keypoint_classification.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import csv\n",
    "\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "RANDOM_SEED = 42"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = 'model/keypoint_classifier/keypoint.csv'\n",
    "model_save_path = 'model/keypoint_classifier/keypoint_classifier.keras'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "NUM_CLASSES = 8"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_dataset = np.loadtxt(dataset, delimiter=',', dtype='float32', usecols=list(range(1, (21 * 2) + 1)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_dataset = np.loadtxt(dataset, delimiter=',', dtype='int32', usecols=(0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X_dataset, y_dataset, train_size=0.75, random_state=RANDOM_SEED)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = tf.keras.models.Sequential([\n",
    "    tf.keras.layers.Input((21 * 2, )),\n",
    "    tf.keras.layers.Dropout(0.2),\n",
    "    tf.keras.layers.Dense(20, activation='relu'),\n",
    "    tf.keras.layers.Dropout(0.4),\n",
    "    tf.keras.layers.Dense(10, activation='relu'),\n",
    "    tf.keras.layers.Dense(NUM_CLASSES, activation='softmax')\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"sequential\"</span>\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1mModel: \"sequential\"\u001b[0m\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
       "┃<span style=\"font-weight: bold\"> Layer (type)                    </span>┃<span style=\"font-weight: bold\"> Output Shape           </span>┃<span style=\"font-weight: bold\">       Param # </span>┃\n",
       "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
       "│ dropout (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dropout</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">42</span>)             │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                   │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">20</span>)             │           <span style=\"color: #00af00; text-decoration-color: #00af00\">860</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dropout_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dropout</span>)             │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">20</span>)             │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">10</span>)             │           <span style=\"color: #00af00; text-decoration-color: #00af00\">210</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense_2 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">8</span>)              │            <span style=\"color: #00af00; text-decoration-color: #00af00\">88</span> │\n",
       "└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
       "</pre>\n"
      ],
      "text/plain": [
       "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
       "┃\u001b[1m \u001b[0m\u001b[1mLayer (type)                   \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape          \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m      Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
       "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
       "│ dropout (\u001b[38;5;33mDropout\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m42\u001b[0m)             │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense (\u001b[38;5;33mDense\u001b[0m)                   │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m20\u001b[0m)             │           \u001b[38;5;34m860\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dropout_1 (\u001b[38;5;33mDropout\u001b[0m)             │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m20\u001b[0m)             │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense_1 (\u001b[38;5;33mDense\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m10\u001b[0m)             │           \u001b[38;5;34m210\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense_2 (\u001b[38;5;33mDense\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m8\u001b[0m)              │            \u001b[38;5;34m88\u001b[0m │\n",
       "└─────────────────────────────────┴────────────────────────┴───────────────┘\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">1,158</span> (4.52 KB)\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m1,158\u001b[0m (4.52 KB)\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">1,158</span> (4.52 KB)\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m1,158\u001b[0m (4.52 KB)\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> (0.00 B)\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model.summary()  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "cp_callback = tf.keras.callbacks.ModelCheckpoint(\n",
    "    model_save_path, verbose=1, save_weights_only=False)\n",
    "\n",
    "es_callback = tf.keras.callbacks.EarlyStopping(patience=20, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "model.compile(\n",
    "    optimizer='adam',\n",
    "    loss='sparse_categorical_crossentropy',\n",
    "    metrics=['accuracy']\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m9s\u001b[0m 288ms/step - accuracy: 0.1406 - loss: 2.0655\n",
      "Epoch 1: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - accuracy: 0.1841 - loss: 2.0446 - val_accuracy: 0.2384 - val_loss: 1.9044\n",
      "Epoch 2/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.2422 - loss: 1.9569\n",
      "Epoch 2: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.2573 - loss: 1.9221 - val_accuracy: 0.3288 - val_loss: 1.7868\n",
      "Epoch 3/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - accuracy: 0.3125 - loss: 1.8246\n",
      "Epoch 3: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.3131 - loss: 1.7992 - val_accuracy: 0.3930 - val_loss: 1.6414\n",
      "Epoch 4/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - accuracy: 0.2812 - loss: 1.7809\n",
      "Epoch 4: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.3286 - loss: 1.6987 - val_accuracy: 0.4004 - val_loss: 1.5138\n",
      "Epoch 5/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - accuracy: 0.3828 - loss: 1.4958\n",
      "Epoch 5: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 991us/step - accuracy: 0.3528 - loss: 1.5949 - val_accuracy: 0.4632 - val_loss: 1.4333\n",
      "Epoch 6/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.4375 - loss: 1.4655\n",
      "Epoch 6: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 977us/step - accuracy: 0.3882 - loss: 1.5333 - val_accuracy: 0.4895 - val_loss: 1.3790\n",
