training_two.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "golden-artwork",
   "metadata": {},
   "outputs": [],
   "source": [
    "from functools import partial\n",
    "\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.callbacks import EarlyStopping, TensorBoard\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "\n",
    "#from nets import freeze_layers, get_model_from_name\n",
    "from utils.callbacks import (ExponentDecayScheduler, LossHistory,\n",
    "                             ModelCheckpoint)\n",
    "from utils.dataloader import ClsDatasets\n",
    "from utils.utils import get_classes\n",
    "from utils.utils_fit import fit_one_epoch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "constant-bathroom",
   "metadata": {},
   "outputs": [],
   "source": [
    "gpus = tf.config.experimental.list_physical_devices(device_type='GPU')\n",
    "for gpu in gpus:\n",
    "    tf.config.experimental.set_memory_growth(gpu, True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "superb-laugh",
   "metadata": {},
   "source": [
    "# 获取数据集，测试集不变"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "declared-missile",
   "metadata": {},
   "outputs": [],
   "source": [
    "class_names = [\"no_sick\", \"sick\"]\n",
    "num_classes = 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "verified-sacramento",
   "metadata": {},
   "outputs": [],
   "source": [
    "annotation_path = \"path_txt//test_stable//train_val_two.txt\"\n",
    "val_split       = 0.2\n",
    "\n",
    "with open(annotation_path, \"r\") as f:\n",
    "    lines = f.readlines()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "entertaining-valuation",
   "metadata": {},
   "outputs": [],
   "source": [
    "lines_0 = lines[:1050]\n",
    "lines_1 = lines[1050:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "fourth-findings",
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(10101)\n",
    "np.random.shuffle(lines_0)\n",
    "np.random.shuffle(lines_1)\n",
    "np.random.seed(None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "radio-gazette",
   "metadata": {},
   "outputs": [],
   "source": [
    "lines = [lines_0, lines_1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "featured-static",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_path = []\n",
    "val_path = []\n",
    "\n",
    "for i in range(2):\n",
    "    line = lines[i]\n",
    "    N = len(line)\n",
    "    num_val     = int(N * val_split)\n",
    "    num_train   = N - num_val\n",
    "    for j in range(N):\n",
    "        if j+1<=num_val:\n",
    "            val_path.append(line[j])\n",
    "        else:\n",
    "            train_path.append(line[j])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "material-tonight",
   "metadata": {},
   "outputs": [],
   "source": [
    "num_train = len(train_path)\n",
    "num_val = len(val_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "opponent-territory",
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(2022)\n",
    "np.random.shuffle(train_path)\n",
    "np.random.shuffle(val_path)\n",
    "np.random.seed(None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "apart-segment",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2671"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "num_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "surprising-wildlife",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "667"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "num_val"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "neutral-teacher",
   "metadata": {},
   "outputs": [],
   "source": [
    "annotation_path = \"path_txt//test_stable//test_two.txt\"\n",
    "\n",
    "with open(annotation_path, \"r\") as f:\n",
    "    test_path = f.readlines()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "amber-terrace",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "832"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(test_path)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "third-metro",
   "metadata": {},
   "source": [
    "# 获取数据的路径"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "arbitrary-equation",
   "metadata": {},
   "outputs": [],
   "source": [
    "class_names = [\"no_sick\", \"sick\"]\n",
    "num_classes = 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "angry-gravity",
   "metadata": {},
   "outputs": [],
   "source": [
    "annotation_path = \"path_txt/cls_two_class_test.txt\"\n",
    "\n",
    "with open(annotation_path, \"r\") as f:\n",
    "    lines_test = f.readlines()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "dominant-missouri",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "952"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(lines_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "standing-harmony",
   "metadata": {},
   "outputs": [],
   "source": [
    "lines_test_0 = lines_test[:436]\n",
    "lines_test_1 = lines_test[436:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "looking-montana",
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(10101)\n",
    "np.random.shuffle(lines_test_0)\n",
    "np.random.shuffle(lines_test_1)\n",
    "np.random.seed(None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "relative-comedy",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "occupational-gates",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "controlling-heath",
   "metadata": {},
   "outputs": [],
   "source": [
    "annotation_path = \"path_txt/cls_two_class_train.txt\"\n",
    "val_split       = 0.2\n",
    "\n",
    "with open(annotation_path, \"r\") as f:\n",
    "    lines = f.readlines()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "blocked-october",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4094"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(lines)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "removable-indicator",
   "metadata": {},
   "outputs": [],
   "source": [
    "lines_0 = lines[:1752]\n",
    "lines_1 = lines[1752:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "anonymous-punch",
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(10101)\n",
    "np.random.shuffle(lines_0)\n",
    "np.random.shuffle(lines_1)\n",
    "np.random.seed(None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "supposed-globe",
   "metadata": {},
   "outputs": [],
   "source": [
    "lines = [lines_0+lines_test_0, lines_1+lines_test_1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "final-ordering",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_path = []\n",
    "val_path = []\n",
    "test_path = []\n",
    "\n",
    "for i in range(2):\n",
    "    line = lines[i]\n",
    "    N = len(line)\n",
    "    num_val     = int(N * val_split)\n",
    "    num_train   = N - num_val\n",
    "    for j in range(N):\n",
    "        if j+1<=num_val:\n",
    "            val_path.append(line[j])\n",
    "        elif (num_val<j+1)and(j+1<=2*num_val):\n",
    "            test_path.append(line[j])\n",
    "        else:\n",
    "            train_path.append(line[j])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "center-knight",
   "metadata": {},
   "outputs": [],
   "source": [
    "num_train = len(train_path)\n",
    "num_val = len(val_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "recent-czech",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3030"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "num_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "wireless-camping",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1008"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "num_val"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "foster-medium",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1008"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(test_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "romantic-slovakia",
