crnn_str_simple.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "68759407",
   "metadata": {},
   "source": [
    "# CRNN\n",
    "\n",
    "paper：\n",
    "[An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Application to Scene Text Recognition](http://arxiv.org/abs/1507.05717)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "b418279a-c930-4f60-a63d-3f3322ce0eee",
   "metadata": {
    "editable": true,
    "slideshow": {
     "slide_type": ""
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import math\n",
    "import torch\n",
    "import numpy as np\n",
    "import torch.nn as nn\n",
    "import matplotlib.pyplot as plt\n",
    "import torchvision.transforms as transforms\n",
    "from torch.utils.data import DataLoader, Dataset\n",
    "from PIL import Image\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "f2abde5b-0beb-489f-aaf7-d51ecd0a4c19",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "65\n"
     ]
    }
   ],
   "source": [
    "batch_size = 32\n",
    "#lr=0.0005  #使用Adadelta优化器，不需要学习率\n",
    "num_epochs = 30\n",
    "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\") \n",
    "char_to_int = {char: i for i, char in enumerate('*0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ- ')}\n",
    "num_classes = len(char_to_int)\n",
    "print(num_classes)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cc285567-c275-4167-915b-ba95116a6a1a",
   "metadata": {},
   "source": [
    "## 1. 定义网络"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "99a5d1b0",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "class CRNN(nn.Module):\n",
    "    def __init__(self, num_classes, input_channels=1):\n",
    "        super(CRNN, self).__init__()\n",
    "\n",
    "        # Convolutional Layers\n",
    "        self.conv = nn.Sequential(\n",
    "            nn.Conv2d(input_channels, 64, kernel_size=3, stride=1, padding=1), #1\n",
    "            nn.ReLU(inplace=True),\n",
    "            nn.MaxPool2d(kernel_size=2, stride=2),\n",
    "\n",
    "            nn.Conv2d(64, 128, kernel_size=3, stride=1, padding=1), #2\n",
    "            nn.ReLU(inplace=True),\n",
    "            nn.MaxPool2d(kernel_size=2, stride=2),\n",
    "\n",
    "            nn.Conv2d(128, 256, kernel_size=3, stride=1, padding=1), #3\n",
    "            nn.BatchNorm2d(256),\n",
    "            nn.ReLU(inplace=True),\n",
    "            \n",
    "            nn.Conv2d(256, 256, kernel_size=3, stride=1, padding=1), #4\n",
    "            nn.ReLU(inplace=True),\n",
    "            nn.MaxPool2d(kernel_size=(2, 2), stride=(2, 1), padding=(0, 1)),\n",
    "\n",
    "            nn.Conv2d(256, 512, kernel_size=3, stride=1, padding=1), #5\n",
    "            nn.BatchNorm2d(512),\n",
    "            nn.ReLU(inplace=True),\n",
    "            \n",
    "            nn.Conv2d(512, 512, kernel_size=3, stride=1, padding=1), #6\n",
    "            nn.ReLU(inplace=True),\n",
    "            nn.MaxPool2d(kernel_size=(2, 2), stride=(2, 1), padding=(0, 1)),\n",
    "\n",
    "            nn.Conv2d(512, 512, kernel_size=2, stride=1, padding=0), #7\n",
    "            nn.BatchNorm2d(512),\n",
    "            nn.ReLU(inplace=True),\n",
    "        )\n",
    "        # Recurrent Laysers\n",
    "        self.lstm1 = nn.LSTM(input_size=512, hidden_size=256, bidirectional=True)\n",
    "        self.lstm2 = nn.LSTM(input_size=512, hidden_size=256, bidirectional=True)\n",
    "        \n",
    "        # Fully connected layer\n",
    "        self.fc = nn.Linear(in_features=512, out_features=num_classes)\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = self.conv(x)\n",
    "        b, c, h, w = x.size()\n",
    "        x = x.squeeze(2)\n",
    "        x = x.permute(2, 0, 1)\n",
    "\n",
    "        x, _ = self.lstm1(x)\n",
    "        x, _ = self.lstm2(x)\n",
    "\n",
    "        x = self.fc(x)\n",
    "\n",
    "        return x\n",
    "\n",
    "crnn = CRNN(num_classes=num_classes).to(device)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "10a32305-a830-45b8-b154-12cf570498f5",
   "metadata": {},
   "source": [
    "## 2. 数据加载"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "310f85b0-34e1-4649-9343-cef3f7257c6e",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "from PIL import Image\n",
    "import torch\n",
    "from torchvision import transforms\n",
    "\n",
    "class CustomDataset(torch.utils.data.Dataset):\n",
    "    def __init__(self, root_dir, transform=None):\n",
    "        self.root_dir = root_dir\n",
    "        self.transform = transform\n",
    "        self.file_list = os.listdir(root_dir)\n",
    "\n",
    "    def __len__(self):\n",
    "        return len(self.file_list)\n",
    "\n",
    "    def __getitem__(self, idx):\n",
    "        # Get file path\n",
    "        img_name = os.path.join(self.root_dir, self.file_list[idx])\n",
    "        # Open image\n",
    "        image = Image.open(img_name).convert('L')\n",
    "        # Extract label from filename\n",
    "        label = self.file_list[idx].split('_')[0]\n",
    "        # Apply transformations if any\n",
    "        if self.transform:\n",
    "            image = self.transform(image)\n",
    "        return image, label\n",
    "\n",
    "# Image transformations\n",
    "transform = transforms.Compose([\n",
    "    transforms.Resize((32, 100)),  # Resize to fixed size (height, width)\n",
    "    transforms.Grayscale(),  # Convert image to grayscale\n",
    "    transforms.ToTensor(),\n",
    "    transforms.Normalize(0.5, 0.5)  # Use tuple for both mean and std\n",
    "])\n",
    "\n",
    "\n",
    "# Create dataset and dataloader\n",
    "train_dataset = CustomDataset(root_dir='/home/han/Documents/SRP/dataset/traindata/', transform=transform)\n",
    "train_loader = torch.utils.data.DataLoader(train_dataset, batch_size=32, shuffle=True, num_workers=4)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "67bfbafe-f603-4ca4-ab89-7781586ff2ab",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def images_show(imgs,labs=[]):\n",
    "    i=0\n",
    "    #w = math.ceil(len(imgs)/2)\n",
    "    for img in imgs:\n",
