hyperparameter_comparison.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from importlib import reload\n",
    "from src import parsing\n",
    "from src import model_path\n",
    "from src import models_path\n",
    "from src import model_goal\n",
    "from src import models_goal\n",
    "from src import model_interface\n",
    "from src import visualization\n",
    "from src import util\n",
    "from src import config\n",
    "reload(parsing)\n",
    "reload(model_path)\n",
    "reload(models_path)\n",
    "reload(model_goal)\n",
    "reload(models_goal)\n",
    "reload(model_interface)\n",
    "reload(visualization)\n",
    "reload(util)\n",
    "reload(config)\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "from tensorflow.keras.utils import plot_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# batch size comparison\n",
    "reload(visualization)\n",
    "\n",
    "history_files = ['path_simple_lstm_batch_64', 'path_simple_lstm_batch_256', 'path_simple_lstm_batch_1024', 'path_simple_lstm_batch_8192']\n",
    "history_dicts = []\n",
    "for history_file in history_files:\n",
    "    history_dicts.append(model_interface.load_history(history_file))\n",
    "\n",
    "visualization.draw_losses(\n",
    "    'val_mean_absolute_error',\n",
    "    'val mean absolute error',\n",
    "    history_dicts,\n",
    "    ['batch size 64', 'batch size 256', 'batch size 1024', 'batch size 8192'],\n",
    "    'batch size comparison',\n",
    "    save_file='batch_size_comparison_lstm_val_mean_absolute_error',\n",
    "    )\n",
    "\n",
    "# visualization.draw_losses(\n",
    "#     'mean_absolute_error',\n",
    "#     'mean absolute error',\n",
    "#     history_dicts,\n",
    "#     ['batch size 64', 'batch size 256', 'batch size 1024', 'batch size 8192'],\n",
    "#     'batch size comparison',\n",
    "#     save_file='batch_size_comparison_lstm_mean_absolute_error',\n",
    "#     )\n",
    "\n",
    "\n",
    "# training time for different batch sizes\n",
    "batch_sizes = [8192, 1024, 256, 64]\n",
    "training_time = [15, 19, 27, 62]\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "ax.xaxis.grid(True)\n",
    "ax.yaxis.grid(True)\n",
    "ax.plot(batch_sizes, training_time, 'bo-')\n",
    "plt.xlabel('batch size')\n",
    "plt.ylabel('training time in minutes')\n",
    "plt.title('Training time for different batch sizes')\n",
    "plt.savefig(f\"batch_size_comparison_lstm_training_time.pdf\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# comparison with time2vec\n",
    "history_files = ['path_no2vec', 'path_yes2vec']\n",
    "history_dicts = []\n",
    "for history_file in history_files:\n",
    "    history_dicts.append(model_interface.load_history(history_file))\n",
    "\n",
    "visualization.draw_histories(\n",
    "    'val_mean_absolute_error',\n",
    "    'val mean absolute error',\n",
    "    history_dicts,\n",
    "    ['no time2vec', 'time2vec'],\n",
    "    'time2vec comparison',\n",
    "    save_file='time2vec_comparison'\n",
    ")\n",
    "\n",
    "visualization.draw_histories(\n",
    "    'val_loss',\n",
    "    'val loss',\n",
    "    history_dicts,\n",
    "    ['no time2vec', 'time2vec'],\n",
    "    'time2vec comparison',\n",
    "    save_file='time2vec_comparison_val_loss'\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# compare head amount\n",
    "history_files = ['tf_head_2', 'path_transformerk_l', 'tf_head_8']\n",
    "history_dicts = []\n",
    "for history_file in history_files:\n",
    "    history_dicts.append(model_interface.load_history(history_file))\n",
    "\n",
    "visualization.draw_histories(\n",
    "    'val_mean_absolute_error',\n",
    "    'val mean absolute error',\n",
    "    history_dicts,\n",
    "    ['2 heads', '4 heads', '8 heads'],\n",
    "    'head amount comparison',\n",
    "    max_x=13,\n",
    "    save_file='head_amount_comparison'\n",
    ")\n",
    "\n",
    "visualization.draw_histories(\n",
    "    'val_loss',\n",
    "    'val loss',\n",
    "    history_dicts,\n",
    "    ['2 heads', '4 heads', '8 heads'],\n",
    "    'head amount comparison',\n",
    "    max_x=13,\n",
    "    save_file='head_amount_comparison_val_loss'\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data, eval_data, test_data = parsing.load_processed_data(\"medium_aa.csv\")\n",
    "input_shape = train_data[0].shape\n",
    "input_shape = (input_shape[1], input_shape[2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model_1\"\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_2 (InputLayer)            [(None, 40, 2)]      0                                            \n",
      "__________________________________________________________________________________________________\n",
      "time_distributed_1 (TimeDistrib (None, 40, 6)        12          input_2[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_1 (Concatenate)     (None, 40, 8)        0           input_2[0][0]                    \n",
