epoch 286/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.5110 - categorical_accuracy: 0.6588\n",
      "Epoch 287/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 2.7177 - categorical_accuracy: 0.6588\n",
      "Epoch 288/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 9.0112 - categorical_accuracy: 0.6235\n",
      "Epoch 289/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 7.1576 - categorical_accuracy: 0.5176\n",
      "Epoch 290/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 4.6238 - categorical_accuracy: 0.6235\n",
      "Epoch 291/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 4.2845 - categorical_accuracy: 0.5647\n",
      "Epoch 292/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 5.6949 - categorical_accuracy: 0.5059\n",
      "Epoch 293/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 2.6111 - categorical_accuracy: 0.6941\n",
      "Epoch 294/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 4.1506 - categorical_accuracy: 0.6824\n",
      "Epoch 295/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 4.4368 - categorical_accuracy: 0.7412\n",
      "Epoch 296/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 5.0074 - categorical_accuracy: 0.6118\n",
      "Epoch 297/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 18.3825 - categorical_accuracy: 0.6353\n",
      "Epoch 298/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 335.6084 - categorical_accuracy: 0.3412\n",
      "Epoch 299/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 312.7125 - categorical_accuracy: 0.3176\n",
      "Epoch 300/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 128.0842 - categorical_accuracy: 0.3412\n",
      "Epoch 301/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 213.1106 - categorical_accuracy: 0.3412\n",
      "Epoch 302/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 177.5598 - categorical_accuracy: 0.3765\n",
      "Epoch 303/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 149.5125 - categorical_accuracy: 0.3176\n",
      "Epoch 304/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 102.2311 - categorical_accuracy: 0.3529\n",
      "Epoch 305/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 67.5707 - categorical_accuracy: 0.3529\n",
      "Epoch 306/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 83.9989 - categorical_accuracy: 0.2706\n",
      "Epoch 307/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 72.5589 - categorical_accuracy: 0.3647\n",
      "Epoch 308/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 53.1550 - categorical_accuracy: 0.3647\n",
      "Epoch 309/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 113.0042 - categorical_accuracy: 0.3412\n",
      "Epoch 310/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 57.8018 - categorical_accuracy: 0.2824\n",
      "Epoch 311/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 85.9874 - categorical_accuracy: 0.3529\n",
      "Epoch 312/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 128.5955 - categorical_accuracy: 0.2588\n",
      "Epoch 313/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 24.8289 - categorical_accuracy: 0.2941\n",
      "Epoch 314/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 19.5953 - categorical_accuracy: 0.3412\n",
      "Epoch 315/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 19.1102 - categorical_accuracy: 0.3176\n",
      "Epoch 316/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 27.3059 - categorical_accuracy: 0.3882\n",
      "Epoch 317/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 43.8559 - categorical_accuracy: 0.3529\n",
      "Epoch 318/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 54.9545 - categorical_accuracy: 0.2941\n",
      "Epoch 319/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 35.0736 - categorical_accuracy: 0.3176\n",
      "Epoch 320/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 33.7465 - categorical_accuracy: 0.3176\n",
      "Epoch 321/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 33.3747 - categorical_accuracy: 0.3294\n",
      "Epoch 322/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 35.3069 - categorical_accuracy: 0.3059\n",
      "Epoch 323/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 19.2512 - categorical_accuracy: 0.3647\n",
      "Epoch 324/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 18.8776 - categorical_accuracy: 0.4118\n",
      "Epoch 325/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 10.3675 - categorical_accuracy: 0.3412\n",
      "Epoch 326/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 11.5816 - categorical_accuracy: 0.4000\n",
      "Epoch 327/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 12.7693 - categorical_accuracy: 0.3529\n",
      "Epoch 328/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 17.2050 - categorical_accuracy: 0.3294\n",
      "Epoch 329/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 17.1587 - categorical_accuracy: 0.2471\n",
      "Epoch 330/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 37.5110 - categorical_accuracy: 0.3059\n",
      "Epoch 331/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 24.6047 - categorical_accuracy: 0.2941\n",
      "Epoch 332/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 40.4947 - categorical_accuracy: 0.4000\n",
      "Epoch 333/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 20.7439 - categorical_accuracy: 0.3882\n",
      "Epoch 334/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 24.8987 - categorical_accuracy: 0.3176\n",
      "Epoch 335/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 4.9566 - categorical_accuracy: 0.3765\n",
      "Epoch 336/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 17.9402 - categorical_accuracy: 0.3412\n",
      "Epoch 337/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 18.2564 - categorical_accuracy: 0.3059\n",
      "Epoch 338/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 8.8753 - categorical_accuracy: 0.3412\n",
      "Epoch 339/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 21.3249 - categorical_accuracy: 0.3412\n",
      "Epoch 340/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 45.6193 - categorical_accuracy: 0.2824\n",
      "Epoch 341/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 39.4768 - categorical_accuracy: 0.3176\n",
      "Epoch 342/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 32.1750 - categorical_accuracy: 0.2824\n",
      "Epoch 343/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 19.7569 - categorical_accuracy: 0.3294\n",
      "Epoch 344/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 17.2178 - categorical_accuracy: 0.3176\n",
      "Epoch 345/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 14.9621 - categorical_accuracy: 0.2941\n",
      "Epoch 346/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 12.6466 - categorical_accuracy: 0.3176\n",
      "Epoch 347/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 16.7752 - categorical_accuracy: 0.4941\n",
      "Epoch 348/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 8.5349 - categorical_accuracy: 0.3882\n",
      "Epoch 349/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 20.5250 - categorical_accuracy: 0.3059\n",
      "Epoch 350/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 15.4389 - categorical_accuracy: 0.3176\n",
      "Epoch 351/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 26.7006 - categorical_accuracy: 0.2471\n",
      "Epoch 352/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 31.7569 - categorical_accuracy: 0.3529\n",
      "Epoch 353/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 15.0473 - categorical_accuracy: 0.3765\n",
      "Epoch 354/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 32.8398 - categorical_accuracy: 0.3412\n",
      "Epoch 355/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 30.7321 - categorical_accuracy: 0.3176\n",
      "Epoch 356/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 41.1209 - categorical_accuracy: 0.3176\n",
      "Epoch 357/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 26.2466 - categorical_accuracy: 0.3765\n",
      "Epoch 358/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 22.1581 - categorical_accuracy: 0.4235\n",
      "Epoch 359/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 33.7270 - categorical_accuracy: 0.3294\n",
      "Epoch 360/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 20.0774 - categorical_accuracy: 0.3529\n",
      "Epoch 361/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 17.6336 - categorical_accuracy: 0.3647\n",
      "Epoch 362/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 14.7828 - categorical_accuracy: 0.3412\n",
      "Epoch 363/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 5.4324 - categorical_accuracy: 0.3647\n",
      "Epoch 364/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 8.0713 - categorical_accuracy: 0.2353\n",
      "Epoch 365/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 16.9555 - categorical_accuracy: 0.3529\n",
      "Epoch 366/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 30.7024 - categorical_accuracy: 0.3412\n",
      "Epoch 367/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 12.5178 - categorical_accuracy: 0.2353\n",
      "Epoch 368/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 12.8258 - categorical_accuracy: 0.4000\n",
      "Epoch 369/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 10.2256 - categorical_accuracy: 0.3294\n",
      "Epoch 370/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 16.9808 - categorical_accuracy: 0.2824\n",
      "Epoch 371/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 14.9854 - categorical_accuracy: 0.2706\n",
      "Epoch 372/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 16.7454 - categorical_accuracy: 0.2353\n",
      "Epoch 373/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 13.8775 - categorical_accuracy: 0.3412\n",
      "Epoch 374/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 13.5142 - categorical_accuracy: 0.3176\n",
      "Epoch 375/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 26.3647 - categorical_accuracy: 0.3412\n",
      "Epoch 376/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 8.0629 - categorical_accuracy: 0.4000\n",
      "Epoch 377/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 12.0640 - categorical_accuracy: 0.2706\n",
      "Epoch 378/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 10.2857 - categorical_accuracy: 0.3529\n",
      "Epoch 379/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 9.9236 - categorical_accuracy: 0.4000\n",
      "Epoch 380/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 7.0187 - categorical_accuracy: 0.3176\n",
      "Epoch 381/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 4.7640 - categorical_accuracy: 0.3294\n",
      "Epoch 382/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 4.4886 - categorical_accuracy: 0.3412\n",
      "Epoch 383/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 6.6511 - categorical_accuracy: 0.3059\n",
      "Epoch 384/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 10.5132 - categorical_accuracy: 0.3176\n",
      "Epoch 385/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 8.5355 - categorical_accuracy: 0.3294\n",
      "Epoch 386/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 7.7405 - categorical_accuracy: 0.3294\n",
      "Epoch 387/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 8.4064 - categorical_accuracy: 0.3412\n",
      "Epoch 388/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 12.6918 - categorical_accuracy: 0.3647\n",
      "Epoch 389/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 15.6179 - categorical_accuracy: 0.2706\n",
      "Epoch 390/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 8.0851 - categorical_accuracy: 0.3529\n",
      "Epoch 391/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 8.3000 - categorical_accuracy: 0.3176\n",
      "Epoch 392/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 6.6808 - categorical_accuracy: 0.3647\n",
      "Epoch 393/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 9.4858 - categorical_accuracy: 0.3765\n",
      "Epoch 394/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 9.9336 - categorical_accuracy: 0.2941\n",
      "Epoch 395/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 6.2100 - categorical_accuracy: 0.3529\n",
      "Epoch 396/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 4.9530 - categorical_accuracy: 0.4000\n",
      "Epoch 397/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 7.9551 - categorical_accuracy: 0.3059\n",
      "Epoch 398/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 4.8131 - categorical_accuracy: 0.3647\n",
      "Epoch 399/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 8.9476 - categorical_accuracy: 0.3765\n",
      "Epoch 400/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 13.2903 - categorical_accuracy: 0.2706\n",
      "Epoch 401/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 15.1600 - categorical_accuracy: 0.3647\n",
      "Epoch 402/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 19.1846 - categorical_accuracy: 0.3059\n",
      "Epoch 403/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 29.2942 - categorical_accuracy: 0.2824\n",
      "Epoch 404/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 26.4678 - categorical_accuracy: 0.3412\n",
      "Epoch 405/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 24.8642 - categorical_accuracy: 0.3412\n",
      "Epoch 406/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 10.6127 - categorical_accuracy: 0.2941\n",
      "Epoch 407/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 16.5267 - categorical_accuracy: 0.3412\n",
      "Epoch 408/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 7.0303 - categorical_accuracy: 0.3294\n",
      "Epoch 409/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 5.0445 - categorical_accuracy: 0.1765\n",
      "Epoch 410/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 2.9165 - categorical_accuracy: 0.3294\n",
      "Epoch 411/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 2.9944 - categorical_accuracy: 0.2706\n",
      "Epoch 412/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 3.7428 - categorical_accuracy: 0.2824\n",
      "Epoch 413/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 3.9603 - categorical_accuracy: 0.3294\n",
      "Epoch 414/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 5.8966 - categorical_accuracy: 0.3294\n",
      "Epoch 415/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 6.6810 - categorical_accuracy: 0.3882\n",
      "Epoch 416/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 7.1641 - categorical_accuracy: 0.4118\n",
      "Epoch 417/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 7.2034 - categorical_accuracy: 0.2824\n",
      "Epoch 418/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 5.9214 - categorical_accuracy: 0.3882\n",
      "Epoch 419/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 3.6848 - categorical_accuracy: 0.2941\n",
      "Epoch 420/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 5.5031 - categorical_accuracy: 0.3529\n",
      "Epoch 421/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 4.5154 - categorical_accuracy: 0.3529\n",
      "Epoch 422/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 5.5364 - categorical_accuracy: 0.4000\n",
      "Epoch 423/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 3.3927 - categorical_accuracy: 0.3059\n",
      "Epoch 424/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 3.0094 - categorical_accuracy: 0.2706\n",
      "Epoch 425/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 4.3992 - categorical_accuracy: 0.3412\n",
      "Epoch 426/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 2.8999 - categorical_accuracy: 0.3059\n",
      "Epoch 427/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 2.8187 - categorical_accuracy: 0.3294\n",
      "Epoch 428/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 3.2684 - categorical_accuracy: 0.3176\n",
      "Epoch 429/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 2.6686 - categorical_accuracy: 0.2824\n",
      "Epoch 430/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 2.8794 - categorical_accuracy: 0.3765\n",
      "Epoch 431/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 3.4044 - categorical_accuracy: 0.2588\n",
      "Epoch 432/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 2.3147 - categorical_accuracy: 0.3529\n",
      "Epoch 433/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 2.2428 - categorical_accuracy: 0.3176\n",
      "Epoch 434/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 2.9567 - categorical_accuracy: 0.2588\n",
      "Epoch 435/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 2.4085 - categorical_accuracy: 0.2588\n",
      "Epoch 436/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 2.5371 - categorical_accuracy: 0.2706\n",
      "Epoch 437/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 2.2150 - categorical_accuracy: 0.3176\n",
      "Epoch 438/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.8292 - categorical_accuracy: 0.2941\n",
      "Epoch 439/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.5976 - categorical_accuracy: 0.3176\n",
      "Epoch 440/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.6700 - categorical_accuracy: 0.3529\n",
      "Epoch 441/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 2.0542 - categorical_accuracy: 0.2000\n",
      "Epoch 442/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 2.2994 - categorical_accuracy: 0.3176\n",
      "Epoch 443/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 2.3766 - categorical_accuracy: 0.3529\n",
      "Epoch 444/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 3.4882 - categorical_accuracy: 0.2941\n",
      "Epoch 445/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 4.2518 - categorical_accuracy: 0.3294\n",
      "Epoch 446/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 4.0857 - categorical_accuracy: 0.2824\n",
      "Epoch 447/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 3.8963 - categorical_accuracy: 0.3059\n",
      "Epoch 448/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 3.6079 - categorical_accuracy: 0.3647\n",
      "Epoch 449/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 3.8580 - categorical_accuracy: 0.3294\n",
      "Epoch 450/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 3.0978 - categorical_accuracy: 0.3294\n",
      "Epoch 451/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 3.6442 - categorical_accuracy: 0.2941\n",
      "Epoch 452/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 3.4101 - categorical_accuracy: 0.2118\n",
      "Epoch 453/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 3.5094 - categorical_accuracy: 0.2706\n",
      "Epoch 454/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 4.1233 - categorical_accuracy: 0.3647\n",
      "Epoch 455/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 5.0277 - categorical_accuracy: 0.2588\n",
      "Epoch 456/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 2.7696 - categorical_accuracy: 0.3882\n",
      "Epoch 457/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 2.3141 - categorical_accuracy: 0.3882\n",
      "Epoch 458/2000\n",
      "3/3 [==============================] - 0s 47ms/step - loss: 2.5786 - categorical_accuracy: 0.3412\n",
      "Epoch 459/2000\n",
      "3/3 [==============================] - 0s 53ms/step - loss: 2.2739 - categorical_accuracy: 0.3882\n",
      "Epoch 460/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 2.7577 - categorical_accuracy: 0.3882\n",
      "Epoch 461/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 4.3498 - categorical_accuracy: 0.2824\n",
      "Epoch 462/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 5.9434 - categorical_accuracy: 0.3412\n",
      "Epoch 463/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 4.1359 - categorical_accuracy: 0.4000\n",
      "Epoch 464/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 5.7933 - categorical_accuracy: 0.2824\n",
      "Epoch 465/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 3.5183 - categorical_accuracy: 0.3294\n",
      "Epoch 466/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 2.4589 - categorical_accuracy: 0.4235\n",
      "Epoch 467/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 4.7646 - categorical_accuracy: 0.2706\n",
      "Epoch 468/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 2.8322 - categorical_accuracy: 0.2588\n",
      "Epoch 469/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 2.0615 - categorical_accuracy: 0.2824\n",
      "Epoch 470/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.2514 - categorical_accuracy: 0.2706\n",
      "Epoch 471/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.4804 - categorical_accuracy: 0.2706\n",
      "Epoch 472/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.5183 - categorical_accuracy: 0.3059\n",
      "Epoch 473/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 2.0795 - categorical_accuracy: 0.3294\n",
      "Epoch 474/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 2.0835 - categorical_accuracy: 0.3294\n",
      "Epoch 475/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.6649 - categorical_accuracy: 0.2824\n",
      "Epoch 476/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.8458 - categorical_accuracy: 0.3412\n",
      "Epoch 477/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 2.5224 - categorical_accuracy: 0.3529\n",
      "Epoch 478/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 2.6321 - categorical_accuracy: 0.3059\n",
      "Epoch 479/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 3.5596 - categorical_accuracy: 0.2941\n",
      "Epoch 480/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.8139 - categorical_accuracy: 0.2824\n",
      "Epoch 481/2000\n",
      "3/3 [==============================] - 0s 47ms/step - loss: 2.0576 - categorical_accuracy: 0.2471\n",
      "Epoch 482/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.7209 - categorical_accuracy: 0.3647\n",
      "Epoch 483/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 2.1856 - categorical_accuracy: 0.3412\n",
      "Epoch 484/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.4252 - categorical_accuracy: 0.3529\n",
      "Epoch 485/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.7467 - categorical_accuracy: 0.3059\n",
      "Epoch 486/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 3.1395 - categorical_accuracy: 0.3176\n",
      "Epoch 487/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 2.6774 - categorical_accuracy: 0.3765\n",
      "Epoch 488/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 2.2863 - categorical_accuracy: 0.3647\n",
      "Epoch 489/2000\n",
      "3/3 [==============================] - 0s 47ms/step - loss: 1.6829 - categorical_accuracy: 0.3412\n",
      "Epoch 490/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 2.7747 - categorical_accuracy: 0.2588\n",
      "Epoch 491/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 3.1980 - categorical_accuracy: 0.3529\n",
      "Epoch 492/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 2.7984 - categorical_accuracy: 0.2941\n",
      "Epoch 493/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 2.0162 - categorical_accuracy: 0.3412\n",
      "Epoch 494/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.6269 - categorical_accuracy: 0.3529\n",
      "Epoch 495/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.5866 - categorical_accuracy: 0.2706\n",
      "Epoch 496/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 2.1030 - categorical_accuracy: 0.3059\n",
      "Epoch 497/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 2.9900 - categorical_accuracy: 0.2353\n",
      "Epoch 498/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 2.7472 - categorical_accuracy: 0.2588\n",
      "Epoch 499/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 2.5605 - categorical_accuracy: 0.3294\n",
      "Epoch 500/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 2.6379 - categorical_accuracy: 0.2706\n",
      "Epoch 501/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 2.9528 - categorical_accuracy: 0.4118\n",
      "Epoch 502/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 3.5494 - categorical_accuracy: 0.3412\n",
      "Epoch 503/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 4.2668 - categorical_accuracy: 0.2588\n",
      "Epoch 504/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 3.5666 - categorical_accuracy: 0.3176\n",
      "Epoch 505/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 5.5038 - categorical_accuracy: 0.3412\n",
      "Epoch 506/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 5.7155 - categorical_accuracy: 0.2941\n",
      "Epoch 507/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 6.2030 - categorical_accuracy: 0.2706\n",
      "Epoch 508/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 4.8790 - categorical_accuracy: 0.2706\n",
