grain_unearth-bilinear-mixup.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "executionInfo": {
     "elapsed": 2716,
     "status": "ok",
     "timestamp": 1624439925606,
     "user": {
      "displayName": "ramdhan wibawa",
      "photoUrl": "",
      "userId": "14519448541243845830"
     },
     "user_tz": -420
    },
    "id": "mW8mQghJfh6D"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    },
    {
     "ename": "AttributeError",
     "evalue": "module 'tensorflow.compat.v2' has no attribute '__internal__'",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-1-a10c542f5d12>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m     27\u001b[0m \u001b[1;31m#from tensorflow.keras.applications.efficientnet import preprocess_input\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     28\u001b[0m \u001b[1;32mfrom\u001b[0m \u001b[0mtensorflow\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mkeras\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mapplications\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mresnet50\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mpreprocess_input\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 29\u001b[1;33m \u001b[1;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpreprocessing\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mimage\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mImageDataGenerator\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     30\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mtensorflow\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mkeras\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mbackend\u001b[0m \u001b[1;32mas\u001b[0m \u001b[0mK\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     31\u001b[0m \u001b[1;32mfrom\u001b[0m \u001b[0msklearn\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmetrics\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0maccuracy_score\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\object_detection_gpu\\lib\\site-packages\\keras\\__init__.py\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m     23\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     24\u001b[0m \u001b[1;31m# See b/110718070#comment18 for more details about this import.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 25\u001b[1;33m \u001b[1;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mmodels\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     26\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     27\u001b[0m \u001b[1;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mengine\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0minput_layer\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mInput\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\object_detection_gpu\\lib\\site-packages\\keras\\models.py\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m     17\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     18\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mtensorflow\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcompat\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mv2\u001b[0m \u001b[1;32mas\u001b[0m \u001b[0mtf\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 19\u001b[1;33m \u001b[1;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mbackend\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     20\u001b[0m \u001b[1;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mmetrics\u001b[0m \u001b[1;32mas\u001b[0m \u001b[0mmetrics_module\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     21\u001b[0m \u001b[1;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0moptimizer_v1\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\object_detection_gpu\\lib\\site-packages\\keras\\backend\\__init__.py\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[1;32mfrom\u001b[0m \u001b[1;33m.\u001b[0m\u001b[0mload_backend\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mepsilon\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      2\u001b[0m \u001b[1;32mfrom\u001b[0m \u001b[1;33m.\u001b[0m\u001b[0mload_backend\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mset_epsilon\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      3\u001b[0m \u001b[1;32mfrom\u001b[0m \u001b[1;33m.\u001b[0m\u001b[0mload_backend\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mfloatx\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      4\u001b[0m \u001b[1;32mfrom\u001b[0m \u001b[1;33m.\u001b[0m\u001b[0mload_backend\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mset_floatx\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      5\u001b[0m \u001b[1;32mfrom\u001b[0m \u001b[1;33m.\u001b[0m\u001b[0mload_backend\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mcast_to_floatx\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\object_detection_gpu\\lib\\site-packages\\keras\\backend\\load_backend.py\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m     88\u001b[0m \u001b[1;32melif\u001b[0m \u001b[0m_BACKEND\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;34m'tensorflow'\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     89\u001b[0m     \u001b[0msys\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstderr\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mwrite\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'Using TensorFlow backend.\\n'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 90\u001b[1;33m     \u001b[1;32mfrom\u001b[0m \u001b[1;33m.\u001b[0m\u001b[0mtensorflow_backend\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[1;33m*\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     91\u001b[0m \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     92\u001b[0m     \u001b[1;31m# Try and load external backend.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\object_detection_gpu\\lib\\site-packages\\keras\\backend\\tensorflow_backend.py\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m     23\u001b[0m \u001b[1;32mfrom\u001b[0m \u001b[0mdistutils\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mversion\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mStrictVersion\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     24\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 25\u001b[1;33m \u001b[1;32mfrom\u001b[0m \u001b[1;33m.\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mutils\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mgeneric_utils\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mtranspose_shape\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     26\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     27\u001b[0m \u001b[0mpy_all\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mall\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\object_detection_gpu\\lib\\site-packages\\keras\\utils\\generic_utils.py\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m    412\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    413\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 414\u001b[1;33m \u001b[1;33m@\u001b[0m\u001b[0mtf_contextlib\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcontextmanager\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    415\u001b[0m \u001b[1;32mdef\u001b[0m \u001b[0mskip_failed_serialization\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    416\u001b[0m   \u001b[1;32mglobal\u001b[0m \u001b[0m_SKIP_FAILED_SERIALIZATION\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\object_detection_gpu\\lib\\site-packages\\keras\\utils\\tf_contextlib.py\u001b[0m in \u001b[0;36mcontextmanager\u001b[1;34m(target)\u001b[0m\n\u001b[0;32m     31\u001b[0m   \"\"\"\n\u001b[0;32m     32\u001b[0m   \u001b[0mcontext_manager\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0m_contextlib\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcontextmanager\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtarget\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 33\u001b[1;33m   \u001b[1;32mreturn\u001b[0m \u001b[0mtf\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m__internal__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdecorator\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmake_decorator\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtarget\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcontext_manager\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m'contextmanager'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;31mAttributeError\u001b[0m: module 'tensorflow.compat.v2' has no attribute '__internal__'"
