widzenie-komputerowe-projekt/graph.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow.python.framework.convert_to_constants import convert_variables_to_constants_v2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = tf.keras.models.load_model('./model')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'./frozen_models\\\\frozen_graph.pb'"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Convert Keras model to ConcreteFunction\n",
    "full_model = tf.function(lambda x: model(x))\n",
    "full_model = full_model.get_concrete_function(\n",
    "    tf.TensorSpec(model.inputs[0].shape, model.inputs[0].dtype))\n",
    "\n",
    "# Get frozen ConcreteFunction\n",
    "frozen_func = convert_variables_to_constants_v2(full_model)\n",
    "frozen_func.graph.as_graph_def()\n",
    "\n",
    "\n",
    "# Save frozen graph from frozen ConcreteFunction to hard drive\n",
    "tf.io.write_graph(graph_or_graph_def=frozen_func.graph,\n",
    "                  logdir=\"./frozen_models\",\n",
    "                  name=\"frozen_graph.pb\",\n",
    "                  as_text=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def wrap_frozen_graph(graph_def, inputs, outputs, print_graph=False):\n",
    "    def _imports_graph_def():\n",
    "        tf.compat.v1.import_graph_def(graph_def, name=\"\")\n",
    "\n",
    "    wrapped_import = tf.compat.v1.wrap_function(_imports_graph_def, [])\n",
    "    import_graph = wrapped_import.graph\n",
    "\n",
    "    if print_graph == True:\n",
    "        print(\"-\" * 50)\n",
    "        print(\"Frozen model layers: \")\n",
    "        layers = [op.name for op in import_graph.get_operations()]\n",
    "        for layer in layers:\n",
    "            print(layer)\n",
    "        print(\"-\" * 50)\n",
    "\n",
    "    return wrapped_import.prune(\n",
    "        tf.nest.map_structure(import_graph.as_graph_element, inputs),\n",
    "        tf.nest.map_structure(import_graph.as_graph_element, outputs))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--------------------------------------------------\n",
      "Frozen model layers: \n",
      "x\n",
      "sequential_1/conv2d_5/Conv2D/ReadVariableOp/resource\n",
      "sequential_1/conv2d_5/BiasAdd/ReadVariableOp/resource\n",
      "sequential_1/batch_normalization/ReadVariableOp/resource\n",
      "sequential_1/batch_normalization/ReadVariableOp_1/resource\n",
      "sequential_1/batch_normalization/FusedBatchNormV3/ReadVariableOp/resource\n",
      "sequential_1/batch_normalization/FusedBatchNormV3/ReadVariableOp_1/resource\n",
      "sequential_1/conv2d_6/Conv2D/ReadVariableOp/resource\n",
      "sequential_1/conv2d_6/BiasAdd/ReadVariableOp/resource\n",
      "sequential_1/batch_normalization_1/ReadVariableOp/resource\n",
      "sequential_1/batch_normalization_1/ReadVariableOp_1/resource\n",
      "sequential_1/batch_normalization_1/FusedBatchNormV3/ReadVariableOp/resource\n",
      "sequential_1/batch_normalization_1/FusedBatchNormV3/ReadVariableOp_1/resource\n",
      "sequential_1/conv2d_7/Conv2D/ReadVariableOp/resource\n",