    "        plt.title(labs[i])\n",
    "        #plt.imshow(img.cpu().squeeze(0))\n",
    "        plt.imshow(img.cpu().squeeze(),plt.cm.gray)\n",
    "        plt.show()\n",
    "        i+=1\n",
    "        if  i >= 8:\n",
    "            print(f\"Maximium Images for images_show: {i}\")\n",
    "            break"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "ff10c7bd-0059-4d60-ba10-1b2a5861d000",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAh8AAADeCAYAAACddmLpAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy81sbWrAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAya0lEQVR4nO2de3CV5Z3HfxFKIBDCJZIQbgYFEZFWQKlohdYSRcdqaTtWK+LuH1tFWykzFdHumnaVUKbj2N0qrR0HmVqK2ylr0basqVpal1URpFBUvIDcY+SWhItQybN/7HCW8z4f9BdO8uaA389MZsgv73nf5/qeh/d8zu8pCCEEE0IIIYRIidPauwBCCCGE+GShxYcQQgghUkWLDyGEEEKkihYfQgghhEgVLT6EEEIIkSpafAghhBAiVbT4EEIIIUSqaPEhhBBCiFTR4kMIIYQQqaLFhxBCCCFSRYsPIYQQQqSKFh9CiJw4cOBAexdBCHGSocWHEMJNdXW1FRQU2KpVq+yrX/2q9ezZ084880z74IMPbNasWVZZWWmdOnWyfv362W233WZ79+7Nen1zc7PNnTvXhg0bZoWFhdanTx+76aabbOvWrVnHTZgwwUaMGGErVqywz33uc1ZUVGSDBw+2OXPmWHNzc4o1FkK0BVp8CCFazOTJk+2ss86yX//61/bTn/7Urr32WvvRj35kU6ZMsd/97nc2Y8YMW7BggX3hC1+wQ4cOZV5366232syZM23ixIm2ZMkS+9d//VdbunSpjRs3znbu3Jl1jbq6OvvGN75hN954oy1ZssQmTZpks2bNsscffzzt6gohWpsghBBO7r333mBm4V/+5V8ysaVLlwYzC3Pnzs069oknnghmFh555JEQQgivv/56MLMwbdq0rONeeumlYGbh7rvvzsTGjx8fzCy89NJLWccOHz48XH755a1dLSFEyujJhxCixXzlK1/J/Pu5554zM7Obb74565ivfe1r1rVrV3v22WfNzOz555/H4y688EI755xzMscdpby83C688MKs2MiRI23Tpk2tUQUhRDuixYcQosX07ds38+9du3ZZx44d7fTTT886pqCgwMrLy23Xrl2Z45KvPUpFRUXm70fp3bt3dFxhYaEdPHgw5/ILIdoXLT6EEC2moKAg8+/evXvbhx9+aO+//37WMSEEq6urs9LS0sxxZmY7duyIzrd9+/bMcUKIUx8tPoQQOXHZZZeZmUUi6G9+8xvbv39/5u9f+MIX8LgVK1bY66+/njlOCHHq07G9CyCEOLmZOHGiXX755TZz5kxrbGy0iy++2NasWWP33nuvnX/++TZlyhQzMzv77LPtn/7pn+zf//3f7bTTTrNJkybZu+++a//8z/9sAwYMsO985zvtXBMhRFpo8SGEyImCggJ78sknrbq62ubPn2/333+/lZaW2pQpU2z27NlWWFiYOXbevHl25pln2qOPPmoPPfSQlZSU2BVXXGE1NTXoeAghTk0KQgihvQshhBBCiE8Ocj6EEEIIkSpafAghhBAiVbT4EEIIIUSqaPEhhBBCiFTR4kMIIYQQqdJmi4+HH37YKisrrXPnzjZ69Gj7y1/+0laXEkIIIcRJRJvk+XjiiSds+vTp9vDDD9vFF19sP/vZz2zSpEn22muv2cCBAz/ytc3NzbZ9+3YrLi7OSuEshBBCiPwlhGBNTU1WUVFhp5320c822iTPx9ixY23UqFE2b968TOycc86xa6+91mpqaj7ytVu3brUBAwa0dpGEEEIIkQJbtmyx/v37f+Qxrf7k4/Dhw7Zy5Uq76667suJVVVW2fPny6PhDhw7ZoUOHMr8fXQu98cYbVlxcHMWPhWKHDx+OYk1NTa6yd+wYNwed7+9//3sUa25ujmK08uvVq1cU6969e9bvVK8jR464YlSODh06RLEPP/wwih04cMB1Pnoi1a1btyjWo0ePKLZnz54otnPnzijWuXPnKHZspsyPKgu1O9WD2pnOR+1Hx3nLkuw3Gk/HzomPihFULxrbn/rUp6JY165dXcdRmWlM0XUJ6h+6hve6NDe8r+3SpYsrduz96aOOa2xszPqd5hmNE2q7Tp06RTEanxSjNqb679+/3xXzlpnmLcWoLN6+9d73vGWmuex9D0qWj8pGdaByUIxeS+Xw9re3fJ7r7t+/36qqqnBuJGn1xcfOnTvtyJEjVlZWlhUvKyuzurq66Piamhr7/ve/H8WLi4uz3pS9He+9QRPexQfFvIsP6pT2WHzQjdh7w6KJ6amXGQ/+Dz74IIrRTTyfFh90DW9Zkm1AfUFvMrQIIKhe9Frv4oPKQnOA+tZbZu9iwbuAoFgui4+ioqIoRuObXpvsj1zeFKkv6DjvXKY2yWVsU39r8ZHfiw/ve4v3umbcfknaTDhNXjyEgAWaNWuWNTQ0ZH62bNnSVkUSQgghRB7Q6k8+SktLrUOHDtFTjvr6+uhpiNn/rYBpFbx169asR/m0iqVVtvd/C97/yRC0oqQnLvS/CipzcjW6b9++6Bh6OkBQXekjETqOoAUjvZbq+t5770UxeuRMr6V+pP+10Mcz3ic93idYuTwh8cS89aIxS+f3zgG6bkNDQxR7//33o5j3iRg9CaDretvY+z9rahf6nxv9T9D7kR+1Ac3d5HW9T6EIus/QRyJelc/bFzR+vDGa88mPosxyG8veJ3bUzt4vNnifpiVjNO68H9vSGCNyeTJFbey9D3pedzxa/clHp06dbPTo0VZbW5sVr62ttXHjxrX25YQQQghxktEmX7WdMWOGTZkyxcaMGWMXXXSRPfLII7Z582a75ZZb2uJyQgghhDiJaJPFx3XXXWe7du2yH/zgB7Zjxw4bMWKE/f73v7dBgwa1xeWEEEIIcRLRJosPM7Np06bZtGnT2ur0QgghhDhJabPFR66UlpZmfX2TpB0SBknIOnjwYBTbvXt3FPPmJCC5x5tHgeTPpOBGohRd05tvhOrvlWZJviOJ0Cso0Vdye/fuHcW8eCUwryDp/ZqulxPN4Ufl8H4FkqB2IrxfWfR+jTGXsZyL6NvaX5+ke82JfjWS+oLmKB1H5SC885bmozcHEN0vKKcSjYGePXtGMa9k7u1br8CZy/z2joEkVDZvWgfvnPce572veGI0F497XfeRQgghhBCtgBYfQgghhEgVLT6EEEIIkSpafAghhBAiVfJWOO3UqVNWVkFv1j3KdEfkIiiRkJXLfghJqYiymeaybwaJYXQ+7/4A3j0yCKq/dx8Kko69IpxXZvNKk956eCTHXPZg8PYZ4c1eS3OKruvNqOith3eMEtQX3mvQ+Ca8+w8l69uSPTKSeDNUUp9RhlwSU73ZlKkeNPdOdM+jllzXu9+LV4z3Stae+0ouonguG116r+vdG8jznuntQzM9+RBCCCFEymjxIYQQQohU0eJDCCGEEKmixYcQQgghUiVvhdOGhoYsOcgrypBARfKVN0ueN+ucVxjzbO1Owpc3Q6dXmqV26t69exQjacm7PT3FaCttOs67HbZnS2uz3Lbr9mYx9GbQTPZHLgKdV/L0CnTerJpUPoLaidrEKznmklnV237euUxtQNvbJ9uUrknz0SvNeoXlhoaGKLZ3717Xa2lO0TgrKyuLYtS3XjHVK1J6x7dX4PSOMyLZHzR/qD29c9n7hQevcErj+ES/QEH9ejz05EMIIYQQqaLFhxBCCCFSRYsPIYQQQqSKFh9CCCGESJW8FU5DCFlylTfTpleGpONIlvFeg6RWyoBIAlEyKyLJZ96t3qmu3np5xUevINqtW7coRluHU5ZSr/DkzfLq3Z7dK5x6xV7PePRum+0V0mj8eIXGffv2RTHqH6qXV4b0Znek8lF9vVk/cxH1CG/21mSM+tvbTtTGXkE2F0Hdm9GW+qxHjx5RzCsiezMdeyVZ733Am/WTzpcsXy6ZW3OZP97jvPdaz/mamppc1zTTkw8hhBBCpIwWH0IIIYRIFS0+hBBCCJEqWnwIIYQQIlXyVjgtKChwCzPHQqIQiZ/eDJpeCZPkSpJQPVule7dE90qElGGQZCzKdujNlkltTPX3bs9O4hJlj6QtzHPZxj2XzKVeiTc5BmhMeCW4XLaE94qa1LcE1dWbPZLmozcrrVdeJNEzF9mZxp5ni3WvKO65V5hxO9H9qKioKIrRfKTxQ6I4lY/mLbUTxbxSPbVBLl8goGt4xx6RPJ/3SwteedwrGOfCiYrN3vcuMz35EEIIIUTKaPEhhBBCiFTR4kMIIYQQqaLFhxBCCCFSJW+F065du2bJUCTokEjpzcjplc9yERVJWvKUha5JwhsJZHRNOs67JTqJgHQ+ahPKlukVyN5///0o1tjYGMX69esXxUiiI2HO2wZe4dQrWyX7iEQz79bfhFdwI8GPREDv1uTeMnslZppT3rbySnm5ZJWkuezZnpzmFI1Zgsaxd9zRcTQGqHy5XKOhoSGKkQxKbUD3Gm8G35KSkuOW8+PIJRtscq7R+xTVn8Y79YV3bHsz1bamwEpz4njoyYcQQgghUkWLDyGEEEKkihYfQgghhEgVLT6EEEIIkSp5K5wmoSyLJBR5ZTYSY+gaXhnHu90yyUfJmHdLb5LP6LVUr1zKS5lGSZbyyoskVZWXl0ex0tLSKOYV3Lxbp3vlylyE0+Rx7ZWxkDJUeqVrEvyoHylG7U7lIymPyufNDEnHde/ePYrt2bMnitGYomvQmE+2szf7qlfg9WblpWvs3LnTdQ3qM++c984fb0Zkuid574XezMQEjWWaQ8lxS9Ls6aefHsVIqN+1a1cU8/aPN4Ox997oEVhbkpVcTz6EEEIIkSpafAghhBAiVbT4EEIIIUSqaPEhhBBCiFTJW+G0oKAgS3ohKYjER292UJJnvJkNKebd2pxiSWmnR48e0TEkO3m3ufZmcfRmxKNreDOXeqHXerJHHi/mFa1IyqPXkgjn3WI8ed3i4uLoGJLqSEijcpBESeej8tJW7DT2SAb1Zhf29g9B44zuA9RWBM1lKl/Pnj2jGLUVSajJdqF7QC6ZVr118M55773Mmy2TJFRqT7oGyb8k0FNfkOhJdauvr3cdR/WgsiSvS+PznXfe+djXmZn16tXLVTbvPdk7H0lWpfon521L7vl68iGEEEKIVNHiQwghhBCposWHEEIIIVKlxYuPP//5z3b11VdbRUWFFRQU2JNPPpn19xCCVVdXW0VFhXXp0sUmTJhg69ata63yCiGEEOIkp8XC6f79++3Tn/60/cM//IN95Stfif4+d+5ce+CBB+yxxx6zoUOH2n333WcTJ0609evXo1h3PLp27Zol05FQQ6JMLnIlCU/eLINeaceTTY8kQhLZqBx0/tbehpyO80pv1D9emY3O5820SW3gzYBIdfNm7iRhLClpUX+TGOYdFyTBUXlJQqU6UP/kIjZ75V+vxOztM5ov3ozIdD56LZEss7edCCqbN+upp2xm3LdeyZ5iXjGVxjdJjiQ7U8y7vbs3AzTNIbrGvn37PrZshPc9icpGsirhzfBK9aL7RbLMJNgfjxYvPiZNmmSTJk3Cv4UQ7MEHH7R77rnHJk+ebGZmCxYssLKyMlu4cKF985vfbOnlhBBCCHGK0arOx8aNG62urs6qqqoyscLCQhs/frwtX74cX3Po0CFrbGzM+hFCCCHEqUurLj7q6urMzKysrCwrXlZWlvlbkpqaGispKcn8DBgwoDWLJIQQQog8o02