restaurant_seoul_open.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "executionInfo": {
     "elapsed": 4,
     "status": "ok",
     "timestamp": 1643174394115,
     "user": {
      "displayName": "Jeyun Yang",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GjmyKFQBrUlKkDabIMU0ePGNHgAebn4xUo1Q2VA-Q=s64",
      "userId": "07012603247442409826"
     },
     "user_tz": -540
    },
    "id": "v2rqFer00yZ_"
   },
   "outputs": [],
   "source": [
    "import os\n",
    "os.environ[\"CUDA_VISIBLE_DEVICE\"] = \"0\"\n",
    "os.environ[\"CUDA_LAUNCH_BLOCKING\"] = \"1\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "FL8stXmg0yaD"
   },
   "source": [
    "# 사용 패키지"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "executionInfo": {
     "elapsed": 8564,
     "status": "ok",
     "timestamp": 1643174414761,
     "user": {
      "displayName": "Jeyun Yang",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GjmyKFQBrUlKkDabIMU0ePGNHgAebn4xUo1Q2VA-Q=s64",
      "userId": "07012603247442409826"
     },
     "user_tz": -540
    },
    "id": "WRsj5THo0yaE"
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import cv2\n",
    "from tqdm import tqdm\n",
    "from glob import glob\n",
    "import os\n",
    "import json \n",
    "import torch\n",
    "from torch import nn\n",
    "import torch.nn.functional as F\n",
    "from torchvision import models, transforms\n",
    "from torchsummary import summary\n",
    "from torch.utils.data import Dataset, DataLoader, TensorDataset\n",
    "from sklearn.metrics import f1_score\n",
    "from sklearn import metrics\n",
    "from sklearn.model_selection import train_test_split\n",
    "import random\n",
    "from datetime import datetime"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "현재:  2023-05-02 10:54:23.353698\n",
      "Device: NVIDIA GeForce RTX 3060\n",
      "CUDA Version: 11.6\n"
     ]
    }
   ],
   "source": [
    "now = datetime.now()\n",
    "\n",
    "print(\"현재: \", now)\n",
    "print('Device: ' + torch.cuda.get_device_name(0))\n",
    "print(\"CUDA Version: \" + torch.version.cuda)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "JpRVSFBs0yaF"
   },
   "source": [
    "# 데이터 불러오기"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 245
    },
    "executionInfo": {
     "elapsed": 284,
     "status": "error",
     "timestamp": 1643174636439,
     "user": {
      "displayName": "Jeyun Yang",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GjmyKFQBrUlKkDabIMU0ePGNHgAebn4xUo1Q2VA-Q=s64",
      "userId": "07012603247442409826"
     },
     "user_tz": -540
    },
    "id": "zKlzPttt0yaG",
    "outputId": "ce63f1a0-1011-4b54-cdbd-c94c986e4ed8"
   },
   "outputs": [],
   "source": [
    "bpath = r'E:\\221226_resv2\\data\\bins4'\n",
    "path1 = bpath + 'm'\n",
    "path2 = bpath + 'r'\n",
    "path3 = bpath + 'b' \n",
    "resv2_path = r'E:\\221226_resv2'\n",
    "\n",
    "train_len = range(819)\n",
    "test_len = range(205)\n",
    "\n",
    "tr_0 = [[bpath + r'\\train\\train' + '{:06d}_img'.format(i) for i in train_len],\n",
    "    [bpath + r'\\train\\train' + '{:06d}_met'.format(i) for i in train_len],\n",
    "    [bpath + r'\\train\\train' + '{:06d}_lab'.format(i) for i in train_len],\n",
    "    [bpath + r'\\train\\train' + '{:06d}_ind'.format(i) for i in train_len]\n",
    "]\n",
    "\n",
    "tr_1 = [[path1 + r'\\train\\train' + '{:06d}_img'.format(i) for i in train_len],\n",
    "    [path1 + r'\\train\\train' + '{:06d}_met'.format(i) for i in train_len],\n",
    "    [path1 + r'\\train\\train' + '{:06d}_lab'.format(i) for i in train_len],\n",
    "    [path1 + r'\\train\\train' + '{:06d}_ind'.format(i) for i in train_len]\n",
    "]\n",
    "\n",
    "tr_2 = [[path2 + r'\\train\\train' + '{:06d}_img'.format(i) for i in train_len],\n",
    "    [path2 + r'\\train\\train' + '{:06d}_met'.format(i) for i in train_len],\n",
    "    [path2 + r'\\train\\train' + '{:06d}_lab'.format(i) for i in train_len],\n",
    "    [path2 + r'\\train\\train' + '{:06d}_ind'.format(i) for i in train_len]\n",
    "]\n",
    "\n",
    "tr_3 = [[path3 + r'\\train\\train' + '{:06d}_img'.format(i) for i in train_len],\n",
    "    [path3 + r'\\train\\train' + '{:06d}_met'.format(i) for i in train_len],\n",
    "    [path3 + r'\\train\\train' + '{:06d}_lab'.format(i) for i in train_len],\n",
    "    [path3 + r'\\train\\train' + '{:06d}_ind'.format(i) for i in train_len]\n",
    "]\n",
    "trl = [tr_0, tr_1, tr_2, tr_3]\n",
    "\n",
    "te_0 = [[bpath + r'\\test\\test' + '{:06d}_img'.format(i) for i in test_len],\n",
    "    [bpath + r'\\test\\test' + '{:06d}_met'.format(i) for i in test_len],\n",
    "    [bpath + r'\\test\\test' + '{:06d}_lab'.format(i) for i in test_len],\n",
    "    [bpath + r'\\test\\test' + '{:06d}_ind'.format(i) for i in test_len]\n",
    "]\n",
    "\n",
    "te_1 = [[path1 + r'\\test\\test' + '{:06d}_img'.format(i) for i in test_len],\n",
    "    [path1 + r'\\test\\test' + '{:06d}_met'.format(i) for i in test_len],\n",
    "    [path1 + r'\\test\\test' + '{:06d}_lab'.format(i) for i in test_len],\n",
    "    [path1 + r'\\test\\test' + '{:06d}_ind'.format(i) for i in test_len]\n",
    "]\n",
    "\n",
    "te_2 = [[path2 + r'\\test\\test' + '{:06d}_img'.format(i) for i in test_len],\n",
    "    [path2 + r'\\test\\test' + '{:06d}_met'.format(i) for i in test_len],\n",
    "    [path2 + r'\\test\\test' + '{:06d}_lab'.format(i) for i in test_len],\n",
    "    [path2 + r'\\test\\test' + '{:06d}_ind'.format(i) for i in test_len]\n",
    "]\n",
    "\n",
    "te_3 = [[path3 + r'\\test\\test' + '{:06d}_img'.format(i) for i in test_len],\n",
    "    [path3 + r'\\test\\test' + '{:06d}_met'.format(i) for i in test_len],\n",
    "    [path3 + r'\\test\\test' + '{:06d}_lab'.format(i) for i in test_len],\n",
    "    [path3 + r'\\test\\test' + '{:06d}_ind'.format(i) for i in test_len]\n",
    "]\n",
    "tel = [te_0, te_1, te_2, te_3]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "BnLSM_WM0yaJ"
   },
   "source": [
    "# 데이터 로드"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "class LoadDataset(Dataset):\n",
    "    def __init__(self, img_dir, met_dir, lab_dir, ind_dir):\n",
    "        self.img_dir = img_dir\n",
    "        self.met_dir = met_dir\n",
    "        self.lab_dir = lab_dir\n",
    "        self.ind_dir = ind_dir\n",
    "\n",
    "    def __len__(self):\n",
    "        return len(self.lab_dir)\n",
    "    \n",
    "    def __getitem__(self, i):\n",
    "        img = torch.load(self.img_dir[i])\n",
    "        # met = torch.load(self.met_dir[i])\n",
    "        lab = torch.load(self.lab_dir[i])\n",
    "        ind = torch.load(self.ind_dir[i])\n",
    "\n",