     ]
    }
   ],
   "source": [
    "import logging\n",
    "import sys\n",
    "import os\n",
    "from os import getenv\n",
    "from os.path import abspath, basename, split,dirname\n",
    "import tarfile\n",
    "from shutil import copyfile\n",
    "import random, glob\n",
    "from skimage.util import random_noise\n",
    "from skimage.transform import rotate\n",
    "import time\n",
    "import numpy as np\n",
    "import math\n",
    "\n",
    "import tensorflow as tf\n",
    "import tensorflow_datasets as tfds\n",
    "#import efficientnet.keras as efn\n",
    "#from efficientnet.tfkeras import EfficientNetB2\n",
    "from tensorflow.keras.applications import EfficientNetB0,EfficientNetB3, EfficientNetB2\n",
    "from tensorflow.keras.applications.resnet50 import ResNet50\n",
    "from tensorflow.keras.applications.resnet import ResNet101\n",
    "from tensorflow.keras.applications.inception_v3 import InceptionV3\n",
    "from tensorflow.keras.applications.mobilenet import MobileNet\n",
    "from tensorflow.keras.applications.densenet import DenseNet121, DenseNet169, DenseNet201\n",
    "from tensorflow.keras import layers\n",
    "from tensorflow import keras\n",
    "#from tensorflow.keras.applications.efficientnet import preprocess_input\n",
    "from tensorflow.keras.applications.resnet50 import preprocess_input\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "import tensorflow.keras.backend as K\n",
    "from sklearn.metrics import accuracy_score\n",
    "from tensorflow.keras.callbacks import ModelCheckpoint"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "logger = logging.getLogger(__name__)\n",
    "logging.basicConfig(level=logging.INFO)\n",
    "ds_build_cmd = \"tfds build grains --manual_dir=\" + \"D:/Ramdhan/Unearth/visual-grain-analysis/data/public2\"\n",
    "logger.info(ds_build_cmd)\n",
    "os.system(ds_build_cmd)"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "model1 = ResNet50(weights='imagenet', include_top=False, input_shape=(224, 224, 3))\n",
    "    \n",
    "for layer in model1.layers:\n",
    "    print(layer.name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "alpha=0.2\n",
    "\n",
    "def sample_beta_distribution(size, concentration_0=alpha, concentration_1=alpha):\n",
    "    gamma_1_sample = tf.random.gamma(shape=[size], alpha=concentration_1)\n",
    "    gamma_2_sample = tf.random.gamma(shape=[size], alpha=concentration_0)\n",
    "    return gamma_1_sample / (gamma_1_sample + gamma_2_sample)\n",
    "\n",
    "\n",
    "def mix_up(ds_one, ds_two, alpha=0.2):\n",
    "    # Unpack two datasets\n",
    "    images_one, labels_one = ds_one\n",
    "    images_two, labels_two = ds_two\n",
    "    batch_size = tf.shape(images_one)[0]\n",
    "\n",
    "    # Sample lambda and reshape it to do the mixup\n",
    "    l = sample_beta_distribution(batch_size, alpha, alpha)\n",
    "    x_l = tf.reshape(l, (batch_size, 1, 1, 1))\n",
    "    y_l = tf.reshape(l, (batch_size, 1))\n",
    "\n",
    "    # Perform mixup on both images and labels by combining a pair of images/labels\n",
    "    # (one from each dataset) into one image/label\n",
    "    images = images_one * x_l + images_two * (1 - x_l)\n",
    "    labels = labels_one * y_l + labels_two * (1 - y_l)\n",
    "    return (images, labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "executionInfo": {
     "elapsed": 493,
     "status": "ok",
     "timestamp": 1624440223471,
     "user": {
      "displayName": "ramdhan wibawa",
      "photoUrl": "",
      "userId": "14519448541243845830"
     },
     "user_tz": -420
    },
    "id": "-GFNqiGNfh6F"
   },
   "outputs": [],
   "source": [
    "logger = logging.getLogger(__name__)\n",
    "logging.basicConfig(level=logging.INFO)\n",
    "IMG_SIZE = 224\n",
    "LBL = dict(zip(['B_BSMUT1', 'B_CLEV5B', 'B_DISTO', 'B_GRMEND', 'B_HDBARL',\n",
    "       'B_PICKLD', 'B_SKINED', 'B_SOUND', 'B_SPRTED', 'B_SPTMLD',\n",
    "       'O_GROAT', 'O_HDOATS', 'O_SEPAFF', 'O_SOUND', 'O_SPOTMA',\n",
    "       'WD_RADPODS', 'WD_RYEGRASS', 'WD_SPEARGRASS', 'WD_WILDOATS',\n",
    "       'W_DISTO', 'W_FLDFUN', 'W_INSDA2', 'W_PICKLE', 'W_SEVERE',\n",
    "       'W_SOUND', 'W_SPROUT', 'W_STAIND', 'W_WHITEG'], range(28)))\n",
    "\n",
    "cls_map = dict(zip(LBL.values(),LBL.keys()))\n",
    "\n",
    "model_version = '001'\n",
    "\n",
    "# Work around for a SageMaker path issue\n",
    "# (see https://github.com/aws/sagemaker-python-sdk/issues/648)\n",
    "# WARNING - removing this may cause the submission process to fail\n",
    "#if abspath(\"D:/Ramdhan/Unearth/visual-grain-analysis\") not in sys.path:\n",
    " #   sys.path.append(abspath(\"D:/Ramdhan/Unearth/visual-grain-analysis\"))\n",
    "\n",
    "def input_preprocess(image, label):\n",
    "    label = tf.one_hot(label, 28)\n",
    "    image = preprocess_input(image)\n",
    "    image = tf.cast(image,'float32')\n",
    "    return image, label\n",
    "\n",
    "def get_subfolders(directory):\n",
    "    subfolders = os.listdir(directory)\n",
    "    subfolders.sort()\n",
    "    return subfolders\n",
    "\n",
    "def save_class_list(class_list):\n",
    "    class_list.sort()\n",
    "    target=open(\"class_list.txt\",'w')\n",
    "    for c in class_list:\n",
    "        target.write(c)\n",
    "        target.write(\"\\n\")\n",
    "\n",
    "def get_num_files(directory):\n",
    "    if not os.path.exists(directory):\n",
    "        return 0\n",
    "    cnt = 0\n",
    "    for r, dirs, files in os.walk(directory):\n",
    "        for dr in dirs:\n",
    "            cnt += len(glob.glob(os.path.join(r, dr + \"/*\")))\n",
    "    return cnt\n",
    "\n",
    "def outer_product(x):\n",
    "    #Einstein Notation  [batch,1,1,depth] x [batch,1,1,depth] -> [batch,depth,depth]\n",
    "    phi_I = tf.einsum('ijkm,ijkn->imn',x[0],x[1])\n",
    "    \n",
    "    # Reshape from [batch_size,depth,depth] to [batch_size, depth*depth]\n",
    "    phi_I = tf.reshape(phi_I,[-1,x[0].shape[3]*x[1].shape[3]])\n",
    "    \n",
    "    # Divide by feature map size [sizexsize]\n",
    "    size1 = int(x[1].shape[1])\n",
    "    size2 = int(x[1].shape[2])\n",
    "    phi_I = tf.divide(phi_I, size1*size2)\n",
    "    \n",
    "    # Take signed square root of phi_I\n",
    "    y_ssqrt = tf.multiply(tf.sign(phi_I),tf.sqrt(tf.abs(phi_I)+1e-12))\n",
    "    \n",
    "    # Apply l2 normalization\n",
    "    z_l2 = tf.nn.l2_normalize(y_ssqrt, axis=1)\n",
    "    return z_l2\n",
    "\n",
    "def get_cosine_schedule_with_warmup(lr, num_warmup_steps, num_training_steps, num_cycles=0.5):\n",
    "    \"\"\"\n",
    "    Modified the get_cosine_schedule_with_warmup from huggingface for tensorflow\n",
    "    (https://huggingface.co/transformers/_modules/transformers/optimization.html#get_cosine_schedule_with_warmup)\n",
    "\n",
    "    Create a schedule with a learning rate that decreases following the\n",
    "    values of the cosine function between 0 and `pi * cycles` after a warmup\n",
    "    period during which it increases linearly between 0 and 1.\n",
    "    \"\"\"\n",
    "\n",
    "    def lrfn(epoch):\n",
    "        if epoch < num_warmup_steps:\n",
    "            return float(epoch) / float(max(1, num_warmup_steps)) * lr\n",
    "        progress = float(epoch - num_warmup_steps) / float(max(1, num_training_steps - num_warmup_steps))\n",
    "        return max(0.0, 0.5 * (1.0 + math.cos(math.pi * float(num_cycles) * 2.0 * progress))) * lr\n",
    "\n",
    "    return tf.keras.callbacks.LearningRateScheduler(lrfn, verbose=True)\n",
    "\n",
    "def random_rotation(image_array):\n",
    "    # pick a random degree of rotation between 25% on the left and 25% on the right\n",
    "    random_degree = random.uniform(-180, 180)\n",
    "    return rotate(image_array, random_degree, mode='edge')\n",
    "\n",
    "def rotation_generator(batches):\n",
    "    while True:\n",
    "        batch_x, batch_y = next(batches)\n",
    "        batch_crops = np.zeros((batch_x.shape[0], 224, 224, 3))\n",
    "        for i in range(batch_x.shape[0]):\n",
    "            #batch_crops[i] = resize_pad(batch_x[i])\n",
    "            batch_crops[i] = random_rotation(batch_x[i])\n",
    "        yield (batch_crops, batch_y)\n",
    "        \n",
    "\n",
    "def train(args):\n",
    "    global NUM_CLASSES\n",
    "    TRAIN_DIR = os.path.join(args['data_dir'],  'train')\n",