      "                                                                 time_distributed_1[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "layer_normalization_1 (LayerNor (None, 40, 8)        16          concatenate_1[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "transformer_block_1 (Transforme (None, 40, 8)        13392       layer_normalization_1[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "tf.math.multiply_2 (TFOpLambda) (None, 40, 8)        0           transformer_block_1[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "tf.math.multiply_3 (TFOpLambda) (None, 40, 8)        0           layer_normalization_1[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "tf.__operators__.add_1 (TFOpLam (None, 40, 8)        0           tf.math.multiply_2[0][0]         \n",
      "                                                                 tf.math.multiply_3[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "transformer_block_2 (Transforme (None, 40, 8)        13392       tf.__operators__.add_1[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "tf.math.multiply_4 (TFOpLambda) (None, 40, 8)        0           transformer_block_2[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "tf.math.multiply_5 (TFOpLambda) (None, 40, 8)        0           tf.__operators__.add_1[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "tf.__operators__.add_2 (TFOpLam (None, 40, 8)        0           tf.math.multiply_4[0][0]         \n",
      "                                                                 tf.math.multiply_5[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "flatten_1 (Flatten)             (None, 320)          0           tf.__operators__.add_2[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "dense_8 (Dense)                 (None, 800)          256800      flatten_1[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "dropout_7 (Dropout)             (None, 800)          0           dense_8[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "dense_9 (Dense)                 (None, 100)          80100       dropout_7[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "reshape_1 (Reshape)             (None, 50, 2)        0           dense_9[0][0]                    \n",
      "==================================================================================================\n",
      "Total params: 363,712\n",
      "Trainable params: 363,648\n",
      "Non-trainable params: 64\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "## generate model numbers/graphs\n",
    "from src import transformer_k\n",
    "reload(transformer_k)\n",
    "\n",
    "m = model_path.ModelPath(uses_goal=False)\n",
    "model = transformer_k.build_model(\n",
    "    input_shape,\n",
    "    head_size=64, # 256\n",
    "    num_heads=4, # 4\n",
    "    ff_dim=256, # 4\n",
    "    num_transformer_blocks=2, # 4\n",
    "    mlp_units=1, # [128]\n",
    "    mlp_dropout=0.4, # 0.4\n",
    "    dropout=0.3, # 0.25\n",
    "    time2vec_dim=3, # 3\n",
    "    bayesian_dropout=True,\n",
    ")\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model_2\"\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_3 (InputLayer)            [(None, 40, 2)]      0                                            \n",
      "__________________________________________________________________________________________________\n",
      "conv1d (Conv1D)                 (None, 40, 32)       64          input_3[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling1d (MaxPooling1D)    (None, 40, 2)        0           input_3[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_1 (Conv1D)               (None, 40, 32)       40960       conv1d[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_2 (Conv1D)               (None, 40, 32)       20480       conv1d[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_3 (Conv1D)               (None, 40, 32)       10240       conv1d[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_4 (Conv1D)               (None, 40, 32)       64          max_pooling1d[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_2 (Concatenate)     (None, 40, 128)      0           conv1d_1[0][0]                   \n",
      "                                                                 conv1d_2[0][0]                   \n",
      "                                                                 conv1d_3[0][0]                   \n",
      "                                                                 conv1d_4[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_6 (BatchNor (None, 40, 128)      512         concatenate_2[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation (Activation)         (None, 40, 128)      0           batch_normalization_6[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_5 (Conv1D)               (None, 40, 32)       4096        activation[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling1d_1 (MaxPooling1D)  (None, 40, 128)      0           activation[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_6 (Conv1D)               (None, 40, 32)       40960       conv1d_5[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_7 (Conv1D)               (None, 40, 32)       20480       conv1d_5[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_8 (Conv1D)               (None, 40, 32)       10240       conv1d_5[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_9 (Conv1D)               (None, 40, 32)       4096        max_pooling1d_1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_3 (Concatenate)     (None, 40, 128)      0           conv1d_6[0][0]                   \n",