      "Epoch 7/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.3281 - loss: 1.5446\n",
      "Epoch 7: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 999us/step - accuracy: 0.4112 - loss: 1.4990 - val_accuracy: 0.5172 - val_loss: 1.3276\n",
      "Epoch 8/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.3750 - loss: 1.4561\n",
      "Epoch 8: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 960us/step - accuracy: 0.4284 - loss: 1.4430 - val_accuracy: 0.5381 - val_loss: 1.2978\n",
      "Epoch 9/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.4844 - loss: 1.3539\n",
      "Epoch 9: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 963us/step - accuracy: 0.4413 - loss: 1.4002 - val_accuracy: 0.5476 - val_loss: 1.2619\n",
      "Epoch 10/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.4297 - loss: 1.4002\n",
      "Epoch 10: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 972us/step - accuracy: 0.4262 - loss: 1.4141 - val_accuracy: 0.5685 - val_loss: 1.2257\n",
      "Epoch 11/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - accuracy: 0.4219 - loss: 1.4934\n",
      "Epoch 11: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 958us/step - accuracy: 0.4511 - loss: 1.3838 - val_accuracy: 0.5942 - val_loss: 1.1872\n",
      "Epoch 12/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - accuracy: 0.4766 - loss: 1.2884\n",
      "Epoch 12: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 965us/step - accuracy: 0.4725 - loss: 1.3248 - val_accuracy: 0.6030 - val_loss: 1.1565\n",
      "Epoch 13/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.4766 - loss: 1.2751\n",
      "Epoch 13: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 986us/step - accuracy: 0.4796 - loss: 1.3084 - val_accuracy: 0.6084 - val_loss: 1.1239\n",
      "Epoch 14/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.5469 - loss: 1.2819\n",
      "Epoch 14: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 970us/step - accuracy: 0.4920 - loss: 1.2673 - val_accuracy: 0.6327 - val_loss: 1.0952\n",
      "Epoch 15/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.5391 - loss: 1.2732\n",
      "Epoch 15: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.5059 - loss: 1.2649 - val_accuracy: 0.6698 - val_loss: 1.0608\n",
      "Epoch 16/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.4922 - loss: 1.2255\n",
      "Epoch 16: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.5181 - loss: 1.2317 - val_accuracy: 0.7340 - val_loss: 1.0414\n",
      "Epoch 17/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.4844 - loss: 1.2455\n",
      "Epoch 17: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 978us/step - accuracy: 0.5114 - loss: 1.2368 - val_accuracy: 0.7319 - val_loss: 1.0068\n",
      "Epoch 18/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.5234 - loss: 1.1676\n",
      "Epoch 18: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 989us/step - accuracy: 0.5261 - loss: 1.2025 - val_accuracy: 0.7488 - val_loss: 0.9817\n",
      "Epoch 19/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - accuracy: 0.6016 - loss: 1.1557\n",
      "Epoch 19: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 984us/step - accuracy: 0.5346 - loss: 1.1963 - val_accuracy: 0.7468 - val_loss: 0.9576\n",
      "Epoch 20/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.5234 - loss: 1.1826\n",
      "Epoch 20: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 985us/step - accuracy: 0.5513 - loss: 1.1664 - val_accuracy: 0.7603 - val_loss: 0.9210\n",
      "Epoch 21/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.5156 - loss: 1.2392\n",
      "Epoch 21: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.5500 - loss: 1.1569 - val_accuracy: 0.7616 - val_loss: 0.8979\n",
      "Epoch 22/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - accuracy: 0.5234 - loss: 1.1494\n",
      "Epoch 22: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 989us/step - accuracy: 0.5511 - loss: 1.1287 - val_accuracy: 0.7725 - val_loss: 0.8744\n",
      "Epoch 23/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6172 - loss: 1.1425\n",
      "Epoch 23: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 978us/step - accuracy: 0.5649 - loss: 1.1200 - val_accuracy: 0.7826 - val_loss: 0.8460\n",
      "Epoch 24/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.5859 - loss: 1.1034\n",
      "Epoch 24: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 984us/step - accuracy: 0.5838 - loss: 1.0941 - val_accuracy: 0.7826 - val_loss: 0.8233\n",
      "Epoch 25/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.5469 - loss: 1.0895\n",
      "Epoch 25: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.5866 - loss: 1.0707 - val_accuracy: 0.7920 - val_loss: 0.7939\n",
      "Epoch 26/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6016 - loss: 1.0385\n",
      "Epoch 26: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 967us/step - accuracy: 0.5813 - loss: 1.0781 - val_accuracy: 0.7887 - val_loss: 0.7729\n",
      "Epoch 27/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6250 - loss: 1.0604\n",
      "Epoch 27: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.5902 - loss: 1.0677 - val_accuracy: 0.7907 - val_loss: 0.7491\n",
      "Epoch 28/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 14ms/step - accuracy: 0.5781 - loss: 1.0607\n",
      "Epoch 28: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.5934 - loss: 1.0602 - val_accuracy: 0.7974 - val_loss: 0.7423\n",
      "Epoch 29/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.5781 - loss: 1.0873\n",
      "Epoch 29: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 963us/step - accuracy: 0.6043 - loss: 1.0484 - val_accuracy: 0.8055 - val_loss: 0.7233\n",
      "Epoch 30/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6875 - loss: 0.9304\n",
      "Epoch 30: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 935us/step - accuracy: 0.6145 - loss: 1.0043 - val_accuracy: 0.8035 - val_loss: 0.7142\n",
      "Epoch 31/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6484 - loss: 0.9213\n",
      "Epoch 31: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 952us/step - accuracy: 0.6092 - loss: 1.0045 - val_accuracy: 0.8143 - val_loss: 0.6973\n",
      "Epoch 32/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6172 - loss: 0.9889\n",