    "        return {\n",
    "            'img' : img,\n",
    "            # 'metadata' : met,\n",
    "            'labels' : lab,\n",
    "            'indices' : ind\n",
    "        }\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "iG9kidq-0yaN"
   },
   "source": [
    "# 하이퍼파라미터 및 변수"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "id": "PByrB6Lk0yaN"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Batch Size : 256\n",
      "Metadata Dimension : 15\n",
      "Weight Mode : Imbalanced\n"
     ]
    }
   ],
   "source": [
    "device = torch.device(\"cuda:0\")\n",
    "batch_size = 256\n",
    "meta_n = 15\n",
    "learning_rate = 2e-4\n",
    "class_n = 2\n",
    "dropout_rate = 0.25\n",
    "epochs = 100\n",
    "best_save_path = 'best_model.pt'\n",
    "shallow = False\n",
    "imbalanced_weight = True\n",
    "\n",
    "torch.cuda.max_split_size_mb = 512\n",
    "\n",
    "print('Batch Size : {:d}'.format(batch_size))\n",
    "print('Metadata Dimension : {:d}'.format(meta_n))\n",
    "print('Weight Mode : ' + ('Imbalanced' if imbalanced_weight else 'Uniform'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "SANtRKsL0yaN"
   },
   "source": [
    "# 데이터셋 구성"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "id": "hvC5iCWD0yaO"
   },
   "outputs": [],
   "source": [
    "tr = [LoadDataset(tr_0[0], tr_0[1], tr_0[2], tr_0[3]),\n",
    "    LoadDataset(tr_1[0], tr_1[1], tr_3[2], tr_3[3]),\n",
    "    LoadDataset(tr_2[0], tr_2[1], tr_2[2], tr_2[3]),\n",
    "    LoadDataset(tr_3[0], tr_3[1], tr_3[2], tr_3[3])\n",
    "]\n",
    "te = [LoadDataset(te_0[0], te_0[1], te_0[2], te_0[3]),\n",
    "    LoadDataset(te_1[0], te_1[1], te_3[2], te_3[3]),\n",
    "    LoadDataset(te_2[0], te_2[1], te_2[2], te_2[3]),\n",
    "    LoadDataset(te_3[0], te_3[1], te_3[2], te_3[3])\n",
    "]\n",
    "\n",
    "wholequote = '''\n",
    "train_dataloader = [DataLoader(tr[0], num_workers=0, shuffle=True),\n",
    "    DataLoader(tr[1], num_workers=0, shuffle=True),\n",
    "    DataLoader(tr[2], num_workers=0, shuffle=True),\n",
    "    DataLoader(tr[3], num_workers=0, shuffle=True)\n",
    "]\n",
    "\n",
    "test_dataloader = [DataLoader(te[0], num_workers=0, shuffle=False),\n",
    "    DataLoader(te[1], num_workers=0, shuffle=False),\n",
    "    DataLoader(te[2], num_workers=0, shuffle=False),\n",
    "    DataLoader(te[3], num_workers=0, shuffle=False)\n",
    "]'''"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "xXVRTN2y0yaO"
   },
   "source": [
    "# 모델\n",
    "\n",
    "## 이미지 분류 모델 : MobileNet-64"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "id": "qKIBUSQg0yaO"
   },
   "outputs": [],
   "source": [
    "class MobileNet64(nn.Module):\n",
    "    def __init__(self, class_n, dropout_rate = 0.5, width_mult=1.0, res_mult=1.0, shallow=False):\n",
    "        super(MobileNet64, self).__init__()\n",
    "        self.width_mult = width_mult\n",
    "        self.res_mult = res_mult\n",
    "        self.class_n = class_n\n",
    "        self.rate = dropout_rate\n",
    "        # self.meta_hidden = [200, 200]\n",
    "        self.conv1 = self._conv(2, self._apply_mult(32), 3, 1)\n",
    "        self.dropout = nn.Dropout(dropout_rate)\n",
    "        if shallow is False:\n",
    "            self.channels = [32, 64, 128, 128, 256,\n",
    "                             256, 512, 512, 512, 512, 512, 512, 1024, 1024]\n",
    "            self.stride = [1, 2, 1, 2, 1, 2, 1, 1, 1, 1, 1, 2, 1]\n",
    "        else:\n",
    "            self.channels = [32, 64, 128, 128, 256,\n",
    "                             256, 512, 1024, 1024]\n",
    "            self.stride = [1, 2, 1, 2, 1, 2, 2, 1]\n",
    "        self.dw = nn.Sequential(\n",
    "            *[self._dw(self._apply_mult(channel),\n",
    "                       self._apply_mult(self.channels[i+1]),\n",
    "                       self.stride[i])\n",
    "              for i, channel in enumerate(self.channels[:-1])]\n",
    "        )\n",
    "        self.avgpool = nn.AdaptiveAvgPool2d(1)\n",
    "        # self.meta_hl = nn.Sequential(\n",
    "        #     nn.Linear(meta_n, self.meta_hidden[0]),\n",
    "        #     nn.ReLU(inplace=True),\n",
    "        #     nn.Linear(self.meta_hidden[0], self.meta_hidden[1]),\n",
    "        #     nn.ReLU(inplace=True),\n",
    "        #     nn.Dropout()\n",
    "        # )\n",
    "        self.fc = nn.Sequential(\n",
    "            nn.Linear(self.channels[-1], 1000),\n",
    "            nn.Linear(1000,self.class_n)\n",
    "        )\n",
    "        self.n_sm = nn.Softmax(dim=1)\n",
    "\n",
    "    def _apply_mult(self, channel):\n",
    "        return int((channel * self.width_mult) // 1)\n",
    "\n",
    "    def _dw(self, in_channels, out_channels, stride=1):\n",
    "        return nn.Sequential(\n",
    "            nn.Conv2d(in_channels, in_channels, 3,\n",
    "                      stride, 1, groups=in_channels),\n",
    "            nn.BatchNorm2d(in_channels),\n",
    "            nn.ReLU(),\n",
    "            nn.Conv2d(in_channels, out_channels, 1, 1, 0),\n",
    "            nn.BatchNorm2d(out_channels),\n",
    "            nn.ReLU()\n",
    "        )\n",
    "\n",
    "    def _conv(self, in_channels, out_channels, kernel_size, stride):\n",
    "        return nn.Sequential(\n",
    "            nn.Conv2d(in_channels, out_channels, kernel_size,\n",
    "                      stride=stride, padding=1),\n",
    "            nn.BatchNorm2d(out_channels),\n",
    "            nn.ReLU()\n",
    "        )\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = self.conv1(x)\n",
    "        x = self.dw(x)\n",
    "        x = self.avgpool(x)\n",
    "        x = torch.flatten(x, 1)\n",
    "        # x = torch.cat((x, self.meta_hl(metadata)), dim=1)\n",
    "        x = self.fc(x)\n",
    "        x = self.n_sm(x)\n",
    "\n",
    "        return x\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = MobileNet64(class_n, dropout_rate, shallow=shallow)\n",
    "model = model.to(device)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "V7OrqLfZ0yaQ"
   },
   "source": [
    "# 학습"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "id": "7FDAPIhZ0yaQ"
   },
   "outputs": [],
   "source": [
    "if imbalanced_weight:\n",
    "    class_weights = torch.FloatTensor([0.0638915, 0.9361085]).to(device) #[ 1 / number of instances for each class]\n",
    "else:\n",
    "    class_weights = torch.FloatTensor([0.1201091, 0.8798909]).to(device)\n",
    "criterion = nn.CrossEntropyLoss(weight = class_weights)\n",
    "optimizer = torch.optim.Adam(model.parameters(), lr=learning_rate)\n",
    "\n",
    "load = True\n",