+7ZL8fDCEcNzd7mbNmmUNDQ2Zny1btrRFkYQQQgiRJ7RqhtOj26DX1dVZ3759M/H6+vroachRCgsLUSras2dPlrzkFYC820snpSAz3q6boLKQ+EcxIin37N271/U6ryDqFWRzkWvpOK9ARSKTt27U31QWEuG856N60BbrhKcsu3fvjo6hOeHNhErbpHuz7XozaHrFR6/s7BVYvRk0SSL0iq50H6AYSXmeunkzCXvnKB1HePvR+1oqs7fPKEZjxZvZ2dtnBM0rr8Dp6Q96HyC51gu1E13DOy6oPQka78l+9N63zVr5yUdlZaWVl5dbbW1tJnb48GFbtmyZjRs3rjUvJYQQQoiTlBY/+di3b5+9/fbbmd83btxoq1evtl69etnAgQNt+vTpNnv2bBsyZIgNGTLEZs+ebUVFRXbDDTe0asGFEEIIcXLS4sXHK6+8Yp///Oczv8+YMcPMzKZOnWqPPfaY3XnnnXbw4EGbNm2a7dmzx8aOHWvPPPNMi3J8CCGEEOLUpcWLjwkTJnzk5zoFBQVWXV1t1dXVuZRLCCGEEKcorSqctiYhhI+VV7zZE70SnVfm8sqvFPMIWbmIXLmQSzu1dhu3ZGvmJNQuJEvRNbyyKm2TTZC4liyfN3svSZQkjZJoR1IZ1b979+5RzFOHlsS8GYdzOV8uAiv1LbUp9QfFklB5cxF9vVkrvfcVIpcsst4YZcckedorNtO4pfvAnj17XGXp1q1bFKP5cuyXLcx47JCw7p0DVAfKj5XLHPC+xyVlb29WWTNtLCeEEEKIlNHiQwghhBCposWHEEIIIVJFiw8hhBBCpEreCqdFRUVZWeBIlPFuB92SayYhkcm7PTvh2ZqbykECEGW3pPp7JVyvwOo9jq5BZfFKhIQ3ayXJUt4t2+kaJSUlrtdSfyTLQtIa9Tedn+rVo0ePKEZtQlt9kzBJ1yDBjaRWb+ZO77bmBEluVJZctqOnNiU84zYXOduLV+r1zjNvWeg4kl9p7HnFa+ozb/notTT/vPdHkkmT85TmLdWLspRS9t5c7tPU31SWE82uS3PneOjJhxBCCCFSRYsPIYQQQqSKFh9CCCGESBUtPoQQQgiRKnkrnO7duzdL1GltKTGXbH/ezH50PhKtksIciVd0fm8WQ29WRBL3vFuxk6Dk3e7eKzcR3syLuWT2y0Vw82yxThk1vRJYLtlwvRIq9bdXdvZu/+3dEpzK4hUpvW3qzVbsmctmcX+0trDtFcq948ebMZbwZqUlSK6kutEYoHYnGZTGY2lpqeu6JFnTGEjOZ6rXsV+mOArVge4N3vb03le9Y8Az5/ft2+cqm5mefAghhBAiZbT4EEIIIUSqaPEhhBBCiFTR4kMIIYQQqZK3wunBgwezRD+SdigznVdmI3mG5B4SjzxbC5v5s6MmRSYqG0mEVC+vkOYVyOg4kqy8mUap/t7slt66kSxFshi1KW2b7c0u6xXLkuPsvffec5WjV69eUYwyeVKMxiy1O9WLBD96bc+ePaMY1d8rgHsFaE/WYDO/gEeiIol01KZ03eQ1aE55t1OnMevNjkrn88qbXrmWztenT58o1rt37yi2a9euKFZfXx/FvFl4vWOP6kZziOpLcz5ZN+pbOhfN+fLyclfZKOaVs2nM0j2P5ndynNF99njoyYcQQgghUkWLDyGEEEKkihYfQgghhEiVvHU+krvaej/Dpc9md+zYEcXoszr6zI0+I6NrkC9Cr6XP95Pl27ZtW3QMfV47cuTIKEaf1b366qtRbMOGDVGMPl99//33o9jOnTuj2OjRo6PYZZddFsWGDx8exSoqKqLY7t27oxh9nkhjgD6zX7NmTRT729/+5opR39Jn1vR5Mn0WO2jQoKzfP/OZz0THnHPOOVGsrKwsijU0NESxPXv2RDEa28XFxVHM+zm5NwEWeUnkPOzduzeKUZ8tW7YsipErQJ9301iuq6uLYtTO/fr1i2KjRo2KYjQnBw8e/LHnojFLdfD6W+Qx0JyieyPdo+jecPbZZ0exMWPGRLF169ZFsaeeeiqKvfvuu1GM7oVUN68LQzuGk1dx1llnRbERI0ZEseRcNot3QPYm6aP7DPVFLkn1ctk9mcZZ8n1JScaEEEIIkbdo8SGEEEKIVNHiQwghhBCposWHEEIIIVKlIHi3yEuJxsZGKykpsa1bt2YlEfMmnSKRh6Q8gqRWbwIa7662mzZtimJvvfVW1u8k2m3dujWKkRzo3amUEsYMHTrUdRxJhCRyUaxv376uGLUd9ePChQuj2PPPPx/FKAkR1S0pB5qxkEZ1o8RBVObXX3896/dVq1a5ykYSXFVVVRSbOHFiFKMxQJIaiakkZ5NER8Icycl//etfo9gf/vCHKEaCJJWPklj1798/ipEcOHDgwCi2fv36KEbS5MaNG6MYyb6f/exns36/4IILomNI2D7jjDOiGI1jkvxI3iRhkNrz2WefjWLz58+PYnT/JRGb5gW97ZDAO2zYsChG0D15y5YtUYzuo5s3b45i1M40/8aNGxfFPv/5z2f9TskwvQnAqM+8uy57k9R5d7L27JTc1NRkw4YNs4aGBqx31ms/8q9CCCGEEK2MFh9CCCGESBUtPoQQQgiRKlp8CCGEECJV8jbDaUNDQ5aAQ3ITiXAkfpaUlEQx7860JAF5ZSGSoCijYFLUI7mNpKiLLrooil188cVR7HOf+1wUKy0tjWKUZZGkJRKPaMdVugZl7CMx7Omnn45ilN2SxE/KGHr99ddHsQsvvDCKeXfqJBHMu0Nosp2p3an+JNfSDp8kIpNYSZk2SbqmuUfHUbbQ//iP/4hi77zzThQ7/fTTo9jll1/uitH8pv4hIZay5lJ22alTp0axtWvXRjGSh5MC53//939Hx5AkPHbs2ChGQjQJt9TfJMOSNEvC+yuvvBLFSAaluZcUMM34nkQiO41vz87BZixSkgBO2VtfeOGFKPaLX/wiitH9LPneQplgSWL3Zhr17pLrvW95d7L2ZEdVhlMhhBBC5C1afAghhBAiVbT4EEIIIUSqaPEhhBBCiFTJW+G0ubk5S6whKY8EKhJqKKsbSUsknNJr6RokupKsSRJqMksclY3kzUsvvTSKffGLX4xilO2RpCUS8qgOJIZ5z+fdAn7Dhg1RjOS4mTNnRrHrrrsuitH4IdHVmwmUMslSu5D8mcz8R2OH+pb6ggTmRYsWRbEbb7wxitH279Qm27dvj2K0Pf0bb7wRxUiupLH81a9+1VU+77xNbmtuxpK5d37v3bs3ilVUVEQxEqCT84DGBM0L6ovzzjsvipFwS5llSQaktqM60HGUlfWb3/xmFKNMlyQvbtu2LYo1NTVFMW+GT+98pMzOJCJTJma6RlLWpL6l+lNWY4p55VK6b9E9hNrEK78mofvicY89oSsIIYQQQpwgWnwIIYQQIlW0+BBCCCFEqmjxIYQQQohUaZFwWlNTY4sXL7Y33njDunTpYuPGjbMf/vCHWRncQgj2/e9/3x555BHbs2ePjR071h566CE799xzW1SwioqKLFGJsrCRQEXiDUk2lLWRpCVvJktvdkuSuZLbUA8YMCA6pqysLIpRm1LWSpJcqbwkHhEkSJKAmNw63oyzCVKWQBIBSfArLi6OYiQlkrhFQhaJhV45jM5HsllSZvOOE8po+9RTT0UxyrJJouKZZ54Zxag9KRvj22+/HcVoDFCmzcrKyihG45vanURKktxIkKS60bb11Gc0Lij7JmUbPf/887N+p/vRn//85yhGoi9t605CJ5WNxhTJ6G+++WYUo6yvlNWY7jXJ+9vxXktziu4D1D90Xa9gvGnTpihG44zurUTytTTuqK7erLw0fgi6D5IgS+1E74+e67ZEVG3Rk49ly5bZbbfdZi+++KLV1tbahx9+aFVVVVkNNHfuXHvggQfsJz/5ia1YscLKy8tt4sSJaC0LIYQQ4pNHi558LF26NOv3+fPnW58+fWzlypV26aWXWgjBHnzwQbvnnnts8uTJZma2YMECKysrs4ULF+LXsA4dOpT19Ub6ipMQQgghTh1ycj6OPv46+ght48aNVldXZ1VVVZljCgsLbfz48bZ8+XI8R01NjZWUlGR+6CMHIYQQQpw6nPDiI4RgM2bMsEsuucRGjBhhZv+/q2Xy89uysjLc8dLMbNasWdbQ0JD5ocQ6QgghhDh1OOEMp7fffrutWbMGtx5OSichhOOKKIWFhSiHbdiwwbp16/aRZSABhuQZkq/omt7MeSQtUZZO8lxIXEu+ll5Hwi1ltyRJjQRRkvRIUPJmDiSBjITYl156KYrRgpMERI+sa8bSEwm2lM3Uuw01SaLeLLzJ7eNpy21arJNU580s690im/qb5hnJ3iTp0bj1tieNUcpc6pXR6Rr0WpLt6LrU9nS+5H2M5gq1O5WX6kX3C5LHd+3aFcXeeustV4zON3jw4Ch2wQUXRDEajySZU3sS1FYk3dK4pbHnlSupnekekhRMSbAeNGhQFPO2E70vUv/QfZ/qSnXw3muS7dQSt/OEnnx861vfsiVLltjzzz+fZUsfteKTN876+np8MxFCCCHEJ48WLT5CCHb77bfb4sWL7bnnnotWdJWVlVZeXm61tbWZ2OHDh23ZsmU2bty41imxEEIIIU5qWvSxy2233WYLFy603/72t1ZcXJx5wlFSUmJdunSxgoICmz59us2ePduGDBliQ4YMsdmzZ1tRUZHdcMMNbVIBIYQQQpxctGjxMW/ePDMzmzBhQlZ8/vz5dvPNN5uZ2Z133mkHDx60adOmZZKMPfPMM5hoRQghhBCfPFq0+PBmOKuurrbq6uoTLZOZ/Z8QeqzYSPIMiY8k1NBrSeYiOZDOR2Kmd4t1EteSPgzJSCT7kKhIr929e3cUI/GKyuvNlEiQmEsiE7U7iVYk627YsCGKUZZJqi/VjTJe0nEkVpHETCJYsp2pTUgCozFGbUxyKc0BKq+3z+h89B8MklppC3iSMKndSZr0Zu6kMhNbt26NYiR3k/jnyZJM91Fqd+pvuibVn/qM5venP/3pKEZzj75YQH1B2VG926x75XYaP3SNbdu2RTFqF7on07ylPqL7cvI9gyTkzZs3RzGqP5XD+wUKeu+i+ejtH5o/yfPRfet4aG8XIYQQQqSKFh9CCCGESBUtPoQQQgiRKlp8CCGEECJVTjjDaVuzadOmLNmGRCtKXEbHebO6kYRKkHBK4hEJSp4smJTdksr7zjvvRDGSsaj+VF4SyEgy8gq3JDeRWJj89pQZb8++du3aKEay2KhRo1xlITGT5FySS0muJDmMSLYptTFtG04yH/VFcgt3szirqhmPAW+mUcpuSVLvK6+8EsUoo+2aNWui2HnnnRfFvNuak5xMsiHFvDI2tT2Nqddffz3r91WrVkXH0Njp27ev6zia33QfoPn99ttvRzGSuGleUEZOuudRmSnmzXxLsia1O/UPvT/QcRQj0ZPuK8nX9uzZMzrmrLPOimJUV+84pr6geUtjxftFAxJik+fzfCklUz73kUIIIYQQrYAWH0IIIYRIFS0+hBBCCJEqWnwIIYQQIlXyVjgdPHhwltRHUh4JMIR3G2WStAgS9agsJDcRSVnKm42ShCKqK8lI3gyVXmmW6krtSVIVCZ1Hd0g+FpIN//jHP0Yx2hJ89OjRUYyywZK4Rm3vzbjr2VaAxlNSUjQzW7RoURQjmY1ETZIXKTMmCbw0zkiApsyYlN2RYiQ0UozmGZWZYlRmyq5bUVERxWjs1dfXRzGSwFeuXJn1+6uvvhodc+WVV0Yx2p6eRE2SEr3ZLSkTLM09EhBpbJeWlkYxKjPNqcbGxihG9zPvVvE0Vqge3vJ5ReTkWOnevXt0jDcDLX2pgu6/NLap/t4YjRWat8n3FrqXHw89+RBCCCFEqmjxIYQQQohU0eJDCCGEEKmixYcQQgghUiVvhdMePXpkiTrercNJvNm3b18U82Zi8wo6JOXRNUhkSp6PxCM6F4lMJCDSa0lyJKGKMv1R+QgSyEhypGx/JIaRSLl06dIo9uyzz0YxEqHOPvvsKEYZWEmcJVGRxoBHJv2f//mf6BjaDpwEzCuuuCKKTZkyJYrRXKH+psyqJPORWDhmzJgoRv24fPnyKHbfffdFsT59+kSxL3/5y1Fs2LBhUYwE0d69e0cxGgPvvvtuFHvxxRejWFIkNeMsvEkpjyRPKsfIkSOjGGWqpXvjjh07ohiJip4t4c34vkLH0fihbMV0nybJka5BMjFJo96s0yTsejO60rxKZvClex7Vi/qWxqz33k3t6X0voNfSlwWS16X2OB568iGEEEKIVNHiQwghhBCposWHEEIIIVJFiw8hhBBCpEreCqedO3fOEhu9Qo03OygdRzGCpB0SGkm+IlExGfNmJCV5ispGEpBX4CWoLF4xlfqCYiQqUuZJkvfWr18fxSgT6nPPPRfFqA1IyKJ2pvFIEmpSmCN580tf+lIUu+aaa6IYiYCUbZbkXxLtKGsnZTYcOHBgFDvzzDOj2JAhQ6LYhAkTotjXv/71KEYi7n/9139FsV/96ldRzCvlUd1oPNL4pqyxgwcPjmLJvvzsZz8bHUNjh8q7ffv2KEbyIpWNxvbOnTujGMmgdF26T5Fc6d2ynvBmoqb7Bd1/qb+9oj2Nb5rfAwYMyPqdspTSnKe280qzdD8iQZRi1MaE532EpObjoScfQgghhEgVLT6EEEIIkSpafAghhBAiVbT4EEIIIUSqFARvqs+UaGxstJKSElu7dm3Wls0kBZEA4xUBvcIpxUgy8mZCJZEpeT4SvmgLZsrCR/IQSZkkLZFURhIUZY/0SkubNm2KYhs2bIhilGmUMhvSuPBKWiTEUlZNusbmzZujGMlsnnFB5aAx6806SLFcttf2bk1OkOhK7UT9Q/OAXkuCLWWGJCmchEsSm0ngpHsNxZLjx5tR0ns/ojbZtWtXFKPrUj96ZXTKKkpz1NtONFZoDBz7vnAUuv+Q3E/nozag+y3dH+m9IDn2KBM1SbM0b6m81Bfe90IaAzSmvPefZKypqcmGDx9uDQ0NWO9j0ZMPIYQQQqSKFh9CCCGESBUtPoQQQgiRKlp8CCGEECJV8lY4XbNmTZZY5BVgSI7LJeseyVwU80pAVOYkJAVR2Si7JZ2fxFQqG7WdNyukd8t2r9Do2b7ZzL+9tlfMpHpQfUmOI3GNshsm+9KbjdPbniSwUjZKqj+JvpTxs66uLopR25WXl7vKR5kRvTGCBERvJl2aLyQverP1emRsGk/UP97sq96MyNRnJHTSvYbGHtXVK1KSEEzH0ZyndqHXEiQsk0hK9wHPPcRbf2o7amPPe4gZC6dELhmrk8c1NjZaWVmZhFMhhBBC5B9afAghhBAiVbT4EEIIIUSqaPEhhBBCiFTxpaVsB7p3754lrJBsR1noSKwjkcm7TbpXBvQKrJ5t5r2ZLN98880oRuUliZCuQVtQ03EkRlFmUJLUqB7Uj97MfiRLURuQhEp9QRkv6bjKysooRnIY1SPZzpSNkqS/008/PYqRROjN+krH0Rig/qZ2J957770o5pWTSQYl2ZDancrnzZJMkvCgQYOiGEmodP9JCo3ee4VXwKQxQOOd7m8kUdKcp7YjSXbv3r1RjIR/mlN0DYKOy2WM0j3JK+d6ZFKSpGmcUNZT6kcaK1R/730gl4y7STGXRN3joScfQgghhEgVLT6EEEIIkSpafAghhBAiVVq0+Jg3b56NHDky42NcdNFF9oc//CHz9xCCVVdXW0VFhXXp0sUmTJhg69ata/VCCyGEEOLkpUXCaf/+/W3OnDl21llnmZnZggUL7JprrrFXX33Vzj33XJs7d6498MAD9thjj9nQoUPtvvvus4kTJ9r69etRoPsoDh8+nCXukCxFcksu2UwpRpCg4xVtPNIbCZj0OhIQvdnqCJKb9u3bF8UoA6RXFvNuE04yG8UGDhwYxWiskfhIsh0JjSRfbdu2LYoRnvFI27UTJK7R+b0ZXqlv6Tiqv/e61Bck7lHGVO888Iriya3OzViUrq+vj2LvvPNOFKPMmCRSJuvhFeC9EiXVn4RGug94RWzqb5o/1D90bySZmI7zbhVPIi6JyAS1lfe6NDeS7UftTu3pFafpfN5MqF7Zm4Rbqn+yfHRPOR4tevJx9dVX25VXXmlDhw61oUOH2v3332/dunWzF1980UII9uCDD9o999xjkydPthEjRtiCBQvswIEDtnDhwpZcRgghhBCnMCfsfBw5csQWLVpk+/fvt4suusg2btxodXV1VlVVlTmmsLDQxo8fb8uXLz/ueQ4dOmSNjY1ZP0IIIYQ4dWnx4mPt2rXWrVs3KywstFtuucX+8z//04YPH555bJr8jnxZWRk+Uj1KTU2NlZSUZH4GDBjQ0iIJIYQQ4iSixYuPs88+21avXm0vvvii3XrrrTZ16lR77bXXMn9Pfs4UQvjIXfhmzZplDQ0NmZ8tW7a0tEhCCCGEOIlocYbTTp06ZYTTMWPG2IoVK+zHP/6xzZw508z+Txw7VqCrr6/HjIFHKSwsxO2QPZCgRFKMd5t0bxZIOp9XAvJkp6NjqK7e8pIERAtC73b3FRUVUYyEL8rc6ZUSzzjjjChGAh6Jn5T5tV+/flGMxiVlraSxQmWhGMl7yfFDwhuJZnSuHTt2RDESGinLLYmuJJwS1Gc0br0yH2XD9W4xTmWm69IcoraietBYofN5JD+vKE/9TW3inXt0v6BxRnWg15Kc7RWMqc8829Ob+TOc0vsKxbwZPknCpOsmZVqqA8nUVA4axwSNKe97F8WoLETyOO8XD8xaIc9HCMEOHTpklZWVVl5ebrW1tZm/HT582JYtW2bjxo3L9TJCCCGEOEVo0ZOPu+++2yZNmmQDBgywpqYmW7Rokf3pT3+ypUuXWkFBgU2fPt1mz55tQ4YMsSFDhtjs2bOtqKjIbrjhhrYqvxBCCCFOMlq0+HjvvfdsypQptmPHDispKbGRI0fa0qVLbeLEiWZmduedd9rBgwdt2rRptmfPHhs7dqw988wzLc7xIYQQQohTl4JAkkQ70tDQYD169LBVq1ZlJfGhYnqL7v38zutQeD9zPFHnw5v0xbtDIX2m53U+PDsZmvHnzuRK0Gfb3mRN1Mbbt2+PYvQ5PvkN5J/QrpxUDxp7ns9/zeL+pc9/vc4HfTWdHB9yKo7dNfooNFYI6jOv85FLQjqv80EJ2bxelnc30BNNLEhjh8ZJLs4HJQKkOpADQe3k9fK8u8FSG3t3sqZr0FihMU/1oPnnva7H8/LuyO7dad2bGM7r0NBxVBbP+0NTU5ONGjXK9u7di7v0HkuLhdO25ugbx6hRo9q5JEIIIYRoKU1NTR+7+Mi7Jx/Nzc22fft2Ky4utqamJhswYIBt2bIF/5cm0qOxsVF9kSeoL/IL9Uf+oL5oX0II1tTUZBUVFR/7ZDDvnnycdtpp1r9/fzP7/48Gjm5kJ9of9UX+oL7IL9Qf+YP6ov34uCceR8n5q7ZCCCGEEC1Biw8hhBBCpEpeLz4KCwvt3nvvPeEMqKL1UF/kD+qL/EL9kT+oL04e8k44FUIIIcSpTV4/+RBCCCHEqYcWH0IIIYRIFS0+hBBCCJEqWnwIIYQQIlW0+BBCCCFEquTt4uPhhx+2yspK69y5s40ePdr+8pe/tHeRTnlqamrsggsusOLiYuvTp49de+21tn79+qxjQghWXV1tFRUV1qVLF5swYYKtW7eunUr8yaGmpsYKCgps+vTpmZj6Il22bdtmN954o/Xu3duKiorsM5/5jK1cuTLzd/VHOnz44Yf2ve99zyorK61Lly42ePBg+8EPfpC1GZr64iQg5CGLFi0Kn/rUp8LPf/7z8Nprr4U77rgjdO3aNWzatKm9i3ZKc/nll4f58+eHv/3tb2H16tXhqquuCgMHDgz79u3LHDNnzpxQXFwcfvOb34S1a9eG6667LvTt2zc0Nja2Y8lPbV5++eVwxhlnhJEjR4Y77rgjE1dfpMfu3bvDoEGDws033xxeeumlsHHjxvDHP/4xvP3225lj1B/pcN9994XevXuHp59+OmzcuDH8+te/Dt26dQsPPvhg5hj1Rf6Tl4uPCy+8MNxyyy1ZsWHDhoW77rqrnUr0yaS+vj6YWVi2bFkIIYTm5uZQXl4e5syZkznmgw8+CCUlJeGnP/1pexXzlKapqSkMGTIk1NbWhvHjx2cWH+qLdJk5c2a45JJLjvt39Ud6XHXVVeEf//Efs2KTJ08ON954YwhBfXGykHcfuxw+fNhWrlxpVVVVWfGqqipbvnx5O5Xqk0lDQ4OZmfXq1cvMzDZu3Gh1dXVZfVNYWGjjx49X37QRt912m1111VX2xS9+MSuuvkiXJUuW2JgxY+xrX/ua9enTx84//3z7+c9/nvm7+iM9LrnkEnv22WftzTffNDOzv/71r/bCCy/YlVdeaWbqi5OFvNvVdufOnXbkyBErKyvLipeVlVldXV07leqTRwjBZsyYYZdccomNGDHCzCzT/tQ3mzZtSr2MpzqLFi2yVatW2YoVK6K/qS/SZcOGDTZv3jybMWOG3X333fbyyy/bt7/9bSssLLSbbrpJ/ZEiM2fOtIaGBhs2bJh16NDBjhw5Yvfff79df/31Zqa5cbKQd4uPoxQUFGT9HkKIYqLtuP32223NmjX2wgsvRH9T37Q9W7ZssTvuuMOeeeYZ69y583GPU1+kQ3Nzs40ZM8Zmz55tZmbnn3++rVu3zubNm2c33XRT5jj1R9vzxBNP2OOPP24LFy60c88911avXm3Tp0+3iooKmzp1auY49UV+k3cfu5SWllqHDh2ipxz19fXRSla0Dd/61rdsyZIl9vzzz1v//v0z8fLycjMz9U0KrFy50urr62306NHWsWNH69ixoy1btsz+7d/+zTp27Jhpb/VFOvTt29eGDx+eFTvnnHNs8+bNZqa5kSbf/e537a677rKvf/3rdt5559mUKVPsO9/5jtXU1JiZ+uJkIe8WH506dbLRo0dbbW1tVry2ttbGjRvXTqX6ZBBCsNtvv90WL15szz33nFVWVmb9vbKy0srLy7P65vDhw7Zs2TL1TStz2WWX2dq1a2316tWZnzFjxtg3vvENW716tQ0ePFh9kSIXX3xx9LXzN9980wYNGmRmmhtpcuDAATvttOy3rg4dOmS+aqu+OEloR9n1uBz9qu2jjz4aXnvttTB9+vTQtWvX8O6777Z30U5pbr311lBSUhL+9Kc/hR07dmR+Dhw4kDlmzpw5oaSkJCxevDisXbs2XH/99foKW0oc+22XENQXafLyyy+Hjh07hvvvvz+89dZb4Ze//GUoKioKjz/+eOYY9Uc6TJ06NfTr1y/zVdvFixeH0tLScOedd2aOUV/kP3m5+AghhIceeigMGjQodOrUKYwaNSrzdU/RdpgZ/syfPz9zTHNzc7j33ntDeXl5KCwsDJdeemlYu3Zt+xX6E0Ry8aG+SJennnoqjBgxIhQWFoZhw4aFRx55JOvv6o90aGxsDHfccUcYOHBg6Ny5cxg8eHC45557wqFDhzLHqC/yn4IQQmjPJy9CCCGE+GSRd86HEEIIIU5ttPgQQgghRKpo8SGEEEKIVNHiQwghhBCposWHEEIIIVJFiw8hhBBCpIoWH0IIIYRIFS0+hBBCCJEqWnwIIYQQIlW0+BBCCCFEqmjxIYQQQohU+V/fjvH1DmzfyQAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Maximium Images for images_show: 8\n"
     ]
    }
   ],
   "source": [
    "for i, (images, labels) in enumerate(train_loader):\n",
    "    images_show(images,labels)\n",
    "    break"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "09c39d33-97f0-4f86-bbd3-cca534ecc15b",
   "metadata": {},
   "source": [
    "## 3. 优化器与损失函数"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "9e2cacb9-09d6-4617-8880-23cf3f09b51b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 优化器\n",
    "#optimizer = torch.optim.Adam(crnn.parameters(), lr=lr)\n",