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(2022)\n",
    "np.random.shuffle(train_path)\n",
    "np.random.shuffle(val_path)\n",
    "np.random.seed(None)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "external-austin",
   "metadata": {},
   "source": [
    "# 模型"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "unexpected-tomato",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.layers import (Dense,Input,GlobalAveragePooling2D)\n",
    "from tensorflow.keras.models import Model\n",
    "from tensorflow.keras import applications\n",
    "\n",
    "def resnet101(input_shape=None):\n",
    "    net = applications.ResNet101(include_top=False,input_shape=input_shape,weights=None)\n",
    "\n",
    "    return net\n",
    "\n",
    "def ResNet101(input_shape=[224,224,3],classes=4):\n",
    "    img_input = Input(input_shape)\n",
    "\n",
    "    x = resnet101(input_shape)(img_input)\n",
    "    x = GlobalAveragePooling2D()(x)\n",
    "    x = Dense(256)(x)\n",
    "    x = Dense(256)(x)\n",
    "    x = Dense(classes,activation='softmax', name='fc1000')(x)\n",
    "    x = Model(img_input, x, name='ResNet101')\n",
    "    return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "critical-prairie",
   "metadata": {},
   "outputs": [],
   "source": [
    "model = ResNet101(input_shape=[224,224,3],classes=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "alpha-heart",
   "metadata": {},
   "outputs": [],
   "source": [
    "#model.load_weights('D:\\class_PR\\classification-tf2-main\\logs\\with_dict-ep015-loss0.060-val_loss0.509.h5')\n",
    "model.load_weights(\"weights/cls_2/cls_2-ep020-loss0.078-val_loss0.243.h5\")\n",
    "#model.load_weights('two_cls_w.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "threaded-seeker",
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(2):\n",
    "    model.layers[i].trainable = True"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "searching-drinking",
   "metadata": {},
   "source": [
    "# 回调函数"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "positive-controversy",
   "metadata": {},
   "outputs": [],
   "source": [
    "logging         = TensorBoard(log_dir = 'logs/')\n",
    "checkpoint      = ModelCheckpoint('logs/ep{epoch:03d}-loss{loss:.3f}-val_loss{val_loss:.3f}.h5',\n",
    "                                    monitor='val_loss', save_weights_only=True, save_best_only=False, period=5)\n",
    "reduce_lr       = ExponentDecayScheduler(decay_rate = 0.95, verbose = 1)\n",
    "early_stopping  = EarlyStopping(monitor='val_loss', min_delta=0, patience=10, verbose=1)\n",
    "loss_history    = LossHistory('logs/')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "dying-reasoning",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 2671 samples, val on 667 samples, with batch size 16.\n"
     ]
    }
   ],
   "source": [
    "batch_size      = 16\n",
    "Lr              = 1e-5\n",
    "Init_Epoch      = 0\n",
    "Freeze_Epoch    = 0\n",
    "Epoch           = 100\n",
    "input_shape     = [224, 224]\n",
    "num_workers     = 1\n",
    "\n",
    "epoch_step          = num_train // batch_size\n",
    "epoch_step_val      = num_val // batch_size\n",
    "\n",
    "if epoch_step == 0 or epoch_step_val == 0:\n",
    "    raise ValueError('数据集过小，无法进行训练，请扩充数据集。')\n",
    "            \n",
    "print('Train on {} samples, val on {} samples, with batch size {}.'.format(num_train, num_val, batch_size))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "presidential-guyana",
   "metadata": {},
   "source": [
    "# 训练集和验证集"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "optional-european",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dataloader    = ClsDatasets(train_path, input_shape, batch_size, num_classes, train = True)\n",
    "val_dataloader      = ClsDatasets(val_path, input_shape, batch_size, num_classes, train = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eight-protest",
   "metadata": {},
   "source": [
    "# 训练模型"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "local-vulnerability",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "  2/166 [..............................] - ETA: 9:24 - loss: 0.0027 - categorical_accuracy: 1.0000WARNING:tensorflow:Callbacks method `on_train_batch_end` is slow compared to the batch time (batch time: 0.6090s vs `on_train_batch_end` time: 6.1588s). Check your callbacks.\n",
      " 62/166 [==========>...................] - ETA: 1:34 - loss: 0.1259 - categorical_accuracy: 0.9587"
     ]
    },
    {
     "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-50-02dc97d35e21>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m     12\u001b[0m     \u001b[0muse_multiprocessing\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;32mTrue\u001b[0m \u001b[1;32mif\u001b[0m \u001b[0mnum_workers\u001b[0m \u001b[1;33m>\u001b[0m \u001b[1;36m1\u001b[0m \u001b[1;32melse\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[0;32m     13\u001b[0m     \u001b[0mworkers\u001b[0m             \u001b[1;33m=\u001b[0m \u001b[0mnum_workers\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 14\u001b[1;33m     \u001b[0mcallbacks\u001b[0m           \u001b[1;33m=\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0mlogging\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcheckpoint\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mreduce_lr\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mearly_stopping\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mloss_history\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     15\u001b[0m     \u001b[1;31m#class_weight= cls_wt_dict\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     16\u001b[0m )\n",
      "\u001b[1;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\util\\deprecation.py\u001b[0m in \u001b[0;36mnew_func\u001b[1;34m(*args, **kwargs)\u001b[0m\n\u001b[0;32m    322\u001b[0m               \u001b[1;34m'in a future version'\u001b[0m \u001b[1;32mif\u001b[0m \u001b[0mdate\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mNone\u001b[0m \u001b[1;32melse\u001b[0m \u001b[1;33m(\u001b[0m\u001b[1;34m'after %s'\u001b[0m \u001b[1;33m%\u001b[0m \u001b[0mdate\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    323\u001b[0m               instructions)\n\u001b[1;32m--> 324\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    325\u001b[0m     return tf_decorator.make_decorator(\n\u001b[0;32m    326\u001b[0m         \u001b[0mfunc\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnew_func\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m'deprecated'\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\keras\\engine\\training.py\u001b[0m in \u001b[0;36mfit_generator\u001b[1;34m(self, generator, steps_per_epoch, epochs, verbose, callbacks, validation_data, validation_steps, validation_freq, class_weight, max_queue_size, workers, use_multiprocessing, shuffle, initial_epoch)\u001b[0m\n\u001b[0;32m   1827\u001b[0m         \u001b[0muse_multiprocessing\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0muse_multiprocessing\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1828\u001b[0m         \u001b[0mshuffle\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mshuffle\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1829\u001b[1;33m         initial_epoch=initial_epoch)\n\u001b[0m\u001b[0;32m   1830\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1831\u001b[0m   @deprecation.deprecated(\n",