      "sequential_1/conv2d_7/BiasAdd/ReadVariableOp/resource\n",
      "sequential_1/batch_normalization_2/ReadVariableOp/resource\n",
      "sequential_1/batch_normalization_2/ReadVariableOp_1/resource\n",
      "sequential_1/batch_normalization_2/FusedBatchNormV3/ReadVariableOp/resource\n",
      "sequential_1/batch_normalization_2/FusedBatchNormV3/ReadVariableOp_1/resource\n",
      "sequential_1/conv2d_8/Conv2D/ReadVariableOp/resource\n",
      "sequential_1/conv2d_8/BiasAdd/ReadVariableOp/resource\n",
      "sequential_1/batch_normalization_3/ReadVariableOp/resource\n",
      "sequential_1/batch_normalization_3/ReadVariableOp_1/resource\n",
      "sequential_1/batch_normalization_3/FusedBatchNormV3/ReadVariableOp/resource\n",
      "sequential_1/batch_normalization_3/FusedBatchNormV3/ReadVariableOp_1/resource\n",
      "sequential_1/conv2d_9/Conv2D/ReadVariableOp/resource\n",
      "sequential_1/conv2d_9/BiasAdd/ReadVariableOp/resource\n",
      "sequential_1/batch_normalization_4/ReadVariableOp/resource\n",
      "sequential_1/batch_normalization_4/ReadVariableOp_1/resource\n",
      "sequential_1/batch_normalization_4/FusedBatchNormV3/ReadVariableOp/resource\n",
      "sequential_1/batch_normalization_4/FusedBatchNormV3/ReadVariableOp_1/resource\n",
      "sequential_1/flatten_1/Const\n",
      "sequential_1/dense_3/MatMul/ReadVariableOp/resource\n",
      "sequential_1/dense_3/BiasAdd/ReadVariableOp/resource\n",
      "sequential_1/dense_4/MatMul/ReadVariableOp/resource\n",
      "sequential_1/dense_4/BiasAdd/ReadVariableOp/resource\n",
      "sequential_1/dense_5/MatMul/ReadVariableOp/resource\n",
      "sequential_1/dense_5/BiasAdd/ReadVariableOp/resource\n",
      "sequential_1/conv2d_5/Conv2D/ReadVariableOp\n",
      "sequential_1/conv2d_5/Conv2D\n",
      "sequential_1/conv2d_5/BiasAdd/ReadVariableOp\n",
      "sequential_1/conv2d_5/BiasAdd\n",
      "sequential_1/conv2d_5/Relu\n",
      "sequential_1/batch_normalization/ReadVariableOp\n",
      "sequential_1/batch_normalization/ReadVariableOp_1\n",
      "sequential_1/batch_normalization/FusedBatchNormV3/ReadVariableOp\n",
      "sequential_1/batch_normalization/FusedBatchNormV3/ReadVariableOp_1\n",
      "sequential_1/batch_normalization/FusedBatchNormV3\n",
      "sequential_1/max_pooling2d_3/MaxPool\n",
      "sequential_1/conv2d_6/Conv2D/ReadVariableOp\n",
      "sequential_1/conv2d_6/Conv2D\n",
      "sequential_1/conv2d_6/BiasAdd/ReadVariableOp\n",
      "sequential_1/conv2d_6/BiasAdd\n",
      "sequential_1/conv2d_6/Relu\n",
      "sequential_1/batch_normalization_1/ReadVariableOp\n",
      "sequential_1/batch_normalization_1/ReadVariableOp_1\n",
      "sequential_1/batch_normalization_1/FusedBatchNormV3/ReadVariableOp\n",
      "sequential_1/batch_normalization_1/FusedBatchNormV3/ReadVariableOp_1\n",
      "sequential_1/batch_normalization_1/FusedBatchNormV3\n",
      "sequential_1/max_pooling2d_4/MaxPool\n",
      "sequential_1/conv2d_7/Conv2D/ReadVariableOp\n",
      "sequential_1/conv2d_7/Conv2D\n",
      "sequential_1/conv2d_7/BiasAdd/ReadVariableOp\n",
      "sequential_1/conv2d_7/BiasAdd\n",