      "Epoch 509/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 4.7141 - categorical_accuracy: 0.3176\n",
      "Epoch 510/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 3.4262 - categorical_accuracy: 0.3882\n",
      "Epoch 511/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 2.4706 - categorical_accuracy: 0.3882\n",
      "Epoch 512/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 3.5232 - categorical_accuracy: 0.3529\n",
      "Epoch 513/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 2.4803 - categorical_accuracy: 0.3412\n",
      "Epoch 514/2000\n",
      "3/3 [==============================] - 0s 49ms/step - loss: 2.7147 - categorical_accuracy: 0.2941\n",
      "Epoch 515/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.9361 - categorical_accuracy: 0.3647\n",
      "Epoch 516/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 2.0051 - categorical_accuracy: 0.3529\n",
      "Epoch 517/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 2.1835 - categorical_accuracy: 0.3176\n",
      "Epoch 518/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 2.3656 - categorical_accuracy: 0.3412\n",
      "Epoch 519/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 2.1847 - categorical_accuracy: 0.2824\n",
      "Epoch 520/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 2.0306 - categorical_accuracy: 0.2353\n",
      "Epoch 521/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.4445 - categorical_accuracy: 0.3294\n",
      "Epoch 522/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.3430 - categorical_accuracy: 0.3294\n",
      "Epoch 523/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.5665 - categorical_accuracy: 0.3294\n",
      "Epoch 524/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.5526 - categorical_accuracy: 0.2941\n",
      "Epoch 525/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.6142 - categorical_accuracy: 0.2941\n",
      "Epoch 526/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.6050 - categorical_accuracy: 0.2588\n",
      "Epoch 527/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.7803 - categorical_accuracy: 0.3647\n",
      "Epoch 528/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 2.7322 - categorical_accuracy: 0.3294\n",
      "Epoch 529/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 3.2631 - categorical_accuracy: 0.3412\n",
      "Epoch 530/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.9659 - categorical_accuracy: 0.3529\n",
      "Epoch 531/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 2.0971 - categorical_accuracy: 0.2706\n",
      "Epoch 532/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.4084 - categorical_accuracy: 0.2353\n",
      "Epoch 533/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.2696 - categorical_accuracy: 0.3176\n",
      "Epoch 534/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.5793 - categorical_accuracy: 0.2235\n",
      "Epoch 535/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.3077 - categorical_accuracy: 0.2824\n",
      "Epoch 536/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.4184 - categorical_accuracy: 0.3059\n",
      "Epoch 537/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.3525 - categorical_accuracy: 0.3294\n",
      "Epoch 538/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.2511 - categorical_accuracy: 0.3176\n",
      "Epoch 539/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.4847 - categorical_accuracy: 0.2235\n",
      "Epoch 540/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.6022 - categorical_accuracy: 0.2118\n",
      "Epoch 541/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.5500 - categorical_accuracy: 0.2706\n",
      "Epoch 542/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.3482 - categorical_accuracy: 0.3294\n",
      "Epoch 543/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.6002 - categorical_accuracy: 0.1765\n",
      "Epoch 544/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.2921 - categorical_accuracy: 0.2824\n",
      "Epoch 545/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.5313 - categorical_accuracy: 0.3412\n",
      "Epoch 546/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.8564 - categorical_accuracy: 0.2941\n",
      "Epoch 547/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.6100 - categorical_accuracy: 0.2824\n",
      "Epoch 548/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.5944 - categorical_accuracy: 0.3294\n",
      "Epoch 549/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.5546 - categorical_accuracy: 0.3294\n",
      "Epoch 550/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.6298 - categorical_accuracy: 0.3294\n",
      "Epoch 551/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.6694 - categorical_accuracy: 0.3176\n",
      "Epoch 552/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 2.2493 - categorical_accuracy: 0.3294\n",
      "Epoch 553/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.6393 - categorical_accuracy: 0.3882\n",
      "Epoch 554/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 2.1077 - categorical_accuracy: 0.2353\n",
      "Epoch 555/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.7365 - categorical_accuracy: 0.3412\n",
      "Epoch 556/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 2.2198 - categorical_accuracy: 0.3647\n",
      "Epoch 557/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.9437 - categorical_accuracy: 0.3059\n",
      "Epoch 558/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.7332 - categorical_accuracy: 0.3529\n",
      "Epoch 559/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.5933 - categorical_accuracy: 0.3529\n",
      "Epoch 560/2000\n",
      "3/3 [==============================] - 0s 49ms/step - loss: 1.4882 - categorical_accuracy: 0.3647\n",
      "Epoch 561/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.5117 - categorical_accuracy: 0.3059\n",
      "Epoch 562/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.3858 - categorical_accuracy: 0.2824\n",
      "Epoch 563/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.5660 - categorical_accuracy: 0.2824\n",
      "Epoch 564/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.8019 - categorical_accuracy: 0.2706\n",
      "Epoch 565/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 2.3202 - categorical_accuracy: 0.3294\n",
      "Epoch 566/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 2.2116 - categorical_accuracy: 0.3412\n",
      "Epoch 567/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.7560 - categorical_accuracy: 0.3882\n",
      "Epoch 568/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 2.3982 - categorical_accuracy: 0.2118\n",
      "Epoch 569/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.7814 - categorical_accuracy: 0.2706\n",
      "Epoch 570/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.8320 - categorical_accuracy: 0.2353\n",
      "Epoch 571/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.8002 - categorical_accuracy: 0.2941\n",
      "Epoch 572/2000\n",
      "3/3 [==============================] - 0s 53ms/step - loss: 1.6725 - categorical_accuracy: 0.3294\n",
      "Epoch 573/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.7626 - categorical_accuracy: 0.3294\n",
      "Epoch 574/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.4653 - categorical_accuracy: 0.4353\n",
      "Epoch 575/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.7285 - categorical_accuracy: 0.3647\n",
      "Epoch 576/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.6950 - categorical_accuracy: 0.3294\n",
      "Epoch 577/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 2.0367 - categorical_accuracy: 0.3176\n",
      "Epoch 578/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.6964 - categorical_accuracy: 0.3529\n",
      "Epoch 579/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.4848 - categorical_accuracy: 0.2706\n",
      "Epoch 580/2000\n",
      "3/3 [==============================] - 0s 52ms/step - loss: 1.2300 - categorical_accuracy: 0.4000\n",
      "Epoch 581/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.4991 - categorical_accuracy: 0.3059\n",
      "Epoch 582/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.3087 - categorical_accuracy: 0.3176\n",
      "Epoch 583/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.2793 - categorical_accuracy: 0.3412\n",
      "Epoch 584/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.2434 - categorical_accuracy: 0.3176\n",
      "Epoch 585/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2532 - categorical_accuracy: 0.3882\n",
      "Epoch 586/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.5940 - categorical_accuracy: 0.3176\n",
      "Epoch 587/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.5785 - categorical_accuracy: 0.2824\n",
      "Epoch 588/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.3330 - categorical_accuracy: 0.3294\n",
      "Epoch 589/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.2442 - categorical_accuracy: 0.3059\n",
      "Epoch 590/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.3470 - categorical_accuracy: 0.3176\n",
      "Epoch 591/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.5594 - categorical_accuracy: 0.3412\n",
      "Epoch 592/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.4235 - categorical_accuracy: 0.3294\n",
      "Epoch 593/2000\n",
      "3/3 [==============================] - 0s 49ms/step - loss: 1.6826 - categorical_accuracy: 0.2824\n",
      "Epoch 594/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.7505 - categorical_accuracy: 0.2941\n",
      "Epoch 595/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.3010 - categorical_accuracy: 0.2824\n",
      "Epoch 596/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.2945 - categorical_accuracy: 0.2471\n",
      "Epoch 597/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.3037 - categorical_accuracy: 0.2941\n",
      "Epoch 598/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.2547 - categorical_accuracy: 0.2588\n",
      "Epoch 599/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.2351 - categorical_accuracy: 0.2588\n",
      "Epoch 600/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1427 - categorical_accuracy: 0.4000\n",
      "Epoch 601/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1969 - categorical_accuracy: 0.3059\n",
      "Epoch 602/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2214 - categorical_accuracy: 0.3882\n",
      "Epoch 603/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.3574 - categorical_accuracy: 0.3176\n",
      "Epoch 604/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.3259 - categorical_accuracy: 0.3176\n",
      "Epoch 605/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.2584 - categorical_accuracy: 0.2588\n",
      "Epoch 606/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.2431 - categorical_accuracy: 0.2941\n",
      "Epoch 607/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.3029 - categorical_accuracy: 0.2706\n",
      "Epoch 608/2000\n",
      "3/3 [==============================] - 0s 31ms/step - loss: 1.7362 - categorical_accuracy: 0.3412\n",
      "Epoch 609/2000\n",
      "3/3 [==============================] - 0s 30ms/step - loss: 1.5167 - categorical_accuracy: 0.2706\n",
      "Epoch 610/2000\n",
      "3/3 [==============================] - 0s 31ms/step - loss: 1.2453 - categorical_accuracy: 0.3294\n",
      "Epoch 611/2000\n",
      "3/3 [==============================] - 0s 31ms/step - loss: 1.4140 - categorical_accuracy: 0.2706\n",
      "Epoch 612/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2553 - categorical_accuracy: 0.3176\n",
      "Epoch 613/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1482 - categorical_accuracy: 0.3059\n",
      "Epoch 614/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 1.2656 - categorical_accuracy: 0.2471\n",
      "Epoch 615/2000\n",
      "3/3 [==============================] - 0s 31ms/step - loss: 1.3458 - categorical_accuracy: 0.2471\n",
      "Epoch 616/2000\n",
      "3/3 [==============================] - 0s 31ms/step - loss: 1.3586 - categorical_accuracy: 0.4353\n",
      "Epoch 617/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.7221 - categorical_accuracy: 0.2941\n",
      "Epoch 618/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.3701 - categorical_accuracy: 0.3176\n",
      "Epoch 619/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.2478 - categorical_accuracy: 0.3882\n",
      "Epoch 620/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 1.3310 - categorical_accuracy: 0.3176\n",
      "Epoch 621/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.1857 - categorical_accuracy: 0.2824\n",
      "Epoch 622/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 1.3366 - categorical_accuracy: 0.2941\n",
      "Epoch 623/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1800 - categorical_accuracy: 0.2941\n",
      "Epoch 624/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 1.1965 - categorical_accuracy: 0.3412\n",
      "Epoch 625/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.1473 - categorical_accuracy: 0.4118\n",
      "Epoch 626/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 1.2373 - categorical_accuracy: 0.2706\n",
      "Epoch 627/2000\n",
      "3/3 [==============================] - 0s 30ms/step - loss: 1.3254 - categorical_accuracy: 0.2706\n",
      "Epoch 628/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.2499 - categorical_accuracy: 0.2118\n",
      "Epoch 629/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.3153 - categorical_accuracy: 0.2941\n",
      "Epoch 630/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.6584 - categorical_accuracy: 0.3059\n",
      "Epoch 631/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.4673 - categorical_accuracy: 0.2824\n",
      "Epoch 632/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1745 - categorical_accuracy: 0.3765\n",
      "Epoch 633/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2992 - categorical_accuracy: 0.3176\n",
      "Epoch 634/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.4371 - categorical_accuracy: 0.3294\n",
      "Epoch 635/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.4423 - categorical_accuracy: 0.3529\n",
      "Epoch 636/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.3869 - categorical_accuracy: 0.3647\n",
      "Epoch 637/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2864 - categorical_accuracy: 0.2941\n",
      "Epoch 638/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.2768 - categorical_accuracy: 0.3059\n",
      "Epoch 639/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.2878 - categorical_accuracy: 0.3294\n",
      "Epoch 640/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.4268 - categorical_accuracy: 0.3176\n",
      "Epoch 641/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.3231 - categorical_accuracy: 0.3176\n",
      "Epoch 642/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1839 - categorical_accuracy: 0.3647\n",
      "Epoch 643/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.4266 - categorical_accuracy: 0.2353\n",
      "Epoch 644/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.3339 - categorical_accuracy: 0.2706\n",
      "Epoch 645/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.2598 - categorical_accuracy: 0.3059\n",
      "Epoch 646/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2065 - categorical_accuracy: 0.3059\n",
      "Epoch 647/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.3736 - categorical_accuracy: 0.3765\n",
      "Epoch 648/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.5811 - categorical_accuracy: 0.2941\n",
      "Epoch 649/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.3852 - categorical_accuracy: 0.2941\n",
      "Epoch 650/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.2020 - categorical_accuracy: 0.2941\n",
      "Epoch 651/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.2867 - categorical_accuracy: 0.3176\n",
      "Epoch 652/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.3803 - categorical_accuracy: 0.3176\n",
      "Epoch 653/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.3552 - categorical_accuracy: 0.2235\n",
      "Epoch 654/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2653 - categorical_accuracy: 0.2706\n",
      "Epoch 655/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.4401 - categorical_accuracy: 0.3412\n",
      "Epoch 656/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.5466 - categorical_accuracy: 0.2706\n",
      "Epoch 657/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.5496 - categorical_accuracy: 0.3765\n",
      "Epoch 658/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.8121 - categorical_accuracy: 0.3294\n",
      "Epoch 659/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.4420 - categorical_accuracy: 0.3176\n",
      "Epoch 660/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.4148 - categorical_accuracy: 0.3176\n",
      "Epoch 661/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2767 - categorical_accuracy: 0.3176\n",
      "Epoch 662/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.2398 - categorical_accuracy: 0.3176\n",
      "Epoch 663/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.3076 - categorical_accuracy: 0.3294\n",
      "Epoch 664/2000\n",
      "3/3 [==============================] - 0s 59ms/step - loss: 1.3887 - categorical_accuracy: 0.3647\n",
      "Epoch 665/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.4317 - categorical_accuracy: 0.3412\n",
      "Epoch 666/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.3090 - categorical_accuracy: 0.2588\n",
      "Epoch 667/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.3976 - categorical_accuracy: 0.3176\n",
      "Epoch 668/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.4457 - categorical_accuracy: 0.3765\n",
      "Epoch 669/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.4420 - categorical_accuracy: 0.2471\n",
      "Epoch 670/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.2304 - categorical_accuracy: 0.3294\n",
      "Epoch 671/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.2879 - categorical_accuracy: 0.3176\n",
      "Epoch 672/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 1.4483 - categorical_accuracy: 0.2588\n",
      "Epoch 673/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.2130 - categorical_accuracy: 0.3882\n",
      "Epoch 674/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.4308 - categorical_accuracy: 0.3765\n",
      "Epoch 675/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.4783 - categorical_accuracy: 0.3059\n",
      "Epoch 676/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.2563 - categorical_accuracy: 0.3529\n",
      "Epoch 677/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.3214 - categorical_accuracy: 0.3176\n",
      "Epoch 678/2000\n",
      "3/3 [==============================] - 0s 31ms/step - loss: 1.1660 - categorical_accuracy: 0.3294\n",
      "Epoch 679/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.4966 - categorical_accuracy: 0.2235\n",
      "Epoch 680/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.6718 - categorical_accuracy: 0.3412\n",
      "Epoch 681/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.5215 - categorical_accuracy: 0.2941\n",
      "Epoch 682/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.8502 - categorical_accuracy: 0.2941\n",
      "Epoch 683/2000\n",
      "3/3 [==============================] - 0s 31ms/step - loss: 1.3083 - categorical_accuracy: 0.3647\n",
      "Epoch 684/2000\n",
      "3/3 [==============================] - 0s 31ms/step - loss: 1.3258 - categorical_accuracy: 0.2588\n",
      "Epoch 685/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 1.2381 - categorical_accuracy: 0.3882\n",
      "Epoch 686/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.2272 - categorical_accuracy: 0.3294\n",
      "Epoch 687/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.4290 - categorical_accuracy: 0.2353\n",
      "Epoch 688/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 1.3895 - categorical_accuracy: 0.2941\n",
      "Epoch 689/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.3901 - categorical_accuracy: 0.2588\n",
      "Epoch 690/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 1.4956 - categorical_accuracy: 0.2471\n",
      "Epoch 691/2000\n",
      "3/3 [==============================] - 0s 31ms/step - loss: 1.4251 - categorical_accuracy: 0.3882\n",
      "Epoch 692/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 1.2613 - categorical_accuracy: 0.3647\n",
      "Epoch 693/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.4059 - categorical_accuracy: 0.3529\n",
      "Epoch 694/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 1.4717 - categorical_accuracy: 0.3412\n",
      "Epoch 695/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.6107 - categorical_accuracy: 0.2941\n",
      "Epoch 696/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.4536 - categorical_accuracy: 0.4000\n",
      "Epoch 697/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 1.2051 - categorical_accuracy: 0.3765\n",
      "Epoch 698/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 1.2433 - categorical_accuracy: 0.3059\n",
      "Epoch 699/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.4601 - categorical_accuracy: 0.2588\n",
      "Epoch 700/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 1.1598 - categorical_accuracy: 0.2706\n",
      "Epoch 701/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.1406 - categorical_accuracy: 0.2941\n",
      "Epoch 702/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1651 - categorical_accuracy: 0.2118\n",
      "Epoch 703/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.3136 - categorical_accuracy: 0.2706\n",
      "Epoch 704/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.2257 - categorical_accuracy: 0.3529\n",
      "Epoch 705/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.2459 - categorical_accuracy: 0.2941\n",
      "Epoch 706/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2455 - categorical_accuracy: 0.3765\n",
      "Epoch 707/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.3369 - categorical_accuracy: 0.3176\n",
      "Epoch 708/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 1.3665 - categorical_accuracy: 0.2706\n",
      "Epoch 709/2000\n",
      "3/3 [==============================] - 0s 31ms/step - loss: 1.5133 - categorical_accuracy: 0.2941\n",
      "Epoch 710/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 1.3364 - categorical_accuracy: 0.3765\n",
      "Epoch 711/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1891 - categorical_accuracy: 0.3647\n",
      "Epoch 712/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.4818 - categorical_accuracy: 0.3059\n",
      "Epoch 713/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.8041 - categorical_accuracy: 0.2941\n",
      "Epoch 714/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 1.4884 - categorical_accuracy: 0.2941\n",
      "Epoch 715/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 1.5415 - categorical_accuracy: 0.3647\n",
      "Epoch 716/2000\n",
      "3/3 [==============================] - 0s 32ms/step - loss: 1.5502 - categorical_accuracy: 0.3294\n",
      "Epoch 717/2000\n",
      "3/3 [==============================] - 0s 31ms/step - loss: 1.2378 - categorical_accuracy: 0.3765\n",
      "Epoch 718/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.3041 - categorical_accuracy: 0.2941\n",
      "Epoch 719/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.2257 - categorical_accuracy: 0.3412\n",
      "Epoch 720/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.4137 - categorical_accuracy: 0.3059\n",
      "Epoch 721/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.4045 - categorical_accuracy: 0.3176\n",
      "Epoch 722/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.2144 - categorical_accuracy: 0.2706\n",
      "Epoch 723/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.2119 - categorical_accuracy: 0.3059\n",
      "Epoch 724/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2849 - categorical_accuracy: 0.2706\n",
      "Epoch 725/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.3069 - categorical_accuracy: 0.2941\n",
      "Epoch 726/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.3921 - categorical_accuracy: 0.3647\n",
      "Epoch 727/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.6091 - categorical_accuracy: 0.3176\n",
      "Epoch 728/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.2190 - categorical_accuracy: 0.3412\n",
      "Epoch 729/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1570 - categorical_accuracy: 0.3412\n",
      "Epoch 730/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1985 - categorical_accuracy: 0.3294\n",
      "Epoch 731/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2410 - categorical_accuracy: 0.2706\n",