      "\u001b[1;32md:\\miniconda\\lib\\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;32md:\\miniconda\\lib\\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   1101\u001b[0m               \u001b[0mlogs\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtmp_logs\u001b[0m  \u001b[1;31m# No error, now safe to assign to logs.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1102\u001b[0m               \u001b[0mend_step\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mstep\u001b[0m \u001b[1;33m+\u001b[0m \u001b[0mdata_handler\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstep_increment\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1103\u001b[1;33m               \u001b[0mcallbacks\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mon_train_batch_end\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mend_step\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlogs\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   1104\u001b[0m         \u001b[0mepoch_logs\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcopy\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcopy\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mlogs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1105\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\keras\\callbacks.py\u001b[0m in \u001b[0;36mon_train_batch_end\u001b[1;34m(self, batch, logs)\u001b[0m\n\u001b[0;32m    438\u001b[0m     \"\"\"\n\u001b[0;32m    439\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_should_call_train_batch_hooks\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 440\u001b[1;33m       \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_call_batch_hook\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mModeKeys\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mTRAIN\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m'end'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mbatch\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlogs\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mlogs\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    441\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    442\u001b[0m   \u001b[1;32mdef\u001b[0m \u001b[0mon_test_batch_begin\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mbatch\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlogs\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mNone\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;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\keras\\callbacks.py\u001b[0m in \u001b[0;36m_call_batch_hook\u001b[1;34m(self, mode, hook, batch, logs)\u001b[0m\n\u001b[0;32m    287\u001b[0m       \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_call_batch_begin_hook\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mmode\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mbatch\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlogs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    288\u001b[0m     \u001b[1;32melif\u001b[0m \u001b[0mhook\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;34m'end'\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 289\u001b[1;33m       \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_call_batch_end_hook\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mmode\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mbatch\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlogs\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    290\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    291\u001b[0m       \u001b[1;32mraise\u001b[0m \u001b[0mValueError\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'Unrecognized hook: {}'\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mhook\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;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\keras\\callbacks.py\u001b[0m in \u001b[0;36m_call_batch_end_hook\u001b[1;34m(self, mode, batch, logs)\u001b[0m\n\u001b[0;32m    307\u001b[0m       \u001b[0mbatch_time\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[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_batch_start_time\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    308\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 309\u001b[1;33m     \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_call_batch_hook_helper\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mhook_name\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mbatch\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlogs\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    310\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    311\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_check_timing\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\keras\\callbacks.py\u001b[0m in \u001b[0;36m_call_batch_hook_helper\u001b[1;34m(self, hook_name, batch, logs)\u001b[0m\n\u001b[0;32m    343\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    344\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mnumpy_logs\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m:\u001b[0m  \u001b[1;31m# Only convert once.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 345\u001b[1;33m           \u001b[0mnumpy_logs\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtf_utils\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mto_numpy_or_python_type\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mlogs\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    346\u001b[0m         \u001b[0mhook\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mbatch\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnumpy_logs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    347\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\keras\\utils\\tf_utils.py\u001b[0m in \u001b[0;36mto_numpy_or_python_type\u001b[1;34m(tensors)\u001b[0m\n\u001b[0;32m    535\u001b[0m     \u001b[1;32mreturn\u001b[0m \u001b[0mt\u001b[0m  \u001b[1;31m# Don't turn ragged or sparse tensors to NumPy.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    536\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 537\u001b[1;33m   \u001b[1;32mreturn\u001b[0m \u001b[0mnest\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmap_structure\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0m_to_single_numpy_or_python_type\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtensors\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    538\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    539\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\util\\nest.py\u001b[0m in \u001b[0;36mmap_structure\u001b[1;34m(func, *structure, **kwargs)\u001b[0m\n\u001b[0;32m    633\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    634\u001b[0m   return pack_sequence_as(\n\u001b[1;32m--> 635\u001b[1;33m       \u001b[0mstructure\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0mfunc\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;32mfor\u001b[0m \u001b[0mx\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mentries\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    636\u001b[0m       expand_composites=expand_composites)\n\u001b[0;32m    637\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\util\\nest.py\u001b[0m in \u001b[0;36m<listcomp>\u001b[1;34m(.0)\u001b[0m\n\u001b[0;32m    633\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    634\u001b[0m   return pack_sequence_as(\n\u001b[1;32m--> 635\u001b[1;33m       \u001b[0mstructure\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0mfunc\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;32mfor\u001b[0m \u001b[0mx\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mentries\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    636\u001b[0m       expand_composites=expand_composites)\n\u001b[0;32m    637\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\keras\\utils\\tf_utils.py\u001b[0m in \u001b[0;36m_to_single_numpy_or_python_type\u001b[1;34m(t)\u001b[0m\n\u001b[0;32m    531\u001b[0m   \u001b[1;32mdef\u001b[0m \u001b[0m_to_single_numpy_or_python_type\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mt\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    532\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mt\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mops\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mTensor\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--> 533\u001b[1;33m       \u001b[0mx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mt\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[0m\n\u001b[0m\u001b[0;32m    534\u001b[0m       \u001b[1;32mreturn\u001b[0m \u001b[0mx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mitem\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;32mif\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mndim\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;36m0\u001b[0m \u001b[1;32melse\u001b[0m \u001b[0mx\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    535\u001b[0m     \u001b[1;32mreturn\u001b[0m \u001b[0mt\u001b[0m  \u001b[1;31m# Don't turn ragged or sparse tensors to NumPy.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32md:\\miniconda\\lib\\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;32md:\\miniconda\\lib\\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": [