    "    VAL_DIR = os.path.join(args['data_dir'], 'val')\n",
    "    size = (IMG_SIZE, IMG_SIZE)\n",
    "    batch_size = 32\n",
    "    train_ds = tf.keras.preprocessing.image_dataset_from_directory(\n",
    "        \"D:/Ramdhan/Unearth/visual-grain-analysis/data/public/train\",\n",
    "        image_size=size,\n",
    "        batch_size=batch_size,\n",
    "        interpolation='nearest'\n",
    "    )\n",
    "\n",
    "    train_ds_one = tf.keras.preprocessing.image_dataset_from_directory(\n",
    "        \"D:/Ramdhan/Unearth/visual-grain-analysis/data/public/train\",\n",
    "        image_size=size,\n",
    "        batch_size=batch_size,\n",
    "        interpolation='nearest'\n",
    "    )\n",
    "\n",
    "    train_ds_two = tf.keras.preprocessing.image_dataset_from_directory(\n",
    "        \"D:/Ramdhan/Unearth/visual-grain-analysis/data/public/train\",\n",
    "        image_size=size,\n",
    "        batch_size=batch_size,\n",
    "        interpolation='nearest'\n",
    "    )\n",
    "        \n",
    "    train_ds_one = train_ds_one.map(input_preprocess)\n",
    "    train_ds_two = train_ds_two.map(input_preprocess)\n",
    "\n",
    "    train_ds_combine = tf.data.Dataset.zip((train_ds_one, train_ds_two))\n",
    "\n",
    "    val_ds = tf.keras.preprocessing.image_dataset_from_directory(\n",
    "        \"D:/Ramdhan/Unearth/visual-grain-analysis/data/public/val\",\n",
    "        image_size=size,\n",
    "        batch_size=batch_size,\n",
    "        interpolation='nearest'\n",
    "    )\n",
    "\n",
    "    val_ds = val_ds.map(input_preprocess)\n",
    "\n",
    "    train_ds_combine_cum = train_ds_combine.map(lambda train_ds_one,train_ds_two: mix_up(train_ds_one,train_ds_two, alpha=0.2))\n",
    "\n",
    "    logger.info(\"Training Model\")\n",
    "    class_list = get_subfolders(TRAIN_DIR)\n",
    "    NUM_CLASSES = len(class_list)\n",
    "\n",
    "    num_train_images = get_num_files(TRAIN_DIR)\n",
    "    num_val_images = get_num_files(VAL_DIR)\n",
    "    \n",
    "    model = build_model_bilinear(num_classes=NUM_CLASSES)\n",
    "    filename = 'best.h5'\n",
    "    #reduce_lr = tf.keras.callbacks.ReduceLROnPlateau(monitor='loss', factor=0.5, patience=50, min_lr=0.0001)\n",
    "    checkpoint = ModelCheckpoint(filename, monitor='val_acc_grain', verbose=1, save_best_only=True, mode='max')\n",
    "    stop = tf.keras.callbacks.EarlyStopping(monitor='val_loss', patience=40, mode= 'min', min_delta=0.00)\n",
    "    lr_schedule= get_cosine_schedule_with_warmup(lr=0.001, num_warmup_steps=4, num_training_steps=10)\n",
    "\n",
    "    # This comes from an environment variable so we can set it to 1 for our\n",
    "    # development pipeline. Feel free to set this to any value.\n",
    "    weights = dict(zip([0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27], [1,1,1,1,1,1,1,8.94,1,1,1,1,1,8.94,1,1,1,1,1,1,1,1,1,1,8.94,1,1,1]))\n",
    "    \n",
    "    history = model.fit(train_ds_combine_cum, epochs=10, validation_data=val_ds, shuffle=True,\n",
    "                        callbacks=[stop,checkpoint,lr_schedule],#steps_per_epoch=num_train_images // batch_size,\n",
    "                         verbose=1, class_weight=weights)\n",
    "    \n",
    "    #save_model(model,args['model_dir'])\n",
    "    #save_model(model,filename)\n",
    "    \n",
    "    return history\n",
    "\n",
    "def acc_grain(y_true, y_pred):\n",
    "    y_true_class = K.argmax(y_true, axis=-1)\n",
    "    y_pred_class = K.argmax(y_pred, axis=-1)\n",
    "    sample_weight = np.hstack((np.where(y_true_class.numpy().reshape(-1,1)==7,True,False),np.where(y_true_class.numpy().reshape(-1,1)==24,True,False),np.where(y_true_class.numpy().reshape(-1,1)==13,True,False)))\n",
    "    sample_weight = np.any(sample_weight,axis=1)\n",
    "    sample_weight = np.where(sample_weight,8.94,1.0)\n",
    "    return K.constant(accuracy_score(y_true_class,y_pred_class,sample_weight=sample_weight))\n",
    "\n",
    "# Add on new FC layers with dropout for fine tuning\n",
    "def build_model_bilinear(num_classes):\n",
    "    \n",
    "    input_tensor = layers.Input(shape=(IMG_SIZE, IMG_SIZE, 3))\n",
    "    \n",
    "    model1 = ResNet50(weights='imagenet', include_top=False, input_tensor=input_tensor,input_shape=(IMG_SIZE, IMG_SIZE, 3))\n",
    "    model2 = ResNet50(weights='imagenet', include_top=False, input_tensor=input_tensor,input_shape=(IMG_SIZE, IMG_SIZE, 3))\n",
    "    \n",
    "    for layer in model1.layers[:]:\n",
    "        layer.trainable = False\n",
    "    for layer in model2.layers[:-1]:\n",
    "        layer.trainable = False\n",
    "        \n",
    "    for i, layer in enumerate(model1.layers):\n",
    "        layer._name = 'model1_' + layer.name\n",
    "\n",
    "    last_layer1 = model1.get_layer('model1_conv5_block3_3_conv')\n",
    "    last_output1 = last_layer1.output\n",
    "\n",
    "    for i, layer in enumerate(model2.layers):\n",
    "        layer._name = 'model2_' + layer.name\n",
    "\n",
    "    last_layer2 = model2.get_layer('model2_conv5_block3_3_conv')\n",
    "    #last_layer2 = model2.get_layer('model2_top_conv')\n",
    "    last_output2 = last_layer2.output\n",
    "    \n",
    "    \n",
    "    model1_ = tf.keras.Model(inputs=model1.input, outputs=last_output1)\n",
    "    model2_ = tf.keras.Model(inputs=model2.input, outputs=last_output2)\n",
    "    \n",
    "    #optimizer = tf.keras.optimizers.Adam(learning_rate=1e-5, decay=1e-3)\n",
    "   \n",
    "    \n",
    "    model1_.compile(optimizer='adam',loss=\"categorical_crossentropy\", metrics=[acc_grain],run_eagerly=True)\n",
    "    model2_.compile(optimizer='adam',loss=\"categorical_crossentropy\", metrics=[acc_grain],run_eagerly=True)\n",
    "    \n",
    "    d1=model1_.output\n",
    "    d2=model2_.output\n",
    "\n",
    "    bilinear = layers.Lambda(outer_product, name='outer_product1')([d1,d2])\n",
    "    \n",
    "    predictions= layers.Dense(num_classes, activation='softmax', name='pred')(bilinear)\n",
    "    model = tf.keras.Model(inputs=model1.input, outputs=predictions)\n",
    "    \n",
    "    model.compile(optimizer='adam', loss=\"categorical_crossentropy\", metrics=[acc_grain],run_eagerly=True)\n",
    "    \n",
    "    return model\n",
    "    \n",
    "def save_model(model, model_dir):\n",
    "    \"\"\"Save model to a binary file.\n",
    "\n",
    "    This function must write the model to disk in a format that can\n",
    "    be loaded from the model_fn.\n",
    "\n",
    "    WARNING - modifying this function may cause the submission process to fail.\n",
    "    \"\"\"\n",
    "    #sm_model_dir = os.path.join(model_dir, model_version)\n",
    "    sm_model_dir = model_dir\n",
    "    if (not os.path.isdir(sm_model_dir)):\n",
    "        os.makedirs(sm_model_dir)\n",
    "        \n",
    "    logger.info(f\" model dir is {model_dir}\")\n",
    "    model.save(sm_model_dir)\n",
    "    \n",
    "    modelPath = os.path.join(sm_model_dir, 'output')\n",
    "    if (not os.path.isdir(modelPath)):\n",
    "        os.makedirs(modelPath)\n",
    "    #if (not os.path.isdir(sm_model_dir + '/code')):\n",
    "        #os.makedirs(sm_model_dir + '/code')\n",
    "\n",
    "    # Move inference.py so it gets picked up in the archive\n",
    "    #copyfile('/content/sample_data/inference.py', sm_model_dir + '/code/inference.py')\n",
    "    #copyfile('/content/sample_data/inference-requirements.txt', sm_model_dir+ '/code/requirements.txt')\n",
    "\n",
    "    with tarfile.open(os.path.join(modelPath, 'model.tar.gz'), mode='x:gz') as archive:\n",
    "        archive.add(sm_model_dir, recursive=True)\n",
    "\n",
    "def model_fn(model_dir):\n",
    "    \"\"\"Load model from binary file.\n",
    "\n",
    "    This function loads the model from disk. It is called by SageMaker.\n",
    "\n",
    "    WARNING - modifying this function may case the submission process to fail.\n",
    "    \"\"\"\n",
    "    #model_filepath = os.path.join(model_dir,  model_version)\n",
    "    #logger.info(\"loading model from \" + model_filepath)\n",
    "    model = keras.models.load_model('best.h5',custom_objects={'acc_grain':acc_grain},compile=False)\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "tf.test.is_gpu_available()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "tf.test.gpu_device_name() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 459
    },