    "optimizer = torch.optim.Adadelta(crnn.parameters())\n",
    "\n",
    "# 损失函数\n",
    "criterion = nn.CTCLoss()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a6c6fef8-d2f5-4882-935b-8af748d8c47a",
   "metadata": {},
   "source": [
    "## 4. 训练"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "018704ae-50d4-4551-a402-4b0462a30de4",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "65\n",
      "Epoch [1/30], Step [10/313], Loss: 3.4756155014038086\n",
      "Epoch [1/30], Step [20/313], Loss: 3.4512743949890137\n",
      "Epoch [1/30], Step [30/313], Loss: 3.388484001159668\n",
      "Epoch [1/30], Step [40/313], Loss: 3.3849270343780518\n",
      "Epoch [1/30], Step [50/313], Loss: 3.407365083694458\n",
      "Epoch [1/30], Step [60/313], Loss: 3.2823128700256348\n",
      "Epoch [1/30], Step [70/313], Loss: 3.1954331398010254\n",
      "Epoch [1/30], Step [80/313], Loss: 3.2822930812835693\n",
      "Epoch [1/30], Step [90/313], Loss: 3.251730442047119\n",
      "Epoch [1/30], Step [100/313], Loss: 3.1958558559417725\n",
      "Epoch [1/30], Step [110/313], Loss: 3.3246164321899414\n",
      "Epoch [1/30], Step [120/313], Loss: 3.316563606262207\n",
      "Epoch [1/30], Step [130/313], Loss: 3.168177604675293\n",
      "Epoch [1/30], Step [140/313], Loss: 3.2563834190368652\n",
      "Epoch [1/30], Step [150/313], Loss: 3.2450475692749023\n",
      "Epoch [1/30], Step [160/313], Loss: 3.294084072113037\n",
      "Epoch [1/30], Step [170/313], Loss: 3.1786892414093018\n",
      "Epoch [1/30], Step [180/313], Loss: 3.2703018188476562\n",
      "Epoch [1/30], Step [190/313], Loss: 3.20400071144104\n",
      "Epoch [1/30], Step [200/313], Loss: 3.0098719596862793\n",
      "Epoch [1/30], Step [210/313], Loss: 3.015619993209839\n",
      "Epoch [1/30], Step [220/313], Loss: 3.336141586303711\n",
      "Epoch [1/30], Step [230/313], Loss: 3.0203819274902344\n",
      "Epoch [1/30], Step [240/313], Loss: 2.9514355659484863\n",
      "Epoch [1/30], Step [250/313], Loss: 3.0576741695404053\n",
      "Epoch [1/30], Step [260/313], Loss: 3.0450124740600586\n",
      "Epoch [1/30], Step [270/313], Loss: 2.7352733612060547\n",
      "Epoch [1/30], Step [280/313], Loss: 2.9474284648895264\n",
      "Epoch [1/30], Step [290/313], Loss: 2.9289419651031494\n",
      "Epoch [1/30], Step [300/313], Loss: 2.9796009063720703\n",
      "Epoch [1/30], Step [310/313], Loss: 2.8893909454345703\n",
      "Epoch [2/30], Step [10/313], Loss: 2.537053108215332\n",
      "Epoch [2/30], Step [20/313], Loss: 2.699399948120117\n",
      "Epoch [2/30], Step [30/313], Loss: 2.7814102172851562\n",
      "Epoch [2/30], Step [40/313], Loss: 2.964231491088867\n",
      "Epoch [2/30], Step [50/313], Loss: 2.3765735626220703\n",
      "Epoch [2/30], Step [60/313], Loss: 2.3989577293395996\n",
      "Epoch [2/30], Step [70/313], Loss: 2.117370367050171\n",
      "Epoch [2/30], Step [80/313], Loss: 1.8633804321289062\n",
      "Epoch [2/30], Step [90/313], Loss: 2.1434032917022705\n",
      "Epoch [2/30], Step [100/313], Loss: 1.795850157737732\n",
      "Epoch [2/30], Step [110/313], Loss: 2.14856219291687\n",
      "Epoch [2/30], Step [120/313], Loss: 1.6398409605026245\n",
      "Epoch [2/30], Step [130/313], Loss: 1.527125597000122\n",
      "Epoch [2/30], Step [140/313], Loss: 1.4854307174682617\n",
      "Epoch [2/30], Step [150/313], Loss: 1.4930143356323242\n",
      "Epoch [2/30], Step [160/313], Loss: 1.2177485227584839\n",
      "Epoch [2/30], Step [170/313], Loss: 1.0547609329223633\n",
      "Epoch [2/30], Step [180/313], Loss: 0.7646936178207397\n",
      "Epoch [2/30], Step [190/313], Loss: 1.3117748498916626\n",
      "Epoch [2/30], Step [200/313], Loss: 0.5988591909408569\n",
      "Epoch [2/30], Step [210/313], Loss: 1.1176514625549316\n",
      "Epoch [2/30], Step [220/313], Loss: 0.7780730724334717\n",
      "Epoch [2/30], Step [230/313], Loss: 0.5135999321937561\n",
      "Epoch [2/30], Step [240/313], Loss: 0.9821073412895203\n",
      "Epoch [2/30], Step [250/313], Loss: 0.6666058301925659\n",
      "Epoch [2/30], Step [260/313], Loss: 0.7456468939781189\n",
      "Epoch [2/30], Step [270/313], Loss: 0.6311333179473877\n",
      "Epoch [2/30], Step [280/313], Loss: 0.5303908586502075\n",
      "Epoch [2/30], Step [290/313], Loss: 0.41763341426849365\n",
      "Epoch [2/30], Step [300/313], Loss: 0.3534128665924072\n",
      "Epoch [2/30], Step [310/313], Loss: 0.6243836283683777\n",
      "Epoch [3/30], Step [10/313], Loss: 0.41272228956222534\n",
      "Epoch [3/30], Step [20/313], Loss: 0.3634149730205536\n",
      "Epoch [3/30], Step [30/313], Loss: 0.5250187516212463\n",
      "Epoch [3/30], Step [40/313], Loss: 0.6313105821609497\n",
      "Epoch [3/30], Step [50/313], Loss: 0.3179798722267151\n",
      "Epoch [3/30], Step [60/313], Loss: 0.35706520080566406\n",
      "Epoch [3/30], Step [70/313], Loss: 0.2936643064022064\n",
      "Epoch [3/30], Step [80/313], Loss: 0.2721777856349945\n",
      "Epoch [3/30], Step [90/313], Loss: 0.39927002787590027\n",
      "Epoch [3/30], Step [100/313], Loss: 0.4575042724609375\n",
      "Epoch [3/30], Step [110/313], Loss: 0.3079488277435303\n",
      "Epoch [3/30], Step [120/313], Loss: 0.5478957891464233\n",
      "Epoch [3/30], Step [130/313], Loss: 0.30758318305015564\n",
      "Epoch [3/30], Step [140/313], Loss: 0.31976622343063354\n",
      "Epoch [3/30], Step [150/313], Loss: 0.22386851906776428\n",
      "Epoch [3/30], Step [160/313], Loss: 0.3315160870552063\n",
      "Epoch [3/30], Step [170/313], Loss: 0.44112661480903625\n",
      "Epoch [3/30], Step [180/313], Loss: 0.48069268465042114\n",
      "Epoch [3/30], Step [190/313], Loss: 0.24443835020065308\n",
      "Epoch [3/30], Step [200/313], Loss: 0.17972443997859955\n",
      "Epoch [3/30], Step [210/313], Loss: 0.3747038245201111\n",
      "Epoch [3/30], Step [220/313], Loss: 0.1473936140537262\n",
      "Epoch [3/30], Step [230/313], Loss: 0.2684433162212372\n",
      "Epoch [3/30], Step [240/313], Loss: 0.345661997795105\n",
      "Epoch [3/30], Step [250/313], Loss: 0.25406935811042786\n",
      "Epoch [3/30], Step [260/313], Loss: 0.158375084400177\n",