    "model.compile(loss = 'categorical_crossentropy',\n",
    "                    optimizer = Adam(lr = Lr),\n",
    "                    metrics = ['categorical_accuracy'])\n",
    "            \n",
    "model.fit_generator(\n",
    "    generator           = train_dataloader,\n",
    "    steps_per_epoch     = epoch_step,\n",
    "    validation_data     = val_dataloader,\n",
    "    validation_steps    = epoch_step_val,\n",
    "    epochs              = Epoch,\n",
    "    initial_epoch       = Init_Epoch,\n",
    "    use_multiprocessing = True if num_workers > 1 else False,\n",
    "    workers             = num_workers,\n",
    "    callbacks           = [logging, checkpoint, reduce_lr, early_stopping, loss_history]\n",
    "    #class_weight= cls_wt_dict\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "failing-engineering",
   "metadata": {},
   "source": [
    "# 画混淆矩阵"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "fleet-creek",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.utils.multiclass import unique_labels\n",
    "from matplotlib import pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "# borrowed from scikit learn  这个是为了画混淆矩阵\n",
    "def plot_confusion_matrix(y_true, y_pred, classes,\n",
    "                          normalize=False,\n",
    "                          title=None,\n",
    "                          cmap=plt.cm.Blues):\n",
    "    \"\"\"\n",
    "    This function prints and plots the confusion matrix.\n",
    "    Normalization can be applied by setting `normalize=True`.\n",
    "    \"\"\"\n",
    "    if not title:\n",
    "        if normalize:\n",
    "            title = 'Normalized confusion matrix'\n",
    "        else:\n",
    "            title = 'Confusion matrix, without normalization'\n",
    "\n",
    "    # Compute confusion matrix\n",
    "    cm = confusion_matrix(y_true, y_pred)\n",
    "    # Only use the labels that appear in the data\n",
    "    classes = classes[unique_labels(y_true, y_pred)]\n",
    "    if normalize:\n",
    "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "        print(\"Normalized confusion matrix\")\n",
    "    else:\n",
    "        print('Confusion matrix, without normalization')\n",
    "\n",
    "    print(cm)\n",
    "\n",
    "    fig, ax = plt.subplots()\n",
    "    im = ax.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    ax.figure.colorbar(im, ax=ax)\n",
    "    # We want to show all ticks...\n",
    "    ax.set(xticks=np.arange(cm.shape[1]),\n",
    "           yticks=np.arange(cm.shape[0]),\n",
    "           # ... and label them with the respective list entries\n",
    "           xticklabels=classes, yticklabels=classes,\n",
    "           title=title,\n",
    "           ylabel='True label',\n",
    "           xlabel='Predicted label')\n",
    "\n",
    "    # Rotate the tick labels and set their alignment.\n",
    "    plt.setp(ax.get_xticklabels(), rotation=45, ha=\"right\",\n",
    "             rotation_mode=\"anchor\")\n",
    "\n",
    "    # Loop over data dimensions and create text annotations.\n",
    "    fmt = '.2f' if normalize else 'd'\n",
    "    thresh = cm.max() / 2.\n",
    "    for i in range(cm.shape[0]):\n",
    "        for j in range(cm.shape[1]):\n",
    "            ax.text(j, i, format(cm[i, j], fmt),\n",
    "                    ha=\"center\", va=\"center\",\n",
    "                    color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "    fig.tight_layout()\n",
    "    return ax"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "pediatric-casting",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "covered-capitol",
   "metadata": {},
   "outputs": [],
   "source": [
    "val_for_test = ClsDatasets(test_path, input_shape, 1, num_classes, train = False)\n",
    "\n",
    "x_val_for_test = []\n",
    "y_val_for_test =[]\n",
    "\n",
    "for x,y in val_for_test:\n",
    "    x_val_for_test.append(x)\n",
    "    y_val_for_test.append(y)\n",
    "\n",
    "x_val_for_test = np.asarray(x_val_for_test)\n",
    "y_val_for_test = np.asarray(y_val_for_test)\n",
    "\n",
    "x_val_for_test = np.squeeze(x_val_for_test, 1)\n",
    "y_val_for_test = np.squeeze(y_val_for_test, 1)\n",
    "\n",
    "y_val_for_test_pred = model.predict(x_val_for_test)\n",
    "\n",
    "Y_pred_val = np.argmax(y_val_for_test_pred,axis=1)\n",
    "Y_actual_val = np.argmax(y_val_for_test,axis=1)\n",
    "\n",
    "#plot_confusion_matrix(Y_actual_val, Y_pred_val, np.array(class_names))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "naval-alcohol",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[243  19]\n",
      " [ 15 555]]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:title={'center':'Confusion matrix, without normalization'}, xlabel='Predicted label', ylabel='True label'>"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion_matrix(Y_actual_val, Y_pred_val, np.array(class_names))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "separate-villa",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cloudy-surprise",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "accessory-tender",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "critical-vietnam",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cardiac-detail",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "international-pioneer",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "advisory-shannon",
   "metadata": {},
   "source": [
    "# 三分类"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "unknown-tourist",
   "metadata": {},
   "outputs": [],
   "source": [
    "class_names_three = [\"Retinopathy_grade_1\", \"Retinopathy_grade_2\",\"Retinopathy_grade_3\"]\n",
    "num_classes_three = 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "applicable-making",
   "metadata": {},
   "outputs": [],
   "source": [
    "annotation_path = \"path_txt//test_stable/train_val_three.txt\"\n",
    "\n",
    "with open(annotation_path, \"r\") as f:\n",
    "    lines_test_three = f.readlines()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "nutritional-cause",
   "metadata": {},
   "outputs": [],
   "source": [
    "lines_test_0_three = lines_test_three[:687]\n",
    "lines_test_1_three = lines_test_three[687:1475]\n",
    "lines_test_2_three = lines_test_three[1475:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "expected-syria",
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(10101)\n",
    "np.random.shuffle(lines_test_0_three)\n",
    "np.random.shuffle(lines_test_1_three)\n",
    "np.random.shuffle(lines_test_2_three)\n",
    "np.random.seed(None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "acute-brighton",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "demographic-defense",