    "executionInfo": {
     "elapsed": 38441,
     "status": "error",
     "timestamp": 1624440265326,
     "user": {
      "displayName": "ramdhan wibawa",
      "photoUrl": "",
      "userId": "14519448541243845830"
     },
     "user_tz": -420
    },
    "id": "RLmWtwmRfh6P",
    "outputId": "e5e2ae19-620c-4d31-f907-12bb2132639e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 19182 files belonging to 28 classes.\n",
      "Found 19182 files belonging to 28 classes.\n",
      "Found 19182 files belonging to 28 classes.\n",
      "Found 7648 files belonging to 28 classes.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:__main__:Training Model\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Epoch 00001: LearningRateScheduler reducing learning rate to 0.0.\n",
      "Epoch 1/10\n",
      " 31/600 [>.............................] - ETA: 7:09 - loss: 21.9479 - acc_grain: 0.1103"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-6-f940733bccd2>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m     13\u001b[0m            'data_dir' : \"D:/Ramdhan/Unearth/visual-grain-analysis/data/public/\"}\n\u001b[0;32m     14\u001b[0m     \u001b[0mstart\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtime\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtime\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 15\u001b[1;33m     \u001b[0mhistory\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtrain\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     16\u001b[0m     \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'duration run'\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtime\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtime\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m-\u001b[0m\u001b[0mstart\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m/\u001b[0m\u001b[1;36m60\u001b[0m\u001b[1;33m/\u001b[0m\u001b[1;36m60\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;34m' hours'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m<ipython-input-3-3630e3fe606d>\u001b[0m in \u001b[0;36mtrain\u001b[1;34m(args)\u001b[0m\n\u001b[0;32m    161\u001b[0m     history = model.fit(train_ds_combine_cum, epochs=10, validation_data=val_ds, shuffle=True,\n\u001b[0;32m    162\u001b[0m                         \u001b[0mcallbacks\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mstop\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mcheckpoint\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mlr_schedule\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;31m#steps_per_epoch=num_train_images // batch_size,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 163\u001b[1;33m                          verbose=1, class_weight=weights)\n\u001b[0m\u001b[0;32m    164\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    165\u001b[0m     \u001b[1;31m#save_model(model,args['model_dir'])\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\keras\\engine\\training.py\u001b[0m in \u001b[0;36m_method_wrapper\u001b[1;34m(self, *args, **kwargs)\u001b[0m\n\u001b[0;32m    106\u001b[0m   \u001b[1;32mdef\u001b[0m \u001b[0m_method_wrapper\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    107\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_in_multi_worker_mode\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m  \u001b[1;31m# pylint: disable=protected-access\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 108\u001b[1;33m       \u001b[1;32mreturn\u001b[0m \u001b[0mmethod\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    109\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    110\u001b[0m     \u001b[1;31m# Running inside `run_distribute_coordinator` already.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\keras\\engine\\training.py\u001b[0m in \u001b[0;36mfit\u001b[1;34m(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, validation_batch_size, validation_freq, max_queue_size, workers, use_multiprocessing)\u001b[0m\n\u001b[0;32m   1096\u001b[0m                 batch_size=batch_size):\n\u001b[0;32m   1097\u001b[0m               \u001b[0mcallbacks\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mon_train_batch_begin\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mstep\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1098\u001b[1;33m               \u001b[0mtmp_logs\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtrain_function\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0miterator\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1099\u001b[0m               \u001b[1;32mif\u001b[0m \u001b[0mdata_handler\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mshould_sync\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1100\u001b[0m                 \u001b[0mcontext\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0masync_wait\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\keras\\engine\\training.py\u001b[0m in \u001b[0;36mtrain_function\u001b[1;34m(iterator)\u001b[0m\n\u001b[0;32m    804\u001b[0m       \u001b[1;32mdef\u001b[0m \u001b[0mtrain_function\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0miterator\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    805\u001b[0m         \u001b[1;34m\"\"\"Runs a training execution with one step.\"\"\"\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 806\u001b[1;33m         \u001b[1;32mreturn\u001b[0m \u001b[0mstep_function\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0miterator\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    807\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    808\u001b[0m     \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\keras\\engine\\training.py\u001b[0m in \u001b[0;36mstep_function\u001b[1;34m(model, iterator)\u001b[0m\n\u001b[0;32m    794\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    795\u001b[0m       \u001b[0mdata\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnext\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0miterator\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 796\u001b[1;33m       \u001b[0moutputs\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdistribute_strategy\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mrun\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mrun_step\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0margs\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdata\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    797\u001b[0m       outputs = reduce_per_replica(\n\u001b[0;32m    798\u001b[0m           outputs, self.distribute_strategy, reduction='first')\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\distribute\\distribute_lib.py\u001b[0m in \u001b[0;36mrun\u001b[1;34m(***failed resolving arguments***)\u001b[0m\n\u001b[0;32m   1209\u001b[0m       fn = autograph.tf_convert(\n\u001b[0;32m   1210\u001b[0m           fn, autograph_ctx.control_status_ctx(), convert_by_default=False)\n\u001b[1;32m-> 1211\u001b[1;33m       \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_extended\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcall_for_each_replica\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfn\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0margs\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkwargs\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1212\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1213\u001b[0m   \u001b[1;31m# TODO(b/151224785): Remove deprecated alias.