      "Epoch [3/30], Step [270/313], Loss: 0.31448179483413696\n",
      "Epoch [3/30], Step [280/313], Loss: 0.20775893330574036\n",
      "Epoch [3/30], Step [290/313], Loss: 0.39908266067504883\n",
      "Epoch [3/30], Step [300/313], Loss: 0.3601880669593811\n",
      "Epoch [3/30], Step [310/313], Loss: 0.2521408200263977\n",
      "Epoch [4/30], Step [10/313], Loss: 0.08542972803115845\n",
      "Epoch [4/30], Step [20/313], Loss: 0.23132894933223724\n",
      "Epoch [4/30], Step [30/313], Loss: 0.11297217011451721\n",
      "Epoch [4/30], Step [40/313], Loss: 0.1625399887561798\n",
      "Epoch [4/30], Step [50/313], Loss: 0.18167643249034882\n",
      "Epoch [4/30], Step [60/313], Loss: 0.2276604175567627\n",
      "Epoch [4/30], Step [70/313], Loss: 0.18280212581157684\n",
      "Epoch [4/30], Step [80/313], Loss: 0.212289959192276\n",
      "Epoch [4/30], Step [90/313], Loss: 0.14039835333824158\n",
      "Epoch [4/30], Step [100/313], Loss: 0.15205277502536774\n",
      "Epoch [4/30], Step [110/313], Loss: 0.1763436198234558\n",
      "Epoch [4/30], Step [120/313], Loss: 0.14651286602020264\n",
      "Epoch [4/30], Step [130/313], Loss: 0.19344739615917206\n",
      "Epoch [4/30], Step [140/313], Loss: 0.288939893245697\n",
      "Epoch [4/30], Step [150/313], Loss: 0.12853588163852692\n",
      "Epoch [4/30], Step [160/313], Loss: 0.1330496370792389\n",
      "Epoch [4/30], Step [170/313], Loss: 0.1395414173603058\n",
      "Epoch [4/30], Step [180/313], Loss: 0.21836934983730316\n",
      "Epoch [4/30], Step [190/313], Loss: 0.34643837809562683\n",
      "Epoch [4/30], Step [200/313], Loss: 0.36623361706733704\n",
      "Epoch [4/30], Step [210/313], Loss: 0.1072249785065651\n",
      "Epoch [4/30], Step [220/313], Loss: 0.11099988222122192\n",
      "Epoch [4/30], Step [230/313], Loss: 0.058067940175533295\n",
      "Epoch [4/30], Step [240/313], Loss: 0.18922965228557587\n",
      "Epoch [4/30], Step [250/313], Loss: 0.08734370768070221\n",
      "Epoch [4/30], Step [260/313], Loss: 0.08052287995815277\n",
      "Epoch [4/30], Step [270/313], Loss: 0.11816612631082535\n",
      "Epoch [4/30], Step [280/313], Loss: 0.1347126066684723\n",
      "Epoch [4/30], Step [290/313], Loss: 0.1418801248073578\n",
      "Epoch [4/30], Step [300/313], Loss: 0.1663600206375122\n",
      "Epoch [4/30], Step [310/313], Loss: 0.1263505071401596\n",
      "Epoch [5/30], Step [10/313], Loss: 0.07413765043020248\n",
      "Epoch [5/30], Step [20/313], Loss: 0.26614487171173096\n",
      "Epoch [5/30], Step [30/313], Loss: 0.04735477268695831\n",
      "Epoch [5/30], Step [40/313], Loss: 0.0783112496137619\n",
      "Epoch [5/30], Step [50/313], Loss: 0.06623736023902893\n",
      "Epoch [5/30], Step [60/313], Loss: 0.11622191965579987\n",
      "Epoch [5/30], Step [70/313], Loss: 0.07555154711008072\n",
      "Epoch [5/30], Step [80/313], Loss: 0.0961269736289978\n",
      "Epoch [5/30], Step [90/313], Loss: 0.048632197082042694\n",
      "Epoch [5/30], Step [100/313], Loss: 0.0247429721057415\n",
      "Epoch [5/30], Step [110/313], Loss: 0.09567306190729141\n",
      "Epoch [5/30], Step [120/313], Loss: 0.07566327601671219\n",
      "Epoch [5/30], Step [130/313], Loss: 0.08290019631385803\n",
      "Epoch [5/30], Step [140/313], Loss: 0.049928564578294754\n",
      "Epoch [5/30], Step [150/313], Loss: 0.03548017889261246\n",
      "Epoch [5/30], Step [160/313], Loss: 0.026962406933307648\n",
      "Epoch [5/30], Step [170/313], Loss: 0.16595947742462158\n",
      "Epoch [5/30], Step [180/313], Loss: 0.07110541313886642\n",
      "Epoch [5/30], Step [190/313], Loss: 0.05958889424800873\n",
      "Epoch [5/30], Step [200/313], Loss: 0.13461554050445557\n",
      "Epoch [5/30], Step [210/313], Loss: 0.19183512032032013\n",
      "Epoch [5/30], Step [220/313], Loss: 0.08407632261514664\n",
      "Epoch [5/30], Step [230/313], Loss: 0.10742630064487457\n",
      "Epoch [5/30], Step [240/313], Loss: 0.13727372884750366\n",
      "Epoch [5/30], Step [250/313], Loss: 0.08160965144634247\n",
      "Epoch [5/30], Step [260/313], Loss: 0.11485307663679123\n",
      "Epoch [5/30], Step [270/313], Loss: 0.11764675378799438\n",
      "Epoch [5/30], Step [280/313], Loss: 0.07662233710289001\n",
      "Epoch [5/30], Step [290/313], Loss: 0.057289108633995056\n",
      "Epoch [5/30], Step [300/313], Loss: 0.0324222706258297\n",
      "Epoch [5/30], Step [310/313], Loss: 0.08624458312988281\n",
      "Epoch [6/30], Step [10/313], Loss: 0.1246640682220459\n",
      "Epoch [6/30], Step [20/313], Loss: 0.08544090390205383\n",
      "Epoch [6/30], Step [30/313], Loss: 0.020466135814785957\n",
      "Epoch [6/30], Step [40/313], Loss: 0.034838393330574036\n",
      "Epoch [6/30], Step [50/313], Loss: 0.08539838343858719\n",
      "Epoch [6/30], Step [60/313], Loss: 0.039343059062957764\n",
      "Epoch [6/30], Step [70/313], Loss: 0.08325247466564178\n",
      "Epoch [6/30], Step [80/313], Loss: 0.08146393299102783\n",
      "Epoch [6/30], Step [90/313], Loss: 0.022304493933916092\n",
      "Epoch [6/30], Step [100/313], Loss: 0.09582893550395966\n",
      "Epoch [6/30], Step [110/313], Loss: 0.0485217347741127\n",
      "Epoch [6/30], Step [120/313], Loss: 0.05828748270869255\n",
      "Epoch [6/30], Step [130/313], Loss: 0.018124520778656006\n",
      "Epoch [6/30], Step [140/313], Loss: 0.045695193111896515\n",
      "Epoch [6/30], Step [150/313], Loss: 0.02714885026216507\n",
      "Epoch [6/30], Step [160/313], Loss: 0.10809831321239471\n",
      "Epoch [6/30], Step [170/313], Loss: 0.024077415466308594\n",
      "Epoch [6/30], Step [180/313], Loss: 0.020233333110809326\n",
      "Epoch [6/30], Step [190/313], Loss: 0.04919932782649994\n",
      "Epoch [6/30], Step [200/313], Loss: 0.036010462790727615\n",
      "Epoch [6/30], Step [210/313], Loss: 0.06872093677520752\n",
      "Epoch [6/30], Step [220/313], Loss: 0.07302659749984741\n",
      "Epoch [6/30], Step [230/313], Loss: 0.010087333619594574\n",
      "Epoch [6/30], Step [240/313], Loss: 0.014652468264102936\n",
      "Epoch [6/30], Step [250/313], Loss: 0.09264696389436722\n",