   "metadata": {},
   "outputs": [],
   "source": [
    "val_split       = 0.2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "turkish-yugoslavia",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "round-thesis",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "another-trauma",
   "metadata": {},
   "outputs": [],
   "source": [
    "lines_three = [lines_test_0_three, lines_test_1_three, lines_test_2_three]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "diagnostic-correlation",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_path_three = []\n",
    "val_path_three = []\n",
    "#test_path_three = []\n",
    "\n",
    "for i in range(3):\n",
    "    line = lines_three[i]\n",
    "    N = len(line)\n",
    "    num_val     = int(N * val_split)\n",
    "    num_train   = N - num_val\n",
    "    for j in range(N):\n",
    "        if j+1<=num_val:\n",
    "            val_path_three.append(line[j])\n",
    "        else:\n",
    "            train_path_three.append(line[j])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "bright-mattress",
   "metadata": {},
   "outputs": [],
   "source": [
    "num_train = len(train_path_three)\n",
    "num_val = len(val_path_three)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "careful-creek",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1832"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "num_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "patient-female",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "456"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "num_val"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "fallen-scanning",
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(2022)\n",
    "np.random.shuffle(train_path_three)\n",
    "np.random.shuffle(val_path_three)\n",
    "np.random.seed(None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "million-cinema",
   "metadata": {},
   "outputs": [],
   "source": [
    "annotation_path = \"path_txt//test_stable//test_three.txt\"\n",
    "\n",
    "with open(annotation_path, \"r\") as f:\n",
    "    test_path_three = f.readlines()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "isolated-philip",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "570"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(test_path_three)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "pretty-burning",
   "metadata": {},
   "source": [
    "# 三分类模型"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "incomplete-african",
   "metadata": {},
   "outputs": [],
   "source": [
    "model_three = ResNet101(input_shape=[224,224,3],classes=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "under-scope",
   "metadata": {},
   "outputs": [],
   "source": [
    "model_three.load_weights(\"weights/cls_3/cls_3-ep020-loss0.062-val_loss0.201.h5\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "actual-steps",
   "metadata": {},
   "source": [
    "# 训练"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "entire-madison",
   "metadata": {},
   "outputs": [],
   "source": [
    "logging         = TensorBoard(log_dir = 'logs/')\n",
    "checkpoint      = ModelCheckpoint('logs/ep{epoch:03d}-loss{loss:.3f}-val_loss{val_loss:.3f}.h5',\n",
    "                                    monitor='val_loss', save_weights_only=True, save_best_only=False, period=5)\n",
    "reduce_lr       = ExponentDecayScheduler(decay_rate = 0.95, verbose = 1)\n",
    "early_stopping  = EarlyStopping(monitor='val_loss', min_delta=0, patience=10, verbose=1)\n",
    "loss_history    = LossHistory('logs/')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "arabic-majority",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 1832 samples, val on 456 samples, with batch size 16.\n"
     ]
    }
   ],
   "source": [
    "batch_size      = 16\n",
    "Lr              = 1e-4\n",
    "Init_Epoch      = 0\n",
    "Epoch           = 100\n",
    "input_shape     = [224, 224]\n",
    "num_workers     = 1\n",
    "\n",
    "epoch_step          = num_train // batch_size\n",
    "epoch_step_val      = num_val // batch_size\n",
    "\n",
    "if epoch_step == 0 or epoch_step_val == 0:\n",
    "    raise ValueError('数据集过小，无法进行训练，请扩充数据集。')\n",
    "            \n",
    "print('Train on {} samples, val on {} samples, with batch size {}.'.format(num_train, num_val, batch_size))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "reflected-yahoo",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dataloader_three    = ClsDatasets(train_path_three, input_shape, batch_size, num_classes_three, train = True)\n",
    "val_dataloader_three      = ClsDatasets(val_path_three, input_shape, batch_size, num_classes_three, train = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "bulgarian-collar",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From <ipython-input-23-7f13874d9692>:14: Model.fit_generator (from tensorflow.python.keras.engine.training) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Please use Model.fit, which supports generators.\n",
      "Epoch 1/100\n",
      "  1/114 [..............................] - ETA: 0s - loss: 0.9508 - categorical_accuracy: 0.4375WARNING:tensorflow:From d:\\miniconda\\lib\\site-packages\\tensorflow\\python\\ops\\summary_ops_v2.py:1277: stop (from tensorflow.python.eager.profiler) is deprecated and will be removed after 2020-07-01.\n",
      "Instructions for updating:\n",
      "use `tf.profiler.experimental.stop` instead.\n",
      "  2/114 [..............................] - ETA: 2:31 - loss: 0.8837 - categorical_accuracy: 0.5312WARNING:tensorflow:Callbacks method `on_train_batch_end` is slow compared to the batch time (batch time: 0.3708s vs `on_train_batch_end` time: 2.2924s). Check your callbacks.\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.9626 - categorical_accuracy: 0.5844Setting learning rate to 9.499999760009814e-05.\n",
      "114/114 [==============================] - 156s 1s/step - loss: 0.9626 - categorical_accuracy: 0.5844 - val_loss: 1.2247 - val_categorical_accuracy: 0.3415\n",
      "Epoch 2/100\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.6521 - categorical_accuracy: 0.6990Setting learning rate to 9.02499959920533e-05.\n",
      "114/114 [==============================] - 104s 910ms/step - loss: 0.6521 - categorical_accuracy: 0.6990 - val_loss: 1.1248 - val_categorical_accuracy: 0.3594\n",
      "Epoch 3/100\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.5801 - categorical_accuracy: 0.7505Setting learning rate to 8.573749619245064e-05.\n",
      "114/114 [==============================] - 79s 695ms/step - loss: 0.5801 - categorical_accuracy: 0.7505 - val_loss: 1.1350 - val_categorical_accuracy: 0.3281\n",
      "Epoch 4/100\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.4885 - categorical_accuracy: 0.7802Setting learning rate to 8.145062311086804e-05.\n",
      "114/114 [==============================] - 81s 708ms/step - loss: 0.4885 - categorical_accuracy: 0.7802 - val_loss: 1.0403 - val_categorical_accuracy: 0.4420\n",
      "Epoch 5/100\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.4677 - categorical_accuracy: 0.8076Setting learning rate to 7.737808919046074e-05.\n",
      "114/114 [==============================] - 82s 721ms/step - loss: 0.4677 - categorical_accuracy: 0.8076 - val_loss: 1.0094 - val_categorical_accuracy: 0.4888\n",
      "Epoch 6/100\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.3793 - categorical_accuracy: 0.8405Setting learning rate to 7.350918749580159e-05.\n",
      "114/114 [==============================] - 82s 716ms/step - loss: 0.3793 - categorical_accuracy: 0.8405 - val_loss: 0.7545 - val_categorical_accuracy: 0.6183\n",