    "if load:\n",
    "    loaded_model = torch.load(r'E:\\221226_resv2\\230310_cnn' + r'\\modelw_100.pt')\n",
    "    model.load_state_dict(loaded_model['model_state_dict'])\n",
    "    optimizer.load_state_dict(loaded_model['optimizer_state_dict'])\n",
    "    loss_plot = loaded_model['loss_history']\n",
    "    metric_plot = loaded_model['metric_history']\n",
    "    val_loss_plot = loaded_model['val_loss_history']\n",
    "    val_metric_plot = loaded_model['val_metric_history']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "id": "jP6jfOvQ0yaQ"
   },
   "outputs": [],
   "source": [
    "def accuracy_function(real, pred):\n",
    "    real = real.cpu()\n",
    "    pred = torch.argmax(pred, dim=1).cpu()\n",
    "    score = f1_score(real, pred, average='macro')\n",
    "    return score\n",
    "\n",
    "def train_step(batch_item, training):\n",
    "    img = batch_item['img'].to(device)\n",
    "    # metadata = batch_item['metadata'].to(device)\n",
    "    label = batch_item['labels'].to(device)\n",
    "    if training is True:\n",
    "        model.train()\n",
    "        optimizer.zero_grad()\n",
    "        output = model(img)\n",
    "        loss = criterion(output, label) \n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "        score = accuracy_function(label, output)\n",
    "        return loss, score\n",
    "    else:\n",
    "        model.eval()\n",
    "        with torch.no_grad():\n",
    "            output = model(img)\n",
    "            loss = criterion(output, label)\n",
    "        score = accuracy_function(label, output)\n",
    "        return loss, score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "id": "ShFB4SCq0yaQ",
    "outputId": "2e7e5861-43bb-4764-bc76-6c9b5505de5c"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "819it [03:35,  4.57it/s, Epoch=1, Loss=0.761162, Mean Loss=0.672470, Mean F-1=0.417270]\n",
      "205it [00:16, 12.73it/s, Epoch=1, Val Loss=0.652698, Mean Val Loss=0.686774, Mean Val F-1=0.480436]\n",
      "819it [03:29,  4.59it/s, Epoch=2, Loss=0.781355, Mean Loss=0.656985, Mean F-1=0.432357]\n",
      "205it [00:17, 11.89it/s, Epoch=2, Val Loss=0.678326, Mean Val Loss=0.721224, Mean Val F-1=0.532773]\n",
      "819it [03:32,  4.61it/s, Epoch=3, Loss=0.712958, Mean Loss=0.648922, Mean F-1=0.441713]\n",
      "205it [00:17, 11.92it/s, Epoch=3, Val Loss=0.664459, Mean Val Loss=0.658672, Mean Val F-1=0.485911]\n",
      "819it [03:22,  4.76it/s, Epoch=4, Loss=0.733249, Mean Loss=0.641681, Mean F-1=0.443201]\n",
      "205it [00:15, 13.00it/s, Epoch=4, Val Loss=0.639414, Mean Val Loss=0.645794, Mean Val F-1=0.428317]\n",
      "819it [03:19,  4.90it/s, Epoch=5, Loss=0.695785, Mean Loss=0.631571, Mean F-1=0.451948]\n",
      "205it [00:14, 13.98it/s, Epoch=5, Val Loss=0.707120, Mean Val Loss=0.643503, Mean Val F-1=0.372698]\n",
      "819it [03:18,  4.91it/s, Epoch=6, Loss=0.710868, Mean Loss=0.627961, Mean F-1=0.457245]\n",
      "205it [00:14, 14.06it/s, Epoch=6, Val Loss=0.734972, Mean Val Loss=0.661516, Mean Val F-1=0.318299]\n",
      "819it [03:19,  4.89it/s, Epoch=7, Loss=0.673562, Mean Loss=0.622516, Mean F-1=0.469154]\n",
      "205it [00:14, 13.98it/s, Epoch=7, Val Loss=0.715821, Mean Val Loss=0.649309, Mean Val F-1=0.341949]\n",
      "819it [03:18,  4.92it/s, Epoch=8, Loss=0.697099, Mean Loss=0.616433, Mean F-1=0.471159]\n",
      "205it [00:14, 14.05it/s, Epoch=8, Val Loss=0.640375, Mean Val Loss=0.695615, Mean Val F-1=0.554517]\n",
      "819it [03:19,  4.88it/s, Epoch=9, Loss=0.653999, Mean Loss=0.607074, Mean F-1=0.480277]\n",
      "205it [00:14, 13.98it/s, Epoch=9, Val Loss=0.607758, Mean Val Loss=0.709081, Mean Val F-1=0.553928]\n",
      "819it [03:19,  4.92it/s, Epoch=10, Loss=0.672378, Mean Loss=0.606040, Mean F-1=0.484548]\n",
      "205it [00:14, 14.04it/s, Epoch=10, Val Loss=0.672785, Mean Val Loss=0.615516, Mean Val F-1=0.480851]\n",
      "819it [03:19,  4.89it/s, Epoch=11, Loss=0.673918, Mean Loss=0.601965, Mean F-1=0.486980]\n",
      "205it [00:14, 13.99it/s, Epoch=11, Val Loss=0.576155, Mean Val Loss=0.629888, Mean Val F-1=0.529013]\n",
      "819it [03:19,  4.90it/s, Epoch=12, Loss=0.659005, Mean Loss=0.596933, Mean F-1=0.491454]\n",
      "205it [00:14, 13.98it/s, Epoch=12, Val Loss=0.896943, Mean Val Loss=0.768923, Mean Val F-1=0.124827]\n",
      "819it [03:18,  4.87it/s, Epoch=13, Loss=0.627357, Mean Loss=0.585310, Mean F-1=0.506312]\n",
      "205it [00:14, 13.98it/s, Epoch=13, Val Loss=0.733586, Mean Val Loss=0.667936, Mean Val F-1=0.358211]\n",
      "819it [03:18,  4.88it/s, Epoch=14, Loss=0.638989, Mean Loss=0.588266, Mean F-1=0.501342]\n",
      "205it [00:14, 14.05it/s, Epoch=14, Val Loss=0.581945, Mean Val Loss=0.725331, Mean Val F-1=0.555128]\n",
      "819it [03:18,  4.91it/s, Epoch=15, Loss=0.624948, Mean Loss=0.584899, Mean F-1=0.507582]\n",
      "205it [00:14, 13.98it/s, Epoch=15, Val Loss=0.565889, Mean Val Loss=0.692199, Mean Val F-1=0.561740]\n",
      "819it [03:19,  4.92it/s, Epoch=16, Loss=0.592589, Mean Loss=0.576959, Mean F-1=0.514262]\n",
      "205it [00:14, 14.04it/s, Epoch=16, Val Loss=0.578016, Mean Val Loss=0.724553, Mean Val F-1=0.553568]\n",
      "819it [03:19,  4.90it/s, Epoch=17, Loss=0.612778, Mean Loss=0.565682, Mean F-1=0.525139]\n",
      "205it [00:14, 13.97it/s, Epoch=17, Val Loss=0.544139, Mean Val Loss=0.712794, Mean Val F-1=0.564370]\n",
      "819it [03:19,  4.89it/s, Epoch=18, Loss=0.581622, Mean Loss=0.572290, Mean F-1=0.516619]\n",
      "205it [00:14, 14.04it/s, Epoch=18, Val Loss=0.547437, Mean Val Loss=0.699250, Mean Val F-1=0.560212]\n",
      "819it [03:19,  4.89it/s, Epoch=19, Loss=0.626044, Mean Loss=0.569696, Mean F-1=0.524139]\n",
      "205it [00:14, 13.98it/s, Epoch=19, Val Loss=0.640004, Mean Val Loss=0.692748, Mean Val F-1=0.555788]\n",
      "819it [03:19,  4.90it/s, Epoch=20, Loss=0.595039, Mean Loss=0.558266, Mean F-1=0.534357]\n",
      "205it [00:14, 14.02it/s, Epoch=20, Val Loss=0.896515, Mean Val Loss=0.769929, Mean Val F-1=0.161684]\n",
      "819it [03:18,  4.90it/s, Epoch=21, Loss=0.580519, Mean Loss=0.546859, Mean F-1=0.545511]\n",
      "205it [00:14, 13.98it/s, Epoch=21, Val Loss=0.628808, Mean Val Loss=0.717029, Mean Val F-1=0.565406]\n",
      "819it [03:19,  4.88it/s, Epoch=22, Loss=0.586520, Mean Loss=0.557266, Mean F-1=0.535170]\n",
      "205it [00:14, 13.99it/s, Epoch=22, Val Loss=0.834209, Mean Val Loss=0.735227, Mean Val F-1=0.235143]\n",
      "819it [03:19,  4.89it/s, Epoch=23, Loss=0.593423, Mean Loss=0.556565, Mean F-1=0.539107]\n",
      "205it [00:14, 13.97it/s, Epoch=23, Val Loss=0.858956, Mean Val Loss=0.744317, Mean Val F-1=0.207756]\n",