      "Epoch [6/30], Step [260/313], Loss: 0.06534775346517563\n",
      "Epoch [6/30], Step [270/313], Loss: 0.018160227686166763\n",
      "Epoch [6/30], Step [280/313], Loss: 0.07750639319419861\n",
      "Epoch [6/30], Step [290/313], Loss: 0.07620817422866821\n",
      "Epoch [6/30], Step [300/313], Loss: 0.05151793360710144\n",
      "Epoch [6/30], Step [310/313], Loss: 0.03562059998512268\n",
      "Epoch [7/30], Step [10/313], Loss: 0.013769972138106823\n",
      "Epoch [7/30], Step [20/313], Loss: 0.01574706844985485\n",
      "Epoch [7/30], Step [30/313], Loss: 0.021523619070649147\n",
      "Epoch [7/30], Step [40/313], Loss: 0.04815588518977165\n",
      "Epoch [7/30], Step [50/313], Loss: 0.05576574429869652\n",
      "Epoch [7/30], Step [60/313], Loss: 0.023262370377779007\n",
      "Epoch [7/30], Step [70/313], Loss: 0.028494499623775482\n",
      "Epoch [7/30], Step [80/313], Loss: 0.039897918701171875\n",
      "Epoch [7/30], Step [90/313], Loss: 0.03564365208148956\n",
      "Epoch [7/30], Step [100/313], Loss: 0.019276706501841545\n",
      "Epoch [7/30], Step [110/313], Loss: 0.021360307931900024\n",
      "Epoch [7/30], Step [120/313], Loss: 0.03813501074910164\n",
      "Epoch [7/30], Step [130/313], Loss: 0.03318513557314873\n",
      "Epoch [7/30], Step [140/313], Loss: 0.05194530636072159\n",
      "Epoch [7/30], Step [150/313], Loss: 0.027724120765924454\n",
      "Epoch [7/30], Step [160/313], Loss: 0.03392777964472771\n",
      "Epoch [7/30], Step [170/313], Loss: 0.03257615119218826\n",
      "Epoch [7/30], Step [180/313], Loss: 0.017113249748945236\n",
      "Epoch [7/30], Step [190/313], Loss: 0.028723498806357384\n",
      "Epoch [7/30], Step [200/313], Loss: 0.03293512016534805\n",
      "Epoch [7/30], Step [210/313], Loss: 0.07762615382671356\n",
      "Epoch [7/30], Step [220/313], Loss: 0.004779702052474022\n",
      "Epoch [7/30], Step [230/313], Loss: 0.01947595179080963\n",
      "Epoch [7/30], Step [240/313], Loss: 0.031023699790239334\n",
      "Epoch [7/30], Step [250/313], Loss: 0.01716390810906887\n",
      "Epoch [7/30], Step [260/313], Loss: 0.02990640327334404\n",
      "Epoch [7/30], Step [270/313], Loss: 0.02575651928782463\n",
      "Epoch [7/30], Step [280/313], Loss: 0.01762041263282299\n",
      "Epoch [7/30], Step [290/313], Loss: 0.02411014586687088\n",
      "Epoch [7/30], Step [300/313], Loss: 0.04555368423461914\n",
      "Epoch [7/30], Step [310/313], Loss: 0.04781798645853996\n",
      "Epoch [8/30], Step [10/313], Loss: 0.024726178497076035\n",
      "Epoch [8/30], Step [20/313], Loss: 0.012763464823365211\n",
      "Epoch [8/30], Step [30/313], Loss: 0.007188020274043083\n",
      "Epoch [8/30], Step [40/313], Loss: 0.006685336120426655\n",
      "Epoch [8/30], Step [50/313], Loss: 0.005187585949897766\n",
      "Epoch [8/30], Step [60/313], Loss: 0.02222178690135479\n",
      "Epoch [8/30], Step [70/313], Loss: 0.006547600030899048\n",
      "Epoch [8/30], Step [80/313], Loss: 0.0060340771451592445\n",
      "Epoch [8/30], Step [90/313], Loss: 0.02824201062321663\n",
      "Epoch [8/30], Step [100/313], Loss: 0.022185254842042923\n",
      "Epoch [8/30], Step [110/313], Loss: 0.005666364450007677\n",
      "Epoch [8/30], Step [120/313], Loss: 0.04002898186445236\n",
      "Epoch [8/30], Step [130/313], Loss: 0.019501322880387306\n",
      "Epoch [8/30], Step [140/313], Loss: 0.08075112104415894\n",
      "Epoch [8/30], Step [150/313], Loss: 0.02067490853369236\n",
      "Epoch [8/30], Step [160/313], Loss: 0.087003692984581\n",
      "Epoch [8/30], Step [170/313], Loss: 0.027343010529875755\n",
      "Epoch [8/30], Step [180/313], Loss: 0.002856450853869319\n",
      "Epoch [8/30], Step [190/313], Loss: 0.026761367917060852\n",
      "Epoch [8/30], Step [200/313], Loss: 0.07579603046178818\n",
      "Epoch [8/30], Step [210/313], Loss: 0.07361599057912827\n",
      "Epoch [8/30], Step [220/313], Loss: 0.06976528465747833\n",
      "Epoch [8/30], Step [230/313], Loss: 0.021572982892394066\n",
      "Epoch [8/30], Step [240/313], Loss: 0.025206351652741432\n",
      "Epoch [8/30], Step [250/313], Loss: 0.018072672188282013\n",
      "Epoch [8/30], Step [260/313], Loss: 0.08741290867328644\n",
      "Epoch [8/30], Step [270/313], Loss: 0.028597379103302956\n",
      "Epoch [8/30], Step [280/313], Loss: 0.10492199659347534\n",
      "Epoch [8/30], Step [290/313], Loss: 0.010160986334085464\n",
      "Epoch [8/30], Step [300/313], Loss: 0.016492392867803574\n",
      "Epoch [8/30], Step [310/313], Loss: 0.09230532497167587\n",
      "Epoch [9/30], Step [10/313], Loss: 0.024626201018691063\n",
      "Epoch [9/30], Step [20/313], Loss: 0.009747136384248734\n",
      "Epoch [9/30], Step [30/313], Loss: 0.016757097095251083\n",
      "Epoch [9/30], Step [40/313], Loss: 0.004830983467400074\n",
      "Epoch [9/30], Step [50/313], Loss: 0.03179458528757095\n",
      "Epoch [9/30], Step [60/313], Loss: 0.008365372195839882\n",
      "Epoch [9/30], Step [70/313], Loss: 0.007673070300370455\n",
      "Epoch [9/30], Step [80/313], Loss: 0.008451328612864017\n",
      "Epoch [9/30], Step [90/313], Loss: 0.022977320477366447\n",
      "Epoch [9/30], Step [100/313], Loss: 0.012225007638335228\n",
      "Epoch [9/30], Step [110/313], Loss: 0.010517828166484833\n",
      "Epoch [9/30], Step [120/313], Loss: 0.02111656591296196\n",
      "Epoch [9/30], Step [130/313], Loss: 0.010590795427560806\n",
      "Epoch [9/30], Step [140/313], Loss: 0.037341564893722534\n",
      "Epoch [9/30], Step [150/313], Loss: 0.01456386037170887\n",
      "Epoch [9/30], Step [160/313], Loss: 0.00830178614705801\n",
      "Epoch [9/30], Step [170/313], Loss: 0.01594408228993416\n",
      "Epoch [9/30], Step [180/313], Loss: 0.013327674940228462\n",
      "Epoch [9/30], Step [190/313], Loss: 0.01800786890089512\n",
      "Epoch [9/30], Step [200/313], Loss: 0.008313998579978943\n",
      "Epoch [9/30], Step [210/313], Loss: 0.027636872604489326\n",
      "Epoch [9/30], Step [220/313], Loss: 0.025663673877716064\n",
      "Epoch [9/30], Step [230/313], Loss: 0.00446098018437624\n",