      "Epoch 7/100\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.3163 - categorical_accuracy: 0.8739Setting learning rate to 6.983372950344346e-05.\n",
      "114/114 [==============================] - 82s 721ms/step - loss: 0.3163 - categorical_accuracy: 0.8739 - val_loss: 0.5551 - val_categorical_accuracy: 0.7545\n",
      "Epoch 8/100\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.2637 - categorical_accuracy: 0.8958Setting learning rate to 6.634204510191921e-05.\n",
      "114/114 [==============================] - 82s 723ms/step - loss: 0.2637 - categorical_accuracy: 0.8958 - val_loss: 0.4176 - val_categorical_accuracy: 0.8393\n",
      "Epoch 9/100\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.2565 - categorical_accuracy: 0.9035Setting learning rate to 6.302494111878331e-05.\n",
      "114/114 [==============================] - 83s 726ms/step - loss: 0.2565 - categorical_accuracy: 0.9035 - val_loss: 0.3005 - val_categorical_accuracy: 0.8862\n",
      "Epoch 10/100\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.2060 - categorical_accuracy: 0.9150Setting learning rate to 5.9873694408452134e-05.\n",
      "114/114 [==============================] - 83s 725ms/step - loss: 0.2060 - categorical_accuracy: 0.9150 - val_loss: 0.3551 - val_categorical_accuracy: 0.8996\n",
      "Epoch 11/100\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.1710 - categorical_accuracy: 0.9326Setting learning rate to 5.68800103792455e-05.\n",
      "114/114 [==============================] - 83s 726ms/step - loss: 0.1710 - categorical_accuracy: 0.9326 - val_loss: 0.3238 - val_categorical_accuracy: 0.8929\n",
      "Epoch 12/100\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.1487 - categorical_accuracy: 0.9413Setting learning rate to 5.4036009169067255e-05.\n",
      "114/114 [==============================] - 82s 722ms/step - loss: 0.1487 - categorical_accuracy: 0.9413 - val_loss: 0.3138 - val_categorical_accuracy: 0.9196\n",
      "Epoch 13/100\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.1567 - categorical_accuracy: 0.9370Setting learning rate to 5.133420836500591e-05.\n",
      "114/114 [==============================] - 82s 722ms/step - loss: 0.1567 - categorical_accuracy: 0.9370 - val_loss: 0.2576 - val_categorical_accuracy: 0.9263\n",
      "Epoch 14/100\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.1079 - categorical_accuracy: 0.9655Setting learning rate to 4.876749881077558e-05.\n",
      "114/114 [==============================] - 83s 725ms/step - loss: 0.1079 - categorical_accuracy: 0.9655 - val_loss: 0.2611 - val_categorical_accuracy: 0.9263\n",
      "Epoch 15/100\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.1196 - categorical_accuracy: 0.9589Setting learning rate to 4.63291238702368e-05.\n",
      "114/114 [==============================] - 83s 726ms/step - loss: 0.1196 - categorical_accuracy: 0.9589 - val_loss: 0.2889 - val_categorical_accuracy: 0.9286\n",
      "Epoch 16/100\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.1056 - categorical_accuracy: 0.9605Setting learning rate to 4.4012669059156905e-05.\n",
      "114/114 [==============================] - 83s 731ms/step - loss: 0.1056 - categorical_accuracy: 0.9605 - val_loss: 0.5351 - val_categorical_accuracy: 0.8438\n",
      "Epoch 17/100\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.1005 - categorical_accuracy: 0.9627Setting learning rate to 4.181203439657111e-05.\n",
      "114/114 [==============================] - 83s 729ms/step - loss: 0.1005 - categorical_accuracy: 0.9627 - val_loss: 0.3128 - val_categorical_accuracy: 0.9040\n",
      "Epoch 18/100\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.0879 - categorical_accuracy: 0.9688Setting learning rate to 3.9721430948702614e-05.\n",
      "114/114 [==============================] - 83s 724ms/step - loss: 0.0879 - categorical_accuracy: 0.9688 - val_loss: 0.2851 - val_categorical_accuracy: 0.9174\n",
      "Epoch 19/100\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.0706 - categorical_accuracy: 0.9759Setting learning rate to 3.773536009248346e-05.\n",
      "114/114 [==============================] - 82s 724ms/step - loss: 0.0706 - categorical_accuracy: 0.9759 - val_loss: 0.2185 - val_categorical_accuracy: 0.9375\n",
      "Epoch 20/100\n",
      "114/114 [==============================] - ETA: 0s - loss: 0.0621 - categorical_accuracy: 0.9786Setting learning rate to 3.584859277907526e-05.\n",
      "114/114 [==============================] - 84s 733ms/step - loss: 0.0621 - categorical_accuracy: 0.9786 - val_loss: 0.2013 - val_categorical_accuracy: 0.9442\n",
      "Epoch 21/100\n",
      " 11/114 [=>............................] - ETA: 1:03 - loss: 0.0227 - categorical_accuracy: 0.9886"
     ]
    },
    {
     "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-23-7f13874d9692>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m     12\u001b[0m     \u001b[0muse_multiprocessing\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;32mTrue\u001b[0m \u001b[1;32mif\u001b[0m \u001b[0mnum_workers\u001b[0m \u001b[1;33m>\u001b[0m \u001b[1;36m1\u001b[0m \u001b[1;32melse\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[0;32m     13\u001b[0m     \u001b[0mworkers\u001b[0m             \u001b[1;33m=\u001b[0m \u001b[0mnum_workers\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 14\u001b[1;33m     \u001b[0mcallbacks\u001b[0m           \u001b[1;33m=\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0mlogging\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcheckpoint\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mreduce_lr\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mearly_stopping\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mloss_history\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     15\u001b[0m     \u001b[1;31m#class_weight= cls_wt_dict\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     16\u001b[0m )\n",
      "\u001b[1;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\util\\deprecation.py\u001b[0m in \u001b[0;36mnew_func\u001b[1;34m(*args, **kwargs)\u001b[0m\n\u001b[0;32m    322\u001b[0m               \u001b[1;34m'in a future version'\u001b[0m \u001b[1;32mif\u001b[0m \u001b[0mdate\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mNone\u001b[0m \u001b[1;32melse\u001b[0m \u001b[1;33m(\u001b[0m\u001b[1;34m'after %s'\u001b[0m \u001b[1;33m%\u001b[0m \u001b[0mdate\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    323\u001b[0m               instructions)\n\u001b[1;32m--> 324\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    325\u001b[0m     return tf_decorator.make_decorator(\n\u001b[0;32m    326\u001b[0m         \u001b[0mfunc\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnew_func\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m'deprecated'\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\keras\\engine\\training.py\u001b[0m in \u001b[0;36mfit_generator\u001b[1;34m(self, generator, steps_per_epoch, epochs, verbose, callbacks, validation_data, validation_steps, validation_freq, class_weight, max_queue_size, workers, use_multiprocessing, shuffle, initial_epoch)\u001b[0m\n\u001b[0;32m   1827\u001b[0m         \u001b[0muse_multiprocessing\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0muse_multiprocessing\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1828\u001b[0m         \u001b[0mshuffle\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mshuffle\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1829\u001b[1;33m         initial_epoch=initial_epoch)\n\u001b[0m\u001b[0;32m   1830\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1831\u001b[0m   @deprecation.deprecated(\n",