      "Epoch 32: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 947us/step - accuracy: 0.6115 - loss: 1.0054 - val_accuracy: 0.8211 - val_loss: 0.6901\n",
      "Epoch 33/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - accuracy: 0.6875 - loss: 0.9389\n",
      "Epoch 33: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 964us/step - accuracy: 0.6270 - loss: 0.9713 - val_accuracy: 0.8217 - val_loss: 0.6801\n",
      "Epoch 34/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6562 - loss: 0.9639\n",
      "Epoch 34: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 971us/step - accuracy: 0.6170 - loss: 0.9828 - val_accuracy: 0.8217 - val_loss: 0.6682\n",
      "Epoch 35/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6484 - loss: 1.0185\n",
      "Epoch 35: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 935us/step - accuracy: 0.6417 - loss: 0.9562 - val_accuracy: 0.8278 - val_loss: 0.6476\n",
      "Epoch 36/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6016 - loss: 0.9535\n",
      "Epoch 36: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 949us/step - accuracy: 0.6162 - loss: 0.9755 - val_accuracy: 0.8204 - val_loss: 0.6514\n",
      "Epoch 37/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6094 - loss: 1.0025\n",
      "Epoch 37: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 960us/step - accuracy: 0.6338 - loss: 0.9671 - val_accuracy: 0.8292 - val_loss: 0.6452\n",
      "Epoch 38/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.5391 - loss: 1.0691\n",
      "Epoch 38: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6230 - loss: 0.9659 - val_accuracy: 0.8305 - val_loss: 0.6353\n",
      "Epoch 39/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6172 - loss: 0.9642\n",
      "Epoch 39: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 952us/step - accuracy: 0.6256 - loss: 0.9659 - val_accuracy: 0.8238 - val_loss: 0.6311\n",
      "Epoch 40/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.5781 - loss: 1.0059\n",
      "Epoch 40: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 952us/step - accuracy: 0.6335 - loss: 0.9403 - val_accuracy: 0.8386 - val_loss: 0.6228\n",
      "Epoch 41/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6562 - loss: 0.9362\n",
      "Epoch 41: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 968us/step - accuracy: 0.6404 - loss: 0.9440 - val_accuracy: 0.8413 - val_loss: 0.6135\n",
      "Epoch 42/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6250 - loss: 0.9760\n",
      "Epoch 42: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 957us/step - accuracy: 0.6452 - loss: 0.9377 - val_accuracy: 0.8420 - val_loss: 0.6014\n",
      "Epoch 43/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7109 - loss: 0.9003\n",
      "Epoch 43: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 955us/step - accuracy: 0.6610 - loss: 0.9086 - val_accuracy: 0.8528 - val_loss: 0.6027\n",
      "Epoch 44/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6172 - loss: 0.9413\n",
      "Epoch 44: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6414 - loss: 0.9247 - val_accuracy: 0.8433 - val_loss: 0.5948\n",
      "Epoch 45/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - accuracy: 0.6953 - loss: 0.8826\n",
      "Epoch 45: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 970us/step - accuracy: 0.6615 - loss: 0.9035 - val_accuracy: 0.8778 - val_loss: 0.5845\n",
      "Epoch 46/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6641 - loss: 0.8716\n",
      "Epoch 46: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 964us/step - accuracy: 0.6495 - loss: 0.9121 - val_accuracy: 0.8548 - val_loss: 0.5882\n",
      "Epoch 47/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6328 - loss: 0.8850\n",
      "Epoch 47: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 978us/step - accuracy: 0.6478 - loss: 0.9083 - val_accuracy: 0.8656 - val_loss: 0.5745\n",
      "Epoch 48/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6406 - loss: 0.9870\n",
      "Epoch 48: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 974us/step - accuracy: 0.6584 - loss: 0.9049 - val_accuracy: 0.8548 - val_loss: 0.5686\n",
      "Epoch 49/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.5938 - loss: 1.0435\n",
      "Epoch 49: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6453 - loss: 0.9414 - val_accuracy: 0.8879 - val_loss: 0.5697\n",
      "Epoch 50/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6016 - loss: 0.9419\n",
      "Epoch 50: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 964us/step - accuracy: 0.6611 - loss: 0.9051 - val_accuracy: 0.8596 - val_loss: 0.5587\n",
      "Epoch 51/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6641 - loss: 0.8246\n",
      "Epoch 51: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 937us/step - accuracy: 0.6611 - loss: 0.8899 - val_accuracy: 0.8771 - val_loss: 0.5610\n",
      "Epoch 52/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7344 - loss: 0.7398\n",
      "Epoch 52: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 958us/step - accuracy: 0.6590 - loss: 0.9098 - val_accuracy: 0.8704 - val_loss: 0.5662\n",
      "Epoch 53/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6406 - loss: 0.8317\n",
      "Epoch 53: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 943us/step - accuracy: 0.6712 - loss: 0.8694 - val_accuracy: 0.8879 - val_loss: 0.5562\n",
      "Epoch 54/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6797 - loss: 0.8893\n",
      "Epoch 54: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 955us/step - accuracy: 0.6597 - loss: 0.8785 - val_accuracy: 0.8663 - val_loss: 0.5528\n",
      "Epoch 55/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6719 - loss: 0.9423\n",
      "Epoch 55: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 957us/step - accuracy: 0.6501 - loss: 0.9028 - val_accuracy: 0.8798 - val_loss: 0.5506\n",
      "Epoch 56/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6094 - loss: 0.9727\n",
      "Epoch 56: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 980us/step - accuracy: 0.6620 - loss: 0.8992 - val_accuracy: 0.8805 - val_loss: 0.5428\n",
      "Epoch 57/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - accuracy: 0.7344 - loss: 0.8731\n",
      "Epoch 57: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 979us/step - accuracy: 0.6751 - loss: 0.8763 - val_accuracy: 0.8845 - val_loss: 0.5329\n",
      "Epoch 58/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6719 - loss: 0.8027\n",