      "Epoch 732/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1739 - categorical_accuracy: 0.3647\n",
      "Epoch 733/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1839 - categorical_accuracy: 0.3529\n",
      "Epoch 734/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2000 - categorical_accuracy: 0.3176\n",
      "Epoch 735/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.2869 - categorical_accuracy: 0.4118\n",
      "Epoch 736/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.4093 - categorical_accuracy: 0.2588\n",
      "Epoch 737/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.3323 - categorical_accuracy: 0.3765\n",
      "Epoch 738/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1147 - categorical_accuracy: 0.4471\n",
      "Epoch 739/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.2897 - categorical_accuracy: 0.3294\n",
      "Epoch 740/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.3282 - categorical_accuracy: 0.2706\n",
      "Epoch 741/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2507 - categorical_accuracy: 0.3529\n",
      "Epoch 742/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.5869 - categorical_accuracy: 0.3529\n",
      "Epoch 743/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2735 - categorical_accuracy: 0.2235\n",
      "Epoch 744/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2856 - categorical_accuracy: 0.3059\n",
      "Epoch 745/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.4927 - categorical_accuracy: 0.2941\n",
      "Epoch 746/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.5313 - categorical_accuracy: 0.3647\n",
      "Epoch 747/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2848 - categorical_accuracy: 0.3765\n",
      "Epoch 748/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.6221 - categorical_accuracy: 0.3412\n",
      "Epoch 749/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.7127 - categorical_accuracy: 0.3412\n",
      "Epoch 750/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2054 - categorical_accuracy: 0.4353\n",
      "Epoch 751/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.2527 - categorical_accuracy: 0.2941\n",
      "Epoch 752/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2101 - categorical_accuracy: 0.3647\n",
      "Epoch 753/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.2979 - categorical_accuracy: 0.2824\n",
      "Epoch 754/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1624 - categorical_accuracy: 0.3882\n",
      "Epoch 755/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1816 - categorical_accuracy: 0.2941\n",
      "Epoch 756/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2072 - categorical_accuracy: 0.3529\n",
      "Epoch 757/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.1561 - categorical_accuracy: 0.3647\n",
      "Epoch 758/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1317 - categorical_accuracy: 0.2235\n",
      "Epoch 759/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.3572 - categorical_accuracy: 0.2235\n",
      "Epoch 760/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.3336 - categorical_accuracy: 0.3529\n",
      "Epoch 761/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.3761 - categorical_accuracy: 0.2941\n",
      "Epoch 762/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.4245 - categorical_accuracy: 0.3294\n",
      "Epoch 763/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.3034 - categorical_accuracy: 0.2941\n",
      "Epoch 764/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.2373 - categorical_accuracy: 0.3882\n",
      "Epoch 765/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.1783 - categorical_accuracy: 0.3294\n",
      "Epoch 766/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1812 - categorical_accuracy: 0.3294\n",
      "Epoch 767/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.2432 - categorical_accuracy: 0.3412\n",
      "Epoch 768/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1479 - categorical_accuracy: 0.3412\n",
      "Epoch 769/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1314 - categorical_accuracy: 0.3412\n",
      "Epoch 770/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2030 - categorical_accuracy: 0.3176\n",
      "Epoch 771/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1903 - categorical_accuracy: 0.2588\n",
      "Epoch 772/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1364 - categorical_accuracy: 0.4000\n",
      "Epoch 773/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1274 - categorical_accuracy: 0.3176\n",
      "Epoch 774/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.1445 - categorical_accuracy: 0.3765\n",
      "Epoch 775/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2456 - categorical_accuracy: 0.3176\n",
      "Epoch 776/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1420 - categorical_accuracy: 0.3294\n",
      "Epoch 777/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2393 - categorical_accuracy: 0.2824\n",
      "Epoch 778/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2170 - categorical_accuracy: 0.3294\n",
      "Epoch 779/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1489 - categorical_accuracy: 0.3529\n",
      "Epoch 780/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2577 - categorical_accuracy: 0.2941\n",
      "Epoch 781/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.3213 - categorical_accuracy: 0.3647\n",
      "Epoch 782/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.2910 - categorical_accuracy: 0.3412\n",
      "Epoch 783/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.2105 - categorical_accuracy: 0.3412\n",
      "Epoch 784/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1606 - categorical_accuracy: 0.3294\n",
      "Epoch 785/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0998 - categorical_accuracy: 0.3882\n",
      "Epoch 786/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1470 - categorical_accuracy: 0.3647\n",
      "Epoch 787/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1210 - categorical_accuracy: 0.4588\n",
      "Epoch 788/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1681 - categorical_accuracy: 0.3412\n",
      "Epoch 789/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1511 - categorical_accuracy: 0.3412\n",
      "Epoch 790/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1007 - categorical_accuracy: 0.3294\n",
      "Epoch 791/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1615 - categorical_accuracy: 0.4118\n",
      "Epoch 792/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.2549 - categorical_accuracy: 0.3294\n",
      "Epoch 793/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.3160 - categorical_accuracy: 0.2471\n",
      "Epoch 794/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0868 - categorical_accuracy: 0.3882\n",
      "Epoch 795/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1844 - categorical_accuracy: 0.3765\n",
      "Epoch 796/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.1948 - categorical_accuracy: 0.3294\n",
      "Epoch 797/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0780 - categorical_accuracy: 0.4706\n",
      "Epoch 798/2000\n",
      "3/3 [==============================] - 0s 62ms/step - loss: 1.0862 - categorical_accuracy: 0.4118\n",
      "Epoch 799/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.2345 - categorical_accuracy: 0.3765\n",
      "Epoch 800/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.3852 - categorical_accuracy: 0.2824\n",
      "Epoch 801/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.3849 - categorical_accuracy: 0.3529\n",
      "Epoch 802/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.3315 - categorical_accuracy: 0.3882\n",
      "Epoch 803/2000\n",
      "3/3 [==============================] - 0s 52ms/step - loss: 1.4522 - categorical_accuracy: 0.2941\n",
      "Epoch 804/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.4105 - categorical_accuracy: 0.3294\n",
      "Epoch 805/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.2834 - categorical_accuracy: 0.3647\n",
      "Epoch 806/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.3334 - categorical_accuracy: 0.3412\n",
      "Epoch 807/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2343 - categorical_accuracy: 0.3412\n",
      "Epoch 808/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.2304 - categorical_accuracy: 0.3176\n",
      "Epoch 809/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.2351 - categorical_accuracy: 0.2824\n",
      "Epoch 810/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1679 - categorical_accuracy: 0.3412\n",
      "Epoch 811/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.2017 - categorical_accuracy: 0.3529\n",
      "Epoch 812/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.1740 - categorical_accuracy: 0.3529\n",
      "Epoch 813/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.2435 - categorical_accuracy: 0.4235\n",
      "Epoch 814/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.3574 - categorical_accuracy: 0.3412\n",
      "Epoch 815/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.4404 - categorical_accuracy: 0.3176\n",
      "Epoch 816/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.2920 - categorical_accuracy: 0.3529\n",
      "Epoch 817/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.4329 - categorical_accuracy: 0.3412\n",
      "Epoch 818/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.4492 - categorical_accuracy: 0.3059\n",
      "Epoch 819/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.3684 - categorical_accuracy: 0.3882\n",
      "Epoch 820/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.4123 - categorical_accuracy: 0.2353\n",
      "Epoch 821/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.2341 - categorical_accuracy: 0.3412\n",
      "Epoch 822/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.1435 - categorical_accuracy: 0.4588\n",
      "Epoch 823/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.4867 - categorical_accuracy: 0.3412\n",
      "Epoch 824/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.3623 - categorical_accuracy: 0.2706\n",
      "Epoch 825/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0965 - categorical_accuracy: 0.3765\n",
      "Epoch 826/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.2085 - categorical_accuracy: 0.2824\n",
      "Epoch 827/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.1571 - categorical_accuracy: 0.3882\n",
      "Epoch 828/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.1730 - categorical_accuracy: 0.3765\n",
      "Epoch 829/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.1430 - categorical_accuracy: 0.3647\n",
      "Epoch 830/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.1143 - categorical_accuracy: 0.4471\n",
      "Epoch 831/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.2503 - categorical_accuracy: 0.3059\n",
      "Epoch 832/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.1878 - categorical_accuracy: 0.3529\n",
      "Epoch 833/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.1784 - categorical_accuracy: 0.2824\n",
      "Epoch 834/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1270 - categorical_accuracy: 0.3647\n",
      "Epoch 835/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.1576 - categorical_accuracy: 0.3412\n",
      "Epoch 836/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.3303 - categorical_accuracy: 0.2471\n",
      "Epoch 837/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1638 - categorical_accuracy: 0.3882\n",
      "Epoch 838/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.2568 - categorical_accuracy: 0.3647\n",
      "Epoch 839/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.2609 - categorical_accuracy: 0.3412\n",
      "Epoch 840/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.5418 - categorical_accuracy: 0.3412\n",
      "Epoch 841/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.5298 - categorical_accuracy: 0.3412\n",
      "Epoch 842/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.3405 - categorical_accuracy: 0.2941\n",
      "Epoch 843/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.3341 - categorical_accuracy: 0.3412\n",
      "Epoch 844/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.3360 - categorical_accuracy: 0.3059\n",
      "Epoch 845/2000\n",
      "3/3 [==============================] - 0s 47ms/step - loss: 1.2158 - categorical_accuracy: 0.3412\n",
      "Epoch 846/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.2413 - categorical_accuracy: 0.3765\n",
      "Epoch 847/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1689 - categorical_accuracy: 0.3412\n",
      "Epoch 848/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.2576 - categorical_accuracy: 0.3647\n",
      "Epoch 849/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.1375 - categorical_accuracy: 0.3412\n",
      "Epoch 850/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.1642 - categorical_accuracy: 0.3647\n",
      "Epoch 851/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.1069 - categorical_accuracy: 0.3529\n",
      "Epoch 852/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0772 - categorical_accuracy: 0.3765\n",
      "Epoch 853/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1206 - categorical_accuracy: 0.4588\n",
      "Epoch 854/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.1378 - categorical_accuracy: 0.3059\n",
      "Epoch 855/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1767 - categorical_accuracy: 0.4235\n",
      "Epoch 856/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.2066 - categorical_accuracy: 0.4000\n",
      "Epoch 857/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.3509 - categorical_accuracy: 0.4235\n",
      "Epoch 858/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1681 - categorical_accuracy: 0.3412\n",
      "Epoch 859/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.6492 - categorical_accuracy: 0.3294\n",
      "Epoch 860/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.4437 - categorical_accuracy: 0.3647\n",
      "Epoch 861/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1885 - categorical_accuracy: 0.3882\n",
      "Epoch 862/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2517 - categorical_accuracy: 0.3765\n",
      "Epoch 863/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0881 - categorical_accuracy: 0.4471\n",
      "Epoch 864/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2563 - categorical_accuracy: 0.3412\n",
      "Epoch 865/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1282 - categorical_accuracy: 0.3765\n",
      "Epoch 866/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.1898 - categorical_accuracy: 0.2706\n",
      "Epoch 867/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0665 - categorical_accuracy: 0.4353\n",
      "Epoch 868/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0623 - categorical_accuracy: 0.4235\n",
      "Epoch 869/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0798 - categorical_accuracy: 0.4000\n",
      "Epoch 870/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1482 - categorical_accuracy: 0.3412\n",
      "Epoch 871/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1251 - categorical_accuracy: 0.3176\n",
      "Epoch 872/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0973 - categorical_accuracy: 0.4000\n",
      "Epoch 873/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1071 - categorical_accuracy: 0.3529\n",
      "Epoch 874/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.1456 - categorical_accuracy: 0.3059\n",
      "Epoch 875/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0989 - categorical_accuracy: 0.5176\n",
      "Epoch 876/2000\n",
      "3/3 [==============================] - 0s 50ms/step - loss: 1.1830 - categorical_accuracy: 0.4824\n",
      "Epoch 877/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.2182 - categorical_accuracy: 0.3412\n",
      "Epoch 878/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.2819 - categorical_accuracy: 0.3059\n",
      "Epoch 879/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2244 - categorical_accuracy: 0.3059\n",
      "Epoch 880/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.2110 - categorical_accuracy: 0.3294\n",
      "Epoch 881/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.1475 - categorical_accuracy: 0.3412\n",
      "Epoch 882/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.2024 - categorical_accuracy: 0.3412\n",
      "Epoch 883/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0836 - categorical_accuracy: 0.4235\n",
      "Epoch 884/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1182 - categorical_accuracy: 0.4235\n",
      "Epoch 885/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.2470 - categorical_accuracy: 0.2588\n",
      "Epoch 886/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.2032 - categorical_accuracy: 0.3176\n",
      "Epoch 887/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1462 - categorical_accuracy: 0.3765\n",
      "Epoch 888/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.2324 - categorical_accuracy: 0.2941\n",
      "Epoch 889/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.2132 - categorical_accuracy: 0.3059\n",
      "Epoch 890/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.2417 - categorical_accuracy: 0.4118\n",
      "Epoch 891/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.2060 - categorical_accuracy: 0.3059\n",
      "Epoch 892/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1989 - categorical_accuracy: 0.3412\n",
      "Epoch 893/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1791 - categorical_accuracy: 0.4000\n",
      "Epoch 894/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.3051 - categorical_accuracy: 0.2706\n",
      "Epoch 895/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.3662 - categorical_accuracy: 0.3294\n",
      "Epoch 896/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1916 - categorical_accuracy: 0.3765\n",
      "Epoch 897/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1946 - categorical_accuracy: 0.3412\n",
      "Epoch 898/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0915 - categorical_accuracy: 0.3765\n",
      "Epoch 899/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1499 - categorical_accuracy: 0.4235\n",
      "Epoch 900/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.1601 - categorical_accuracy: 0.3647\n",
      "Epoch 901/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.1592 - categorical_accuracy: 0.3765\n",
      "Epoch 902/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1328 - categorical_accuracy: 0.3765\n",
      "Epoch 903/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.2051 - categorical_accuracy: 0.3412\n",
      "Epoch 904/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.2838 - categorical_accuracy: 0.3647\n",
      "Epoch 905/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.2201 - categorical_accuracy: 0.3412\n",
      "Epoch 906/2000\n",
      "3/3 [==============================] - 0s 51ms/step - loss: 1.2446 - categorical_accuracy: 0.3059\n",
      "Epoch 907/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1033 - categorical_accuracy: 0.3529\n",
      "Epoch 908/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1272 - categorical_accuracy: 0.4118\n",
      "Epoch 909/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2407 - categorical_accuracy: 0.3647\n",
      "Epoch 910/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2983 - categorical_accuracy: 0.2353\n",
      "Epoch 911/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.3638 - categorical_accuracy: 0.4000\n",
      "Epoch 912/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.3747 - categorical_accuracy: 0.3176\n",
      "Epoch 913/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.4240 - categorical_accuracy: 0.2941\n",
      "Epoch 914/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.3806 - categorical_accuracy: 0.3529\n",
      "Epoch 915/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.4713 - categorical_accuracy: 0.2824\n",
      "Epoch 916/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.3010 - categorical_accuracy: 0.3412\n",
      "Epoch 917/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.3617 - categorical_accuracy: 0.3765\n",
      "Epoch 918/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.5472 - categorical_accuracy: 0.4000\n",
      "Epoch 919/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.2330 - categorical_accuracy: 0.3647\n",
      "Epoch 920/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2330 - categorical_accuracy: 0.2824\n",
      "Epoch 921/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.4448 - categorical_accuracy: 0.2706\n",
      "Epoch 922/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.4577 - categorical_accuracy: 0.2706\n",
      "Epoch 923/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.1948 - categorical_accuracy: 0.4824\n",
      "Epoch 924/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2587 - categorical_accuracy: 0.2706\n",
      "Epoch 925/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2352 - categorical_accuracy: 0.3059\n",
      "Epoch 926/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.1818 - categorical_accuracy: 0.3529\n",
      "Epoch 927/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1724 - categorical_accuracy: 0.2706\n",
      "Epoch 928/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1613 - categorical_accuracy: 0.3882\n",
      "Epoch 929/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2752 - categorical_accuracy: 0.3647\n",
      "Epoch 930/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1667 - categorical_accuracy: 0.3647\n",
      "Epoch 931/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.2207 - categorical_accuracy: 0.2941\n",
      "Epoch 932/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0893 - categorical_accuracy: 0.3176\n",
      "Epoch 933/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1179 - categorical_accuracy: 0.3059\n",
      "Epoch 934/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1093 - categorical_accuracy: 0.3059\n",
      "Epoch 935/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1816 - categorical_accuracy: 0.2706\n",
      "Epoch 936/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.1920 - categorical_accuracy: 0.3412\n",
      "Epoch 937/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.1482 - categorical_accuracy: 0.3059\n",
      "Epoch 938/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.1761 - categorical_accuracy: 0.2706\n",
      "Epoch 939/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1918 - categorical_accuracy: 0.2235\n",
      "Epoch 940/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1609 - categorical_accuracy: 0.3059\n",
      "Epoch 941/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.1990 - categorical_accuracy: 0.2471\n",
      "Epoch 942/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1211 - categorical_accuracy: 0.3412\n",
      "Epoch 943/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0895 - categorical_accuracy: 0.3647\n",
      "Epoch 944/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.1447 - categorical_accuracy: 0.2000\n",
      "Epoch 945/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.2778 - categorical_accuracy: 0.3647\n",
      "Epoch 946/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.2613 - categorical_accuracy: 0.3059\n",
      "Epoch 947/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1708 - categorical_accuracy: 0.2824\n",
      "Epoch 948/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1439 - categorical_accuracy: 0.3647\n",
      "Epoch 949/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2224 - categorical_accuracy: 0.3294\n",
      "Epoch 950/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.3783 - categorical_accuracy: 0.3294\n",
      "Epoch 951/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.4181 - categorical_accuracy: 0.2824\n",
      "Epoch 952/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2545 - categorical_accuracy: 0.3294\n",
      "Epoch 953/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1963 - categorical_accuracy: 0.3647\n",
      "Epoch 954/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1046 - categorical_accuracy: 0.3765\n",
      "Epoch 955/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.2169 - categorical_accuracy: 0.4000\n",
      "Epoch 956/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.8163 - categorical_accuracy: 0.3529\n",
      "Epoch 957/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.4369 - categorical_accuracy: 0.3294\n",
      "Epoch 958/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.4562 - categorical_accuracy: 0.3294\n",
      "Epoch 959/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2507 - categorical_accuracy: 0.3647\n",