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\distribute\\distribute_lib.py\u001b[0m in \u001b[0;36mcall_for_each_replica\u001b[1;34m(self, fn, args, kwargs)\u001b[0m\n\u001b[0;32m   2583\u001b[0m       \u001b[0mkwargs\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m{\u001b[0m\u001b[1;33m}\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   2584\u001b[0m     \u001b[1;32mwith\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_container_strategy\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mscope\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 2585\u001b[1;33m       \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_call_for_each_replica\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfn\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   2586\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   2587\u001b[0m   \u001b[1;32mdef\u001b[0m \u001b[0m_call_for_each_replica\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfn\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\distribute\\distribute_lib.py\u001b[0m in \u001b[0;36m_call_for_each_replica\u001b[1;34m(self, fn, args, kwargs)\u001b[0m\n\u001b[0;32m   2943\u001b[0m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_container_strategy\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   2944\u001b[0m         replica_id_in_sync_group=constant_op.constant(0, dtypes.int32)):\n\u001b[1;32m-> 2945\u001b[1;33m       \u001b[1;32mreturn\u001b[0m \u001b[0mfn\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   2946\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   2947\u001b[0m   \u001b[1;32mdef\u001b[0m \u001b[0m_reduce_to\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mreduce_op\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mvalue\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdestinations\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mexperimental_hints\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\autograph\\impl\\api.py\u001b[0m in \u001b[0;36mwrapper\u001b[1;34m(*args, **kwargs)\u001b[0m\n\u001b[0;32m    273\u001b[0m   \u001b[1;32mdef\u001b[0m \u001b[0mwrapper\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    274\u001b[0m     \u001b[1;32mwith\u001b[0m \u001b[0mag_ctx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mControlStatusCtx\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mstatus\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mag_ctx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mStatus\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mUNSPECIFIED\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 275\u001b[1;33m       \u001b[1;32mreturn\u001b[0m \u001b[0mfunc\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    276\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    277\u001b[0m   \u001b[1;32mif\u001b[0m \u001b[0minspect\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0misfunction\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfunc\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;32mor\u001b[0m \u001b[0minspect\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mismethod\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfunc\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\keras\\engine\\training.py\u001b[0m in \u001b[0;36mrun_step\u001b[1;34m(data)\u001b[0m\n\u001b[0;32m    787\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    788\u001b[0m       \u001b[1;32mdef\u001b[0m \u001b[0mrun_step\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdata\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 789\u001b[1;33m         \u001b[0moutputs\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtrain_step\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdata\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    790\u001b[0m         \u001b[1;31m# Ensure counter is updated only if `train_step` succeeds.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    791\u001b[0m         \u001b[1;32mwith\u001b[0m \u001b[0mops\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcontrol_dependencies\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0m_minimum_control_deps\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0moutputs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\keras\\engine\\training.py\u001b[0m in \u001b[0;36mtrain_step\u001b[1;34m(self, data)\u001b[0m\n\u001b[0;32m    757\u001b[0m               self.trainable_variables)\n\u001b[0;32m    758\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 759\u001b[1;33m     \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcompiled_metrics\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mupdate_state\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my_pred\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0msample_weight\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    760\u001b[0m     \u001b[1;32mreturn\u001b[0m \u001b[1;33m{\u001b[0m\u001b[0mm\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mname\u001b[0m\u001b[1;33m:\u001b[0m \u001b[0mm\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mresult\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;32mfor\u001b[0m \u001b[0mm\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmetrics\u001b[0m\u001b[1;33m}\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    761\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\keras\\engine\\compile_utils.py\u001b[0m in \u001b[0;36mupdate_state\u001b[1;34m(self, y_true, y_pred, sample_weight)\u001b[0m\n\u001b[0;32m    407\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mmetric_obj\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    408\u001b[0m           \u001b[1;32mcontinue\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 409\u001b[1;33m         \u001b[0mmetric_obj\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mupdate_state\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_t\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my_p\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0msample_weight\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mmask\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    410\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    411\u001b[0m       \u001b[1;32mfor\u001b[0m \u001b[0mweighted_metric_obj\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mweighted_metric_objs\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\keras\\utils\\metrics_utils.py\u001b[0m in \u001b[0;36mdecorated\u001b[1;34m(metric_obj, *args, **kwargs)\u001b[0m\n\u001b[0;32m     88\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     89\u001b[0m     \u001b[1;32mwith\u001b[0m \u001b[0mtf_utils\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mgraph_context_for_symbolic_tensors\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 90\u001b[1;33m       \u001b[0mupdate_op\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mupdate_state_fn\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     91\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[0mupdate_op\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m:\u001b[0m  \u001b[1;31m# update_op will be None in eager execution.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     92\u001b[0m       \u001b[0mmetric_obj\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0madd_update\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mupdate_op\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\keras\\metrics.py\u001b[0m in \u001b[0;36mupdate_state_fn\u001b[1;34m(*args, **kwargs)\u001b[0m\n\u001b[0;32m    174\u001b[0m         \u001b[0mcontrol_status\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mag_ctx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcontrol_status_ctx\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    175\u001b[0m         \u001b[0mag_update_state\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mautograph\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtf_convert\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mobj_update_state\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcontrol_status\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 176\u001b[1;33m         \u001b[1;32mreturn\u001b[0m \u001b[0mag_update_state\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    177\u001b[0m     \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    178\u001b[0m       \u001b[1;32mif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mupdate_state\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdef_function\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mFunction\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\autograph\\impl\\api.py\u001b[0m in \u001b[0;36mwrapper\u001b[1;34m(*args, **kwargs)\u001b[0m\n\u001b[0;32m    253\u001b[0m       \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    254\u001b[0m         \u001b[1;32mwith\u001b[0m \u001b[0mconversion_ctx\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 255\u001b[1;33m           \u001b[1;32mreturn\u001b[0m \u001b[0mconverted_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mf\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkwargs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0moptions\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0moptions\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    256\u001b[0m       \u001b[1;32mexcept\u001b[0m \u001b[0mException\u001b[0m \u001b[1;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[1;33m:\u001b[0m  \u001b[1;31m# pylint:disable=broad-except\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    