      "819it [03:19,  4.89it/s, Epoch=24, Loss=0.575634, Mean Loss=0.541017, Mean F-1=0.553151]\n",
      "205it [00:14, 13.98it/s, Epoch=24, Val Loss=0.592497, Mean Val Loss=0.649352, Mean Val F-1=0.515511]\n",
      "819it [03:19,  4.89it/s, Epoch=25, Loss=0.559821, Mean Loss=0.530896, Mean F-1=0.563981]\n",
      "205it [00:14, 13.97it/s, Epoch=25, Val Loss=0.805951, Mean Val Loss=0.667576, Mean Val F-1=0.361844]\n",
      "819it [03:19,  4.89it/s, Epoch=26, Loss=0.571294, Mean Loss=0.545067, Mean F-1=0.549198]\n",
      "205it [00:14, 13.97it/s, Epoch=26, Val Loss=0.709926, Mean Val Loss=0.637461, Mean Val F-1=0.426277]\n",
      "819it [03:17,  4.93it/s, Epoch=27, Loss=0.595486, Mean Loss=0.546106, Mean F-1=0.551485]\n",
      "205it [00:14, 14.04it/s, Epoch=27, Val Loss=0.904269, Mean Val Loss=0.783475, Mean Val F-1=0.145710]\n",
      "819it [03:18,  4.93it/s, Epoch=28, Loss=0.564446, Mean Loss=0.528844, Mean F-1=0.568023]\n",
      "205it [00:14, 14.04it/s, Epoch=28, Val Loss=0.606623, Mean Val Loss=0.642229, Mean Val F-1=0.472288]\n",
      "819it [03:18,  4.91it/s, Epoch=29, Loss=0.558422, Mean Loss=0.521779, Mean F-1=0.575670]\n",
      "205it [00:14, 14.03it/s, Epoch=29, Val Loss=0.561577, Mean Val Loss=0.628498, Mean Val F-1=0.504393]\n",
      "819it [03:18,  4.93it/s, Epoch=30, Loss=0.578921, Mean Loss=0.534688, Mean F-1=0.559958]\n",
      "205it [00:14, 14.04it/s, Epoch=30, Val Loss=0.687150, Mean Val Loss=0.633874, Mean Val F-1=0.444973]\n",
      "819it [03:17,  4.91it/s, Epoch=31, Loss=0.610038, Mean Loss=0.538022, Mean F-1=0.561284]\n",
      "205it [00:14, 14.04it/s, Epoch=31, Val Loss=0.604554, Mean Val Loss=0.678740, Mean Val F-1=0.563731]\n",
      "819it [03:18,  4.91it/s, Epoch=32, Loss=0.565650, Mean Loss=0.519962, Mean F-1=0.577330]\n",
      "205it [00:14, 14.05it/s, Epoch=32, Val Loss=0.832471, Mean Val Loss=0.729742, Mean Val F-1=0.270356]\n",
      "819it [03:18,  4.89it/s, Epoch=33, Loss=0.552015, Mean Loss=0.512764, Mean F-1=0.585554]\n",
      "205it [00:14, 14.03it/s, Epoch=33, Val Loss=0.600546, Mean Val Loss=0.629350, Mean Val F-1=0.515052]\n",
      "819it [03:18,  4.92it/s, Epoch=34, Loss=0.616940, Mean Loss=0.526012, Mean F-1=0.571286]\n",
      "205it [00:14, 14.05it/s, Epoch=34, Val Loss=0.555687, Mean Val Loss=0.659450, Mean Val F-1=0.545452]\n",
      "819it [03:17,  4.92it/s, Epoch=35, Loss=0.553897, Mean Loss=0.528260, Mean F-1=0.573725]\n",
      "205it [00:14, 14.03it/s, Epoch=35, Val Loss=0.681216, Mean Val Loss=0.641516, Mean Val F-1=0.433389]\n",
      "819it [03:18,  4.91it/s, Epoch=36, Loss=0.562039, Mean Loss=0.510084, Mean F-1=0.590072]\n",
      "205it [00:14, 14.04it/s, Epoch=36, Val Loss=0.616074, Mean Val Loss=0.761794, Mean Val F-1=0.542536]\n",
      "819it [03:18,  4.91it/s, Epoch=37, Loss=0.557405, Mean Loss=0.506002, Mean F-1=0.592260]\n",
      "205it [00:14, 14.03it/s, Epoch=37, Val Loss=0.534308, Mean Val Loss=0.651672, Mean Val F-1=0.560965]\n",
      "819it [03:18,  4.90it/s, Epoch=38, Loss=0.545955, Mean Loss=0.520216, Mean F-1=0.577276]\n",
      "205it [00:14, 14.04it/s, Epoch=38, Val Loss=0.546200, Mean Val Loss=0.699812, Mean Val F-1=0.565812]\n",
      "819it [03:18,  4.91it/s, Epoch=39, Loss=0.555058, Mean Loss=0.522773, Mean F-1=0.579825]\n",
      "205it [00:14, 14.01it/s, Epoch=39, Val Loss=0.629821, Mean Val Loss=0.734377, Mean Val F-1=0.561826]\n",
      "819it [03:18,  4.91it/s, Epoch=40, Loss=0.555195, Mean Loss=0.506231, Mean F-1=0.593421]\n",
      "205it [00:14, 14.03it/s, Epoch=40, Val Loss=0.549857, Mean Val Loss=0.665990, Mean Val F-1=0.562578]\n",
      "819it [03:17,  4.93it/s, Epoch=41, Loss=0.557118, Mean Loss=0.498206, Mean F-1=0.603284]\n",
      "205it [00:14, 14.00it/s, Epoch=41, Val Loss=0.553157, Mean Val Loss=0.656887, Mean Val F-1=0.559374]\n",
      "819it [03:18,  4.91it/s, Epoch=42, Loss=0.564431, Mean Loss=0.514566, Mean F-1=0.584389]\n",
      "205it [00:14, 14.04it/s, Epoch=42, Val Loss=0.790338, Mean Val Loss=0.763674, Mean Val F-1=0.217828]\n",
      "819it [03:18,  4.91it/s, Epoch=43, Loss=0.632355, Mean Loss=0.518432, Mean F-1=0.585070]\n",
      "205it [00:14, 14.03it/s, Epoch=43, Val Loss=0.716000, Mean Val Loss=0.672271, Mean Val F-1=0.358595]\n",
      "819it [03:18,  4.91it/s, Epoch=44, Loss=0.507819, Mean Loss=0.495477, Mean F-1=0.605254]\n",
      "205it [00:14, 14.03it/s, Epoch=44, Val Loss=0.521095, Mean Val Loss=0.688751, Mean Val F-1=0.572982]\n",
      "819it [03:18,  4.92it/s, Epoch=45, Loss=0.527190, Mean Loss=0.489806, Mean F-1=0.614501]\n",
      "205it [00:14, 14.02it/s, Epoch=45, Val Loss=0.764838, Mean Val Loss=0.710236, Mean Val F-1=0.293566]\n",
      "819it [03:17,  4.90it/s, Epoch=46, Loss=0.560899, Mean Loss=0.507651, Mean F-1=0.593301]\n",
      "205it [00:14, 14.03it/s, Epoch=46, Val Loss=0.609331, Mean Val Loss=0.641568, Mean Val F-1=0.537094]\n",
      "819it [03:18,  4.91it/s, Epoch=47, Loss=0.560174, Mean Loss=0.512062, Mean F-1=0.592521]\n",
      "205it [00:14, 14.02it/s, Epoch=47, Val Loss=0.727606, Mean Val Loss=0.664538, Mean Val F-1=0.402007]\n",
      "819it [03:18,  4.91it/s, Epoch=48, Loss=0.506782, Mean Loss=0.489090, Mean F-1=0.613343]\n",
      "205it [00:14, 14.04it/s, Epoch=48, Val Loss=0.589785, Mean Val Loss=0.653815, Mean Val F-1=0.515818]\n",
      "819it [03:18,  4.92it/s, Epoch=49, Loss=0.525172, Mean Loss=0.483477, Mean F-1=0.623195]\n",
      "205it [00:14, 14.02it/s, Epoch=49, Val Loss=0.596906, Mean Val Loss=0.630477, Mean Val F-1=0.528210]\n",
      "819it [03:18,  4.92it/s, Epoch=50, Loss=0.538131, Mean Loss=0.502716, Mean F-1=0.600487]\n",
      "205it [00:14, 14.02it/s, Epoch=50, Val Loss=0.628295, Mean Val Loss=0.677184, Mean Val F-1=0.554097]\n",
      "819it [03:17,  4.86it/s, Epoch=51, Loss=0.560071, Mean Loss=0.507129, Mean F-1=0.598383]\n",
      "205it [00:14, 14.01it/s, Epoch=51, Val Loss=0.555212, Mean Val Loss=0.638924, Mean Val F-1=0.517006]\n",
      "819it [03:18,  4.91it/s, Epoch=52, Loss=0.513806, Mean Loss=0.483036, Mean F-1=0.623550]\n",
      "205it [00:14, 14.03it/s, Epoch=52, Val Loss=0.758739, Mean Val Loss=0.676474, Mean Val F-1=0.369816]\n",
      "819it [03:18,  4.91it/s, Epoch=53, Loss=0.582298, Mean Loss=0.478023, Mean F-1=0.630900]\n",
      "205it [00:14, 14.00it/s, Epoch=53, Val Loss=0.455106, Mean Val Loss=0.687332, Mean Val F-1=0.573886]\n",
      "819it [03:18,  4.91it/s, Epoch=54, Loss=0.542834, Mean Loss=0.501154, Mean F-1=0.602658]\n",
      "205it [00:14, 14.02it/s, Epoch=54, Val Loss=0.568906, Mean Val Loss=0.670685, Mean Val F-1=0.554619]\n",
      "819it [03:18,  4.90it/s, Epoch=55, Loss=0.546275, Mean Loss=0.501467, Mean F-1=0.604879]\n",