      "Epoch 960/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 2.0916 - categorical_accuracy: 0.3412\n",
      "Epoch 961/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.5741 - categorical_accuracy: 0.2824\n",
      "Epoch 962/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.4420 - categorical_accuracy: 0.3059\n",
      "Epoch 963/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.5774 - categorical_accuracy: 0.3412\n",
      "Epoch 964/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.3317 - categorical_accuracy: 0.3176\n",
      "Epoch 965/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1847 - categorical_accuracy: 0.2941\n",
      "Epoch 966/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.2309 - categorical_accuracy: 0.3176\n",
      "Epoch 967/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1786 - categorical_accuracy: 0.3294\n",
      "Epoch 968/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.3090 - categorical_accuracy: 0.3412\n",
      "Epoch 969/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.3551 - categorical_accuracy: 0.2824\n",
      "Epoch 970/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.2270 - categorical_accuracy: 0.3294\n",
      "Epoch 971/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.2663 - categorical_accuracy: 0.3059\n",
      "Epoch 972/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.2328 - categorical_accuracy: 0.4235\n",
      "Epoch 973/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.5204 - categorical_accuracy: 0.3294\n",
      "Epoch 974/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.3977 - categorical_accuracy: 0.3059\n",
      "Epoch 975/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1651 - categorical_accuracy: 0.3647\n",
      "Epoch 976/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.1967 - categorical_accuracy: 0.3412\n",
      "Epoch 977/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2205 - categorical_accuracy: 0.3059\n",
      "Epoch 978/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.2275 - categorical_accuracy: 0.2706\n",
      "Epoch 979/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1098 - categorical_accuracy: 0.3765\n",
      "Epoch 980/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1220 - categorical_accuracy: 0.3412\n",
      "Epoch 981/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.1948 - categorical_accuracy: 0.2941\n",
      "Epoch 982/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0736 - categorical_accuracy: 0.4235\n",
      "Epoch 983/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1989 - categorical_accuracy: 0.3765\n",
      "Epoch 984/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1839 - categorical_accuracy: 0.4588\n",
      "Epoch 985/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1526 - categorical_accuracy: 0.3765\n",
      "Epoch 986/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.1212 - categorical_accuracy: 0.4706\n",
      "Epoch 987/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1970 - categorical_accuracy: 0.3176\n",
      "Epoch 988/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1130 - categorical_accuracy: 0.4000\n",
      "Epoch 989/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1818 - categorical_accuracy: 0.3882\n",
      "Epoch 990/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1886 - categorical_accuracy: 0.2706\n",
      "Epoch 991/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0967 - categorical_accuracy: 0.3765\n",
      "Epoch 992/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.1127 - categorical_accuracy: 0.3059\n",
      "Epoch 993/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1123 - categorical_accuracy: 0.3412\n",
      "Epoch 994/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.1744 - categorical_accuracy: 0.3294\n",
      "Epoch 995/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1321 - categorical_accuracy: 0.3529\n",
      "Epoch 996/2000\n",
      "3/3 [==============================] - 0s 50ms/step - loss: 1.1850 - categorical_accuracy: 0.3176\n",
      "Epoch 997/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1351 - categorical_accuracy: 0.3765\n",
      "Epoch 998/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0882 - categorical_accuracy: 0.3647\n",
      "Epoch 999/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1802 - categorical_accuracy: 0.2588\n",
      "Epoch 1000/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1084 - categorical_accuracy: 0.3412\n",
      "Epoch 1001/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.1447 - categorical_accuracy: 0.2471\n",
      "Epoch 1002/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1503 - categorical_accuracy: 0.2824\n",
      "Epoch 1003/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1069 - categorical_accuracy: 0.3529\n",
      "Epoch 1004/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1916 - categorical_accuracy: 0.3412\n",
      "Epoch 1005/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1218 - categorical_accuracy: 0.3647\n",
      "Epoch 1006/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.2608 - categorical_accuracy: 0.3176\n",
      "Epoch 1007/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.2072 - categorical_accuracy: 0.3529\n",
      "Epoch 1008/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1700 - categorical_accuracy: 0.3176\n",
      "Epoch 1009/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0956 - categorical_accuracy: 0.3412\n",
      "Epoch 1010/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1256 - categorical_accuracy: 0.3294\n",
      "Epoch 1011/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.1664 - categorical_accuracy: 0.3412\n",
      "Epoch 1012/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0816 - categorical_accuracy: 0.4235\n",
      "Epoch 1013/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1122 - categorical_accuracy: 0.3412\n",
      "Epoch 1014/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1501 - categorical_accuracy: 0.3412\n",
      "Epoch 1015/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1647 - categorical_accuracy: 0.3059\n",
      "Epoch 1016/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1073 - categorical_accuracy: 0.3412\n",
      "Epoch 1017/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1186 - categorical_accuracy: 0.3529\n",
      "Epoch 1018/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.0970 - categorical_accuracy: 0.2824\n",
      "Epoch 1019/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1210 - categorical_accuracy: 0.3176\n",
      "Epoch 1020/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0858 - categorical_accuracy: 0.3647\n",
      "Epoch 1021/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1409 - categorical_accuracy: 0.3176\n",
      "Epoch 1022/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0931 - categorical_accuracy: 0.4235\n",
      "Epoch 1023/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1011 - categorical_accuracy: 0.4000\n",
      "Epoch 1024/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1079 - categorical_accuracy: 0.3294\n",
      "Epoch 1025/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0941 - categorical_accuracy: 0.3647\n",
      "Epoch 1026/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0882 - categorical_accuracy: 0.4941\n",
      "Epoch 1027/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1082 - categorical_accuracy: 0.3294\n",
      "Epoch 1028/2000\n",
      "3/3 [==============================] - 0s 67ms/step - loss: 1.1097 - categorical_accuracy: 0.3765\n",
      "Epoch 1029/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1177 - categorical_accuracy: 0.3529\n",
      "Epoch 1030/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.2078 - categorical_accuracy: 0.4353\n",
      "Epoch 1031/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1206 - categorical_accuracy: 0.3765\n",
      "Epoch 1032/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1504 - categorical_accuracy: 0.2706\n",
      "Epoch 1033/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1930 - categorical_accuracy: 0.3412\n",
      "Epoch 1034/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1641 - categorical_accuracy: 0.3059\n",
      "Epoch 1035/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.1396 - categorical_accuracy: 0.4353\n",
      "Epoch 1036/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.3202 - categorical_accuracy: 0.3412\n",
      "Epoch 1037/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1334 - categorical_accuracy: 0.3294\n",
      "Epoch 1038/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.2282 - categorical_accuracy: 0.3294\n",
      "Epoch 1039/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.1273 - categorical_accuracy: 0.3294\n",
      "Epoch 1040/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.2091 - categorical_accuracy: 0.2941\n",
      "Epoch 1041/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1748 - categorical_accuracy: 0.3059\n",
      "Epoch 1042/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1126 - categorical_accuracy: 0.3529\n",
      "Epoch 1043/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.1515 - categorical_accuracy: 0.2824\n",
      "Epoch 1044/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2025 - categorical_accuracy: 0.2353\n",
      "Epoch 1045/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1060 - categorical_accuracy: 0.3882\n",
      "Epoch 1046/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1209 - categorical_accuracy: 0.3294\n",
      "Epoch 1047/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.1715 - categorical_accuracy: 0.3294\n",
      "Epoch 1048/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1498 - categorical_accuracy: 0.3412\n",
      "Epoch 1049/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.2365 - categorical_accuracy: 0.3176\n",
      "Epoch 1050/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1504 - categorical_accuracy: 0.3647\n",
      "Epoch 1051/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1940 - categorical_accuracy: 0.3529\n",
      "Epoch 1052/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2203 - categorical_accuracy: 0.2941\n",
      "Epoch 1053/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1782 - categorical_accuracy: 0.2941\n",
      "Epoch 1054/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1084 - categorical_accuracy: 0.4235\n",
      "Epoch 1055/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.2019 - categorical_accuracy: 0.2824\n",
      "Epoch 1056/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1906 - categorical_accuracy: 0.3412\n",
      "Epoch 1057/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.1434 - categorical_accuracy: 0.3294\n",
      "Epoch 1058/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0959 - categorical_accuracy: 0.4000\n",
      "Epoch 1059/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1469 - categorical_accuracy: 0.2824\n",
      "Epoch 1060/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1927 - categorical_accuracy: 0.4235\n",
      "Epoch 1061/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.1564 - categorical_accuracy: 0.3412\n",
      "Epoch 1062/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2693 - categorical_accuracy: 0.2824\n",
      "Epoch 1063/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1867 - categorical_accuracy: 0.4353\n",
      "Epoch 1064/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1818 - categorical_accuracy: 0.2941\n",
      "Epoch 1065/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.1507 - categorical_accuracy: 0.3412\n",
      "Epoch 1066/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1200 - categorical_accuracy: 0.4235\n",
      "Epoch 1067/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1488 - categorical_accuracy: 0.2941\n",
      "Epoch 1068/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.0941 - categorical_accuracy: 0.3412\n",
      "Epoch 1069/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0892 - categorical_accuracy: 0.3294\n",
      "Epoch 1070/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.0975 - categorical_accuracy: 0.3176\n",
      "Epoch 1071/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1160 - categorical_accuracy: 0.3647\n",
      "Epoch 1072/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1954 - categorical_accuracy: 0.3882\n",
      "Epoch 1073/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0943 - categorical_accuracy: 0.3882\n",
      "Epoch 1074/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0823 - categorical_accuracy: 0.3412\n",
      "Epoch 1075/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1335 - categorical_accuracy: 0.2941\n",
      "Epoch 1076/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1260 - categorical_accuracy: 0.3176\n",
      "Epoch 1077/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1370 - categorical_accuracy: 0.3176\n",
      "Epoch 1078/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1230 - categorical_accuracy: 0.4353\n",
      "Epoch 1079/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.1002 - categorical_accuracy: 0.3412\n",
      "Epoch 1080/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1464 - categorical_accuracy: 0.3294\n",
      "Epoch 1081/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0780 - categorical_accuracy: 0.4588\n",
      "Epoch 1082/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.3425 - categorical_accuracy: 0.3412\n",
      "Epoch 1083/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.1608 - categorical_accuracy: 0.3882\n",
      "Epoch 1084/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.2446 - categorical_accuracy: 0.3412\n",
      "Epoch 1085/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.2365 - categorical_accuracy: 0.3412\n",
      "Epoch 1086/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0733 - categorical_accuracy: 0.3765\n",
      "Epoch 1087/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.1692 - categorical_accuracy: 0.3176\n",
      "Epoch 1088/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1647 - categorical_accuracy: 0.3176\n",
      "Epoch 1089/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1610 - categorical_accuracy: 0.3294\n",
      "Epoch 1090/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.4023 - categorical_accuracy: 0.3412\n",
      "Epoch 1091/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.2269 - categorical_accuracy: 0.3529\n",
      "Epoch 1092/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2208 - categorical_accuracy: 0.3176\n",
      "Epoch 1093/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1054 - categorical_accuracy: 0.3882\n",
      "Epoch 1094/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1171 - categorical_accuracy: 0.3412\n",
      "Epoch 1095/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.1490 - categorical_accuracy: 0.3882\n",
      "Epoch 1096/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.1594 - categorical_accuracy: 0.3882\n",
      "Epoch 1097/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1071 - categorical_accuracy: 0.3412\n",
      "Epoch 1098/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0757 - categorical_accuracy: 0.4471\n",
      "Epoch 1099/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.1002 - categorical_accuracy: 0.3647\n",
      "Epoch 1100/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1712 - categorical_accuracy: 0.2941\n",
      "Epoch 1101/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1039 - categorical_accuracy: 0.4706\n",
      "Epoch 1102/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0979 - categorical_accuracy: 0.3529\n",
      "Epoch 1103/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.0925 - categorical_accuracy: 0.3294\n",
      "Epoch 1104/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.0899 - categorical_accuracy: 0.3882\n",
      "Epoch 1105/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0808 - categorical_accuracy: 0.4235\n",
      "Epoch 1106/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0957 - categorical_accuracy: 0.3412\n",
      "Epoch 1107/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0816 - categorical_accuracy: 0.4353\n",
      "Epoch 1108/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1074 - categorical_accuracy: 0.4824\n",
      "Epoch 1109/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.0993 - categorical_accuracy: 0.4588\n",
      "Epoch 1110/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0975 - categorical_accuracy: 0.2941\n",
      "Epoch 1111/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0887 - categorical_accuracy: 0.2824\n",
      "Epoch 1112/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.1416 - categorical_accuracy: 0.3294\n",
      "Epoch 1113/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1286 - categorical_accuracy: 0.3294\n",
      "Epoch 1114/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1120 - categorical_accuracy: 0.3882\n",
      "Epoch 1115/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0899 - categorical_accuracy: 0.3647\n",
      "Epoch 1116/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1429 - categorical_accuracy: 0.2471\n",
      "Epoch 1117/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1246 - categorical_accuracy: 0.3294\n",
      "Epoch 1118/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1217 - categorical_accuracy: 0.3529\n",
      "Epoch 1119/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0883 - categorical_accuracy: 0.3412\n",
      "Epoch 1120/2000\n",
      "3/3 [==============================] - 0s 33ms/step - loss: 1.0864 - categorical_accuracy: 0.2941\n",
      "Epoch 1121/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0822 - categorical_accuracy: 0.4353\n",
      "Epoch 1122/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1543 - categorical_accuracy: 0.3529\n",
      "Epoch 1123/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.1616 - categorical_accuracy: 0.3176\n",
      "Epoch 1124/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.1104 - categorical_accuracy: 0.3882\n",
      "Epoch 1125/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1795 - categorical_accuracy: 0.3529\n",
      "Epoch 1126/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1889 - categorical_accuracy: 0.2706\n",
      "Epoch 1127/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.2053 - categorical_accuracy: 0.2941\n",
      "Epoch 1128/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.2824 - categorical_accuracy: 0.4235\n",
      "Epoch 1129/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1676 - categorical_accuracy: 0.3529\n",
      "Epoch 1130/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.7569 - categorical_accuracy: 0.2706\n",
      "Epoch 1131/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.5688 - categorical_accuracy: 0.2824\n",
      "Epoch 1132/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 3.2960 - categorical_accuracy: 0.3529\n",
      "Epoch 1133/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 6.6583 - categorical_accuracy: 0.3412\n",
      "Epoch 1134/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 4.7392 - categorical_accuracy: 0.3176\n",
      "Epoch 1135/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 6.8176 - categorical_accuracy: 0.3882\n",
      "Epoch 1136/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 3.7137 - categorical_accuracy: 0.3882\n",
      "Epoch 1137/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 5.0721 - categorical_accuracy: 0.4000\n",
      "Epoch 1138/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 3.4907 - categorical_accuracy: 0.2824\n",
      "Epoch 1139/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 8.0216 - categorical_accuracy: 0.3412\n",
      "Epoch 1140/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 5.1875 - categorical_accuracy: 0.3412\n",
      "Epoch 1141/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 6.1507 - categorical_accuracy: 0.3529\n",
      "Epoch 1142/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 6.4426 - categorical_accuracy: 0.3765\n",
      "Epoch 1143/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 3.5653 - categorical_accuracy: 0.3765\n",
      "Epoch 1144/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 4.3704 - categorical_accuracy: 0.2824\n",
      "Epoch 1145/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 2.9833 - categorical_accuracy: 0.2706\n",
      "Epoch 1146/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.8338 - categorical_accuracy: 0.3412\n",
      "Epoch 1147/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.7116 - categorical_accuracy: 0.2941\n",
      "Epoch 1148/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1597 - categorical_accuracy: 0.3412\n",
      "Epoch 1149/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.4004 - categorical_accuracy: 0.3412\n",
      "Epoch 1150/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.3385 - categorical_accuracy: 0.2941\n",
      "Epoch 1151/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.2922 - categorical_accuracy: 0.3176\n",
      "Epoch 1152/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2715 - categorical_accuracy: 0.2941\n",
      "Epoch 1153/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1092 - categorical_accuracy: 0.3765\n",
      "Epoch 1154/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2220 - categorical_accuracy: 0.3294\n",
      "Epoch 1155/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.1023 - categorical_accuracy: 0.3765\n",
      "Epoch 1156/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1729 - categorical_accuracy: 0.2706\n",
      "Epoch 1157/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1328 - categorical_accuracy: 0.3412\n",
      "Epoch 1158/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.3676 - categorical_accuracy: 0.2353\n",
      "Epoch 1159/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.6442 - categorical_accuracy: 0.3647\n",
      "Epoch 1160/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 2.1724 - categorical_accuracy: 0.3176\n",
      "Epoch 1161/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 2.1968 - categorical_accuracy: 0.3176\n",
      "Epoch 1162/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.7458 - categorical_accuracy: 0.3176\n",
      "Epoch 1163/2000\n",
      "3/3 [==============================] - 0s 47ms/step - loss: 1.7139 - categorical_accuracy: 0.2824\n",
      "Epoch 1164/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.4216 - categorical_accuracy: 0.2824\n",
      "Epoch 1165/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.9131 - categorical_accuracy: 0.3412\n",
      "Epoch 1166/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 2.2167 - categorical_accuracy: 0.3412\n",
      "Epoch 1167/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.5676 - categorical_accuracy: 0.3412\n",
      "Epoch 1168/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.3093 - categorical_accuracy: 0.3529\n",
      "Epoch 1169/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.2749 - categorical_accuracy: 0.3412\n",
      "Epoch 1170/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.2881 - categorical_accuracy: 0.3176\n",
      "Epoch 1171/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.1392 - categorical_accuracy: 0.3765\n",
      "Epoch 1172/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.3102 - categorical_accuracy: 0.3059\n",
      "Epoch 1173/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.3651 - categorical_accuracy: 0.3176\n",
      "Epoch 1174/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.3768 - categorical_accuracy: 0.3176\n",
      "Epoch 1175/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.2899 - categorical_accuracy: 0.3176\n",
      "Epoch 1176/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1818 - categorical_accuracy: 0.3412\n",
      "Epoch 1177/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1444 - categorical_accuracy: 0.3412\n",
      "Epoch 1178/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1393 - categorical_accuracy: 0.2235\n",
      "Epoch 1179/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.1442 - categorical_accuracy: 0.3529\n",
      "Epoch 1180/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1273 - categorical_accuracy: 0.2941\n",
      "Epoch 1181/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.1336 - categorical_accuracy: 0.3176\n",
      "Epoch 1182/2000\n",
      "3/3 [==============================] - 0s 70ms/step - loss: 1.1078 - categorical_accuracy: 0.2824\n",
      "Epoch 1183/2000\n",
      "3/3 [==============================] - 0s 54ms/step - loss: 1.0980 - categorical_accuracy: 0.3294\n",
      "Epoch 1184/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1114 - categorical_accuracy: 0.3412\n",
      "Epoch 1185/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0973 - categorical_accuracy: 0.3294\n",