      "Epoch 58: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 967us/step - accuracy: 0.6593 - loss: 0.8860 - val_accuracy: 0.8710 - val_loss: 0.5431\n",
      "Epoch 59/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6562 - loss: 0.8632\n",
      "Epoch 59: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 979us/step - accuracy: 0.6500 - loss: 0.8936 - val_accuracy: 0.8859 - val_loss: 0.5332\n",
      "Epoch 60/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6094 - loss: 0.9528\n",
      "Epoch 60: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 946us/step - accuracy: 0.6660 - loss: 0.8741 - val_accuracy: 0.8852 - val_loss: 0.5250\n",
      "Epoch 61/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6484 - loss: 0.8465\n",
      "Epoch 61: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 943us/step - accuracy: 0.6562 - loss: 0.8947 - val_accuracy: 0.8940 - val_loss: 0.5276\n",
      "Epoch 62/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 14ms/step - accuracy: 0.7344 - loss: 0.7732\n",
      "Epoch 62: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 987us/step - accuracy: 0.6786 - loss: 0.8529 - val_accuracy: 0.8845 - val_loss: 0.5247\n",
      "Epoch 63/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6406 - loss: 0.8612\n",
      "Epoch 63: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 978us/step - accuracy: 0.6658 - loss: 0.8743 - val_accuracy: 0.8872 - val_loss: 0.5220\n",
      "Epoch 64/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6641 - loss: 0.7641\n",
      "Epoch 64: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 951us/step - accuracy: 0.6730 - loss: 0.8612 - val_accuracy: 0.8852 - val_loss: 0.5116\n",
      "Epoch 65/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6875 - loss: 0.7463\n",
      "Epoch 65: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 972us/step - accuracy: 0.6540 - loss: 0.8724 - val_accuracy: 0.8893 - val_loss: 0.5186\n",
      "Epoch 66/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7344 - loss: 0.8026\n",
      "Epoch 66: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 960us/step - accuracy: 0.6729 - loss: 0.8855 - val_accuracy: 0.8872 - val_loss: 0.5162\n",
      "Epoch 67/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6484 - loss: 0.9222\n",
      "Epoch 67: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 971us/step - accuracy: 0.6667 - loss: 0.8751 - val_accuracy: 0.8913 - val_loss: 0.5112\n",
      "Epoch 68/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - accuracy: 0.6719 - loss: 0.8839\n",
      "Epoch 68: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6761 - loss: 0.8606 - val_accuracy: 0.8940 - val_loss: 0.4997\n",
      "Epoch 69/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - accuracy: 0.7109 - loss: 0.8560\n",
      "Epoch 69: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 989us/step - accuracy: 0.6799 - loss: 0.8533 - val_accuracy: 0.8980 - val_loss: 0.5053\n",
      "Epoch 70/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6875 - loss: 0.8398\n",
      "Epoch 70: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 977us/step - accuracy: 0.6724 - loss: 0.8547 - val_accuracy: 0.8913 - val_loss: 0.5004\n",
      "Epoch 71/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6562 - loss: 0.9069\n",
      "Epoch 71: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 942us/step - accuracy: 0.6692 - loss: 0.8724 - val_accuracy: 0.8866 - val_loss: 0.5013\n",
      "Epoch 72/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7109 - loss: 0.7881\n",
      "Epoch 72: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 979us/step - accuracy: 0.6819 - loss: 0.8317 - val_accuracy: 0.8798 - val_loss: 0.5059\n",
      "Epoch 73/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6484 - loss: 0.8511\n",
      "Epoch 73: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6565 - loss: 0.8929 - val_accuracy: 0.8899 - val_loss: 0.5059\n",
      "Epoch 74/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7031 - loss: 0.8624\n",
      "Epoch 74: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 953us/step - accuracy: 0.6925 - loss: 0.8550 - val_accuracy: 0.8893 - val_loss: 0.4993\n",
      "Epoch 75/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6406 - loss: 0.8622\n",
      "Epoch 75: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6795 - loss: 0.8396 - val_accuracy: 0.8980 - val_loss: 0.4830\n",
      "Epoch 76/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6328 - loss: 0.8948\n",
      "Epoch 76: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 939us/step - accuracy: 0.6688 - loss: 0.8563 - val_accuracy: 0.8764 - val_loss: 0.4948\n",
      "Epoch 77/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6641 - loss: 1.0990\n",
      "Epoch 77: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 926us/step - accuracy: 0.6759 - loss: 0.8792 - val_accuracy: 0.8920 - val_loss: 0.4883\n",
      "Epoch 78/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7344 - loss: 0.8201\n",
      "Epoch 78: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 944us/step - accuracy: 0.6941 - loss: 0.8445 - val_accuracy: 0.8832 - val_loss: 0.4882\n",
      "Epoch 79/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6797 - loss: 0.9500\n",
      "Epoch 79: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 956us/step - accuracy: 0.6809 - loss: 0.8624 - val_accuracy: 0.8920 - val_loss: 0.4902\n",
      "Epoch 80/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7422 - loss: 0.7677\n",
      "Epoch 80: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6810 - loss: 0.8339 - val_accuracy: 0.8926 - val_loss: 0.4851\n",
      "Epoch 81/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6641 - loss: 0.8454\n",
      "Epoch 81: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 948us/step - accuracy: 0.6866 - loss: 0.8294 - val_accuracy: 0.8994 - val_loss: 0.4772\n",
      "Epoch 82/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6406 - loss: 0.9221\n",
      "Epoch 82: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 967us/step - accuracy: 0.6615 - loss: 0.8690 - val_accuracy: 0.8926 - val_loss: 0.4763\n",
      "Epoch 83/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6953 - loss: 0.8288\n",
      "Epoch 83: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 966us/step - accuracy: 0.6910 - loss: 0.8525 - val_accuracy: 0.9028 - val_loss: 0.4770\n",
      "Epoch 84/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7109 - loss: 0.8151\n",