      "sequential_1/conv2d_7/Relu\n",
      "sequential_1/batch_normalization_2/ReadVariableOp\n",
      "sequential_1/batch_normalization_2/ReadVariableOp_1\n",
      "sequential_1/batch_normalization_2/FusedBatchNormV3/ReadVariableOp\n",
      "sequential_1/batch_normalization_2/FusedBatchNormV3/ReadVariableOp_1\n",
      "sequential_1/batch_normalization_2/FusedBatchNormV3\n",
      "sequential_1/conv2d_8/Conv2D/ReadVariableOp\n",
      "sequential_1/conv2d_8/Conv2D\n",
      "sequential_1/conv2d_8/BiasAdd/ReadVariableOp\n",
      "sequential_1/conv2d_8/BiasAdd\n",
      "sequential_1/conv2d_8/Relu\n",
      "sequential_1/batch_normalization_3/ReadVariableOp\n",
      "sequential_1/batch_normalization_3/ReadVariableOp_1\n",
      "sequential_1/batch_normalization_3/FusedBatchNormV3/ReadVariableOp\n",
      "sequential_1/batch_normalization_3/FusedBatchNormV3/ReadVariableOp_1\n",
      "sequential_1/batch_normalization_3/FusedBatchNormV3\n",
      "sequential_1/conv2d_9/Conv2D/ReadVariableOp\n",
      "sequential_1/conv2d_9/Conv2D\n",
      "sequential_1/conv2d_9/BiasAdd/ReadVariableOp\n",
      "sequential_1/conv2d_9/BiasAdd\n",
      "sequential_1/conv2d_9/Relu\n",
      "sequential_1/batch_normalization_4/ReadVariableOp\n",
      "sequential_1/batch_normalization_4/ReadVariableOp_1\n",
      "sequential_1/batch_normalization_4/FusedBatchNormV3/ReadVariableOp\n",
      "sequential_1/batch_normalization_4/FusedBatchNormV3/ReadVariableOp_1\n",
      "sequential_1/batch_normalization_4/FusedBatchNormV3\n",
      "sequential_1/max_pooling2d_5/MaxPool\n",
      "sequential_1/flatten_1/Reshape\n",
      "sequential_1/dense_3/MatMul/ReadVariableOp\n",
      "sequential_1/dense_3/MatMul\n",
      "sequential_1/dense_3/BiasAdd/ReadVariableOp\n",
      "sequential_1/dense_3/BiasAdd\n",
      "sequential_1/dense_3/Relu\n",
      "sequential_1/dense_4/MatMul/ReadVariableOp\n",
      "sequential_1/dense_4/MatMul\n",
      "sequential_1/dense_4/BiasAdd/ReadVariableOp\n",
      "sequential_1/dense_4/BiasAdd\n",
      "sequential_1/dense_4/Relu\n",
      "sequential_1/dense_5/MatMul/ReadVariableOp\n",
      "sequential_1/dense_5/MatMul\n",
      "sequential_1/dense_5/BiasAdd/ReadVariableOp\n",
      "sequential_1/dense_5/BiasAdd\n",
      "sequential_1/dense_5/Softmax\n",
      "NoOp\n",
      "Identity\n",
      "--------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "   # Load frozen graph using TensorFlow 1.x functions\n",
    "with tf.io.gfile.GFile(\"./frozen_models/frozen_graph.pb\", \"rb\") as f:\n",
    "    graph_def = tf.compat.v1.GraphDef()\n",
    "    loaded = graph_def.ParseFromString(f.read())\n",
    "\n",
    "# Wrap frozen graph to ConcreteFunctions\n",
    "frozen_func = wrap_frozen_graph(graph_def=graph_def,\n",
    "                                inputs=[\"x:0\"],\n",
    "                                outputs=[\"Identity:0\"],\n",