      "\u001b[1;32md:\\miniconda\\lib\\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;32md:\\miniconda\\lib\\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   1101\u001b[0m               \u001b[0mlogs\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtmp_logs\u001b[0m  \u001b[1;31m# No error, now safe to assign to logs.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1102\u001b[0m               \u001b[0mend_step\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mstep\u001b[0m \u001b[1;33m+\u001b[0m \u001b[0mdata_handler\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstep_increment\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1103\u001b[1;33m               \u001b[0mcallbacks\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mon_train_batch_end\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mend_step\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlogs\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   1104\u001b[0m         \u001b[0mepoch_logs\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcopy\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcopy\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mlogs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1105\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\keras\\callbacks.py\u001b[0m in \u001b[0;36mon_train_batch_end\u001b[1;34m(self, batch, logs)\u001b[0m\n\u001b[0;32m    438\u001b[0m     \"\"\"\n\u001b[0;32m    439\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_should_call_train_batch_hooks\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 440\u001b[1;33m       \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_call_batch_hook\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mModeKeys\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mTRAIN\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m'end'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mbatch\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlogs\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mlogs\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    441\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    442\u001b[0m   \u001b[1;32mdef\u001b[0m \u001b[0mon_test_batch_begin\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mbatch\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlogs\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mNone\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;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\keras\\callbacks.py\u001b[0m in \u001b[0;36m_call_batch_hook\u001b[1;34m(self, mode, hook, batch, logs)\u001b[0m\n\u001b[0;32m    287\u001b[0m       \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_call_batch_begin_hook\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mmode\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mbatch\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlogs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    288\u001b[0m     \u001b[1;32melif\u001b[0m \u001b[0mhook\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;34m'end'\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 289\u001b[1;33m       \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_call_batch_end_hook\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mmode\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mbatch\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlogs\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    290\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    291\u001b[0m       \u001b[1;32mraise\u001b[0m \u001b[0mValueError\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'Unrecognized hook: {}'\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mhook\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;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\keras\\callbacks.py\u001b[0m in \u001b[0;36m_call_batch_end_hook\u001b[1;34m(self, mode, batch, logs)\u001b[0m\n\u001b[0;32m    307\u001b[0m       \u001b[0mbatch_time\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[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_batch_start_time\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    308\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 309\u001b[1;33m     \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_call_batch_hook_helper\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mhook_name\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mbatch\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlogs\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    310\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    311\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_check_timing\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\keras\\callbacks.py\u001b[0m in \u001b[0;36m_call_batch_hook_helper\u001b[1;34m(self, hook_name, batch, logs)\u001b[0m\n\u001b[0;32m    343\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    344\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mnumpy_logs\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m:\u001b[0m  \u001b[1;31m# Only convert once.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 345\u001b[1;33m           \u001b[0mnumpy_logs\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtf_utils\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mto_numpy_or_python_type\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mlogs\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    346\u001b[0m         \u001b[0mhook\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mbatch\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnumpy_logs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    347\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\keras\\utils\\tf_utils.py\u001b[0m in \u001b[0;36mto_numpy_or_python_type\u001b[1;34m(tensors)\u001b[0m\n\u001b[0;32m    535\u001b[0m     \u001b[1;32mreturn\u001b[0m \u001b[0mt\u001b[0m  \u001b[1;31m# Don't turn ragged or sparse tensors to NumPy.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    536\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 537\u001b[1;33m   \u001b[1;32mreturn\u001b[0m \u001b[0mnest\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmap_structure\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0m_to_single_numpy_or_python_type\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtensors\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    538\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    539\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\util\\nest.py\u001b[0m in \u001b[0;36mmap_structure\u001b[1;34m(func, *structure, **kwargs)\u001b[0m\n\u001b[0;32m    633\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    634\u001b[0m   return pack_sequence_as(\n\u001b[1;32m--> 635\u001b[1;33m       \u001b[0mstructure\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0mfunc\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;32mfor\u001b[0m \u001b[0mx\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mentries\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    636\u001b[0m       expand_composites=expand_composites)\n\u001b[0;32m    637\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\util\\nest.py\u001b[0m in \u001b[0;36m<listcomp>\u001b[1;34m(.0)\u001b[0m\n\u001b[0;32m    633\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    634\u001b[0m   return pack_sequence_as(\n\u001b[1;32m--> 635\u001b[1;33m       \u001b[0mstructure\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0mfunc\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;32mfor\u001b[0m \u001b[0mx\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mentries\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    636\u001b[0m       expand_composites=expand_composites)\n\u001b[0;32m    637\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32md:\\miniconda\\lib\\site-packages\\tensorflow\\python\\keras\\utils\\tf_utils.py\u001b[0m in \u001b[0;36m_to_single_numpy_or_python_type\u001b[1;34m(t)\u001b[0m\n\u001b[0;32m    531\u001b[0m   \u001b[1;32mdef\u001b[0m \u001b[0m_to_single_numpy_or_python_type\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mt\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    532\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mt\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mops\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mTensor\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--> 533\u001b[1;33m       \u001b[0mx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mt\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[0m\n\u001b[0m\u001b[0;32m    534\u001b[0m       \u001b[1;32mreturn\u001b[0m \u001b[0mx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mitem\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;32mif\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mndim\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;36m0\u001b[0m \u001b[1;32melse\u001b[0m \u001b[0mx\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    535\u001b[0m     \u001b[1;32mreturn\u001b[0m \u001b[0mt\u001b[0m  \u001b[1;31m# Don't turn ragged or sparse tensors to NumPy.