      "Epoch 84: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 947us/step - accuracy: 0.6920 - loss: 0.8212 - val_accuracy: 0.8967 - val_loss: 0.4655\n",
      "Epoch 85/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6875 - loss: 0.9032\n",
      "Epoch 85: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 947us/step - accuracy: 0.6776 - loss: 0.8407 - val_accuracy: 0.9007 - val_loss: 0.4666\n",
      "Epoch 86/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6875 - loss: 0.8702\n",
      "Epoch 86: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6911 - loss: 0.8367 - val_accuracy: 0.8960 - val_loss: 0.4647\n",
      "Epoch 87/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6172 - loss: 0.9051\n",
      "Epoch 87: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 972us/step - accuracy: 0.6826 - loss: 0.8424 - val_accuracy: 0.8886 - val_loss: 0.4669\n",
      "Epoch 88/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7266 - loss: 0.8411\n",
      "Epoch 88: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 960us/step - accuracy: 0.6844 - loss: 0.8447 - val_accuracy: 0.8940 - val_loss: 0.4620\n",
      "Epoch 89/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6406 - loss: 0.8905\n",
      "Epoch 89: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 955us/step - accuracy: 0.6866 - loss: 0.8212 - val_accuracy: 0.8866 - val_loss: 0.4702\n",
      "Epoch 90/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6250 - loss: 0.8518\n",
      "Epoch 90: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 948us/step - accuracy: 0.6877 - loss: 0.8228 - val_accuracy: 0.8886 - val_loss: 0.4722\n",
      "Epoch 91/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6641 - loss: 0.9019\n",
      "Epoch 91: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6997 - loss: 0.8082 - val_accuracy: 0.8879 - val_loss: 0.4634\n",
      "Epoch 92/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6562 - loss: 0.8906\n",
      "Epoch 92: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 972us/step - accuracy: 0.6845 - loss: 0.8394 - val_accuracy: 0.8825 - val_loss: 0.4680\n",
      "Epoch 93/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - accuracy: 0.6016 - loss: 0.8297\n",
      "Epoch 93: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6834 - loss: 0.8061 - val_accuracy: 0.9014 - val_loss: 0.4609\n",
      "Epoch 94/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7109 - loss: 0.7197\n",
      "Epoch 94: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 959us/step - accuracy: 0.6865 - loss: 0.8112 - val_accuracy: 0.8920 - val_loss: 0.4564\n",
      "Epoch 95/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7422 - loss: 0.7246\n",
      "Epoch 95: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 961us/step - accuracy: 0.6985 - loss: 0.8150 - val_accuracy: 0.8899 - val_loss: 0.4516\n",
      "Epoch 96/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7344 - loss: 0.7268\n",
      "Epoch 96: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 964us/step - accuracy: 0.6944 - loss: 0.8187 - val_accuracy: 0.8974 - val_loss: 0.4530\n",
      "Epoch 97/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6875 - loss: 0.8137\n",
      "Epoch 97: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 963us/step - accuracy: 0.6958 - loss: 0.7982 - val_accuracy: 0.8960 - val_loss: 0.4487\n",
      "Epoch 98/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6875 - loss: 0.8664\n",
      "Epoch 98: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6891 - loss: 0.8216 - val_accuracy: 0.8933 - val_loss: 0.4492\n",
      "Epoch 99/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7109 - loss: 0.7880\n",
      "Epoch 99: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 949us/step - accuracy: 0.7010 - loss: 0.7815 - val_accuracy: 0.8872 - val_loss: 0.4472\n",
      "Epoch 100/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6719 - loss: 0.8330\n",
      "Epoch 100: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 966us/step - accuracy: 0.6811 - loss: 0.8338 - val_accuracy: 0.8994 - val_loss: 0.4502\n",
      "Epoch 101/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7344 - loss: 0.6953\n",
      "Epoch 101: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 961us/step - accuracy: 0.7018 - loss: 0.7959 - val_accuracy: 0.9115 - val_loss: 0.4402\n",
      "Epoch 102/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6719 - loss: 0.8636\n",
      "Epoch 102: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 940us/step - accuracy: 0.6884 - loss: 0.8311 - val_accuracy: 0.8994 - val_loss: 0.4411\n",
      "Epoch 103/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7578 - loss: 0.6968\n",
      "Epoch 103: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 984us/step - accuracy: 0.7012 - loss: 0.7946 - val_accuracy: 0.9055 - val_loss: 0.4445\n",
      "Epoch 104/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6953 - loss: 0.8648\n",
      "Epoch 104: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 941us/step - accuracy: 0.6971 - loss: 0.8023 - val_accuracy: 0.9007 - val_loss: 0.4390\n",
      "Epoch 105/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6719 - loss: 0.9012\n",
      "Epoch 105: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 936us/step - accuracy: 0.6970 - loss: 0.8081 - val_accuracy: 0.8994 - val_loss: 0.4443\n",
      "Epoch 106/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7031 - loss: 0.8159\n",
      "Epoch 106: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 941us/step - accuracy: 0.7109 - loss: 0.7746 - val_accuracy: 0.8953 - val_loss: 0.4433\n",
      "Epoch 107/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6875 - loss: 0.8234\n",
      "Epoch 107: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 947us/step - accuracy: 0.6977 - loss: 0.8118 - val_accuracy: 0.8980 - val_loss: 0.4463\n",
      "Epoch 108/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6797 - loss: 0.8446\n",
      "Epoch 108: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 933us/step - accuracy: 0.7033 - loss: 0.7956 - val_accuracy: 0.8899 - val_loss: 0.4462\n",
      "Epoch 109/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7188 - loss: 0.7480\n",
      "Epoch 109: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 959us/step - accuracy: 0.7049 - loss: 0.8022 - val_accuracy: 0.8933 - val_loss: 0.4412\n",
      "Epoch 110/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6562 - loss: 0.8605\n",