    "                                print_graph=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import cv2 as cv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "fish = cv.imread('./new_data/train/Fish/acanthopagrus_latus_2.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "fish = cv.resize(fish, (227, 227), interpolation=cv.INTER_AREA)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "class_names=sorted(['Fish', \"Jellyfish\", 'Penguin', 'Puffin', 'Shark', 'Starfish', 'Stingray'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0., 0., 0., 0., 0., 0., 1., 0., 0., 0.]], dtype=float32)"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "frozen_func(x=tf.convert_to_tensor(fish[None, :], dtype=\"float32\"))[0].numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from PIL import Image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "fishes = [\n",
    "    tf.convert_to_tensor(cv.resize(cv.imread('./new_data/train/Shark/D3U6ZGZZCQTF.jpg'), (227,227),interpolation=cv.INTER_AREA)[None, :], dtype='float32'),\n",
    "    tf.convert_to_tensor(cv.resize(cv.imread('./new_data/train/Shark/08XY6WGTVFYN.jpg'), (227,227), interpolation=cv.INTER_AREA)[None, :], dtype='float32')\n",
    "    ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<tf.Tensor: shape=(1, 10), dtype=float32, numpy=array([[1., 0., 0., 0., 0., 0., 0., 0., 0., 0.]], dtype=float32)>]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "frozen_func(x=fishes[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Category: Fish\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABAAAAAKrCAIAAADgdGjDAAEAAElEQVR4nOz93a8sSZIfiP3MPCLz3Hurqns+ej5EzuwMKe0u9UFyIUjgQlhgAT3pSW/6S7XQ00IPBARCgFYLLQXuasnZGc5wONPsqa6PW/eckxFupgeLsLDwCI/MyJPn41aX9e1TkZGRHubm5vbl5ub0f/k//28BEBERxQuD2v0lMPP6F5RX2xGR1cebptl4ywo+mlafPBwOxZP296E7MTNzAzDT4dC+e//+q7vjh5999buH9v3d8csm3RE1UGZuUmpx6GkEZvZr67LfAaCqqvrp/tf39/fffPPNL3/5y6+//vrTp0+qak/6M/EC1O+is56qzzdNczwe7+7u7u7ujsdj0zTM/P0333769PGHT9/1/UNqJDUKyqKnwyEFNEiVVAgA+HEXPlU8VSOhvKmcs/fdqcrMbdv6i1RVRERkINHtwF7NOtGfwhs4Xlf42fj2cvo80vrz1jvvo3NX7b1d12HkYQPnutXnc87MnFJKKVnL9iSnTpWcwoaADdZqO0zH1fvxvc7SAPDu/Wp/DYHlez/wbGbpCNbfy6EmT0BpRJVAxNwgMSEd7t4pEcBKUEEe8fryiw8iknM2RiWilJJhmFJq2zalZOTtus7+Ym1e+/gCYOamaZqmSSn1Dz9EnnEeIKK+763NlJJN5MPh4POloNvxeFzlH5FPKaWGOOf8eP/w+PgoOavqXXtA4HZn9dxW5PYzQ43Pe10fx9rzXJEQNT1ySndExNSMDbKNmAo1TdO2JjYb4wHJ2j/8B6zKbQDr8/q08fzy/qGhyjjKRjuX0wfc7nr+3emghExQQs+Q8VpESNGAEqgRMKgBkeKb5PpoEukbeDbp3didsYPIAER6QPzj2F9Nst5adb7LPrnRV9pfanabfcZvy3HZtn+eDtX+7m0nl2I56uhlv7btsSV0XRdtJA2w+vwp97vab5q7VfxdDmOum1zxxa5tv2J13GsPS/fgcxZBnh8OB6zqhbzOJ6fTKaVkagKAiPR9LyKZSuS35Q+AnLOIuOYyUeZarMCzQbAqrjP1zsJzt78KOeelCeLfqirRjDtHMEEsBBun3bPOLdrC8HpuWNLWMDGzw9DIOROD00xaEdHgiYyoro5XrRe1MY2N6GhfbpDCWHbVcNz13tpc9VcTEcZuvszQbMDlU2P5wDbyq2Onqmb9L39ba01RuT9/fi8lHb3oVJ/VFjcBIiKQvYMIRAQCE2NOtIL93CCzuePmeHxeA1/ZD13suihwZ9h/aHdyzgiuss3W0+lkzyzlvt0pZhnmmsCumfnV+fxyuJWOqHXZ5Ay4hzThSQIm9ptCD3SeblG4vaSCeyZQghIAmGGsVH4LhRBYy692vCKIo/GeVO6/XSjm+2ujcwNYpf/eri31wjPJ82LeLUV38djyVzdBo3g1rpVg0euY/a03tmo8UAVqjTSFnvOPNyTQavvPLSs9UhtfFzjGSDyIexHJuc+cRURThvJgGdQnw4Zhata/x/xeTDoUfGBmx+FwOJ3alJKIqPTElJqUmoZoiDsGKWbG0Pp41QzrjXEsvtpWkOaw+VKJGy61n1QjLhXwOGthpcXW34L2vvm88F67TFna2evjqwRAKyHWOL8uRMPfVZi/hYC+7cpPMWcHm5jY74CIExPU+G1pRjvdVqeAM9LS+ucRYmtLB8BARNxRtwcsAuSRufhGzOkZOzu8DcO0Za6FyN8o2ALLEqosUV8BW7+veZgOA90GT5BZVdmiZcx0lg990KMV8gKi/rn1ZmaxAJgwBCyjSzB4BaNjkAmYCUy5UHxG+wbAPN4/WTyund+aab0x3z9rKMdl/FibRzUwB9u1s8+jmr6uka428LV5V9iZ8flCePrA1fDf6N3TYcOeKRyAbdav2dV78Zk5ANFQfg4HILb/YnOmoJFCCElVVYXQi/Yifc657/vEOeeckjCpUgYSUdJxSJZst/o6TxU4HA6WMLB3Cj2xm46tYXI4HA6Hg+ctcCZVXnNLBtMfN+KHYojP/nyIzAEIxv2G47FXoMy9o3Af69cvAIXDtj2Hl/bospGzP1FV0Cy1YHoYC3pOmngFapKXxr/FxfArQI2dAGPTOI7uljyHFB5Mf2ZiJkwygZgJPEYEMK4Briyh6FqqxsZ4eQqWEcp+W0w990l8xTK+KL4O8zlVUN4x9I8GQweJkJ9Xsd0KtlPv1mDLQV15WjKIlFVZiZJqhrKqUtMCorlThZAAbHSs4flalumzOwCjuS+jM6AEIQgJgxUwZUYACKTAkLK1LxAw2feI5s709y07AK+NwrNAVCvRH7tuAApHYgOucADOthZNPpOrGqJgt3XUXYBHBM5iuISYBIuL6ba00yJKEWrtNLVWbgjP3f4qxLzzeAHJCoFClQDOA3R93yc+ZelEjsQZSACItGDlswMTHYDD4dA0Td/3L+ADRO5BsD+oHSCl1PWas6g2NMb7g8VQOsdLrtqFz2riQbTyC4gPR5r3/XqO4FlDefWrwSSil4vY1cDpE2XHNvJYo9LZcYmi3NRquAmAzIJiro3LFp2XUuasmek/fGHiRzyZqnFxNxyd4AbmQheDFfviz9tHywD0PEDjfB+yuJPEPIS2bf3OuCaZixSjiIA1FddPfHuAiEDhGwaYiJnlM3EAbqUdatwl2hMRwEBDpACr9irmofX2Q6YGyGcx0R9jJFgYgFn/cOt/dAmEQArKBFIiIqXRIdgDMyufSmNx7gy8RVid76+N1A1gaf3jAoNnFaIuuzlxVuddtHziX88juHCe7jXVCjvzQqW8hLjyGZT1FtTstFWoNdLE3xc65iZQa/8FZGXU1n5TNRM1Rl5FL9KI9CKSpTNXQKRnaoApVrqc6hvIe/6Pm90voxWK+UajA4As5o00TYPHySxYYDXGaivjtTflpsDH/9aoUUwhnwO1SF7NMahtWorvvWQF4Lll+nK8tudqvO+m3tn2Y+RjDCeX9BzHd51uNQfAfxuT+ImoU2BtBWDo5nBz4DQAInk+NGRt3moRILIcjSsATJx1+FZFCJCR4wizjFINgMUs0PrStk/8+HMigk53nHpux8epZy0vU4AM3JcuZs3gaYiO0kxo58x9XahxdW1e1OZAdR7pwJMEIU2qmcCAQEWFRDpwoyTMDQIb7MGz9ovbQJUON5JXmQSIRr/BwEJiU1dR2cp4HqJ9Hwz9aY4o4v3qGszLaNXL3/u5+wDR7o8qey+dXXYVcmnj+b14rt63SH+UnDTF+55xBQCLRYANJLEZuIw2TyB+VR7W7OpVqKFU3QNwvt+XQa39557Am3sAMsBGYSO3qUtpepFeJGsaCoAQq8pKZHoDiMgdAN+A+0x9LN5rF9FYYWYZtwJbUSBFti6ntMxlBFF1D0AtN3fDYy7Ybpt6RdUFT4G4Itd/9b6lXtAiEvBaDkANNgSEXRRi+izE5wcTd63Zeu711qKEC1wfvq4/Y7gXHZRcbv62mM3zZQHZK3Ie9x0qRLJgrIqQypVDjCtRntOPsAnYqvEs+cr4LYryoWu9RifcH/CLOLJE1LatLsAfi3i68gBAorbwKCK033t/RXh2B4CUyP6SFXtQFSIelIKQakeUNJRuq+FJa5Hgz90BKPb+smX3q5AKAB4zRZVIqmk/Mv60iqdju2J0UnF//R0vo1VX37s6338EDgDWxmWv6IjxIIR5UbMT9o5jbd55vTWaxyujM3D1S28I2/q9EPJn23m63T6F/Yq2bg7P3X4BxnArDgALkABRJSCr5tEm7kVyoLvUlcsWRJOo2OD4AhBNBwpBRC8H2Y/1DlOaKhTRWAio6Ei82NuR2Lgbcxts7WOEcRoYnpYacTlsvyJK7V3N3hx8wsc7G0RexfnyXiwFy4gAjRJzfQVgew+GB7DdWmIVrK8AjC/1CwBAHzaNPdNkMTW2ws+TzIWMZTYLB8Ch7/vozJtKyzkfj8dVvnKL3Mv7eCEgXdsYZ+U+l3sMzJHw6eBvjy+KPR1XAMTdjEvUyduBvQ5ADWrPJ1ICMYhUiBJIFaQqJARkgKFEDFIlGZm1jmq0Qnah94ZB3Pq
      "text/plain": [
       "<PIL.JpegImagePlugin.JpegImageFile image mode=RGB size=1024x683>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fish_path = './new_data/train/Shark/D3U6ZGZZCQTF.jpg'\n",
    "fish = cv.imread(fish_path)\n",
    "fish = cv.resize(fish, (227, 227), interpolation=cv.INTER_AREA)\n",
    "print(f\"Category: {class_names[np.argmax(frozen_func(x=tf.convert_to_tensor(fish[None, :], dtype='float32'))[0].numpy())]}\")\n",
    "Image.open(fish_path)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "um",
   "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": "3.9.15"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "876e189cbbe99a9a838ece62aae1013186c4bb7e0254a10cfa2f9b2381853efb"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}