      "Epoch 1186/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.1234 - categorical_accuracy: 0.2941\n",
      "Epoch 1187/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1024 - categorical_accuracy: 0.3059\n",
      "Epoch 1188/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1124 - categorical_accuracy: 0.3529\n",
      "Epoch 1189/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1043 - categorical_accuracy: 0.3647\n",
      "Epoch 1190/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1005 - categorical_accuracy: 0.2706\n",
      "Epoch 1191/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1052 - categorical_accuracy: 0.3176\n",
      "Epoch 1192/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0907 - categorical_accuracy: 0.3882\n",
      "Epoch 1193/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0999 - categorical_accuracy: 0.3529\n",
      "Epoch 1194/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0927 - categorical_accuracy: 0.3647\n",
      "Epoch 1195/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0847 - categorical_accuracy: 0.3294\n",
      "Epoch 1196/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1170 - categorical_accuracy: 0.3176\n",
      "Epoch 1197/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.0836 - categorical_accuracy: 0.3529\n",
      "Epoch 1198/2000\n",
      "3/3 [==============================] - 0s 50ms/step - loss: 1.0882 - categorical_accuracy: 0.3882\n",
      "Epoch 1199/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0997 - categorical_accuracy: 0.4353\n",
      "Epoch 1200/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0880 - categorical_accuracy: 0.3412\n",
      "Epoch 1201/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0869 - categorical_accuracy: 0.4000\n",
      "Epoch 1202/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.0857 - categorical_accuracy: 0.5059\n",
      "Epoch 1203/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0887 - categorical_accuracy: 0.5176\n",
      "Epoch 1204/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0836 - categorical_accuracy: 0.5176\n",
      "Epoch 1205/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.0834 - categorical_accuracy: 0.4706\n",
      "Epoch 1206/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0840 - categorical_accuracy: 0.4118\n",
      "Epoch 1207/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0877 - categorical_accuracy: 0.4471\n",
      "Epoch 1208/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0864 - categorical_accuracy: 0.5176\n",
      "Epoch 1209/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0808 - categorical_accuracy: 0.4588\n",
      "Epoch 1210/2000\n",
      "3/3 [==============================] - 0s 49ms/step - loss: 1.0934 - categorical_accuracy: 0.2941\n",
      "Epoch 1211/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0953 - categorical_accuracy: 0.3882\n",
      "Epoch 1212/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0925 - categorical_accuracy: 0.3647\n",
      "Epoch 1213/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1029 - categorical_accuracy: 0.3294\n",
      "Epoch 1214/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.0913 - categorical_accuracy: 0.3412\n",
      "Epoch 1215/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1012 - categorical_accuracy: 0.4118\n",
      "Epoch 1216/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1341 - categorical_accuracy: 0.3294\n",
      "Epoch 1217/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0800 - categorical_accuracy: 0.4353\n",
      "Epoch 1218/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0845 - categorical_accuracy: 0.4824\n",
      "Epoch 1219/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0831 - categorical_accuracy: 0.4471\n",
      "Epoch 1220/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0865 - categorical_accuracy: 0.4353\n",
      "Epoch 1221/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0733 - categorical_accuracy: 0.4118\n",
      "Epoch 1222/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.1032 - categorical_accuracy: 0.3176\n",
      "Epoch 1223/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0961 - categorical_accuracy: 0.3176\n",
      "Epoch 1224/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0881 - categorical_accuracy: 0.4000\n",
      "Epoch 1225/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0916 - categorical_accuracy: 0.3647\n",
      "Epoch 1226/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1249 - categorical_accuracy: 0.3176\n",
      "Epoch 1227/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0774 - categorical_accuracy: 0.4471\n",
      "Epoch 1228/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0896 - categorical_accuracy: 0.3647\n",
      "Epoch 1229/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0943 - categorical_accuracy: 0.4000\n",
      "Epoch 1230/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0914 - categorical_accuracy: 0.3059\n",
      "Epoch 1231/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0907 - categorical_accuracy: 0.3176\n",
      "Epoch 1232/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0953 - categorical_accuracy: 0.5059\n",
      "Epoch 1233/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0867 - categorical_accuracy: 0.4471\n",
      "Epoch 1234/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0837 - categorical_accuracy: 0.5294\n",
      "Epoch 1235/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0889 - categorical_accuracy: 0.3059\n",
      "Epoch 1236/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0884 - categorical_accuracy: 0.3882\n",
      "Epoch 1237/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0864 - categorical_accuracy: 0.5059\n",
      "Epoch 1238/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0831 - categorical_accuracy: 0.3529\n",
      "Epoch 1239/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.0993 - categorical_accuracy: 0.2941\n",
      "Epoch 1240/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0963 - categorical_accuracy: 0.3882\n",
      "Epoch 1241/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0882 - categorical_accuracy: 0.4353\n",
      "Epoch 1242/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1000 - categorical_accuracy: 0.3294\n",
      "Epoch 1243/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1030 - categorical_accuracy: 0.3176\n",
      "Epoch 1244/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0762 - categorical_accuracy: 0.4471\n",
      "Epoch 1245/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.1329 - categorical_accuracy: 0.2588\n",
      "Epoch 1246/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0982 - categorical_accuracy: 0.2824\n",
      "Epoch 1247/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0912 - categorical_accuracy: 0.3765\n",
      "Epoch 1248/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.1154 - categorical_accuracy: 0.3176\n",
      "Epoch 1249/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0984 - categorical_accuracy: 0.3647\n",
      "Epoch 1250/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0980 - categorical_accuracy: 0.3882\n",
      "Epoch 1251/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0998 - categorical_accuracy: 0.3176\n",
      "Epoch 1252/2000\n",
      "3/3 [==============================] - 0s 47ms/step - loss: 1.0981 - categorical_accuracy: 0.3176\n",
      "Epoch 1253/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0933 - categorical_accuracy: 0.3059\n",
      "Epoch 1254/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0951 - categorical_accuracy: 0.3412\n",
      "Epoch 1255/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0859 - categorical_accuracy: 0.3647\n",
      "Epoch 1256/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0854 - categorical_accuracy: 0.4118\n",
      "Epoch 1257/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0912 - categorical_accuracy: 0.2706\n",
      "Epoch 1258/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.0813 - categorical_accuracy: 0.5176\n",
      "Epoch 1259/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0936 - categorical_accuracy: 0.3529\n",
      "Epoch 1260/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.1009 - categorical_accuracy: 0.4000\n",
      "Epoch 1261/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0942 - categorical_accuracy: 0.4000\n",
      "Epoch 1262/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.0848 - categorical_accuracy: 0.5059\n",
      "Epoch 1263/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0847 - categorical_accuracy: 0.4353\n",
      "Epoch 1264/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0841 - categorical_accuracy: 0.3765\n",
      "Epoch 1265/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0834 - categorical_accuracy: 0.5059\n",
      "Epoch 1266/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0888 - categorical_accuracy: 0.3882\n",
      "Epoch 1267/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0854 - categorical_accuracy: 0.5059\n",
      "Epoch 1268/2000\n",
      "3/3 [==============================] - 0s 49ms/step - loss: 1.0844 - categorical_accuracy: 0.3176\n",
      "Epoch 1269/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0854 - categorical_accuracy: 0.3294\n",
      "Epoch 1270/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0835 - categorical_accuracy: 0.4941\n",
      "Epoch 1271/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0853 - categorical_accuracy: 0.4824\n",
      "Epoch 1272/2000\n",
      "3/3 [==============================] - 0s 49ms/step - loss: 1.0836 - categorical_accuracy: 0.5176\n",
      "Epoch 1273/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0814 - categorical_accuracy: 0.5059\n",
      "Epoch 1274/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0891 - categorical_accuracy: 0.3059\n",
      "Epoch 1275/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.0965 - categorical_accuracy: 0.4000\n",
      "Epoch 1276/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0887 - categorical_accuracy: 0.4941\n",
      "Epoch 1277/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0850 - categorical_accuracy: 0.4941\n",
      "Epoch 1278/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0828 - categorical_accuracy: 0.5059\n",
      "Epoch 1279/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0895 - categorical_accuracy: 0.3765\n",
      "Epoch 1280/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0798 - categorical_accuracy: 0.4000\n",
      "Epoch 1281/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0806 - categorical_accuracy: 0.3882\n",
      "Epoch 1282/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0956 - categorical_accuracy: 0.3529\n",
      "Epoch 1283/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0916 - categorical_accuracy: 0.3529\n",
      "Epoch 1284/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0862 - categorical_accuracy: 0.4118\n",
      "Epoch 1285/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0908 - categorical_accuracy: 0.4000\n",
      "Epoch 1286/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0962 - categorical_accuracy: 0.2706\n",
      "Epoch 1287/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0864 - categorical_accuracy: 0.3176\n",
      "Epoch 1288/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0859 - categorical_accuracy: 0.4941\n",
      "Epoch 1289/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0866 - categorical_accuracy: 0.3882\n",
      "Epoch 1290/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.1239 - categorical_accuracy: 0.3412\n",
      "Epoch 1291/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.1051 - categorical_accuracy: 0.2118\n",
      "Epoch 1292/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1150 - categorical_accuracy: 0.3176\n",
      "Epoch 1293/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.1134 - categorical_accuracy: 0.3647\n",
      "Epoch 1294/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0963 - categorical_accuracy: 0.3647\n",
      "Epoch 1295/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.1476 - categorical_accuracy: 0.4000\n",
      "Epoch 1296/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1274 - categorical_accuracy: 0.2941\n",
      "Epoch 1297/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.2696 - categorical_accuracy: 0.3412\n",
      "Epoch 1298/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.1195 - categorical_accuracy: 0.3059\n",
      "Epoch 1299/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.2096 - categorical_accuracy: 0.2706\n",
      "Epoch 1300/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1017 - categorical_accuracy: 0.3647\n",
      "Epoch 1301/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1120 - categorical_accuracy: 0.4471\n",
      "Epoch 1302/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.2017 - categorical_accuracy: 0.2471\n",
      "Epoch 1303/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.1328 - categorical_accuracy: 0.3176\n",
      "Epoch 1304/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.1079 - categorical_accuracy: 0.3765\n",
      "Epoch 1305/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1825 - categorical_accuracy: 0.3176\n",
      "Epoch 1306/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1370 - categorical_accuracy: 0.3059\n",
      "Epoch 1307/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1205 - categorical_accuracy: 0.3529\n",
      "Epoch 1308/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.1179 - categorical_accuracy: 0.3176\n",
      "Epoch 1309/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.1120 - categorical_accuracy: 0.3176\n",
      "Epoch 1310/2000\n",
      "3/3 [==============================] - 0s 49ms/step - loss: 1.0923 - categorical_accuracy: 0.4353\n",
      "Epoch 1311/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0975 - categorical_accuracy: 0.3412\n",
      "Epoch 1312/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1056 - categorical_accuracy: 0.3529\n",
      "Epoch 1313/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0999 - categorical_accuracy: 0.3882\n",
      "Epoch 1314/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0946 - categorical_accuracy: 0.3529\n",
      "Epoch 1315/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0857 - categorical_accuracy: 0.3412\n",
      "Epoch 1316/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1011 - categorical_accuracy: 0.2471\n",
      "Epoch 1317/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0865 - categorical_accuracy: 0.3647\n",
      "Epoch 1318/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0881 - categorical_accuracy: 0.5176\n",
      "Epoch 1319/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0899 - categorical_accuracy: 0.4941\n",
      "Epoch 1320/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0850 - categorical_accuracy: 0.3059\n",
      "Epoch 1321/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.0858 - categorical_accuracy: 0.3412\n",
      "Epoch 1322/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0851 - categorical_accuracy: 0.2824\n",
      "Epoch 1323/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.0882 - categorical_accuracy: 0.3176\n",
      "Epoch 1324/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0852 - categorical_accuracy: 0.4941\n",
      "Epoch 1325/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0932 - categorical_accuracy: 0.3412\n",
      "Epoch 1326/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0827 - categorical_accuracy: 0.4353\n",
      "Epoch 1327/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0957 - categorical_accuracy: 0.2706\n",
      "Epoch 1328/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0987 - categorical_accuracy: 0.3176\n",
      "Epoch 1329/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0883 - categorical_accuracy: 0.4118\n",
      "Epoch 1330/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1010 - categorical_accuracy: 0.3529\n",
      "Epoch 1331/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0878 - categorical_accuracy: 0.3882\n",
      "Epoch 1332/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0831 - categorical_accuracy: 0.3882\n",
      "Epoch 1333/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0938 - categorical_accuracy: 0.3176\n",
      "Epoch 1334/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0831 - categorical_accuracy: 0.4235\n",
      "Epoch 1335/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0873 - categorical_accuracy: 0.3412\n",
      "Epoch 1336/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0894 - categorical_accuracy: 0.3765\n",
      "Epoch 1337/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0864 - categorical_accuracy: 0.3882\n",
      "Epoch 1338/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0969 - categorical_accuracy: 0.3176\n",
      "Epoch 1339/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0872 - categorical_accuracy: 0.5059\n",
      "Epoch 1340/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0960 - categorical_accuracy: 0.4235\n",
      "Epoch 1341/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0825 - categorical_accuracy: 0.5059\n",
      "Epoch 1342/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0924 - categorical_accuracy: 0.3765\n",
      "Epoch 1343/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0846 - categorical_accuracy: 0.4118\n",
      "Epoch 1344/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0885 - categorical_accuracy: 0.3176\n",
      "Epoch 1345/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1037 - categorical_accuracy: 0.3176\n",
      "Epoch 1346/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0954 - categorical_accuracy: 0.3176\n",
      "Epoch 1347/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1232 - categorical_accuracy: 0.3294\n",
      "Epoch 1348/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0915 - categorical_accuracy: 0.3529\n",
      "Epoch 1349/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0992 - categorical_accuracy: 0.3176\n",
      "Epoch 1350/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0849 - categorical_accuracy: 0.4000\n",
      "Epoch 1351/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0883 - categorical_accuracy: 0.3765\n",
      "Epoch 1352/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0869 - categorical_accuracy: 0.5294\n",
      "Epoch 1353/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.1012 - categorical_accuracy: 0.3176\n",
      "Epoch 1354/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0863 - categorical_accuracy: 0.3765\n",
      "Epoch 1355/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0898 - categorical_accuracy: 0.3765\n",
      "Epoch 1356/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0886 - categorical_accuracy: 0.3529\n",
      "Epoch 1357/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.0827 - categorical_accuracy: 0.4588\n",
      "Epoch 1358/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0964 - categorical_accuracy: 0.2941\n",
      "Epoch 1359/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0831 - categorical_accuracy: 0.4118\n",
      "Epoch 1360/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.0936 - categorical_accuracy: 0.3882\n",
      "Epoch 1361/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.1085 - categorical_accuracy: 0.3882\n",
      "Epoch 1362/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0812 - categorical_accuracy: 0.4235\n",
      "Epoch 1363/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1074 - categorical_accuracy: 0.3176\n",
      "Epoch 1364/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1203 - categorical_accuracy: 0.2471\n",
      "Epoch 1365/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.1357 - categorical_accuracy: 0.3176\n",
      "Epoch 1366/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0967 - categorical_accuracy: 0.3647\n",
      "Epoch 1367/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1168 - categorical_accuracy: 0.3176\n",
      "Epoch 1368/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.0987 - categorical_accuracy: 0.3059\n",
      "Epoch 1369/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.1320 - categorical_accuracy: 0.3529\n",
      "Epoch 1370/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.1672 - categorical_accuracy: 0.3294\n",
      "Epoch 1371/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0949 - categorical_accuracy: 0.3059\n",
      "Epoch 1372/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.1531 - categorical_accuracy: 0.3176\n",
      "Epoch 1373/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.3605 - categorical_accuracy: 0.3765\n",
      "Epoch 1374/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.1912 - categorical_accuracy: 0.4000\n",
      "Epoch 1375/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.1719 - categorical_accuracy: 0.3176\n",
      "Epoch 1376/2000\n",
      "3/3 [==============================] - 0s 55ms/step - loss: 1.1423 - categorical_accuracy: 0.3176\n",
      "Epoch 1377/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.1000 - categorical_accuracy: 0.3176\n",
      "Epoch 1378/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0799 - categorical_accuracy: 0.4824\n",
      "Epoch 1379/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0760 - categorical_accuracy: 0.4235\n",
      "Epoch 1380/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0765 - categorical_accuracy: 0.4471\n",
      "Epoch 1381/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0817 - categorical_accuracy: 0.3412\n",
      "Epoch 1382/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0824 - categorical_accuracy: 0.3412\n",
      "Epoch 1383/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0816 - categorical_accuracy: 0.4471\n",
      "Epoch 1384/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0814 - categorical_accuracy: 0.4824\n",
      "Epoch 1385/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.0769 - categorical_accuracy: 0.4824\n",
      "Epoch 1386/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0713 - categorical_accuracy: 0.4118\n",
      "Epoch 1387/2000\n",
      "3/3 [==============================] - 0s 50ms/step - loss: 1.0828 - categorical_accuracy: 0.3412\n",
      "Epoch 1388/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0779 - categorical_accuracy: 0.3882\n",
      "Epoch 1389/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0750 - categorical_accuracy: 0.4706\n",
      "Epoch 1390/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.0751 - categorical_accuracy: 0.3647\n",
      "Epoch 1391/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0759 - categorical_accuracy: 0.4118\n",
      "Epoch 1392/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0707 - categorical_accuracy: 0.4941\n",
      "Epoch 1393/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0722 - categorical_accuracy: 0.4471\n",
      "Epoch 1394/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0730 - categorical_accuracy: 0.4588\n",
      "Epoch 1395/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0720 - categorical_accuracy: 0.4000\n",
      "Epoch 1396/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0714 - categorical_accuracy: 0.4941\n",
      "Epoch 1397/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0733 - categorical_accuracy: 0.4941\n",
      "Epoch 1398/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0718 - categorical_accuracy: 0.5059\n",
      "Epoch 1399/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0720 - categorical_accuracy: 0.5059\n",
      "Epoch 1400/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0708 - categorical_accuracy: 0.5059\n",
      "Epoch 1401/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.0706 - categorical_accuracy: 0.5059\n",
      "Epoch 1402/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0706 - categorical_accuracy: 0.4941\n",
      "Epoch 1403/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0694 - categorical_accuracy: 0.4824\n",
      "Epoch 1404/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0693 - categorical_accuracy: 0.4824\n",
      "Epoch 1405/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0723 - categorical_accuracy: 0.4353\n",
      "Epoch 1406/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0719 - categorical_accuracy: 0.4353\n",
      "Epoch 1407/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0705 - categorical_accuracy: 0.5059\n",
      "Epoch 1408/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0694 - categorical_accuracy: 0.5059\n",
      "Epoch 1409/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0678 - categorical_accuracy: 0.4706\n",
      "Epoch 1410/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 1.0691 - categorical_accuracy: 0.4118\n",
      "Epoch 1411/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0703 - categorical_accuracy: 0.4353\n",