      "Epoch 110: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 952us/step - accuracy: 0.7023 - loss: 0.7827 - val_accuracy: 0.8920 - val_loss: 0.4382\n",
      "Epoch 111/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6562 - loss: 0.9088\n",
      "Epoch 111: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6983 - loss: 0.8117 - val_accuracy: 0.9014 - val_loss: 0.4384\n",
      "Epoch 112/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6406 - loss: 0.8285\n",
      "Epoch 112: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 956us/step - accuracy: 0.6962 - loss: 0.8054 - val_accuracy: 0.8967 - val_loss: 0.4404\n",
      "Epoch 113/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6641 - loss: 0.9793\n",
      "Epoch 113: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 929us/step - accuracy: 0.6845 - loss: 0.8410 - val_accuracy: 0.9095 - val_loss: 0.4398\n",
      "Epoch 114/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6953 - loss: 0.7681\n",
      "Epoch 114: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 964us/step - accuracy: 0.6958 - loss: 0.7951 - val_accuracy: 0.8839 - val_loss: 0.4510\n",
      "Epoch 115/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7031 - loss: 0.8406\n",
      "Epoch 115: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 970us/step - accuracy: 0.6977 - loss: 0.8373 - val_accuracy: 0.8994 - val_loss: 0.4440\n",
      "Epoch 116/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7656 - loss: 0.7541\n",
      "Epoch 116: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 933us/step - accuracy: 0.7088 - loss: 0.7952 - val_accuracy: 0.8974 - val_loss: 0.4312\n",
      "Epoch 117/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6641 - loss: 0.8237\n",
      "Epoch 117: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7106 - loss: 0.7787 - val_accuracy: 0.8940 - val_loss: 0.4338\n",
      "Epoch 118/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - accuracy: 0.7109 - loss: 0.7753\n",
      "Epoch 118: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 963us/step - accuracy: 0.7051 - loss: 0.7737 - val_accuracy: 0.8987 - val_loss: 0.4258\n",
      "Epoch 119/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7734 - loss: 0.6819\n",
      "Epoch 119: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 946us/step - accuracy: 0.7163 - loss: 0.7805 - val_accuracy: 0.8866 - val_loss: 0.4339\n",
      "Epoch 120/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7109 - loss: 0.7318\n",
      "Epoch 120: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 979us/step - accuracy: 0.7133 - loss: 0.7782 - val_accuracy: 0.8960 - val_loss: 0.4253\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 121/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6797 - loss: 0.7427\n",
      "Epoch 121: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7039 - loss: 0.7923 - val_accuracy: 0.9088 - val_loss: 0.4248\n",
      "Epoch 122/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - accuracy: 0.7266 - loss: 0.7400\n",
      "Epoch 122: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7054 - loss: 0.7717 - val_accuracy: 0.8980 - val_loss: 0.4314\n",
      "Epoch 123/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - accuracy: 0.7109 - loss: 0.7671\n",
      "Epoch 123: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7058 - loss: 0.7870 - val_accuracy: 0.9001 - val_loss: 0.4335\n",
      "Epoch 124/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7891 - loss: 0.6948\n",
      "Epoch 124: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 961us/step - accuracy: 0.7259 - loss: 0.7659 - val_accuracy: 0.8886 - val_loss: 0.4276\n",
      "Epoch 125/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7266 - loss: 0.7081\n",
      "Epoch 125: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 930us/step - accuracy: 0.7079 - loss: 0.7864 - val_accuracy: 0.8960 - val_loss: 0.4227\n",
      "Epoch 126/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7578 - loss: 0.7504\n",
      "Epoch 126: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7139 - loss: 0.7846 - val_accuracy: 0.9156 - val_loss: 0.4158\n",
      "Epoch 127/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7109 - loss: 0.9387\n",
      "Epoch 127: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 920us/step - accuracy: 0.7059 - loss: 0.8255 - val_accuracy: 0.8953 - val_loss: 0.4308\n",
      "Epoch 128/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7188 - loss: 0.7761\n",
      "Epoch 128: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 967us/step - accuracy: 0.7033 - loss: 0.7856 - val_accuracy: 0.8960 - val_loss: 0.4196\n",
      "Epoch 129/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7422 - loss: 0.7834\n",
      "Epoch 129: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 969us/step - accuracy: 0.7136 - loss: 0.7837 - val_accuracy: 0.8886 - val_loss: 0.4288\n",
      "Epoch 130/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7188 - loss: 0.7305\n",
      "Epoch 130: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 975us/step - accuracy: 0.6922 - loss: 0.7874 - val_accuracy: 0.8913 - val_loss: 0.4319\n",
      "Epoch 131/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7109 - loss: 0.8089\n",
      "Epoch 131: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 944us/step - accuracy: 0.7136 - loss: 0.7497 - val_accuracy: 0.8906 - val_loss: 0.4207\n",
      "Epoch 132/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6953 - loss: 0.7253\n",
      "Epoch 132: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 925us/step - accuracy: 0.7154 - loss: 0.7557 - val_accuracy: 0.8940 - val_loss: 0.4195\n",
      "Epoch 133/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6797 - loss: 0.7741\n",
      "Epoch 133: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 975us/step - accuracy: 0.7046 - loss: 0.7807 - val_accuracy: 0.9122 - val_loss: 0.4159\n",
      "Epoch 134/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7422 - loss: 0.7938\n",
      "Epoch 134: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 938us/step - accuracy: 0.7071 - loss: 0.8077 - val_accuracy: 0.9055 - val_loss: 0.4173\n",
      "Epoch 135/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6875 - loss: 0.7129\n",
      "Epoch 135: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 938us/step - accuracy: 0.6899 - loss: 0.8032 - val_accuracy: 0.8987 - val_loss: 0.4208\n",