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32md:\\miniconda\\lib\\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;32md:\\miniconda\\lib\\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": [
    "model_three.compile(loss = 'categorical_crossentropy',\n",
    "                    optimizer = Adam(lr = Lr),\n",
    "                    metrics = ['categorical_accuracy'])\n",
    "            \n",
    "model_three.fit_generator(\n",
    "    generator           = train_dataloader_three,\n",
    "    steps_per_epoch     = epoch_step,\n",
    "    validation_data     = val_dataloader_three,\n",
    "    validation_steps    = epoch_step_val,\n",
    "    epochs              = Epoch,\n",
    "    initial_epoch       = Init_Epoch,\n",
    "    use_multiprocessing = True if num_workers > 1 else False,\n",
    "    workers             = num_workers,\n",
    "    callbacks           = [logging, checkpoint, reduce_lr, early_stopping, loss_history]\n",
    "    #class_weight= cls_wt_dict\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "virtual-telescope",
   "metadata": {},
   "source": [
    "# 画三分类的混淆矩阵"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "complex-cycling",
   "metadata": {},
   "outputs": [],
   "source": [
    "val_for_test = ClsDatasets(test_path_three, input_shape, 1, num_classes_three, train = False)\n",
    "\n",
    "x_val_for_test = []\n",
    "y_val_for_test =[]\n",
    "\n",
    "for x,y in val_for_test:\n",
    "    x_val_for_test.append(x)\n",
    "    y_val_for_test.append(y)\n",
    "\n",
    "x_val_for_test = np.asarray(x_val_for_test)\n",
    "y_val_for_test = np.asarray(y_val_for_test)\n",
    "\n",
    "x_val_for_test = np.squeeze(x_val_for_test, 1)\n",
    "y_val_for_test = np.squeeze(y_val_for_test, 1)\n",
    "\n",
    "y_val_for_test_pred = model_three.predict(x_val_for_test)\n",
    "\n",
    "Y_pred_val = np.argmax(y_val_for_test_pred,axis=1)\n",
    "Y_actual_val = np.argmax(y_val_for_test,axis=1)\n",
    "\n",
    "#plot_confusion_matrix(Y_actual_val, Y_pred_val, np.array(class_names))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "indoor-wrong",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[158  13   0]\n",
      " [  8 183   5]\n",
      " [  0  13 190]]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:title={'center':'Confusion matrix, without normalization'}, xlabel='Predicted label', ylabel='True label'>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion_matrix(Y_actual_val, Y_pred_val, np.array(class_names_three))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "actual-faculty",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "uniform-waterproof",
   "metadata": {},
   "source": [
    "# 使用两个模型一起预测"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "toxic-canal",
   "metadata": {},
   "outputs": [],
   "source": [
    "annotation_path = \"path_txt//test_stable//test.txt\"\n",
    "\n",
    "with open(annotation_path, \"r\") as f:\n",
    "    test_path = f.readlines()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "suspended-pitch",
   "metadata": {},
   "outputs": [],
   "source": [
    "model = ResNet101(input_shape=[224,224,3],classes=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "unexpected-ghost",
   "metadata": {},
   "outputs": [],
   "source": [
    "model.load_weights(\"weights/cls_2/cls_2-ep020-loss0.078-val_loss0.243.h5\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "inclusive-criterion",
   "metadata": {},
   "outputs": [],
   "source": [
    "num_classes_final=4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "similar-johnston",
   "metadata": {},
   "outputs": [],
   "source": [
    "input_shape = [224, 224]\n",
    "val_for_test = ClsDatasets(test_path, input_shape, 1, num_classes_final, train = False)\n",
    "\n",
    "x_val_for_test = []\n",
    "y_val_for_test =[]\n",
    "\n",
    "for x,y in val_for_test:\n",
    "    x_val_for_test.append(x)\n",
    "    y_val_for_test.append(y)\n",
    "\n",
    "x_val_for_test = np.asarray(x_val_for_test)\n",
    "y_val_for_test = np.asarray(y_val_for_test)\n",
    "\n",
    "x_val_for_test = np.squeeze(x_val_for_test, 1)\n",
    "y_val_for_test = np.squeeze(y_val_for_test, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "advanced-handling",
   "metadata": {},
   "outputs": [],
   "source": [
    "two_cls = model.predict(x_val_for_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "indirect-grace",
   "metadata": {},
   "outputs": [],
   "source": [
    "two_cls = np.argmax(two_cls,axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "impressed-noise",
   "metadata": {},
   "outputs": [],
   "source": [
    "#two_cls"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "premier-tattoo",
   "metadata": {},
   "outputs": [],
   "source": [
    "three_cls = model_three.predict(x_val_for_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "marked-buying",
   "metadata": {},
   "outputs": [],
   "source": [
    "three_cls = np.argmax(three_cls,axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "straight-warning",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(832,)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "three_cls.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "chemical-parts",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_val_for_test_pred = []\n",
    "for i,x in enumerate(two_cls):\n",
    "    if x==0:\n",
    "        y_val_for_test_pred.append(0)\n",
    "    else:\n",
    "        y_val_for_test_pred.append(three_cls[i]+1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "fifth-mumbai",
   "metadata": {},
   "outputs": [],
   "source": [
    "Y_pred_val =  np.asarray(y_val_for_test_pred) \n",
    "        \n",
    "#Y_pred_val = np.argmax(y_val_for_test_pred,axis=1)\n",
    "Y_actual_val = np.argmax(y_val_for_test,axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "pretty-pilot",
   "metadata": {},
   "outputs": [],
   "source": [
    "class_names_final = [\"Retinopathy_grade_0\",\"Retinopathy_grade_1\", \"Retinopathy_grade_2\",\"Retinopathy_grade_3\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "federal-jewelry",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[251   5   6   0]\n",
      " [ 14 150   7   0]\n",
      " [  8  12 172   4]\n",
      " [  2   1  12 188]]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:title={'center':'Confusion matrix, without normalization'}, xlabel='Predicted label', ylabel='True label'>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion_matrix(Y_actual_val, Y_pred_val, np.array(class_names_final))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "round-advice",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}