      "205it [00:14, 14.00it/s, Epoch=55, Val Loss=0.588129, Mean Val Loss=0.655576, Mean Val F-1=0.539771]\n",
      "819it [03:17,  4.90it/s, Epoch=56, Loss=0.497893, Mean Loss=0.477742, Mean F-1=0.629451]\n",
      "205it [00:14, 14.02it/s, Epoch=56, Val Loss=0.551381, Mean Val Loss=0.671934, Mean Val F-1=0.557510]\n",
      "819it [03:18,  4.90it/s, Epoch=57, Loss=0.501882, Mean Loss=0.472513, Mean F-1=0.635393]\n",
      "205it [00:14, 14.01it/s, Epoch=57, Val Loss=0.459967, Mean Val Loss=0.678329, Mean Val F-1=0.570996]\n",
      "819it [03:18,  4.89it/s, Epoch=58, Loss=0.522879, Mean Loss=0.495151, Mean F-1=0.609990]\n",
      "205it [00:14, 14.03it/s, Epoch=58, Val Loss=0.565952, Mean Val Loss=0.762571, Mean Val F-1=0.542932]\n",
      "819it [03:18,  4.91it/s, Epoch=59, Loss=0.530438, Mean Loss=0.498223, Mean F-1=0.608222]\n",
      "205it [00:14, 14.01it/s, Epoch=59, Val Loss=0.550551, Mean Val Loss=0.694179, Mean Val F-1=0.569913]\n",
      "819it [03:18,  4.92it/s, Epoch=60, Loss=0.511778, Mean Loss=0.474865, Mean F-1=0.633735]\n",
      "205it [00:14, 14.02it/s, Epoch=60, Val Loss=0.602017, Mean Val Loss=0.789511, Mean Val F-1=0.514898]\n",
      "819it [03:17,  4.89it/s, Epoch=61, Loss=0.527921, Mean Loss=0.469014, Mean F-1=0.641513]\n",
      "205it [00:14, 14.01it/s, Epoch=61, Val Loss=0.508260, Mean Val Loss=0.688504, Mean Val F-1=0.564277]\n",
      "819it [03:18,  4.91it/s, Epoch=62, Loss=0.528917, Mean Loss=0.490837, Mean F-1=0.615997]\n",
      "205it [00:14, 14.01it/s, Epoch=62, Val Loss=0.741502, Mean Val Loss=0.674464, Mean Val F-1=0.369123]\n",
      "819it [03:18,  4.92it/s, Epoch=63, Loss=0.536992, Mean Loss=0.494880, Mean F-1=0.612954]\n",
      "205it [00:14, 14.01it/s, Epoch=63, Val Loss=0.582712, Mean Val Loss=0.627994, Mean Val F-1=0.521141]\n",
      "819it [03:18,  4.91it/s, Epoch=64, Loss=0.542473, Mean Loss=0.470032, Mean F-1=0.640005]\n",
      "205it [00:14, 14.02it/s, Epoch=64, Val Loss=0.678189, Mean Val Loss=0.759187, Mean Val F-1=0.548681]\n",
      "819it [03:18,  4.90it/s, Epoch=65, Loss=0.503766, Mean Loss=0.461673, Mean F-1=0.652070]\n",
      "205it [00:14, 13.97it/s, Epoch=65, Val Loss=0.757944, Mean Val Loss=0.663511, Mean Val F-1=0.417284]\n",
      "819it [03:18,  4.92it/s, Epoch=66, Loss=0.544153, Mean Loss=0.487915, Mean F-1=0.619921]\n",
      "205it [00:14, 14.01it/s, Epoch=66, Val Loss=0.586929, Mean Val Loss=0.637489, Mean Val F-1=0.534621]\n",
      "819it [03:18,  4.91it/s, Epoch=67, Loss=0.558253, Mean Loss=0.489305, Mean F-1=0.618872]\n",
      "205it [00:14, 14.00it/s, Epoch=67, Val Loss=0.559720, Mean Val Loss=0.739363, Mean Val F-1=0.558681]\n",
      "819it [03:18,  4.90it/s, Epoch=68, Loss=0.508526, Mean Loss=0.463125, Mean F-1=0.648996]\n",
      "205it [00:14, 14.00it/s, Epoch=68, Val Loss=0.869472, Mean Val Loss=0.777573, Mean Val F-1=0.233996]\n",
      "819it [03:18,  4.90it/s, Epoch=69, Loss=0.507123, Mean Loss=0.458181, Mean F-1=0.657001]\n",
      "205it [00:14, 13.99it/s, Epoch=69, Val Loss=0.877302, Mean Val Loss=0.752530, Mean Val F-1=0.229046]\n",
      "819it [03:17,  4.92it/s, Epoch=70, Loss=0.524324, Mean Loss=0.484491, Mean F-1=0.623295]\n",
      "205it [00:14, 14.01it/s, Epoch=70, Val Loss=0.644471, Mean Val Loss=0.649139, Mean Val F-1=0.541814]\n",
      "819it [03:18,  4.90it/s, Epoch=71, Loss=0.522453, Mean Loss=0.485626, Mean F-1=0.622602]\n",
      "205it [00:14, 13.98it/s, Epoch=71, Val Loss=0.657105, Mean Val Loss=0.678264, Mean Val F-1=0.538788]\n",
      "819it [03:18,  4.90it/s, Epoch=72, Loss=0.499564, Mean Loss=0.457780, Mean F-1=0.658511]\n",
      "205it [00:14, 14.00it/s, Epoch=72, Val Loss=0.559214, Mean Val Loss=0.747790, Mean Val F-1=0.554523]\n",
      "819it [03:18,  4.91it/s, Epoch=73, Loss=0.500089, Mean Loss=0.455313, Mean F-1=0.659532]\n",
      "205it [00:14, 13.99it/s, Epoch=73, Val Loss=0.620139, Mean Val Loss=0.653572, Mean Val F-1=0.457185]\n",
      "819it [03:18,  4.91it/s, Epoch=74, Loss=0.508432, Mean Loss=0.480668, Mean F-1=0.629506]\n",
      "205it [00:14, 14.01it/s, Epoch=74, Val Loss=0.638663, Mean Val Loss=0.662737, Mean Val F-1=0.508179]\n",
      "819it [03:18,  4.91it/s, Epoch=75, Loss=0.507731, Mean Loss=0.481417, Mean F-1=0.630444]\n",
      "205it [00:14, 13.98it/s, Epoch=75, Val Loss=0.604969, Mean Val Loss=0.761168, Mean Val F-1=0.545465]\n",
      "819it [03:20,  4.49it/s, Epoch=76, Loss=0.482639, Mean Loss=0.454204, Mean F-1=0.661561]\n",
      "205it [00:15, 13.57it/s, Epoch=76, Val Loss=0.653557, Mean Val Loss=0.647760, Mean Val F-1=0.441843]\n",
      "819it [03:20,  4.87it/s, Epoch=77, Loss=0.495828, Mean Loss=0.449492, Mean F-1=0.668882]\n",
      "205it [00:14, 13.92it/s, Epoch=77, Val Loss=0.522362, Mean Val Loss=0.657594, Mean Val F-1=0.565290]\n",
      "819it [03:19,  4.87it/s, Epoch=78, Loss=0.536475, Mean Loss=0.476917, Mean F-1=0.633148]\n",
      "205it [00:14, 13.92it/s, Epoch=78, Val Loss=0.620186, Mean Val Loss=0.629062, Mean Val F-1=0.476442]\n",
      "819it [03:19,  4.85it/s, Epoch=79, Loss=0.523104, Mean Loss=0.478602, Mean F-1=0.631653]\n",
      "205it [00:14, 13.92it/s, Epoch=79, Val Loss=0.567184, Mean Val Loss=0.625394, Mean Val F-1=0.513194]\n",
      "819it [03:19,  4.88it/s, Epoch=80, Loss=0.490563, Mean Loss=0.450737, Mean F-1=0.666573]\n",
      "205it [00:14, 13.94it/s, Epoch=80, Val Loss=0.703748, Mean Val Loss=0.671847, Mean Val F-1=0.408328]\n",
      "819it [03:19,  4.87it/s, Epoch=81, Loss=0.494094, Mean Loss=0.445187, Mean F-1=0.673216]\n",
      "205it [00:14, 13.91it/s, Epoch=81, Val Loss=0.577453, Mean Val Loss=0.730040, Mean Val F-1=0.562686]\n",
      "819it [03:19,  4.87it/s, Epoch=82, Loss=0.521193, Mean Loss=0.473365, Mean F-1=0.638717]\n",
      "205it [00:14, 13.90it/s, Epoch=82, Val Loss=0.709105, Mean Val Loss=0.665339, Mean Val F-1=0.404145]\n",
      "819it [03:19,  4.88it/s, Epoch=83, Loss=0.520738, Mean Loss=0.475735, Mean F-1=0.637589]\n",
      "205it [00:14, 13.90it/s, Epoch=83, Val Loss=0.611244, Mean Val Loss=0.635988, Mean Val F-1=0.523120]\n",
      "819it [03:20,  4.88it/s, Epoch=84, Loss=0.489345, Mean Loss=0.448663, Mean F-1=0.673608]\n",
      "205it [00:14, 13.92it/s, Epoch=84, Val Loss=0.541314, Mean Val Loss=0.636115, Mean Val F-1=0.520008]\n",
      "819it [03:20,  4.85it/s, Epoch=85, Loss=0.540233, Mean Loss=0.439135, Mean F-1=0.685668]\n",
      "205it [00:14, 13.90it/s, Epoch=85, Val Loss=0.640557, Mean Val Loss=0.641362, Mean Val F-1=0.451620]\n",
      "819it [03:20,  4.87it/s, Epoch=86, Loss=0.506846, Mean Loss=0.469099, Mean F-1=0.645290]\n",
      "205it [00:14, 13.90it/s, Epoch=86, Val Loss=0.736530, Mean Val Loss=0.676156, Mean Val F-1=0.426464]\n",
      "819it [03:19,  4.88it/s, Epoch=87, Loss=0.519247, Mean Loss=0.472294, Mean F-1=0.643002]\n",