      "Epoch 136/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6797 - loss: 0.8465\n",
      "Epoch 136: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6963 - loss: 0.7961 - val_accuracy: 0.8879 - val_loss: 0.4194\n",
      "Epoch 137/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6484 - loss: 0.8776\n",
      "Epoch 137: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 948us/step - accuracy: 0.6926 - loss: 0.8031 - val_accuracy: 0.8933 - val_loss: 0.4231\n",
      "Epoch 138/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7578 - loss: 0.6998\n",
      "Epoch 138: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 986us/step - accuracy: 0.7206 - loss: 0.7667 - val_accuracy: 0.9102 - val_loss: 0.4095\n",
      "Epoch 139/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7031 - loss: 0.8375\n",
      "Epoch 139: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 980us/step - accuracy: 0.7075 - loss: 0.7907 - val_accuracy: 0.8987 - val_loss: 0.4132\n",
      "Epoch 140/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7891 - loss: 0.6572\n",
      "Epoch 140: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 964us/step - accuracy: 0.7165 - loss: 0.7499 - val_accuracy: 0.8987 - val_loss: 0.4095\n",
      "Epoch 141/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.8125 - loss: 0.5747\n",
      "Epoch 141: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 940us/step - accuracy: 0.7315 - loss: 0.7411 - val_accuracy: 0.9163 - val_loss: 0.3959\n",
      "Epoch 142/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 14ms/step - accuracy: 0.7422 - loss: 0.6451\n",
      "Epoch 142: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7152 - loss: 0.7684 - val_accuracy: 0.8974 - val_loss: 0.4051\n",
      "Epoch 143/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6875 - loss: 0.7805\n",
      "Epoch 143: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 981us/step - accuracy: 0.7130 - loss: 0.7628 - val_accuracy: 0.9061 - val_loss: 0.3974\n",
      "Epoch 144/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - accuracy: 0.7266 - loss: 0.7172\n",
      "Epoch 144: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 933us/step - accuracy: 0.7207 - loss: 0.7501 - val_accuracy: 0.8893 - val_loss: 0.4130\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 145/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6797 - loss: 0.8655\n",
      "Epoch 145: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 947us/step - accuracy: 0.7032 - loss: 0.7728 - val_accuracy: 0.9014 - val_loss: 0.4119\n",
      "Epoch 146/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7656 - loss: 0.6946\n",
      "Epoch 146: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 938us/step - accuracy: 0.7212 - loss: 0.7817 - val_accuracy: 0.9068 - val_loss: 0.4020\n",
      "Epoch 147/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6406 - loss: 0.9380\n",
      "Epoch 147: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7010 - loss: 0.7830 - val_accuracy: 0.8960 - val_loss: 0.4078\n",
      "Epoch 148/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6562 - loss: 0.8722\n",
      "Epoch 148: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 989us/step - accuracy: 0.7053 - loss: 0.7673 - val_accuracy: 0.9088 - val_loss: 0.4104\n",
      "Epoch 149/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6875 - loss: 0.7996\n",
      "Epoch 149: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 935us/step - accuracy: 0.7062 - loss: 0.7698 - val_accuracy: 0.8987 - val_loss: 0.4112\n",
      "Epoch 150/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7344 - loss: 0.7632\n",
      "Epoch 150: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 986us/step - accuracy: 0.7166 - loss: 0.7529 - val_accuracy: 0.9034 - val_loss: 0.4143\n",
      "Epoch 151/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7109 - loss: 0.7225\n",
      "Epoch 151: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 944us/step - accuracy: 0.7024 - loss: 0.7931 - val_accuracy: 0.8967 - val_loss: 0.4151\n",
      "Epoch 152/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7109 - loss: 0.8300\n",
      "Epoch 152: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 948us/step - accuracy: 0.7202 - loss: 0.7681 - val_accuracy: 0.9115 - val_loss: 0.3981\n",
      "Epoch 153/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.8047 - loss: 0.6497\n",
      "Epoch 153: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 925us/step - accuracy: 0.7241 - loss: 0.7637 - val_accuracy: 0.9095 - val_loss: 0.4074\n",
      "Epoch 154/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7266 - loss: 0.7600\n",
      "Epoch 154: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 955us/step - accuracy: 0.7098 - loss: 0.7839 - val_accuracy: 0.9075 - val_loss: 0.4018\n",
      "Epoch 155/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.6719 - loss: 0.8255\n",
      "Epoch 155: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 957us/step - accuracy: 0.7131 - loss: 0.7715 - val_accuracy: 0.8933 - val_loss: 0.4071\n",
      "Epoch 156/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7266 - loss: 0.7631\n",
      "Epoch 156: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 951us/step - accuracy: 0.7204 - loss: 0.7714 - val_accuracy: 0.9021 - val_loss: 0.4014\n",
      "Epoch 157/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7188 - loss: 0.8163\n",
      "Epoch 157: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 963us/step - accuracy: 0.7148 - loss: 0.7678 - val_accuracy: 0.8974 - val_loss: 0.4111\n",
      "Epoch 158/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7344 - loss: 0.7511\n",
      "Epoch 158: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 994us/step - accuracy: 0.7092 - loss: 0.7571 - val_accuracy: 0.9021 - val_loss: 0.4037\n",
      "Epoch 159/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7109 - loss: 0.6887\n",
      "Epoch 159: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7161 - loss: 0.7463 - val_accuracy: 0.9014 - val_loss: 0.3975\n",
      "Epoch 160/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7031 - loss: 0.8590\n",
      "Epoch 160: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 938us/step - accuracy: 0.7076 - loss: 0.7972 - val_accuracy: 0.8825 - val_loss: 0.4139\n",
      "Epoch 161/1000\n",
      "\u001b[1m 1/35\u001b[0m \u001b[37m━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - accuracy: 0.7578 - loss: 0.6854\n",