257\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mhasattr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0me\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m'ag_error_metadata'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\autograph\\impl\\api.py\u001b[0m in \u001b[0;36mconverted_call\u001b[1;34m(f, args, kwargs, caller_fn_scope, options)\u001b[0m\n\u001b[0;32m    455\u001b[0m   \u001b[1;32mif\u001b[0m \u001b[0mconversion\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mis_in_whitelist_cache\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mf\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0moptions\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    456\u001b[0m     \u001b[0mlogging\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mlog\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m'Whitelisted %s: from cache'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mf\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 457\u001b[1;33m     \u001b[1;32mreturn\u001b[0m \u001b[0m_call_unconverted\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mf\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkwargs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0moptions\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mFalse\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    458\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    459\u001b[0m   \u001b[1;32mif\u001b[0m \u001b[0mag_ctx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcontrol_status_ctx\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstatus\u001b[0m \u001b[1;33m==\u001b[0m \u001b[0mag_ctx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mStatus\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mDISABLED\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\autograph\\impl\\api.py\u001b[0m in \u001b[0;36m_call_unconverted\u001b[1;34m(f, args, kwargs, options, update_cache)\u001b[0m\n\u001b[0;32m    337\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    338\u001b[0m   \u001b[1;32mif\u001b[0m \u001b[0mkwargs\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 339\u001b[1;33m     \u001b[1;32mreturn\u001b[0m \u001b[0mf\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    340\u001b[0m   \u001b[1;32mreturn\u001b[0m \u001b[0mf\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    341\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\keras\\metrics.py\u001b[0m in \u001b[0;36mupdate_state\u001b[1;34m(self, y_true, y_pred, sample_weight)\u001b[0m\n\u001b[0;32m    610\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    611\u001b[0m     \u001b[0mag_fn\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mautograph\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtf_convert\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_fn\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mag_ctx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcontrol_status_ctx\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 612\u001b[1;33m     \u001b[0mmatches\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mag_fn\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_true\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my_pred\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_fn_kwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    613\u001b[0m     return super(MeanMetricWrapper, self).update_state(\n\u001b[0;32m    614\u001b[0m         matches, sample_weight=sample_weight)\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\autograph\\impl\\api.py\u001b[0m in \u001b[0;36mwrapper\u001b[1;34m(*args, **kwargs)\u001b[0m\n\u001b[0;32m    253\u001b[0m       \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    254\u001b[0m         \u001b[1;32mwith\u001b[0m \u001b[0mconversion_ctx\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 255\u001b[1;33m           \u001b[1;32mreturn\u001b[0m \u001b[0mconverted_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mf\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkwargs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0moptions\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0moptions\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    256\u001b[0m       \u001b[1;32mexcept\u001b[0m \u001b[0mException\u001b[0m \u001b[1;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[1;33m:\u001b[0m  \u001b[1;31m# pylint:disable=broad-except\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    257\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mhasattr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0me\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m'ag_error_metadata'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\autograph\\impl\\api.py\u001b[0m in \u001b[0;36mconverted_call\u001b[1;34m(f, args, kwargs, caller_fn_scope, options)\u001b[0m\n\u001b[0;32m    594\u001b[0m     \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    595\u001b[0m       \u001b[1;32mif\u001b[0m \u001b[0mkwargs\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 596\u001b[1;33m         \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mconverted_f\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0meffective_args\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    597\u001b[0m       \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    598\u001b[0m         \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mconverted_f\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0meffective_args\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Temp\\tmpe6bdwilf.py\u001b[0m in \u001b[0;36mtf__acc_grain\u001b[1;34m(y_true, y_pred)\u001b[0m\n\u001b[0;32m     10\u001b[0m                 \u001b[0my_true_class\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconverted_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mld\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mK\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0margmax\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mld\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_true\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdict\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0maxis\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m-\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfscope\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     11\u001b[0m                 \u001b[0my_pred_class\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconverted_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mld\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mK\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0margmax\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mld\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_pred\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdict\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0maxis\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m-\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfscope\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 12\u001b[1;33m                 \u001b[0msample_weight\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconverted_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mld\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mnp\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mhstack\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconverted_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mld\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mnp\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mwhere\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconverted_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconverted_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mld\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_true_class\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mnumpy\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfscope\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mreshape\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m-\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfscope\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;36m7\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mTrue\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mFalse\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfscope\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconverted_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