      "205it [00:14, 13.90it/s, Epoch=87, Val Loss=0.562566, Mean Val Loss=0.694547, Mean Val F-1=0.567553]\n",
      "819it [03:20,  4.88it/s, Epoch=88, Loss=0.468295, Mean Loss=0.442784, Mean F-1=0.679449]\n",
      "205it [00:14, 13.89it/s, Epoch=88, Val Loss=0.542525, Mean Val Loss=0.671095, Mean Val F-1=0.565960]\n",
      "819it [03:20,  4.89it/s, Epoch=89, Loss=0.504733, Mean Loss=0.436300, Mean F-1=0.688443]\n",
      "205it [00:14, 13.90it/s, Epoch=89, Val Loss=0.745453, Mean Val Loss=0.710628, Mean Val F-1=0.348872]\n",
      "819it [03:20,  4.87it/s, Epoch=90, Loss=0.521289, Mean Loss=0.467057, Mean F-1=0.648783]\n",
      "205it [00:14, 13.90it/s, Epoch=90, Val Loss=0.562835, Mean Val Loss=0.755363, Mean Val F-1=0.547369]\n",
      "819it [03:20,  4.87it/s, Epoch=91, Loss=0.519810, Mean Loss=0.470606, Mean F-1=0.647987]\n",
      "205it [00:14, 13.88it/s, Epoch=91, Val Loss=0.715433, Mean Val Loss=0.674260, Mean Val F-1=0.440644]\n",
      "819it [03:19,  4.88it/s, Epoch=92, Loss=0.491431, Mean Loss=0.439920, Mean F-1=0.685543]\n",
      "205it [00:14, 13.91it/s, Epoch=92, Val Loss=0.557868, Mean Val Loss=0.669227, Mean Val F-1=0.562598]\n",
      "819it [03:20,  4.87it/s, Epoch=93, Loss=0.506467, Mean Loss=0.433240, Mean F-1=0.693072]\n",
      "205it [00:14, 13.90it/s, Epoch=93, Val Loss=0.674415, Mean Val Loss=0.746830, Mean Val F-1=0.558312]\n",
      "819it [03:20,  4.86it/s, Epoch=94, Loss=0.511414, Mean Loss=0.465716, Mean F-1=0.651849]\n",
      "205it [00:14, 13.88it/s, Epoch=94, Val Loss=0.627841, Mean Val Loss=0.739707, Mean Val F-1=0.558105]\n",
      "819it [03:20,  4.86it/s, Epoch=95, Loss=0.525944, Mean Loss=0.466697, Mean F-1=0.649709]\n",
      "205it [00:14, 13.89it/s, Epoch=95, Val Loss=0.671440, Mean Val Loss=0.663125, Mean Val F-1=0.550711]\n",
      "819it [03:19,  4.85it/s, Epoch=96, Loss=0.490194, Mean Loss=0.435973, Mean F-1=0.690499]\n",
      "205it [00:14, 13.89it/s, Epoch=96, Val Loss=0.507247, Mean Val Loss=0.651284, Mean Val F-1=0.539202]\n",
      "819it [03:20,  4.87it/s, Epoch=97, Loss=0.494420, Mean Loss=0.431465, Mean F-1=0.695277]\n",
      "205it [00:14, 13.90it/s, Epoch=97, Val Loss=0.671592, Mean Val Loss=0.683132, Mean Val F-1=0.442677]\n",
      "819it [03:19,  4.91it/s, Epoch=98, Loss=0.632025, Mean Loss=0.463389, Mean F-1=0.656989]\n",
      "205it [00:14, 13.97it/s, Epoch=98, Val Loss=0.588326, Mean Val Loss=0.714414, Mean Val F-1=0.568045]\n",
      "819it [03:19,  4.90it/s, Epoch=99, Loss=0.541912, Mean Loss=0.464225, Mean F-1=0.653561]\n",
      "205it [00:14, 13.97it/s, Epoch=99, Val Loss=0.804917, Mean Val Loss=0.733657, Mean Val F-1=0.269366]\n",
      "819it [03:18,  4.90it/s, Epoch=100, Loss=0.474517, Mean Loss=0.432572, Mean F-1=0.696286]\n",
      "205it [00:14, 13.96it/s, Epoch=100, Val Loss=0.601380, Mean Val Loss=0.647779, Mean Val F-1=0.477555]\n"
     ]
    }
   ],
   "source": [
    "loss_plot, val_loss_plot = [], []\n",
    "metric_plot, val_metric_plot = [], []\n",
    "\n",
    "for epoch in range(epochs):\n",
    "    if epoch == 0 and load:\n",
    "        val_loss_plot =  loaded_model['val_loss_history']\n",
    "        val_metric_plot = loaded_model['val_metric_history']\n",
    "        loss_plot = loaded_model['loss_history']\n",
    "        loss_metric = loaded_model['metric_history']\n",
    "        break\n",
    "\n",
    "    total_loss, total_val_loss = 0, 0\n",
    "    total_acc, total_val_acc = 0, 0\n",
    "\n",
    "    toggle = epoch % 4\n",
    "\n",
    "    torch.cuda.empty_cache() # corrected\n",
    "    \n",
    "    tqdm_dataset = tqdm(enumerate(tr[toggle]))\n",
    "    training = True\n",
    "    for batch, batch_item in tqdm_dataset:\n",
    "        batch_loss, batch_acc = train_step(batch_item, training)\n",
    "        total_loss += batch_loss\n",
    "        total_acc += batch_acc\n",
    "        tqdm_dataset.set_postfix({\n",
    "            'Epoch': epoch + 1,\n",
    "            'Loss': '{:06f}'.format(batch_loss.item()),\n",
    "            'Mean Loss' : '{:06f}'.format(total_loss/(batch+1)),\n",
    "            'Mean F-1' : '{:06f}'.format(total_acc/(batch+1)),\n",
    "        })\n",
    "        del batch_item\n",
    "\n",
    "    loss_plot.append(total_loss/(batch+1))\n",
    "    metric_plot.append(total_acc/(batch+1))\n",
    "    tqdm_dataset = tqdm(enumerate(te[toggle]))\n",
    "    training = False\n",
    "    for batch, batch_item in tqdm_dataset:\n",
    "        batch_loss, batch_acc = train_step(batch_item, training)\n",
    "        total_val_loss += batch_loss\n",
    "        total_val_acc += batch_acc\n",
    "        \n",
    "        tqdm_dataset.set_postfix({\n",
    "            'Epoch': epoch + 1,\n",
    "            'Val Loss': '{:06f}'.format(batch_loss.item()),\n",
    "            'Mean Val Loss' : '{:06f}'.format(total_val_loss/(batch+1)),\n",
    "            'Mean Val F-1' : '{:06f}'.format(total_val_acc/(batch+1))\n",
    "        })\n",
    "    val_loss_plot.append(total_val_loss/(batch+1))\n",
    "    val_metric_plot.append(total_val_acc/(batch+1))\n",
    "    \n",
    "    if np.max(val_metric_plot) == val_metric_plot[-1]:\n",
    "        if imbalanced_weight:\n",
    "            save_path = resv2_path + '/best_modelw.pt'\n",
    "        else:\n",
    "            save_path = resv2_path + '/best_model.pt'\n",
    "        torch.save(model.state_dict(), save_path)\n",
    "\n",
    "    if imbalanced_weight:\n",
    "        save_path = resv2_path + '/modelw_' + '{:03d}'.format(epoch+1) + '.pt'\n",
    "    else:\n",
    "        save_path = resv2_path + '/model_' + '{:03d}'.format(epoch+1) + '.pt'\n",
    "    torch.save({\n",
    "        'epoch': epochs,\n",
    "        'model_state_dict': model.state_dict(),\n",
    "        'optimizer_state_dict': optimizer.state_dict(),\n",
    "        'loss_history': loss_plot,\n",
    "        'metric_history' : metric_plot,\n",
    "        'val_loss_history': val_loss_plot,\n",
    "        'val_metric_history': val_metric_plot\n",
    "        }, save_path\n",
    "    )\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "HaL9XP8F0yaQ"
   },
   "source": [
    "# 학습 결과"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "id": "PRMz6mo70yaR",
    "outputId": "1b84898e-4b54-42d7-9cf2-1c9e4969c717"
   },
   "outputs": [],
   "source": [
    "loss_plot = torch.tensor(loss_plot)\n",
    "val_loss_plot = torch.tensor(val_loss_plot)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([0.6725, 0.6570, 0.6489, 0.6417, 0.6316, 0.6280, 0.6225, 0.6164, 0.6071,\n",
      "        0.6060, 0.6020, 0.5969, 0.5853, 0.5883, 0.5849, 0.5770, 0.5657, 0.5723,\n",