      "Epoch 1412/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0692 - categorical_accuracy: 0.4000\n",
      "Epoch 1413/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0686 - categorical_accuracy: 0.4824\n",
      "Epoch 1414/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0710 - categorical_accuracy: 0.4941\n",
      "Epoch 1415/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0686 - categorical_accuracy: 0.5059\n",
      "Epoch 1416/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0727 - categorical_accuracy: 0.4471\n",
      "Epoch 1417/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0723 - categorical_accuracy: 0.3647\n",
      "Epoch 1418/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0692 - categorical_accuracy: 0.4588\n",
      "Epoch 1419/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0696 - categorical_accuracy: 0.4941\n",
      "Epoch 1420/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0681 - categorical_accuracy: 0.5059\n",
      "Epoch 1421/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0678 - categorical_accuracy: 0.4824\n",
      "Epoch 1422/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0683 - categorical_accuracy: 0.4706\n",
      "Epoch 1423/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0675 - categorical_accuracy: 0.4824\n",
      "Epoch 1424/2000\n",
      "3/3 [==============================] - 0s 47ms/step - loss: 1.0682 - categorical_accuracy: 0.4941\n",
      "Epoch 1425/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.0694 - categorical_accuracy: 0.4706\n",
      "Epoch 1426/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0713 - categorical_accuracy: 0.3765\n",
      "Epoch 1427/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0660 - categorical_accuracy: 0.4706\n",
      "Epoch 1428/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.0730 - categorical_accuracy: 0.5059\n",
      "Epoch 1429/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0676 - categorical_accuracy: 0.5059\n",
      "Epoch 1430/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0686 - categorical_accuracy: 0.4706\n",
      "Epoch 1431/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0686 - categorical_accuracy: 0.4941\n",
      "Epoch 1432/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0675 - categorical_accuracy: 0.4941\n",
      "Epoch 1433/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0680 - categorical_accuracy: 0.4941\n",
      "Epoch 1434/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0679 - categorical_accuracy: 0.4941\n",
      "Epoch 1435/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0665 - categorical_accuracy: 0.5059\n",
      "Epoch 1436/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.0663 - categorical_accuracy: 0.4824\n",
      "Epoch 1437/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0657 - categorical_accuracy: 0.4706\n",
      "Epoch 1438/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0655 - categorical_accuracy: 0.4824\n",
      "Epoch 1439/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.0658 - categorical_accuracy: 0.5059\n",
      "Epoch 1440/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0647 - categorical_accuracy: 0.4941\n",
      "Epoch 1441/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0645 - categorical_accuracy: 0.4824\n",
      "Epoch 1442/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0735 - categorical_accuracy: 0.4118\n",
      "Epoch 1443/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0666 - categorical_accuracy: 0.4118\n",
      "Epoch 1444/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0671 - categorical_accuracy: 0.5059\n",
      "Epoch 1445/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0669 - categorical_accuracy: 0.5059\n",
      "Epoch 1446/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0661 - categorical_accuracy: 0.5059\n",
      "Epoch 1447/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0646 - categorical_accuracy: 0.4941\n",
      "Epoch 1448/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0653 - categorical_accuracy: 0.5059\n",
      "Epoch 1449/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0648 - categorical_accuracy: 0.4941\n",
      "Epoch 1450/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0644 - categorical_accuracy: 0.4824\n",
      "Epoch 1451/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0662 - categorical_accuracy: 0.4941\n",
      "Epoch 1452/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0649 - categorical_accuracy: 0.5059\n",
      "Epoch 1453/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0662 - categorical_accuracy: 0.4235\n",
      "Epoch 1454/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0649 - categorical_accuracy: 0.4118\n",
      "Epoch 1455/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0628 - categorical_accuracy: 0.4824\n",
      "Epoch 1456/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0644 - categorical_accuracy: 0.5059\n",
      "Epoch 1457/2000\n",
      "3/3 [==============================] - 0s 47ms/step - loss: 1.0669 - categorical_accuracy: 0.5059\n",
      "Epoch 1458/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0637 - categorical_accuracy: 0.4824\n",
      "Epoch 1459/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0646 - categorical_accuracy: 0.4235\n",
      "Epoch 1460/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0645 - categorical_accuracy: 0.4235\n",
      "Epoch 1461/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0639 - categorical_accuracy: 0.4000\n",
      "Epoch 1462/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0632 - categorical_accuracy: 0.4706\n",
      "Epoch 1463/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0655 - categorical_accuracy: 0.4941\n",
      "Epoch 1464/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0623 - categorical_accuracy: 0.5059\n",
      "Epoch 1465/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0627 - categorical_accuracy: 0.4941\n",
      "Epoch 1466/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0633 - categorical_accuracy: 0.4588\n",
      "Epoch 1467/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0632 - categorical_accuracy: 0.4824\n",
      "Epoch 1468/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0621 - categorical_accuracy: 0.4941\n",
      "Epoch 1469/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0628 - categorical_accuracy: 0.4941\n",
      "Epoch 1470/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0622 - categorical_accuracy: 0.5059\n",
      "Epoch 1471/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.0620 - categorical_accuracy: 0.4941\n",
      "Epoch 1472/2000\n",
      "3/3 [==============================] - 0s 47ms/step - loss: 1.0617 - categorical_accuracy: 0.4941\n",
      "Epoch 1473/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0621 - categorical_accuracy: 0.5059\n",
      "Epoch 1474/2000\n",
      "3/3 [==============================] - 0s 51ms/step - loss: 1.0617 - categorical_accuracy: 0.4941\n",
      "Epoch 1475/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0622 - categorical_accuracy: 0.4824\n",
      "Epoch 1476/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0625 - categorical_accuracy: 0.5059\n",
      "Epoch 1477/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.0619 - categorical_accuracy: 0.5059\n",
      "Epoch 1478/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0619 - categorical_accuracy: 0.5059\n",
      "Epoch 1479/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0604 - categorical_accuracy: 0.4824\n",
      "Epoch 1480/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0640 - categorical_accuracy: 0.4235\n",
      "Epoch 1481/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0624 - categorical_accuracy: 0.4588\n",
      "Epoch 1482/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0621 - categorical_accuracy: 0.5059\n",
      "Epoch 1483/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0607 - categorical_accuracy: 0.5059\n",
      "Epoch 1484/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0600 - categorical_accuracy: 0.4941\n",
      "Epoch 1485/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0658 - categorical_accuracy: 0.3882\n",
      "Epoch 1486/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0610 - categorical_accuracy: 0.4471\n",
      "Epoch 1487/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0601 - categorical_accuracy: 0.5059\n",
      "Epoch 1488/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0609 - categorical_accuracy: 0.5059\n",
      "Epoch 1489/2000\n",
      "3/3 [==============================] - 0s 49ms/step - loss: 1.0588 - categorical_accuracy: 0.5176\n",
      "Epoch 1490/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0606 - categorical_accuracy: 0.4118\n",
      "Epoch 1491/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0619 - categorical_accuracy: 0.4118\n",
      "Epoch 1492/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0620 - categorical_accuracy: 0.4235\n",
      "Epoch 1493/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0592 - categorical_accuracy: 0.4353\n",
      "Epoch 1494/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.0599 - categorical_accuracy: 0.5059\n",
      "Epoch 1495/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0605 - categorical_accuracy: 0.4941\n",
      "Epoch 1496/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0584 - categorical_accuracy: 0.4941\n",
      "Epoch 1497/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0592 - categorical_accuracy: 0.4824\n",
      "Epoch 1498/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0593 - categorical_accuracy: 0.5176\n",
      "Epoch 1499/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0601 - categorical_accuracy: 0.4941\n",
      "Epoch 1500/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0580 - categorical_accuracy: 0.5176\n",
      "Epoch 1501/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0591 - categorical_accuracy: 0.5176\n",
      "Epoch 1502/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0581 - categorical_accuracy: 0.5176\n",
      "Epoch 1503/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0582 - categorical_accuracy: 0.4941\n",
      "Epoch 1504/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.0643 - categorical_accuracy: 0.4706\n",
      "Epoch 1505/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0588 - categorical_accuracy: 0.4706\n",
      "Epoch 1506/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0588 - categorical_accuracy: 0.4235\n",
      "Epoch 1507/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.0559 - categorical_accuracy: 0.4824\n",
      "Epoch 1508/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0610 - categorical_accuracy: 0.5176\n",
      "Epoch 1509/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0590 - categorical_accuracy: 0.5059\n",
      "Epoch 1510/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.0574 - categorical_accuracy: 0.4588\n",
      "Epoch 1511/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0602 - categorical_accuracy: 0.4353\n",
      "Epoch 1512/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.0593 - categorical_accuracy: 0.4706\n",
      "Epoch 1513/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0609 - categorical_accuracy: 0.4941\n",
      "Epoch 1514/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0562 - categorical_accuracy: 0.5059\n",
      "Epoch 1515/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0556 - categorical_accuracy: 0.4588\n",
      "Epoch 1516/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0610 - categorical_accuracy: 0.4588\n",
      "Epoch 1517/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0588 - categorical_accuracy: 0.4471\n",
      "Epoch 1518/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0553 - categorical_accuracy: 0.4941\n",
      "Epoch 1519/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0592 - categorical_accuracy: 0.4941\n",
      "Epoch 1520/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0567 - categorical_accuracy: 0.5059\n",
      "Epoch 1521/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0618 - categorical_accuracy: 0.4118\n",
      "Epoch 1522/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.0614 - categorical_accuracy: 0.4000\n",
      "Epoch 1523/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0557 - categorical_accuracy: 0.4941\n",
      "Epoch 1524/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0576 - categorical_accuracy: 0.4941\n",
      "Epoch 1525/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0588 - categorical_accuracy: 0.5059\n",
      "Epoch 1526/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0561 - categorical_accuracy: 0.4941\n",
      "Epoch 1527/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0560 - categorical_accuracy: 0.5176\n",
      "Epoch 1528/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0573 - categorical_accuracy: 0.5059\n",
      "Epoch 1529/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0569 - categorical_accuracy: 0.5176\n",
      "Epoch 1530/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0589 - categorical_accuracy: 0.5059\n",
      "Epoch 1531/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0540 - categorical_accuracy: 0.4824\n",
      "Epoch 1532/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0540 - categorical_accuracy: 0.4941\n",
      "Epoch 1533/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0552 - categorical_accuracy: 0.4824\n",
      "Epoch 1534/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0533 - categorical_accuracy: 0.4824\n",
      "Epoch 1535/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0525 - categorical_accuracy: 0.5176\n",
      "Epoch 1536/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0551 - categorical_accuracy: 0.5059\n",
      "Epoch 1537/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.0581 - categorical_accuracy: 0.4941\n",
      "Epoch 1538/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0534 - categorical_accuracy: 0.4706\n",
      "Epoch 1539/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0568 - categorical_accuracy: 0.4235\n",
      "Epoch 1540/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0547 - categorical_accuracy: 0.4000\n",
      "Epoch 1541/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.0596 - categorical_accuracy: 0.4941\n",
      "Epoch 1542/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0581 - categorical_accuracy: 0.4941\n",
      "Epoch 1543/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0544 - categorical_accuracy: 0.4118\n",
      "Epoch 1544/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0558 - categorical_accuracy: 0.4353\n",
      "Epoch 1545/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0560 - categorical_accuracy: 0.5059\n",
      "Epoch 1546/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.0569 - categorical_accuracy: 0.5059\n",
      "Epoch 1547/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0530 - categorical_accuracy: 0.4235\n",
      "Epoch 1548/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0537 - categorical_accuracy: 0.4471\n",
      "Epoch 1549/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0579 - categorical_accuracy: 0.4353\n",
      "Epoch 1550/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0622 - categorical_accuracy: 0.4824\n",
      "Epoch 1551/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0522 - categorical_accuracy: 0.5176\n",
      "Epoch 1552/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0578 - categorical_accuracy: 0.4235\n",
      "Epoch 1553/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0646 - categorical_accuracy: 0.3882\n",
      "Epoch 1554/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0515 - categorical_accuracy: 0.4706\n",
      "Epoch 1555/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0544 - categorical_accuracy: 0.5059\n",
      "Epoch 1556/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.0577 - categorical_accuracy: 0.4941\n",
      "Epoch 1557/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0514 - categorical_accuracy: 0.4471\n",
      "Epoch 1558/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0500 - categorical_accuracy: 0.4824\n",
      "Epoch 1559/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.0554 - categorical_accuracy: 0.4824\n",
      "Epoch 1560/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0514 - categorical_accuracy: 0.5059\n",
      "Epoch 1561/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0550 - categorical_accuracy: 0.4000\n",
      "Epoch 1562/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0493 - categorical_accuracy: 0.5059\n",
      "Epoch 1563/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0519 - categorical_accuracy: 0.5176\n",
      "Epoch 1564/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.0501 - categorical_accuracy: 0.5059\n",
      "Epoch 1565/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0500 - categorical_accuracy: 0.4824\n",
      "Epoch 1566/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0552 - categorical_accuracy: 0.4353\n",
      "Epoch 1567/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.0470 - categorical_accuracy: 0.4824\n",
      "Epoch 1568/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0588 - categorical_accuracy: 0.5059\n",
      "Epoch 1569/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0503 - categorical_accuracy: 0.5176\n",
      "Epoch 1570/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0491 - categorical_accuracy: 0.4235\n",
      "Epoch 1571/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0534 - categorical_accuracy: 0.4471\n",
      "Epoch 1572/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0481 - categorical_accuracy: 0.5176\n",
      "Epoch 1573/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0550 - categorical_accuracy: 0.5059\n",
      "Epoch 1574/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.0460 - categorical_accuracy: 0.5059\n",
      "Epoch 1575/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0594 - categorical_accuracy: 0.3765\n",
      "Epoch 1576/2000\n",
      "3/3 [==============================] - 0s 47ms/step - loss: 1.0558 - categorical_accuracy: 0.4471\n",
      "Epoch 1577/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0511 - categorical_accuracy: 0.4941\n",
      "Epoch 1578/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0528 - categorical_accuracy: 0.5059\n",
      "Epoch 1579/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.0510 - categorical_accuracy: 0.4353\n",
      "Epoch 1580/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0493 - categorical_accuracy: 0.4118\n",
      "Epoch 1581/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0477 - categorical_accuracy: 0.4353\n",
      "Epoch 1582/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0593 - categorical_accuracy: 0.4706\n",
      "Epoch 1583/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0454 - categorical_accuracy: 0.5176\n",
      "Epoch 1584/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0505 - categorical_accuracy: 0.4471\n",
      "Epoch 1585/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0524 - categorical_accuracy: 0.4471\n",
      "Epoch 1586/2000\n",
      "3/3 [==============================] - 0s 48ms/step - loss: 1.0456 - categorical_accuracy: 0.5176\n",
      "Epoch 1587/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0491 - categorical_accuracy: 0.5176\n",
      "Epoch 1588/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0530 - categorical_accuracy: 0.4235\n",
      "Epoch 1589/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0459 - categorical_accuracy: 0.4235\n",
      "Epoch 1590/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0505 - categorical_accuracy: 0.5059\n",
      "Epoch 1591/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0502 - categorical_accuracy: 0.5176\n",
      "Epoch 1592/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0489 - categorical_accuracy: 0.4471\n",
      "Epoch 1593/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0507 - categorical_accuracy: 0.4000\n",
      "Epoch 1594/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0628 - categorical_accuracy: 0.5059\n",
      "Epoch 1595/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0470 - categorical_accuracy: 0.5176\n",
      "Epoch 1596/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0444 - categorical_accuracy: 0.4471\n",
      "Epoch 1597/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0569 - categorical_accuracy: 0.3765\n",
      "Epoch 1598/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0427 - categorical_accuracy: 0.4824\n",
      "Epoch 1599/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0546 - categorical_accuracy: 0.5059\n",
      "Epoch 1600/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0505 - categorical_accuracy: 0.5059\n",
      "Epoch 1601/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0429 - categorical_accuracy: 0.5059\n",
      "Epoch 1602/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0440 - categorical_accuracy: 0.5059\n",
      "Epoch 1603/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0461 - categorical_accuracy: 0.4824\n",
      "Epoch 1604/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0434 - categorical_accuracy: 0.5059\n",
      "Epoch 1605/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0451 - categorical_accuracy: 0.5059\n",
      "Epoch 1606/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0433 - categorical_accuracy: 0.4941\n",
      "Epoch 1607/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0436 - categorical_accuracy: 0.4353\n",
      "Epoch 1608/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0445 - categorical_accuracy: 0.5176\n",
      "Epoch 1609/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0448 - categorical_accuracy: 0.5059\n",
      "Epoch 1610/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0403 - categorical_accuracy: 0.4941\n",
      "Epoch 1611/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0451 - categorical_accuracy: 0.4118\n",
      "Epoch 1612/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0411 - categorical_accuracy: 0.4824\n",
      "Epoch 1613/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0473 - categorical_accuracy: 0.5059\n",
      "Epoch 1614/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.0431 - categorical_accuracy: 0.4941\n",
      "Epoch 1615/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0484 - categorical_accuracy: 0.5176\n",
      "Epoch 1616/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0408 - categorical_accuracy: 0.5059\n",
      "Epoch 1617/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0427 - categorical_accuracy: 0.4353\n",
      "Epoch 1618/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0498 - categorical_accuracy: 0.4941\n",
      "Epoch 1619/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 1.0408 - categorical_accuracy: 0.5176\n",
      "Epoch 1620/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0421 - categorical_accuracy: 0.4353\n",
      "Epoch 1621/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0409 - categorical_accuracy: 0.4824\n",
      "Epoch 1622/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0417 - categorical_accuracy: 0.5176\n",
      "Epoch 1623/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0401 - categorical_accuracy: 0.5176\n",
      "Epoch 1624/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0390 - categorical_accuracy: 0.4824\n",
      "Epoch 1625/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0413 - categorical_accuracy: 0.4471\n",
      "Epoch 1626/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0407 - categorical_accuracy: 0.4706\n",
      "Epoch 1627/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0561 - categorical_accuracy: 0.4235\n",
      "Epoch 1628/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0483 - categorical_accuracy: 0.4824\n",
      "Epoch 1629/2000\n",
      "3/3 [==============================] - 0s 47ms/step - loss: 1.0483 - categorical_accuracy: 0.5059\n",
      "Epoch 1630/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0546 - categorical_accuracy: 0.4588\n",
      "Epoch 1631/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0417 - categorical_accuracy: 0.4588\n",
      "Epoch 1632/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0498 - categorical_accuracy: 0.4824\n",
      "Epoch 1633/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 1.0452 - categorical_accuracy: 0.4941\n",
      "Epoch 1634/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0461 - categorical_accuracy: 0.5176\n",
      "Epoch 1635/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0412 - categorical_accuracy: 0.4941\n",
      "Epoch 1636/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0494 - categorical_accuracy: 0.5059\n",
      "Epoch 1637/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0398 - categorical_accuracy: 0.4941\n",
      "Epoch 1638/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0422 - categorical_accuracy: 0.4118\n",