mld\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mnp\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mwhere\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconverted_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconverted_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mld\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_true_class\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mnumpy\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfscope\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mreshape\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m-\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfscope\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;36m24\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mTrue\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mFalse\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfscope\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconverted_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mld\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mnp\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mwhere\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconverted_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconverted_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mld\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_true_class\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mnumpy\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfscope\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mreshape\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m-\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfscope\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;36m13\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mTrue\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mFalse\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfscope\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfscope\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     13\u001b[0m                 \u001b[0msample_weight\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconverted_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mld\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mnp\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0many\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mld\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0msample_weight\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdict\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0maxis\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfscope\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     14\u001b[0m                 \u001b[0msample_weight\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconverted_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mld\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mnp\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mwhere\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[0mag__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mld\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0msample_weight\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m8.94\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m1.0\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfscope\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\autograph\\impl\\api.py\u001b[0m in \u001b[0;36mconverted_call\u001b[1;34m(f, args, kwargs, caller_fn_scope, options)\u001b[0m\n\u001b[0;32m    455\u001b[0m   \u001b[1;32mif\u001b[0m \u001b[0mconversion\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mis_in_whitelist_cache\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mf\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0moptions\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    456\u001b[0m     \u001b[0mlogging\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mlog\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m'Whitelisted %s: from cache'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mf\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 457\u001b[1;33m     \u001b[1;32mreturn\u001b[0m \u001b[0m_call_unconverted\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mf\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkwargs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0moptions\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mFalse\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    458\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    459\u001b[0m   \u001b[1;32mif\u001b[0m \u001b[0mag_ctx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcontrol_status_ctx\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstatus\u001b[0m \u001b[1;33m==\u001b[0m \u001b[0mag_ctx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mStatus\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mDISABLED\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\autograph\\impl\\api.py\u001b[0m in \u001b[0;36m_call_unconverted\u001b[1;34m(f, args, kwargs, options, update_cache)\u001b[0m\n\u001b[0;32m    338\u001b[0m   \u001b[1;32mif\u001b[0m \u001b[0mkwargs\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    339\u001b[0m     \u001b[1;32mreturn\u001b[0m \u001b[0mf\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 340\u001b[1;33m   \u001b[1;32mreturn\u001b[0m \u001b[0mf\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    341\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    342\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\framework\\ops.py\u001b[0m in \u001b[0;36mnumpy\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m   1061\u001b[0m     \"\"\"\n\u001b[0;32m   1062\u001b[0m     \u001b[1;31m# TODO(slebedev): Consider avoiding a copy for non-CPU or remote tensors.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1063\u001b[1;33m     \u001b[0mmaybe_arr\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_numpy\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m  \u001b[1;31m# pylint: disable=protected-access\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1064\u001b[0m     \u001b[1;32mreturn\u001b[0m \u001b[0mmaybe_arr\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcopy\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;32mif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mmaybe_arr\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mndarray\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;32melse\u001b[0m \u001b[0mmaybe_arr\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1065\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Roaming\\Python\\Python36\\site-packages\\tensorflow\\python\\framework\\ops.py\u001b[0m in \u001b[0;36m_numpy\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m   1027\u001b[0m   \u001b[1;32mdef\u001b[0m \u001b[0m_numpy\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1028\u001b[0m     \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1029\u001b[1;33m       \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_numpy_internal\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1030\u001b[0m     \u001b[1;32mexcept\u001b[0m \u001b[0mcore\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_NotOkStatusException\u001b[0m \u001b[1;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[1;33m:\u001b[0m  \u001b[1;31m# pylint: disable=protected-access\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1031\u001b[0m       \u001b[0msix\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mraise_from\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mcore\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_status_to_exception\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0me\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcode\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0me\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmessage\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m)\u001b[0m  \u001b[1;31m# pylint: disable=protected-access\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "if __name__ == \"__main__\":\n",
    "    \"\"\"Training Main\n",
    "\n",
    "    The main function is called by both Unearthed's SageMaker pipeline and the\n",
    "    Unearthed CLI's \"unearthed train\" command.\n",
    "    \n",
    "    WARNING - modifying this function may cause the submission process to fail.\n",
    "\n",
    "    The main function must call preprocess, arrange th\n",