      "        0.5697, 0.5583, 0.5469, 0.5573, 0.5566, 0.5410, 0.5309, 0.5451, 0.5461,\n",
      "        0.5288, 0.5218, 0.5347, 0.5380, 0.5200, 0.5128, 0.5260, 0.5283, 0.5101,\n",
      "        0.5060, 0.5202, 0.5228, 0.5062, 0.4982, 0.5146, 0.5184, 0.4955, 0.4898,\n",
      "        0.5077, 0.5121, 0.4891, 0.4835, 0.5027, 0.5071, 0.4830, 0.4780, 0.5012,\n",
      "        0.5015, 0.4777, 0.4725, 0.4952, 0.4982, 0.4749, 0.4690, 0.4908, 0.4949,\n",
      "        0.4700, 0.4617, 0.4879, 0.4893, 0.4631, 0.4582, 0.4845, 0.4856, 0.4578,\n",
      "        0.4553, 0.4807, 0.4814, 0.4542, 0.4495, 0.4769, 0.4786, 0.4507, 0.4452,\n",
      "        0.4734, 0.4757, 0.4487, 0.4391, 0.4691, 0.4723, 0.4428, 0.4363, 0.4671,\n",
      "        0.4706, 0.4399, 0.4332, 0.4657, 0.4667, 0.4360, 0.4315, 0.4634, 0.4642,\n",
      "        0.4326])\n",
      "tensor([0.6868, 0.7212, 0.6587, 0.6458, 0.6435, 0.6615, 0.6493, 0.6956, 0.7091,\n",
      "        0.6155, 0.6299, 0.7689, 0.6679, 0.7253, 0.6922, 0.7246, 0.7128, 0.6993,\n",
      "        0.6927, 0.7699, 0.7170, 0.7352, 0.7443, 0.6494, 0.6676, 0.6375, 0.7835,\n",
      "        0.6422, 0.6285, 0.6339, 0.6787, 0.7297, 0.6293, 0.6594, 0.6415, 0.7618,\n",
      "        0.6517, 0.6998, 0.7344, 0.6660, 0.6569, 0.7637, 0.6723, 0.6888, 0.7102,\n",
      "        0.6416, 0.6645, 0.6538, 0.6305, 0.6772, 0.6389, 0.6765, 0.6873, 0.6707,\n",
      "        0.6556, 0.6719, 0.6783, 0.7626, 0.6942, 0.7895, 0.6885, 0.6745, 0.6280,\n",
      "        0.7592, 0.6635, 0.6375, 0.7394, 0.7776, 0.7525, 0.6491, 0.6783, 0.7478,\n",
      "        0.6536, 0.6627, 0.7612, 0.6478, 0.6576, 0.6291, 0.6254, 0.6718, 0.7300,\n",
      "        0.6653, 0.6360, 0.6361, 0.6414, 0.6762, 0.6945, 0.6711, 0.7106, 0.7554,\n",
      "        0.6743, 0.6692, 0.7468, 0.7397, 0.6631, 0.6513, 0.6831, 0.7144, 0.7337,\n",
      "        0.6478])\n",
      "[0.48043622355965454, 0.5327730245588522, 0.48591073646658867, 0.4283171289227598, 0.3726976724451781, 0.3182992919371278, 0.34194932975160036, 0.5545171395993219, 0.5539278252645549, 0.4808511052142727, 0.5290134024098088, 0.12482744481732423, 0.35821054354130716, 0.5551278052046887, 0.5617401707483654, 0.5535679040827017, 0.5643698752012549, 0.5602121436413591, 0.5557878317457543, 0.16168368881216066, 0.5654060984189638, 0.23514329531740555, 0.2077564254857413, 0.5155108841147117, 0.3618435766400902, 0.42627694930881066, 0.14570952455785338, 0.4722875587929097, 0.504393473747863, 0.4449729486656728, 0.5637309179055187, 0.27035626275286156, 0.5150521201316494, 0.5454521702616937, 0.433388782928386, 0.5425363430877307, 0.5609647153854833, 0.5658124936675928, 0.5618258030629766, 0.5625779458984591, 0.5593741446840302, 0.21782808205862392, 0.35859540181161914, 0.5729817370985074, 0.2935659274754636, 0.5370943205852305, 0.40200711049548665, 0.5158180769029554, 0.5282099659477785, 0.5540968060186023, 0.5170060128145532, 0.3698162522436563, 0.5738862566915931, 0.5546191912259063, 0.5397713494929088, 0.5575103083859154, 0.5709964631420592, 0.5429317391302727, 0.5699128889017835, 0.514897883056525, 0.5642765751149252, 0.3691234532491492, 0.5211407906276311, 0.5486805325833783, 0.4172835238807062, 0.5346210042830327, 0.5586814115848818, 0.2339964524185439, 0.22904622346455916, 0.5418144331076771, 0.538788373698748, 0.5545234176498943, 0.4571852898983669, 0.5081793681431127, 0.5454649674412152, 0.44184286812043716, 0.5652901022568466, 0.4764416482324665, 0.5131941327111033, 0.408327904443956, 0.56268556879444, 0.4041449807791804, 0.5231195292946834, 0.5200078589994332, 0.4516197561483948, 0.4264640753123832, 0.5675534032195834, 0.5659597122174893, 0.34887154308753987, 0.5473691318045815, 0.4406442283233495, 0.5625982966988846, 0.5583115481619709, 0.5581051217986662, 0.550710744451634, 0.5392021267393893, 0.44267725305225747, 0.5680449129121593, 0.26936560307360213, 0.47755495475518184]\n",
      "[0.48043622355965454, 0.5327730245588522, 0.48591073646658867, 0.4283171289227598, 0.3726976724451781, 0.3182992919371278, 0.34194932975160036, 0.5545171395993219, 0.5539278252645549, 0.4808511052142727, 0.5290134024098088, 0.12482744481732423, 0.35821054354130716, 0.5551278052046887, 0.5617401707483654, 0.5535679040827017, 0.5643698752012549, 0.5602121436413591, 0.5557878317457543, 0.16168368881216066, 0.5654060984189638, 0.23514329531740555, 0.2077564254857413, 0.5155108841147117, 0.3618435766400902, 0.42627694930881066, 0.14570952455785338, 0.4722875587929097, 0.504393473747863, 0.4449729486656728, 0.5637309179055187, 0.27035626275286156, 0.5150521201316494, 0.5454521702616937, 0.433388782928386, 0.5425363430877307, 0.5609647153854833, 0.5658124936675928, 0.5618258030629766, 0.5625779458984591, 0.5593741446840302, 0.21782808205862392, 0.35859540181161914, 0.5729817370985074, 0.2935659274754636, 0.5370943205852305, 0.40200711049548665, 0.5158180769029554, 0.5282099659477785, 0.5540968060186023, 0.5170060128145532, 0.3698162522436563, 0.5738862566915931, 0.5546191912259063, 0.5397713494929088, 0.5575103083859154, 0.5709964631420592, 0.5429317391302727, 0.5699128889017835, 0.514897883056525, 0.5642765751149252, 0.3691234532491492, 0.5211407906276311, 0.5486805325833783, 0.4172835238807062, 0.5346210042830327, 0.5586814115848818, 0.2339964524185439, 0.22904622346455916, 0.5418144331076771, 0.538788373698748, 0.5545234176498943, 0.4571852898983669, 0.5081793681431127, 0.5454649674412152, 0.44184286812043716, 0.5652901022568466, 0.4764416482324665, 0.5131941327111033, 0.408327904443956, 0.56268556879444, 0.4041449807791804, 0.5231195292946834, 0.5200078589994332, 0.4516197561483948, 0.4264640753123832, 0.5675534032195834, 0.5659597122174893, 0.34887154308753987, 0.5473691318045815, 0.4406442283233495, 0.5625982966988846, 0.5583115481619709, 0.5581051217986662, 0.550710744451634, 0.5392021267393893, 0.44267725305225747, 0.5680449129121593, 0.26936560307360213, 0.47755495475518184]\n"
     ]
    }
   ],
   "source": [
    "print(loss_plot)\n",
    "print(val_loss_plot)\n",
    "print(metric_plot)\n",
    "print(val_metric_plot)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "id": "5Gnq2Uz70yaR",
    "outputId": "9716e8e1-91f0-4db1-ee9a-ffb234f954ee"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x700 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,7))\n",
    "plt.grid()\n",
    "plt.plot(loss_plot, label='train_loss')\n",
    "plt.plot(val_loss_plot, label='val_loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.ylabel('loss')\n",
    "plt.title(\"Loss\", fontsize=25)\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "id": "nSCCMBK40yaR",