      "Epoch 1639/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0385 - categorical_accuracy: 0.4588\n",
      "Epoch 1640/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0600 - categorical_accuracy: 0.5059\n",
      "Epoch 1641/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0369 - categorical_accuracy: 0.4588\n",
      "Epoch 1642/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0462 - categorical_accuracy: 0.4235\n",
      "Epoch 1643/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0307 - categorical_accuracy: 0.4824\n",
      "Epoch 1644/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0839 - categorical_accuracy: 0.4118\n",
      "Epoch 1645/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0402 - categorical_accuracy: 0.5176\n",
      "Epoch 1646/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 1.0629 - categorical_accuracy: 0.3647\n",
      "Epoch 1647/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0388 - categorical_accuracy: 0.5294\n",
      "Epoch 1648/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0462 - categorical_accuracy: 0.5059\n",
      "Epoch 1649/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.0427 - categorical_accuracy: 0.5176\n",
      "Epoch 1650/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 1.0413 - categorical_accuracy: 0.4588\n",
      "Epoch 1651/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0354 - categorical_accuracy: 0.4824\n",
      "Epoch 1652/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0361 - categorical_accuracy: 0.4824\n",
      "Epoch 1653/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0452 - categorical_accuracy: 0.4118\n",
      "Epoch 1654/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0535 - categorical_accuracy: 0.4941\n",
      "Epoch 1655/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0467 - categorical_accuracy: 0.5059\n",
      "Epoch 1656/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0445 - categorical_accuracy: 0.4471\n",
      "Epoch 1657/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0324 - categorical_accuracy: 0.4824\n",
      "Epoch 1658/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 1.0319 - categorical_accuracy: 0.4824\n",
      "Epoch 1659/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0405 - categorical_accuracy: 0.5294\n",
      "Epoch 1660/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0291 - categorical_accuracy: 0.4941\n",
      "Epoch 1661/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0445 - categorical_accuracy: 0.4471\n",
      "Epoch 1662/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0394 - categorical_accuracy: 0.4706\n",
      "Epoch 1663/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.0298 - categorical_accuracy: 0.5176\n",
      "Epoch 1664/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0298 - categorical_accuracy: 0.4706\n",
      "Epoch 1665/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 1.0285 - categorical_accuracy: 0.5059\n",
      "Epoch 1666/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0314 - categorical_accuracy: 0.4941\n",
      "Epoch 1667/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0284 - categorical_accuracy: 0.5294\n",
      "Epoch 1668/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0268 - categorical_accuracy: 0.4824\n",
      "Epoch 1669/2000\n",
      "3/3 [==============================] - 0s 63ms/step - loss: 1.0220 - categorical_accuracy: 0.4824\n",
      "Epoch 1670/2000\n",
      "3/3 [==============================] - 0s 50ms/step - loss: 1.0215 - categorical_accuracy: 0.5294\n",
      "Epoch 1671/2000\n",
      "3/3 [==============================] - 0s 52ms/step - loss: 1.0370 - categorical_accuracy: 0.4353\n",
      "Epoch 1672/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0197 - categorical_accuracy: 0.5176\n",
      "Epoch 1673/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0235 - categorical_accuracy: 0.5176\n",
      "Epoch 1674/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0100 - categorical_accuracy: 0.4706\n",
      "Epoch 1675/2000\n",
      "3/3 [==============================] - 0s 65ms/step - loss: 1.0295 - categorical_accuracy: 0.4353\n",
      "Epoch 1676/2000\n",
      "3/3 [==============================] - 0s 48ms/step - loss: 1.0444 - categorical_accuracy: 0.5294\n",
      "Epoch 1677/2000\n",
      "3/3 [==============================] - 0s 48ms/step - loss: 1.0105 - categorical_accuracy: 0.5294\n",
      "Epoch 1678/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0326 - categorical_accuracy: 0.4824\n",
      "Epoch 1679/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0049 - categorical_accuracy: 0.4824\n",
      "Epoch 1680/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.0246 - categorical_accuracy: 0.5059\n",
      "Epoch 1681/2000\n",
      "3/3 [==============================] - 0s 68ms/step - loss: 1.0057 - categorical_accuracy: 0.5176\n",
      "Epoch 1682/2000\n",
      "3/3 [==============================] - 0s 47ms/step - loss: 1.0236 - categorical_accuracy: 0.4471\n",
      "Epoch 1683/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0106 - categorical_accuracy: 0.5176\n",
      "Epoch 1684/2000\n",
      "3/3 [==============================] - 0s 48ms/step - loss: 1.0174 - categorical_accuracy: 0.5294\n",
      "Epoch 1685/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0201 - categorical_accuracy: 0.5059\n",
      "Epoch 1686/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0203 - categorical_accuracy: 0.4588\n",
      "Epoch 1687/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.0039 - categorical_accuracy: 0.5294\n",
      "Epoch 1688/2000\n",
      "3/3 [==============================] - 0s 55ms/step - loss: 1.0013 - categorical_accuracy: 0.5176\n",
      "Epoch 1689/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.0013 - categorical_accuracy: 0.5412\n",
      "Epoch 1690/2000\n",
      "3/3 [==============================] - 0s 48ms/step - loss: 1.0048 - categorical_accuracy: 0.5059\n",
      "Epoch 1691/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0018 - categorical_accuracy: 0.5059\n",
      "Epoch 1692/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 0.9927 - categorical_accuracy: 0.5176\n",
      "Epoch 1693/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 0.9990 - categorical_accuracy: 0.5059\n",
      "Epoch 1694/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 0.9862 - categorical_accuracy: 0.5059\n",
      "Epoch 1695/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0263 - categorical_accuracy: 0.5294\n",
      "Epoch 1696/2000\n",
      "3/3 [==============================] - 0s 47ms/step - loss: 0.9805 - categorical_accuracy: 0.5412\n",
      "Epoch 1697/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.0328 - categorical_accuracy: 0.4000\n",
      "Epoch 1698/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 1.0132 - categorical_accuracy: 0.4824\n",
      "Epoch 1699/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 1.0002 - categorical_accuracy: 0.4941\n",
      "Epoch 1700/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0305 - categorical_accuracy: 0.4118\n",
      "Epoch 1701/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 0.9920 - categorical_accuracy: 0.5176\n",
      "Epoch 1702/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 0.9825 - categorical_accuracy: 0.5294\n",
      "Epoch 1703/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 0.9869 - categorical_accuracy: 0.5412\n",
      "Epoch 1704/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.9934 - categorical_accuracy: 0.5059\n",
      "Epoch 1705/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.9805 - categorical_accuracy: 0.4824\n",
      "Epoch 1706/2000\n",
      "3/3 [==============================] - 0s 49ms/step - loss: 0.9922 - categorical_accuracy: 0.5176\n",
      "Epoch 1707/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 0.9789 - categorical_accuracy: 0.5294\n",
      "Epoch 1708/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.9879 - categorical_accuracy: 0.5294\n",
      "Epoch 1709/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 0.9952 - categorical_accuracy: 0.5176\n",
      "Epoch 1710/2000\n",
      "3/3 [==============================] - 0s 50ms/step - loss: 0.9793 - categorical_accuracy: 0.5176\n",
      "Epoch 1711/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0014 - categorical_accuracy: 0.4588\n",
      "Epoch 1712/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.9923 - categorical_accuracy: 0.5176\n",
      "Epoch 1713/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 0.9899 - categorical_accuracy: 0.5294\n",
      "Epoch 1714/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.9804 - categorical_accuracy: 0.4941\n",
      "Epoch 1715/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 0.9740 - categorical_accuracy: 0.4824\n",
      "Epoch 1716/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 0.9696 - categorical_accuracy: 0.5294\n",
      "Epoch 1717/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.9712 - categorical_accuracy: 0.5412\n",
      "Epoch 1718/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 0.9900 - categorical_accuracy: 0.5059\n",
      "Epoch 1719/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.9724 - categorical_accuracy: 0.5294\n",
      "Epoch 1720/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.9736 - categorical_accuracy: 0.5294\n",
      "Epoch 1721/2000\n",
      "3/3 [==============================] - 0s 47ms/step - loss: 0.9684 - categorical_accuracy: 0.5765\n",
      "Epoch 1722/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.9586 - categorical_accuracy: 0.5294\n",
      "Epoch 1723/2000\n",
      "3/3 [==============================] - 0s 49ms/step - loss: 0.9683 - categorical_accuracy: 0.5412\n",
      "Epoch 1724/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.9646 - categorical_accuracy: 0.5059\n",
      "Epoch 1725/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.9617 - categorical_accuracy: 0.5647\n",
      "Epoch 1726/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.9654 - categorical_accuracy: 0.5412\n",
      "Epoch 1727/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.9558 - categorical_accuracy: 0.5294\n",
      "Epoch 1728/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 0.9526 - categorical_accuracy: 0.5294\n",
      "Epoch 1729/2000\n",
      "3/3 [==============================] - 0s 51ms/step - loss: 0.9546 - categorical_accuracy: 0.5294\n",
      "Epoch 1730/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.9642 - categorical_accuracy: 0.5294\n",
      "Epoch 1731/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.9449 - categorical_accuracy: 0.5294\n",
      "Epoch 1732/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 0.9682 - categorical_accuracy: 0.5176\n",
      "Epoch 1733/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.9559 - categorical_accuracy: 0.5294\n",
      "Epoch 1734/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.9705 - categorical_accuracy: 0.5176\n",
      "Epoch 1735/2000\n",
      "3/3 [==============================] - 0s 47ms/step - loss: 0.9429 - categorical_accuracy: 0.5765\n",
      "Epoch 1736/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.9447 - categorical_accuracy: 0.5294\n",
      "Epoch 1737/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 0.9592 - categorical_accuracy: 0.4706\n",
      "Epoch 1738/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.9413 - categorical_accuracy: 0.5059\n",
      "Epoch 1739/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.9517 - categorical_accuracy: 0.5412\n",
      "Epoch 1740/2000\n",
      "3/3 [==============================] - 0s 49ms/step - loss: 0.9487 - categorical_accuracy: 0.5294\n",
      "Epoch 1741/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 1.0620 - categorical_accuracy: 0.4588\n",
      "Epoch 1742/2000\n",
      "3/3 [==============================] - 0s 48ms/step - loss: 0.9345 - categorical_accuracy: 0.5412\n",
      "Epoch 1743/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.9762 - categorical_accuracy: 0.4941\n",
      "Epoch 1744/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.9460 - categorical_accuracy: 0.5412\n",
      "Epoch 1745/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.9479 - categorical_accuracy: 0.5765\n",
      "Epoch 1746/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.9419 - categorical_accuracy: 0.5765\n",
      "Epoch 1747/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.9289 - categorical_accuracy: 0.5294\n",
      "Epoch 1748/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.9266 - categorical_accuracy: 0.5412\n",
      "Epoch 1749/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.9306 - categorical_accuracy: 0.5294\n",
      "Epoch 1750/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.9135 - categorical_accuracy: 0.5529\n",
      "Epoch 1751/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 0.9265 - categorical_accuracy: 0.5412\n",
      "Epoch 1752/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 0.8997 - categorical_accuracy: 0.5294\n",
      "Epoch 1753/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.8839 - categorical_accuracy: 0.5294\n",
      "Epoch 1754/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.9603 - categorical_accuracy: 0.5176\n",
      "Epoch 1755/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 1.1966 - categorical_accuracy: 0.4118\n",
      "Epoch 1756/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.9406 - categorical_accuracy: 0.5529\n",
      "Epoch 1757/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 1.0579 - categorical_accuracy: 0.3529\n",
      "Epoch 1758/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 0.9438 - categorical_accuracy: 0.5294\n",
      "Epoch 1759/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.9855 - categorical_accuracy: 0.5176\n",
      "Epoch 1760/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.9680 - categorical_accuracy: 0.5294\n",
      "Epoch 1761/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 1.0088 - categorical_accuracy: 0.4471\n",
      "Epoch 1762/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 0.9772 - categorical_accuracy: 0.5412\n",
      "Epoch 1763/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 0.9772 - categorical_accuracy: 0.5529\n",
      "Epoch 1764/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.9714 - categorical_accuracy: 0.5647\n",
      "Epoch 1765/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.9509 - categorical_accuracy: 0.5294\n",
      "Epoch 1766/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.9518 - categorical_accuracy: 0.5294\n",
      "Epoch 1767/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.9447 - categorical_accuracy: 0.5412\n",
      "Epoch 1768/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.9390 - categorical_accuracy: 0.5294\n",
      "Epoch 1769/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.9374 - categorical_accuracy: 0.5412\n",
      "Epoch 1770/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.9347 - categorical_accuracy: 0.5294\n",
      "Epoch 1771/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.9322 - categorical_accuracy: 0.5294\n",
      "Epoch 1772/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.9308 - categorical_accuracy: 0.5294\n",
      "Epoch 1773/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.9285 - categorical_accuracy: 0.5529\n",
      "Epoch 1774/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.9250 - categorical_accuracy: 0.5294\n",
      "Epoch 1775/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.9318 - categorical_accuracy: 0.5412\n",
      "Epoch 1776/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.9253 - categorical_accuracy: 0.5294\n",
      "Epoch 1777/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.9232 - categorical_accuracy: 0.5294\n",
      "Epoch 1778/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.9205 - categorical_accuracy: 0.5412\n",
      "Epoch 1779/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.9223 - categorical_accuracy: 0.5176\n",
      "Epoch 1780/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.9064 - categorical_accuracy: 0.5294\n",
      "Epoch 1781/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.9133 - categorical_accuracy: 0.5882\n",
      "Epoch 1782/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.9051 - categorical_accuracy: 0.5176\n",
      "Epoch 1783/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.9150 - categorical_accuracy: 0.5412\n",
      "Epoch 1784/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.9397 - categorical_accuracy: 0.5294\n",
      "Epoch 1785/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.9157 - categorical_accuracy: 0.5176\n",
      "Epoch 1786/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.9116 - categorical_accuracy: 0.5294\n",
      "Epoch 1787/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.9481 - categorical_accuracy: 0.5176\n",
      "Epoch 1788/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.9109 - categorical_accuracy: 0.5294\n",
      "Epoch 1789/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.8780 - categorical_accuracy: 0.5765\n",
      "Epoch 1790/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.8948 - categorical_accuracy: 0.5412\n",
      "Epoch 1791/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.9261 - categorical_accuracy: 0.5412\n",
      "Epoch 1792/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.9028 - categorical_accuracy: 0.4588\n",
      "Epoch 1793/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.8907 - categorical_accuracy: 0.5647\n",
      "Epoch 1794/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.9429 - categorical_accuracy: 0.5294\n",
      "Epoch 1795/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.8701 - categorical_accuracy: 0.5765\n",
      "Epoch 1796/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.9076 - categorical_accuracy: 0.4824\n",
      "Epoch 1797/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.8693 - categorical_accuracy: 0.5412\n",
      "Epoch 1798/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.9232 - categorical_accuracy: 0.5294\n",
      "Epoch 1799/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.8707 - categorical_accuracy: 0.5647\n",
      "Epoch 1800/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.8826 - categorical_accuracy: 0.5294\n",
      "Epoch 1801/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.8867 - categorical_accuracy: 0.5647\n",
      "Epoch 1802/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 0.8579 - categorical_accuracy: 0.5294\n",
      "Epoch 1803/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.8614 - categorical_accuracy: 0.5412\n",
      "Epoch 1804/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.8745 - categorical_accuracy: 0.5412\n",
      "Epoch 1805/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.8393 - categorical_accuracy: 0.5647\n",
      "Epoch 1806/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.8301 - categorical_accuracy: 0.6235\n",
      "Epoch 1807/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.8400 - categorical_accuracy: 0.5412\n",
      "Epoch 1808/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.8059 - categorical_accuracy: 0.6235\n",
      "Epoch 1809/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.8246 - categorical_accuracy: 0.6353\n",
      "Epoch 1810/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 0.8004 - categorical_accuracy: 0.5412\n",
      "Epoch 1811/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.7807 - categorical_accuracy: 0.6353\n",
      "Epoch 1812/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.7868 - categorical_accuracy: 0.6588\n",
      "Epoch 1813/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.7801 - categorical_accuracy: 0.5412\n",
      "Epoch 1814/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.7594 - categorical_accuracy: 0.5647\n",
      "Epoch 1815/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.7579 - categorical_accuracy: 0.6471\n",
      "Epoch 1816/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.7407 - categorical_accuracy: 0.5882\n",
      "Epoch 1817/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.7165 - categorical_accuracy: 0.6235\n",
      "Epoch 1818/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.7379 - categorical_accuracy: 0.6353\n",
      "Epoch 1819/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.7103 - categorical_accuracy: 0.6235\n",
      "Epoch 1820/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.6984 - categorical_accuracy: 0.6471\n",
      "Epoch 1821/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.6885 - categorical_accuracy: 0.6588\n",
      "Epoch 1822/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.6967 - categorical_accuracy: 0.6353\n",
      "Epoch 1823/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 0.6922 - categorical_accuracy: 0.6353\n",
      "Epoch 1824/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.6972 - categorical_accuracy: 0.6118\n",
      "Epoch 1825/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.7016 - categorical_accuracy: 0.6471\n",
      "Epoch 1826/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.6666 - categorical_accuracy: 0.6353\n",
      "Epoch 1827/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.6714 - categorical_accuracy: 0.6471\n",
      "Epoch 1828/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.6627 - categorical_accuracy: 0.6941\n",
      "Epoch 1829/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.6638 - categorical_accuracy: 0.5882\n",
      "Epoch 1830/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 0.6651 - categorical_accuracy: 0.6588\n",
      "Epoch 1831/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.6509 - categorical_accuracy: 0.6353\n",
      "Epoch 1832/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.6469 - categorical_accuracy: 0.6706\n",
      "Epoch 1833/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.6447 - categorical_accuracy: 0.6824\n",
      "Epoch 1834/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.6350 - categorical_accuracy: 0.6706\n",
      "Epoch 1835/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.6318 - categorical_accuracy: 0.6706\n",
      "Epoch 1836/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.6666 - categorical_accuracy: 0.6353\n",
      "Epoch 1837/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 0.6591 - categorical_accuracy: 0.6353\n",
      "Epoch 1838/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.6393 - categorical_accuracy: 0.6471\n",
      "Epoch 1839/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.6596 - categorical_accuracy: 0.5765\n",
      "Epoch 1840/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.6676 - categorical_accuracy: 0.6353\n",
      "Epoch 1841/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.6247 - categorical_accuracy: 0.7059\n",
      "Epoch 1842/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.6542 - categorical_accuracy: 0.6941\n",
      "Epoch 1843/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.6411 - categorical_accuracy: 0.6706\n",
      "Epoch 1844/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.6302 - categorical_accuracy: 0.6471\n",
      "Epoch 1845/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.6376 - categorical_accuracy: 0.6588\n",
      "Epoch 1846/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.6124 - categorical_accuracy: 0.6706\n",
      "Epoch 1847/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.6091 - categorical_accuracy: 0.6706\n",
      "Epoch 1848/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.6265 - categorical_accuracy: 0.6588\n",
      "Epoch 1849/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.6144 - categorical_accuracy: 0.7059\n",
      "Epoch 1850/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.6111 - categorical_accuracy: 0.6706\n",
      "Epoch 1851/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.6386 - categorical_accuracy: 0.6471\n",
      "Epoch 1852/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.5947 - categorical_accuracy: 0.6353\n",
      "Epoch 1853/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.6515 - categorical_accuracy: 0.6353\n",
      "Epoch 1854/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.6730 - categorical_accuracy: 0.6588\n",
      "Epoch 1855/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.6471 - categorical_accuracy: 0.6353\n",
      "Epoch 1856/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.6395 - categorical_accuracy: 0.6353\n",
      "Epoch 1857/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 0.6070 - categorical_accuracy: 0.6706\n",
      "Epoch 1858/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.6076 - categorical_accuracy: 0.6706\n",
      "Epoch 1859/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.6161 - categorical_accuracy: 0.6824\n",