      "                                                                 conv1d_7[0][0]                   \n",
      "                                                                 conv1d_8[0][0]                   \n",
      "                                                                 conv1d_9[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_7 (BatchNor (None, 40, 128)      512         concatenate_3[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_1 (Activation)       (None, 40, 128)      0           batch_normalization_7[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_10 (Conv1D)              (None, 40, 32)       4096        activation_1[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling1d_2 (MaxPooling1D)  (None, 40, 128)      0           activation_1[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_11 (Conv1D)              (None, 40, 32)       40960       conv1d_10[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_12 (Conv1D)              (None, 40, 32)       20480       conv1d_10[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_13 (Conv1D)              (None, 40, 32)       10240       conv1d_10[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_14 (Conv1D)              (None, 40, 32)       4096        max_pooling1d_2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_4 (Concatenate)     (None, 40, 128)      0           conv1d_11[0][0]                  \n",
      "                                                                 conv1d_12[0][0]                  \n",
      "                                                                 conv1d_13[0][0]                  \n",
      "                                                                 conv1d_14[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_15 (Conv1D)              (None, 40, 128)      256         input_3[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_8 (BatchNor (None, 40, 128)      512         concatenate_4[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_9 (BatchNor (None, 40, 128)      512         conv1d_15[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_2 (Activation)       (None, 40, 128)      0           batch_normalization_8[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "add (Add)                       (None, 40, 128)      0           batch_normalization_9[0][0]      \n",
      "                                                                 activation_2[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "activation_3 (Activation)       (None, 40, 128)      0           add[0][0]                        \n",
      "__________________________________________________________________________________________________\n",
      "global_average_pooling1d (Globa (None, 128)          0           activation_3[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "dense_10 (Dense)                (None, 25)           3225        global_average_pooling1d[0][0]   \n",
      "==================================================================================================\n",
      "Total params: 237,081\n",
      "Trainable params: 236,057\n",
      "Non-trainable params: 1,024\n",
      "__________________________________________________________________________________________________\n",
      "Model: \"model_2\"\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_3 (InputLayer)            [(None, 40, 2)]      0                                            \n",
      "__________________________________________________________________________________________________\n",
      "conv1d (Conv1D)                 (None, 40, 32)       64          input_3[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling1d (MaxPooling1D)    (None, 40, 2)        0           input_3[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_1 (Conv1D)               (None, 40, 32)       40960       conv1d[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_2 (Conv1D)               (None, 40, 32)       20480       conv1d[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_3 (Conv1D)               (None, 40, 32)       10240       conv1d[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_4 (Conv1D)               (None, 40, 32)       64          max_pooling1d[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_2 (Concatenate)     (None, 40, 128)      0           conv1d_1[0][0]                   \n",
      "                                                                 conv1d_2[0][0]                   \n",
      "                                                                 conv1d_3[0][0]                   \n",