    "outputId": "47b11b1a-ec42-49c5-a0f7-7a2f5f6a89ec"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x700 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,7))\n",
    "plt.grid()\n",
    "plt.plot(metric_plot, label='train_metric')\n",
    "plt.plot(val_metric_plot, label='val_metric')\n",
    "plt.xlabel('epoch')\n",
    "plt.ylabel('metric')\n",
    "plt.title(\"F-1\", fontsize=25)\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "metric = pd.DataFrame([loss_plot.numpy(), val_loss_plot.numpy(), metric_plot, val_metric_plot]).transpose()\n",
    "if imbalanced_weight:\n",
    "    metric.to_csv(resv2_path + r'\\metricw.csv')\n",
    "else:\n",
    "    metric.to_csv(resv2_path + r'\\metric.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Og22z0s50yaR"
   },
   "source": [
    "# 추론"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "id": "2dvxhh4C0yaS",
    "outputId": "bfeed36c-22bf-426d-98ac-1ced95b33a9a"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "0it [00:00, ?it/s]C:\\Users\\Jeyun\\AppData\\Roaming\\Python\\Python37\\site-packages\\ipykernel_launcher.py:19: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).\n",
      "205it [00:21,  9.33it/s]\n"
     ]
    }
   ],
   "source": [
    "def predict(dataset):\n",
    "    model.eval()\n",
    "    tqdm_dataset = tqdm(enumerate(dataset))\n",
    "    indices = []\n",
    "    outputs = []\n",
    "    preds = []\n",
    "    labels = []\n",
    "\n",
    "    for batch, batch_item in tqdm_dataset:\n",
    "\n",
    "        img = batch_item['img'].to(device)\n",
    "        # metadata = batch_item['metadata'].to(device)\n",
    "        label = batch_item['labels']\n",
    "        indice = batch_item['indices']\n",
    "\n",
    "\n",
    "        with torch.no_grad():\n",
    "            output = model(img)\n",
    "        pred = torch.tensor(torch.argmax(output, dim=1), dtype=torch.int32).cpu().numpy()\n",
    "        indices.extend(indice)\n",
    "        outputs.extend(output)\n",
    "        preds.extend(pred)\n",
    "        labels.extend(label)\n",
    "        \n",
    "    return indices, outputs, preds, labels\n",
    "\n",
    "# model = CNN2RNN(max_len=max_len, embedding_dim=embedding_dim, num_features=num_features, class_n=class_n, rate=dropout_rate)\n",
    "# model.load_state_dict(torch.load(save_path, map_location=device))\n",
    "# model.to(device)\n",
    "\n",
    "\n",
    "pr_indices, pr_outputs, pr_preds, pr_labels = predict(te[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "id": "kVD54C5q0yaS"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "소요시간:  0:04:02.472961\n",
      "현재:  2023-05-02 10:58:25.826659\n"
     ]
    }
   ],
   "source": [
    "# preds = np.array([label_decoder[int(val)] for val in preds])\n",
    "\n",
    "elapsed_time = datetime.now() - now\n",
    "now = datetime.now()\n",
    "\n",
    "print(\"소요시간: \", elapsed_time)\n",
    "print(\"현재: \", now)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "52370\n",
      "             output pred label\n",
      "index                         \n",
      "127551  1.05755e-13    0     0\n",
      "65804    2.1566e-11    0     0\n",
      "164412   0.00135993    0     0\n",
      "242979  3.42142e-09    0     0\n",
      "228357  5.54957e-18    0     0\n",
      "111734   3.5964e-08    0     0\n",
      "121639  3.69918e-11    0     0\n",
      "197907     0.988079    1     0\n",
      "298614            1    1     1\n",
      "491859  4.86568e-08    0     0\n",
      "127542  1.47479e-15    0     0\n",
      "75455      0.314689    0     0\n",
      "535509   0.00187287    0     0\n",
      "5431    2.30117e-08    0     1\n",
      "93916   8.37269e-05    0     0\n",
      "141392  1.27703e-22    0     0\n",
      "335770  2.65112e-16    0     0\n",
      "378516   0.00729046    0     0\n",
      "234542  1.99787e-11    0     0\n",
      "549473  1.09942e-09    0     1\n",
      "307532  1.09116e-07    0     1\n",
      "507508   3.1832e-09    0     1\n",
      "57662   3.53026e-13    0     0\n",
      "380940  1.22529e-06    0     0\n",
      "470034  8.17229e-23    0     0\n",
      "87937   4.47019e-13    0     0\n",
      "371121  8.91784e-13    0     0\n",
      "158160    0.0192636    0     0\n",
      "91181     0.0190851    0     0\n",
      "330128   0.00826942    0     0\n",
      "...             ...  ...   ...\n",
      "277596  7.33118e-19    0     0\n",
      "90528   5.36801e-07    0     0\n",
      "433238   0.00921586    0     0\n",
      "347151     0.676518    1     0\n",
      "532286  3.22736e-14    0     0\n",
      "203030  3.15748e-07    0     0\n",
      "355841  1.85266e-20    0     0\n",
      "191652  1.12762e-13    0     0\n",
      "439091  4.63148e-12    0     0\n",
      "212100  7.39268e-18    0     0\n",
      "67539   7.30209e-21    0     0\n",
      "12315     0.0593153    0     0\n",
      "303512  0.000400221    0     0\n",
      "234309  6.55716e-06    0     0\n",
      "529762    0.0010159    0     0\n",
      "4982       0.998005    1     0\n",
      "544471  4.92519e-05    0     0\n",
      "86954    0.00651239    0     0\n",
      "499141  2.10283e-11    0     0\n",
      "314649  8.09605e-16    0     0\n",
      "475249  1.65375e-12    0     0\n",
      "425620  7.80851e-14    0     0\n",
      "69977   2.60314e-10    0     0\n",
      "228683  0.000873777    0     0\n",
      "207287   4.0785e-20    0     0\n",
      "492871  4.45861e-11    0     0\n",
      "345206  2.52982e-05    0     0\n",
      "343423  8.82875e-06    0     0\n",
      "274542   2.6238e-05    0     0\n",
      "267810  9.78435e-08    0     0\n",
      "\n",
      "[1000 rows x 3 columns]\n",
      "0.12562535802940614 0.06389154095856406\n"
     ]
    }
   ],
   "source": [
    "print(len(pr_preds))\n",
    "pred_plt = pd.DataFrame([[i.numpy() for i in pr_indices], [i.cpu().numpy()[1] for i in pr_outputs], pr_preds, [i.numpy() for i in pr_labels]], index = [\"index\", \"output\", \"pred\", \"label\"]).transpose()\n",
    "\n",
    "pred_plt.set_index(\"index\", inplace=True)\n",
    "print(pred_plt.iloc[0:1000])\n",
    "print(pred_plt.pred.mean(), pred_plt.label.mean())\n",
    "if imbalanced_weight:\n",
    "    pred_plt.to_csv(resv2_path + r'\\pred_pltw.csv')\n",
    "else:\n",
    "    pred_plt.to_csv(resv2_path + r'\\pred_plt.csv')\n"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [],
   "name": "Playground Baseline.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "base",
   "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": "3.7.2"
  },
  "vscode": {
   "interpreter": {
    "hash": "269fb8e673d61fa21fa3a9e68af6f5ef6e50007c1460be8f4db77c309d1a71b1"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}