      "Epoch 161: saving model to model/keypoint_classifier/keypoint_classifier.keras\n",
      "\u001b[1m35/35\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 956us/step - accuracy: 0.7356 - loss: 0.7518 - val_accuracy: 0.8940 - val_loss: 0.4117\n",
      "Epoch 161: early stopping\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.src.callbacks.history.History at 0x3094a9f60>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(\n",
    "    X_train,\n",
    "    y_train,\n",
    "    epochs=1000,\n",
    "    batch_size=128,\n",
    "    validation_data=(X_test, y_test),\n",
    "    callbacks=[cp_callback, es_callback]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m12/12\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 459us/step - accuracy: 0.8918 - loss: 0.4160\n"
     ]
    }
   ],
   "source": [
    "\n",
    "val_loss, val_acc = model.evaluate(X_test, y_test, batch_size=128)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "model = tf.keras.models.load_model(model_save_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 18ms/step\n",
      "[0.00656847 0.6989198  0.04854181 0.00194381 0.18333077 0.00545762\n",
      " 0.05387838 0.00135925]\n",
      "1\n"
     ]
    }
   ],
   "source": [
    "\n",
    "predict_result = model.predict(np.array([X_test[0]]))\n",
    "print(np.squeeze(predict_result))\n",
    "print(np.argmax(np.squeeze(predict_result)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m47/47\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 339us/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification Report\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.96      0.94      0.95       405\n",
      "           1       0.91      0.86      0.89       375\n",
      "           2       0.87      0.99      0.93       334\n",
      "           3       0.82      0.97      0.89        78\n",
      "           4       1.00      0.13      0.23        31\n",
      "           5       0.70      0.98      0.82       105\n",
      "           6       0.98      0.66      0.79        74\n",
      "           7       0.97      0.78      0.87        79\n",
      "\n",
      "    accuracy                           0.89      1481\n",
      "   macro avg       0.90      0.79      0.79      1481\n",
      "weighted avg       0.91      0.89      0.89      1481\n",
      "\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.metrics import confusion_matrix, classification_report\n",
    "\n",
    "def print_confusion_matrix(y_true, y_pred, report=True):\n",
    "    labels = sorted(list(set(y_true)))\n",
    "    cmx_data = confusion_matrix(y_true, y_pred, labels=labels)\n",
    "    \n",
    "    df_cmx = pd.DataFrame(cmx_data, index=labels, columns=labels)\n",
    " \n",
    "    fig, ax = plt.subplots(figsize=(7, 6))\n",
    "    sns.heatmap(df_cmx, annot=True, fmt='g' ,square=False)\n",
    "    ax.set_ylim(len(set(y_true)), 0)\n",
    "    plt.show()\n",
    "    \n",
    "    if report:\n",
    "        print('Classification Report')\n",
    "        print(classification_report(y_test, y_pred))\n",
    "\n",
    "Y_pred = model.predict(X_test)\n",
    "y_pred = np.argmax(Y_pred, axis=1)\n",
    "\n",
    "print_confusion_matrix(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "model.save(model_save_path, include_optimizer=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: /var/folders/zg/1m9wz_hn11xgbtmm_1t209hr0000gn/T/tmpmus2gxup/assets\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: /var/folders/zg/1m9wz_hn11xgbtmm_1t209hr0000gn/T/tmpmus2gxup/assets\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saved artifact at '/var/folders/zg/1m9wz_hn11xgbtmm_1t209hr0000gn/T/tmpmus2gxup'. The following endpoints are available:\n",
      "\n",
      "* Endpoint 'serve'\n",
      "  args_0 (POSITIONAL_ONLY): TensorSpec(shape=(None, 42), dtype=tf.float32, name='input_layer')\n",
      "Output Type:\n",
      "  TensorSpec(shape=(None, 8), dtype=tf.float32, name=None)\n",
      "Captures:\n",
      "  13220775888: TensorSpec(shape=(), dtype=tf.resource, name=None)\n",
      "  13220789440: TensorSpec(shape=(), dtype=tf.resource, name=None)\n",
      "  13221450448: TensorSpec(shape=(), dtype=tf.resource, name=None)\n",
      "  13221458544: TensorSpec(shape=(), dtype=tf.resource, name=None)\n",
      "  13221451152: TensorSpec(shape=(), dtype=tf.resource, name=None)\n",
      "  13221460656: TensorSpec(shape=(), dtype=tf.resource, name=None)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING: All log messages before absl::InitializeLog() is called are written to STDERR\n",
      "W0000 00:00:1722989768.547459 16120026 tf_tfl_flatbuffer_helpers.cc:361] Ignored output_format.\n",
      "W0000 00:00:1722989768.547473 16120026 tf_tfl_flatbuffer_helpers.cc:364] Ignored drop_control_dependency.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "6728"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "tflite_save_path = 'model/keypoint_classifier/keypoint_classifier.tflite'\n",
    "\n",
    "converter = tf.lite.TFLiteConverter.from_keras_model(model)\n",
    "converter.optimizations = [tf.lite.Optimize.DEFAULT]\n",
    "tflite_quantized_model = converter.convert()\n",
    "\n",
    "open(tflite_save_path, 'wb').write(tflite_quantized_model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO: Created TensorFlow Lite XNNPACK delegate for CPU.\n"
     ]
    }
   ],
   "source": [
    "interpreter = tf.lite.Interpreter(model_path=tflite_save_path)\n",
    "interpreter.allocate_tensors()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "input_details = interpreter.get_input_details()\n",
    "output_details = interpreter.get_output_details()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "interpreter.set_tensor(input_details[0]['index'], np.array([X_test[0]]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 75 µs, sys: 33 µs, total: 108 µs\n",
      "Wall time: 107 µs\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "interpreter.invoke()\n",
    "tflite_results = interpreter.get_tensor(output_details[0]['index'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.00656847 0.69891983 0.04854176 0.00194381 0.18333079 0.00545762\n",
      " 0.05387836 0.00135925]\n",
      "1\n"
     ]
    }
   ],
   "source": [
    "print(np.squeeze(tflite_results))\n",
    "print(np.argmax(np.squeeze(tflite_results)))"
   ]
  }
 ],
 "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": "3.10.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}