      "                                                                 conv1d_4[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_6 (BatchNor (None, 40, 128)      512         concatenate_2[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation (Activation)         (None, 40, 128)      0           batch_normalization_6[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_5 (Conv1D)               (None, 40, 32)       4096        activation[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling1d_1 (MaxPooling1D)  (None, 40, 128)      0           activation[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_6 (Conv1D)               (None, 40, 32)       40960       conv1d_5[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_7 (Conv1D)               (None, 40, 32)       20480       conv1d_5[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_8 (Conv1D)               (None, 40, 32)       10240       conv1d_5[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_9 (Conv1D)               (None, 40, 32)       4096        max_pooling1d_1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_3 (Concatenate)     (None, 40, 128)      0           conv1d_6[0][0]                   \n",
      "                                                                 conv1d_7[0][0]                   \n",
      "                                                                 conv1d_8[0][0]                   \n",
      "                                                                 conv1d_9[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_7 (BatchNor (None, 40, 128)      512         concatenate_3[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_1 (Activation)       (None, 40, 128)      0           batch_normalization_7[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_10 (Conv1D)              (None, 40, 32)       4096        activation_1[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling1d_2 (MaxPooling1D)  (None, 40, 128)      0           activation_1[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_11 (Conv1D)              (None, 40, 32)       40960       conv1d_10[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_12 (Conv1D)              (None, 40, 32)       20480       conv1d_10[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_13 (Conv1D)              (None, 40, 32)       10240       conv1d_10[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_14 (Conv1D)              (None, 40, 32)       4096        max_pooling1d_2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_4 (Concatenate)     (None, 40, 128)      0           conv1d_11[0][0]                  \n",
      "                                                                 conv1d_12[0][0]                  \n",
      "                                                                 conv1d_13[0][0]                  \n",
      "                                                                 conv1d_14[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_15 (Conv1D)              (None, 40, 128)      256         input_3[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_8 (BatchNor (None, 40, 128)      512         concatenate_4[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_9 (BatchNor (None, 40, 128)      512         conv1d_15[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_2 (Activation)       (None, 40, 128)      0           batch_normalization_8[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "add (Add)                       (None, 40, 128)      0           batch_normalization_9[0][0]      \n",
      "                                                                 activation_2[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "activation_3 (Activation)       (None, 40, 128)      0           add[0][0]                        \n",
      "__________________________________________________________________________________________________\n",
      "global_average_pooling1d (Globa (None, 128)          0           activation_3[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "dense_10 (Dense)                (None, 25)           3225        global_average_pooling1d[0][0]   \n",
      "==================================================================================================\n",
      "Total params: 237,081\n",
      "Trainable params: 236,057\n",
      "Non-trainable params: 1,024\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "from src import inception\n",
    "reload(inception)\n",
    "\n",
    "m = model_goal.ModelGoal()\n",
    "model_name = 'goal_inception_e'\n",
    "\n",
    "model = inception.InceptionTime(\n",
    "    depth=3,\n",
    "    # nb_filters=64,\n",
    "    # kernel_size=81,\n",
    "    use_residual=True,\n",
    ").build(\n",
    "    input_shape=input_shape,\n",
    "    number_output_classes=config.GOAL_GRID_SIZE**2\n",
    ")\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('You must install pydot (`pip install pydot`) and install graphviz (see instructions at https://graphviz.gitlab.io/download/) ', 'for plot_model/model_to_dot to work.')\n"
     ]
    }
   ],
   "source": [
    "model=models_path.simple_cnn(input_shape)\n",
    "plot_model(model)"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "e7370f93d1d0cde622a1f8e1c04877d8463912d04d973331ad4851f04de6915a"
  },
  "kernelspec": {
   "display_name": "Python 3.8.10 64-bit",
   "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": "3.8.10"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}