    "    \"\"\"\n",
    "    \n",
    "    args = {'model_dir': \"D:/Ramdhan/Unearth/visual-grain-analysis/models/\",\n",
    "           'data_dir' : \"D:/Ramdhan/Unearth/visual-grain-analysis/data/public/\"}\n",
    "    start = time.time()\n",
    "    history = train(args)\n",
    "    print('duration run',(time.time()-start)/60/60,' hours')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "out 0.9617"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2160x1440 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "# Plot training & validation iou_score values\n",
    "plt.figure(figsize=(30, 20))\n",
    "plt.subplot(121)\n",
    "plt.plot(history.history['acc_grain'])\n",
    "plt.plot(history.history['val_acc_grain'])\n",
    "plt.title('Model acc_score')\n",
    "plt.ylabel('acc_score')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Train', 'Test'], loc='upper left')\n",
    "\n",
    "# Plot training & validation loss values\n",
    "plt.subplot(122)\n",
    "plt.plot(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])\n",
    "plt.title('Model loss')\n",
    "plt.ylabel('Loss')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Train', 'Test'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 350
    },
    "executionInfo": {
     "elapsed": 2298,
     "status": "error",
     "timestamp": 1624373965288,
     "user": {
      "displayName": "ramdhan wibawa",
      "photoUrl": "",
      "userId": "14519448541243845830"
     },
     "user_tz": -420
    },
    "id": "UejZgfeXZ5C7",
    "outputId": "fbfcc699-a9ac-4b16-ccbb-0a8c4a25222e"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:__main__:loading the model\n",
      "INFO:__main__:Reading the test set\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.keras.preprocessing import image\n",
    "import pandas as pd\n",
    "import zipfile\n",
    "import csv\n",
    "\n",
    "start = time.time()\n",
    "\n",
    "def load_class_list(class_list_file):\n",
    "    class_list = []\n",
    "    with open(class_list_file, 'r') as csvfile:\n",
    "        file_reader = csv.reader(csvfile)\n",
    "        for row in file_reader:\n",
    "            class_list.append(row[0])\n",
    "    class_list.sort()\n",
    "    return class_list\n",
    "\n",
    "class_list = load_class_list('class_list.txt')\n",
    "\n",
    "IMG_SIZE=224\n",
    "LBL = dict(zip(class_list, range(28)))\n",
    "\n",
    "cls_map = dict(zip(LBL.values(),LBL.keys()))\n",
    "\n",
    "def init_dir(pth):\n",
    "    if os.path.exists(pth):\n",
    "        shutil.rmtree(pth)\n",
    "    os.mkdir(pth)\n",
    "\n",
    "logger.info(\"loading the model\")\n",
    "model = model_fn(args['model_dir'])\n",
    "logger.info(\"Reading the test set\")\n",
    "\n",
    "res = []\n",
    "for fn in glob.iglob(args['data_dir'] + 'val/**/*.png', recursive=True):\n",
    "    file_name = os.path.basename(fn)\n",
    "    path = os.path.abspath(fn)\n",
    "    folder = os.path.split(os.path.dirname(path))[1]\n",
    "    if len(file_name.split(\"-\")) > 2:  # ignore master image with may grains, raw image names are in guid format\n",
    "        im = image.load_img(path, target_size=(IMG_SIZE, IMG_SIZE))\n",
    "        img_array = image.img_to_array(im)\n",
    "        img_batch = np.expand_dims(img_array, axis=0)\n",
    "        img_preprocessed = preprocess_input(img_batch)\n",
    "        img_preprocessed = img_batch\n",
    "        pred = model.predict(img_preprocessed)\n",
    "        top3 = (-pred[0]).argsort()[:3]\n",
    "        res.append({'file_name': file_name, 'path': path, 'cls': folder, 'prediction':top3[0],  'proba_1':pred[0][top3[0]], 'prediction2':top3[1], 'proba_2':pred[0][top3[1]],  'prediction3':top3[2], 'proba_3':pred[0][top3[2]]})\n",
    "\n",
    "df = pd.DataFrame(res)\n",
    "df['prediction'] = df.prediction.map(cls_map)\n",
    "df['prediction2'] = df.prediction2.map(cls_map)\n",
    "df['prediction3'] = df.prediction3.map(cls_map)\n",
    "\n",
    "from sklearn.metrics import accuracy_score\n",
    "\n",
    "def w(x):\n",
    "    if \"_SOUND\" in x:\n",
    "        return 8.94 # weighting  Healthy grains more important than defective grains\n",
    "    return 1\n",
    "\n",
    "def scoring_fn(df):\n",
    "    \"\"\"\n",
    "    Weighted Accuracy Metric 90% Healthy grains, 10% unhealthy grains\n",
    "    \"\"\"\n",
    "    df['weight'] = df.cls.apply(w)\n",
    "    return accuracy_score(df.cls,df.prediction,sample_weight=df.weight)\n",
    "\n",
    "logger.info(f\"predictions have shape of {df.shape}\")\n",
    "\n",
    "score = scoring_fn(df)\n",
    "print (\"Accuracy Score:\" + str(score))\n",
    "# write to the output location\n",
    "\n",
    "print('Duration Run',(time.time()-start)/3600,' hours')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "               precision    recall  f1-score   support\n",
      "\n",
      "     B_BSMUT1       0.98      0.90      0.94       105\n",
      "     B_CLEV5B       0.95      0.41      0.57       100\n",
      "      B_DISTO       0.93      0.43      0.59       100\n",
      "     B_GRMEND       0.98      0.51      0.67        88\n",
      "     B_HDBARL       0.67      0.33      0.44        30\n",
      "     B_PICKLD       0.96      0.47      0.63        92\n",
      "     B_SKINED       1.00      0.20      0.34        88\n",
      "      B_SOUND       0.81      1.00      0.89      1653\n",
      "     B_SPRTED       1.00      0.87      0.93        93\n",
      "     B_SPTMLD       0.90      0.23      0.37       112\n",
      "      O_GROAT       1.00      0.78      0.88        90\n",
      "     O_HDOATS       0.76      0.26      0.39        99\n",
      "     O_SEPAFF       0.94      0.72      0.81       109\n",
      "      O_SOUND       0.92      1.00      0.96      1988\n",
      "     O_SPOTMA       0.83      0.29      0.43       103\n",
      "   WD_RADPODS       0.99      0.99      0.99        94\n",
      "  WD_RYEGRASS       0.96      0.94      0.95        87\n",
      "WD_SPEARGRASS       0.99      0.95      0.97        94\n",
      "  WD_WILDOATS       1.00      0.99      0.99        92\n",
      "      W_DISTO       0.87      0.66      0.75       102\n",
      "     W_FLDFUN       0.74      0.16      0.27       105\n",
      "     W_INSDA2       0.89      0.60      0.72        97\n",
      "     W_PICKLE       0.96      0.88      0.92       104\n",
      "     W_SEVERE       0.97      0.93      0.95       101\n",
      "      W_SOUND       0.83      1.00      0.90      1611\n",
      "     W_SPROUT       0.97      0.33      0.50       108\n",
      "     W_STAIND       0.75      0.43      0.55        92\n",
      "     W_WHITEG       0.93      0.47      0.62       111\n",
      "\n",
      "     accuracy                           0.87      7648\n",
      "    macro avg       0.91      0.63      0.71      7648\n",
      " weighted avg       0.88      0.87      0.85      7648\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "print(classification_report(df['cls'], df['prediction']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x1080 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "from sklearn.metrics import confusion_matrix\n",
    "ticks=np.linspace(0, 27,num=28)\n",
    "plt.figure(figsize=(15,15))\n",
    "plt.imshow(confusion_matrix(df['cls'], df['prediction'],normalize='true'), interpolation='none')\n",
    "plt.colorbar()\n",
    "plt.xticks(ticks,fontsize=6)\n",
    "plt.yticks(ticks,fontsize=6)\n",
    "plt.grid(True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "WZXP_9Kzfh6Q"
   },
   "outputs": [],
   "source": [
    "from keras_preprocessing import image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "s0bgHcc9rskI"
   },
   "outputs": [],
   "source": [
    "NUM_CLASSES=28\n",
    "model = build_model(num_classes=NUM_CLASSES)\n",
    "model.fit?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "3Fcx6Qvnfh6Q",
    "outputId": "fc7bceb9-5ffb-46bc-ffaf-7e1cf07f05c0"
   },
   "outputs": [],
   "source": [
    "image.random_rotation(90)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "saWJd9X3fh6R"
   },
   "outputs": [],
   "source": [
    " rotation_range=args.rotation,\n",
    "      shear_range=args.shear,\n",
    "      zoom_range=args.zoom,\n",
    "      horizontal_flip=args.h_flip,\n",
    "      vertical_flip=args.v_flip"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [],
   "name": "grain_unearth.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}