      "Epoch 1860/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 0.6077 - categorical_accuracy: 0.7176\n",
      "Epoch 1861/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.6098 - categorical_accuracy: 0.6706\n",
      "Epoch 1862/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.6137 - categorical_accuracy: 0.6353\n",
      "Epoch 1863/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.6157 - categorical_accuracy: 0.6353\n",
      "Epoch 1864/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.5997 - categorical_accuracy: 0.6471\n",
      "Epoch 1865/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.6027 - categorical_accuracy: 0.7059\n",
      "Epoch 1866/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.5987 - categorical_accuracy: 0.7176\n",
      "Epoch 1867/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.6051 - categorical_accuracy: 0.7176\n",
      "Epoch 1868/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.5986 - categorical_accuracy: 0.6824\n",
      "Epoch 1869/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.5966 - categorical_accuracy: 0.6706\n",
      "Epoch 1870/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.5956 - categorical_accuracy: 0.6824\n",
      "Epoch 1871/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.5905 - categorical_accuracy: 0.7059\n",
      "Epoch 1872/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.5957 - categorical_accuracy: 0.6588\n",
      "Epoch 1873/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.6039 - categorical_accuracy: 0.6824\n",
      "Epoch 1874/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.6063 - categorical_accuracy: 0.6706\n",
      "Epoch 1875/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.6017 - categorical_accuracy: 0.6824\n",
      "Epoch 1876/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.5858 - categorical_accuracy: 0.7176\n",
      "Epoch 1877/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.5941 - categorical_accuracy: 0.7176\n",
      "Epoch 1878/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 0.6161 - categorical_accuracy: 0.6706\n",
      "Epoch 1879/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.6107 - categorical_accuracy: 0.6824\n",
      "Epoch 1880/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.6056 - categorical_accuracy: 0.7176\n",
      "Epoch 1881/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.6081 - categorical_accuracy: 0.6471\n",
      "Epoch 1882/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.5970 - categorical_accuracy: 0.6706\n",
      "Epoch 1883/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.5953 - categorical_accuracy: 0.7176\n",
      "Epoch 1884/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.5873 - categorical_accuracy: 0.7176\n",
      "Epoch 1885/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.5999 - categorical_accuracy: 0.6941\n",
      "Epoch 1886/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.5819 - categorical_accuracy: 0.6824\n",
      "Epoch 1887/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.6012 - categorical_accuracy: 0.7059\n",
      "Epoch 1888/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.5865 - categorical_accuracy: 0.7176\n",
      "Epoch 1889/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.5967 - categorical_accuracy: 0.6941\n",
      "Epoch 1890/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.5741 - categorical_accuracy: 0.7176\n",
      "Epoch 1891/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.5825 - categorical_accuracy: 0.6706\n",
      "Epoch 1892/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.5781 - categorical_accuracy: 0.6824\n",
      "Epoch 1893/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.5964 - categorical_accuracy: 0.6824\n",
      "Epoch 1894/2000\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 0.5754 - categorical_accuracy: 0.7176\n",
      "Epoch 1895/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.5897 - categorical_accuracy: 0.6941\n",
      "Epoch 1896/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.5768 - categorical_accuracy: 0.7412\n",
      "Epoch 1897/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.5778 - categorical_accuracy: 0.6824\n",
      "Epoch 1898/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.5952 - categorical_accuracy: 0.6941\n",
      "Epoch 1899/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.5690 - categorical_accuracy: 0.7176\n",
      "Epoch 1900/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.5881 - categorical_accuracy: 0.6588\n",
      "Epoch 1901/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.6016 - categorical_accuracy: 0.6471\n",
      "Epoch 1902/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.6000 - categorical_accuracy: 0.6941\n",
      "Epoch 1903/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.5820 - categorical_accuracy: 0.6471\n",
      "Epoch 1904/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 0.6158 - categorical_accuracy: 0.6235\n",
      "Epoch 1905/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.5899 - categorical_accuracy: 0.7059\n",
      "Epoch 1906/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.5883 - categorical_accuracy: 0.7294\n",
      "Epoch 1907/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 0.5696 - categorical_accuracy: 0.6588\n",
      "Epoch 1908/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.6075 - categorical_accuracy: 0.6471\n",
      "Epoch 1909/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.5720 - categorical_accuracy: 0.7294\n",
      "Epoch 1910/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.5722 - categorical_accuracy: 0.7176\n",
      "Epoch 1911/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.5738 - categorical_accuracy: 0.6941\n",
      "Epoch 1912/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.5767 - categorical_accuracy: 0.7294\n",
      "Epoch 1913/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.5845 - categorical_accuracy: 0.7176\n",
      "Epoch 1914/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.5964 - categorical_accuracy: 0.7176\n",
      "Epoch 1915/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 0.5666 - categorical_accuracy: 0.7059\n",
      "Epoch 1916/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.5866 - categorical_accuracy: 0.7176\n",
      "Epoch 1917/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.5847 - categorical_accuracy: 0.7176\n",
      "Epoch 1918/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.5709 - categorical_accuracy: 0.6706\n",
      "Epoch 1919/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.5984 - categorical_accuracy: 0.7176\n",
      "Epoch 1920/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.6052 - categorical_accuracy: 0.7294\n",
      "Epoch 1921/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.5884 - categorical_accuracy: 0.7176\n",
      "Epoch 1922/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.5740 - categorical_accuracy: 0.7176\n",
      "Epoch 1923/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.6269 - categorical_accuracy: 0.6588\n",
      "Epoch 1924/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.6150 - categorical_accuracy: 0.6471\n",
      "Epoch 1925/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.6231 - categorical_accuracy: 0.6941\n",
      "Epoch 1926/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.5729 - categorical_accuracy: 0.6941\n",
      "Epoch 1927/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.5865 - categorical_accuracy: 0.7059\n",
      "Epoch 1928/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.5749 - categorical_accuracy: 0.7059\n",
      "Epoch 1929/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.5717 - categorical_accuracy: 0.7059\n",
      "Epoch 1930/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.5805 - categorical_accuracy: 0.7176\n",
      "Epoch 1931/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.5738 - categorical_accuracy: 0.7059\n",
      "Epoch 1932/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.6252 - categorical_accuracy: 0.6941\n",
      "Epoch 1933/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.5931 - categorical_accuracy: 0.7059\n",
      "Epoch 1934/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.6019 - categorical_accuracy: 0.6706\n",
      "Epoch 1935/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.5922 - categorical_accuracy: 0.6824\n",
      "Epoch 1936/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 0.5806 - categorical_accuracy: 0.6706\n",
      "Epoch 1937/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.5822 - categorical_accuracy: 0.7059\n",
      "Epoch 1938/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.5700 - categorical_accuracy: 0.7176\n",
      "Epoch 1939/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 0.5746 - categorical_accuracy: 0.7059\n",
      "Epoch 1940/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.5741 - categorical_accuracy: 0.6941\n",
      "Epoch 1941/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.5809 - categorical_accuracy: 0.7059\n",
      "Epoch 1942/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 0.5817 - categorical_accuracy: 0.7059\n",
      "Epoch 1943/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.6055 - categorical_accuracy: 0.6824\n",
      "Epoch 1944/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.5702 - categorical_accuracy: 0.7294\n",
      "Epoch 1945/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.6084 - categorical_accuracy: 0.6706\n",
      "Epoch 1946/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.5862 - categorical_accuracy: 0.6824\n",
      "Epoch 1947/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.5938 - categorical_accuracy: 0.7059\n",
      "Epoch 1948/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.5884 - categorical_accuracy: 0.6824\n",
      "Epoch 1949/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.5893 - categorical_accuracy: 0.6941\n",
      "Epoch 1950/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 0.5834 - categorical_accuracy: 0.6824\n",
      "Epoch 1951/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.5762 - categorical_accuracy: 0.6588\n",
      "Epoch 1952/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.5966 - categorical_accuracy: 0.6706\n",
      "Epoch 1953/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.5872 - categorical_accuracy: 0.7294\n",
      "Epoch 1954/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.5756 - categorical_accuracy: 0.7529\n",
      "Epoch 1955/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.5759 - categorical_accuracy: 0.7294\n",
      "Epoch 1956/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.5703 - categorical_accuracy: 0.7176\n",
      "Epoch 1957/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.5632 - categorical_accuracy: 0.7059\n",
      "Epoch 1958/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.5843 - categorical_accuracy: 0.6941\n",
      "Epoch 1959/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.6080 - categorical_accuracy: 0.6706\n",
      "Epoch 1960/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 0.5894 - categorical_accuracy: 0.7647\n",
      "Epoch 1961/2000\n",
      "3/3 [==============================] - 0s 35ms/step - loss: 0.5973 - categorical_accuracy: 0.6941\n",
      "Epoch 1962/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.6583 - categorical_accuracy: 0.6471\n",
      "Epoch 1963/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.6148 - categorical_accuracy: 0.7176\n",
      "Epoch 1964/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.6078 - categorical_accuracy: 0.7294\n",
      "Epoch 1965/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.5793 - categorical_accuracy: 0.7294\n",
      "Epoch 1966/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.5860 - categorical_accuracy: 0.7059\n",
      "Epoch 1967/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.5929 - categorical_accuracy: 0.6941\n",
      "Epoch 1968/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.5551 - categorical_accuracy: 0.7412\n",
      "Epoch 1969/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.6045 - categorical_accuracy: 0.6353\n",
      "Epoch 1970/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.6233 - categorical_accuracy: 0.6353\n",
      "Epoch 1971/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.6142 - categorical_accuracy: 0.7059\n",
      "Epoch 1972/2000\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 0.5997 - categorical_accuracy: 0.6941\n",
      "Epoch 1973/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.5745 - categorical_accuracy: 0.6824\n",
      "Epoch 1974/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.5722 - categorical_accuracy: 0.7176\n",
      "Epoch 1975/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.5791 - categorical_accuracy: 0.6941\n",
      "Epoch 1976/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.5662 - categorical_accuracy: 0.7176\n",
      "Epoch 1977/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.5741 - categorical_accuracy: 0.7059\n",
      "Epoch 1978/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.5645 - categorical_accuracy: 0.7059\n",
      "Epoch 1979/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.5670 - categorical_accuracy: 0.7176\n",
      "Epoch 1980/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.5732 - categorical_accuracy: 0.7059\n",
      "Epoch 1981/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.5802 - categorical_accuracy: 0.7176\n",
      "Epoch 1982/2000\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.5586 - categorical_accuracy: 0.7176\n",
      "Epoch 1983/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.5969 - categorical_accuracy: 0.6706\n",
      "Epoch 1984/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.6070 - categorical_accuracy: 0.6706\n",
      "Epoch 1985/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.5674 - categorical_accuracy: 0.6824\n",
      "Epoch 1986/2000\n",
      "3/3 [==============================] - 0s 38ms/step - loss: 0.5903 - categorical_accuracy: 0.6824\n",
      "Epoch 1987/2000\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.6109 - categorical_accuracy: 0.7059\n",
      "Epoch 1988/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.5726 - categorical_accuracy: 0.6706\n",
      "Epoch 1989/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.5943 - categorical_accuracy: 0.6824\n",
      "Epoch 1990/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.5629 - categorical_accuracy: 0.7176\n",
      "Epoch 1991/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.5808 - categorical_accuracy: 0.7059\n",
      "Epoch 1992/2000\n",
      "3/3 [==============================] - 0s 36ms/step - loss: 0.5777 - categorical_accuracy: 0.7176\n",
      "Epoch 1993/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.5716 - categorical_accuracy: 0.7059\n",
      "Epoch 1994/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.5722 - categorical_accuracy: 0.6824\n",
      "Epoch 1995/2000\n",
      "3/3 [==============================] - 0s 37ms/step - loss: 0.5748 - categorical_accuracy: 0.7176\n",
      "Epoch 1996/2000\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.5648 - categorical_accuracy: 0.7176\n",
      "Epoch 1997/2000\n",
      "3/3 [==============================] - 0s 34ms/step - loss: 0.5623 - categorical_accuracy: 0.7176\n",
      "Epoch 1998/2000\n",
      "3/3 [==============================] - 0s 39ms/step - loss: 0.5617 - categorical_accuracy: 0.7176\n",
      "Epoch 1999/2000\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 0.5657 - categorical_accuracy: 0.7294\n",
      "Epoch 2000/2000\n",
      "3/3 [==============================] - 0s 40ms/step - loss: 0.5587 - categorical_accuracy: 0.7412\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.src.callbacks.History at 0x2d06a5a50>"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(X_train, y_train, epochs=2000, callbacks=[tb_callback])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "db87246a-995c-49c9-ab0d-e5dfbed667f1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " lstm (LSTM)                 (None, 30, 64)            442112    \n",
      "                                                                 \n",
      " lstm_1 (LSTM)               (None, 30, 128)           98816     \n",
      "                                                                 \n",
      " lstm_2 (LSTM)               (None, 64)                49408     \n",
      "                                                                 \n",
      " dense (Dense)               (None, 64)                4160      \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 32)                2080      \n",
      "                                                                 \n",
      " dense_2 (Dense)             (None, 3)                 99        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 596675 (2.28 MB)\n",
      "Trainable params: 596675 (2.28 MB)\n",
      "Non-trainable params: 0 (0.00 Byte)\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()\n",
    "# the params seen below would be in millions had we used a CNN model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "8f91bcd9-e28a-48a5-acb7-9c688e3eec97",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 227ms/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[8.6059589e-03, 9.6244854e-01, 2.8945509e-02],\n",
       "       [3.6067215e-01, 1.4687327e-01, 4.9245456e-01],\n",
       "       [4.4419071e-01, 4.6995528e-02, 5.0881374e-01],\n",
       "       [1.1632606e-01, 6.5038449e-01, 2.3328944e-01],\n",
       "       [7.0249194e-01, 6.3162005e-08, 2.9750800e-01]], dtype=float32)"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res = model.predict(X_test)\n",
    "res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "46e1469e-18ca-426d-affb-b47f5005213e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'iloveyou'"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "actions[np.argmax(res[1])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "4f00808a-9d8e-4273-9861-f2121c13d6a0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'iloveyou'"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "actions[np.argmax(y_test[1])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "ed989766-13df-435e-8b57-903085de37c0",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/keras/src/engine/training.py:3103: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
      "  saving_api.save_model(\n"
     ]
    }
   ],
   "source": [
    "model.save('action.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "cca5717f-4a44-454b-a335-ad61a051b17e",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import multilabel_confusion_matrix, accuracy_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "b9b8a99b-a9e4-4bbe-8a6c-33a5e4cd2104",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 18ms/step\n"
     ]
    }
   ],
   "source": [
    "ytemp = model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "c1b1c95e-1fd0-46f3-8b29-925b01b696ea",
   "metadata": {},
   "outputs": [],
   "source": [
    "# this converts the predictions from their one hot encoded representation to categorical labels\n",
    "ytrue = np.argmax(y_test, axis=1).tolist()\n",
    "ytemp = np.argmax(ytemp, axis=1).tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "a791dc78-d8dc-4d76-8ed5-764191f8d15d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[1, 2, 2, 1, 0]"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ytemp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "88d3bd30-635f-44b2-bd14-286c9e6a088f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[1, 2, 2, 0, 2]"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ytrue"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "2f2b183b-28ba-46c6-844a-74fd68086aa7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[3, 1],\n",
       "        [1, 0]],\n",
       "\n",
       "       [[3, 1],\n",
       "        [0, 1]],\n",
       "\n",
       "       [[2, 0],\n",
       "        [1, 2]]])"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "multilabel_confusion_matrix(ytrue, ytemp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "286965f7-ed37-4e75-8af6-53e1e055fa68",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(ytrue, ytemp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "ed770693-c926-408f-86aa-a58bd28547b0",
   "metadata": {},
   "outputs": [],
   "source": [
    "colors = [(245,117, 16), (117, 245, 16), (16, 117, 245)]\n",
    "def probablity_visulaization(res, actions, input_frame, colors):\n",
    "    output_frame = input_frame.copy()\n",
    "    for num, prob in enumerate(res):\n",
    "        cv2.rectangle(output_frame, (0,60+num*40), (int(prob*100), 90+num*40), colors[num], -1)\n",
    "        cv2.putText(output_frame, actions[num], (0, 85+num*40), cv2.FONT_HERSHEY_SIMPLEX, 1, (255,255,255), 2, cv2.LINE_AA)\n",
    "        \n",
    "    return output_frame"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "3a919b06-93e1-4fb2-aa4f-c9f4828c38ee",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "I0000 00:00:1701642521.793478       1 gl_context.cc:344] GL version: 2.1 (2.1 Metal - 86), renderer: Apple M2\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 78ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "iloveyou\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "iloveyou\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "iloveyou\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 13ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "iloveyou\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "iloveyou\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 9ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 13ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "thanks\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "iloveyou\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "iloveyou\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "hello\n",
      "<class 'mediapipe.python.solution_base.SolutionOutputs'>\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "hello\n"
     ]
    }
   ],
   "source": [
    "sequence = [] # once we get 30 frames in this is when we pass it to our prediction logic\n",
    "sentence = [] # this concatenates our history of detections together\n",
    "threshold = 0.7 # we will render a result if they are above a certain threshold\n",
    "\n",
    "capture = cv2.VideoCapture(1) # this line effectively accesses our webcam\n",
    "# set mediapipe model\n",
    "with mp_holistic.Holistic(min_detection_confidence=0.5, min_tracking_confidence=0.5) as holistic:\n",
    "    # loops through the video capture as long as we are still accessing our webcam\n",
    "    while capture.isOpened():\n",
    "      # reads the feed\n",
    "      ret, frame = capture.read()\n",
    "\n",
    "      # make detections  \n",
    "      image, results = mp_detection(frame, holistic)\n",
    "      print(results)\n",
    "    \n",
    "      # draw landmarks\n",
    "      draw_landmarks(image, results)\n",
    "\n",
    "      # prediction logic\n",
    "      keypoints = extract_values(results)\n",
    "      sequence.insert(0, keypoints)\n",
    "      sequence = sequence[:30]\n",
    "\n",
    "      if len(sequence) == 30:\n",
    "          res = model.predict(np.expand_dims(sequence, axis=0))[0]\n",
    "          print(actions[np.argmax(res)])\n",
    "\n",
    "      # visualisation logic\n",
    "      if res[np.argmax(res)] > threshold:\n",
    "          if len(sentence) > 0:\n",
    "              if actions[np.argmax(res)] != sentence[-1]:\n",
    "                  sentence.append(actions[np.argmax(res)])\n",
    "          else:\n",
    "              sentence.append(actions[np.argmax(res)])\n",
    "            \n",
    "      if len(sentence) > 5:\n",
    "          sentence = sentence[-5:]\n",
    "\n",
    "      cv2.rectangle(image, (0, 0), (640, 40), (245, 117, 16), -1)\n",
    "      cv2.putText(image, ' '.join(sentence), (3, 30), cv2.FONT_HERSHEY_SIMPLEX, 1, (255, 255, 255), 2, cv2.LINE_AA)\n",
    "\n",
    "      # shows to screen\n",
    "      cv2.imshow('OpenCV Feed', image)\n",
    "    \n",
    "      # it breaks out of our loop if we press 'q' on our keyboard\n",
    "      if cv2.waitKey(10) & 0xFF == ord('q'):\n",
    "        break\n",
    "    capture.release()\n",
    "    cv2.destroyAllWindows()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "e8d5f5ea-b0aa-422e-a25e-faea4dcb035e",
   "metadata": {},
   "outputs": [],
   "source": [
    "capture.release()\n",
    "cv2.destroyAllWindows()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "id": "36c52268-059f-4466-af6b-3a15b3e9136e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.36067215, 0.14687327, 0.49245456], dtype=float32)"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res[np.argmax(res)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0496a32b-3717-413c-a11f-39fa4c72f87b",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}