      "Epoch [9/30], Step [240/313], Loss: 0.0030853431671857834\n",
      "Epoch [9/30], Step [250/313], Loss: 0.015224932692945004\n",
      "Epoch [9/30], Step [260/313], Loss: 0.013990222476422787\n",
      "Epoch [9/30], Step [270/313], Loss: 0.005249008536338806\n",
      "Epoch [9/30], Step [280/313], Loss: 0.020915646106004715\n",
      "Epoch [9/30], Step [290/313], Loss: 0.029975729063153267\n",
      "Epoch [9/30], Step [300/313], Loss: 0.02178576961159706\n",
      "Epoch [9/30], Step [310/313], Loss: 0.00688173295930028\n",
      "Epoch [10/30], Step [10/313], Loss: 0.0112040089443326\n",
      "Epoch [10/30], Step [20/313], Loss: 0.012617247179150581\n",
      "Epoch [10/30], Step [30/313], Loss: 0.02209964022040367\n",
      "Epoch [10/30], Step [40/313], Loss: 0.004680768586695194\n",
      "Epoch [10/30], Step [50/313], Loss: 0.004095830954611301\n",
      "Epoch [10/30], Step [60/313], Loss: 0.005446525290608406\n",
      "Epoch [10/30], Step [70/313], Loss: 0.007967228069901466\n",
      "Epoch [10/30], Step [80/313], Loss: 0.004748117178678513\n",
      "Epoch [10/30], Step [90/313], Loss: 0.007501916028559208\n",
      "Epoch [10/30], Step [100/313], Loss: 0.0016614743508398533\n",
      "Epoch [10/30], Step [110/313], Loss: 0.0073442235589027405\n",
      "Epoch [10/30], Step [120/313], Loss: 0.01897839829325676\n",
      "Epoch [10/30], Step [130/313], Loss: 0.001753626624122262\n",
      "Epoch [10/30], Step [140/313], Loss: 0.002661836799234152\n",
      "Epoch [10/30], Step [150/313], Loss: 0.0067362659610807896\n",
      "Epoch [10/30], Step [160/313], Loss: 0.04440152272582054\n",
      "Epoch [10/30], Step [170/313], Loss: 0.00446053221821785\n",
      "Epoch [10/30], Step [180/313], Loss: 0.0033341068774461746\n",
      "Epoch [10/30], Step [190/313], Loss: 0.005700664594769478\n",
      "Epoch [10/30], Step [200/313], Loss: 0.011818531900644302\n",
      "Epoch [10/30], Step [210/313], Loss: 0.02756347507238388\n",
      "Epoch [10/30], Step [220/313], Loss: 0.033969610929489136\n",
      "Epoch [10/30], Step [230/313], Loss: 0.002356700599193573\n",
      "Epoch [10/30], Step [240/313], Loss: 0.0030781652312725782\n",
      "Epoch [10/30], Step [250/313], Loss: 0.0022394671104848385\n",
      "Epoch [10/30], Step [260/313], Loss: 0.011035263538360596\n",
      "Epoch [10/30], Step [270/313], Loss: 0.005411163903772831\n",
      "Epoch [10/30], Step [280/313], Loss: 0.022031038999557495\n",
      "Epoch [10/30], Step [290/313], Loss: 0.003905423916876316\n",
      "Epoch [10/30], Step [300/313], Loss: 0.013191738165915012\n",
      "Epoch [10/30], Step [310/313], Loss: 0.005261418409645557\n",
      "Epoch [11/30], Step [10/313], Loss: 0.0029919249936938286\n",
      "Epoch [11/30], Step [20/313], Loss: 0.0020012417808175087\n",
      "Epoch [11/30], Step [30/313], Loss: 0.014432751573622227\n",
      "Epoch [11/30], Step [40/313], Loss: 0.002290239091962576\n",
      "Epoch [11/30], Step [50/313], Loss: 0.016201263293623924\n",
      "Epoch [11/30], Step [60/313], Loss: 0.002734797541052103\n",
      "Epoch [11/30], Step [70/313], Loss: 0.0068087391555309296\n",
      "Epoch [11/30], Step [80/313], Loss: 0.00415307329967618\n",
      "Epoch [11/30], Step [90/313], Loss: 0.005097340792417526\n",
      "Epoch [11/30], Step [100/313], Loss: 0.002423526020720601\n",
      "Epoch [11/30], Step [110/313], Loss: 0.009634503163397312\n",
      "Epoch [11/30], Step [120/313], Loss: 0.004445008933544159\n",
      "Epoch [11/30], Step [130/313], Loss: 0.004184530582278967\n",
      "Epoch [11/30], Step [140/313], Loss: 0.013655390590429306\n",
      "Epoch [11/30], Step [150/313], Loss: 0.00904710590839386\n",
      "Epoch [11/30], Step [160/313], Loss: 0.0011590534122660756\n",
      "Epoch [11/30], Step [170/313], Loss: 0.004501918330788612\n",
      "Epoch [11/30], Step [180/313], Loss: 0.01060808077454567\n",
      "Epoch [11/30], Step [190/313], Loss: 0.0013250804040580988\n",
      "Epoch [11/30], Step [200/313], Loss: 0.004983981139957905\n",
      "Epoch [11/30], Step [210/313], Loss: 0.0015056191477924585\n",
      "Epoch [11/30], Step [220/313], Loss: 0.007760481908917427\n",
      "Epoch [11/30], Step [230/313], Loss: 0.037708964198827744\n",
      "Epoch [11/30], Step [240/313], Loss: 0.02881431393325329\n",
      "Epoch [11/30], Step [250/313], Loss: 0.014116638340055943\n",
      "Epoch [11/30], Step [260/313], Loss: 0.008036964572966099\n",
      "Epoch [11/30], Step [270/313], Loss: 0.002667809836566448\n",
      "Epoch [11/30], Step [280/313], Loss: 0.010394190438091755\n",
      "Epoch [11/30], Step [290/313], Loss: 0.0033000572584569454\n",
      "Epoch [11/30], Step [300/313], Loss: 0.00442512147128582\n",
      "Epoch [11/30], Step [310/313], Loss: 0.001800642698071897\n",
      "Epoch [12/30], Step [10/313], Loss: 0.01376696489751339\n",
      "Epoch [12/30], Step [20/313], Loss: 0.006268342491239309\n",
      "Epoch [12/30], Step [30/313], Loss: 0.008225628174841404\n",
      "Epoch [12/30], Step [40/313], Loss: 0.002142513170838356\n",
      "Epoch [12/30], Step [50/313], Loss: 0.0027094692923128605\n",
      "Epoch [12/30], Step [60/313], Loss: 0.002189655089750886\n",
      "Epoch [12/30], Step [70/313], Loss: 0.0006485742051154375\n",
      "Epoch [12/30], Step [80/313], Loss: 0.0034646005369722843\n",
      "Epoch [12/30], Step [90/313], Loss: 0.004037179984152317\n",
      "Epoch [12/30], Step [100/313], Loss: 0.004525944124907255\n",
      "Epoch [12/30], Step [110/313], Loss: 0.005315289832651615\n",
      "Epoch [12/30], Step [120/313], Loss: 0.010672066360712051\n",
      "Epoch [12/30], Step [130/313], Loss: 0.002728663617745042\n",
      "Epoch [12/30], Step [140/313], Loss: 0.004912044852972031\n",
      "Epoch [12/30], Step [150/313], Loss: 0.015227064490318298\n",
      "Epoch [12/30], Step [160/313], Loss: 0.002770309103652835\n",
      "Epoch [12/30], Step [170/313], Loss: 0.006167515646666288\n",
      "Epoch [12/30], Step [180/313], Loss: 0.00553114153444767\n",
      "Epoch [12/30], Step [190/313], Loss: 0.0064948988147079945\n",
      "Epoch [12/30], Step [200/313], Loss: 0.005484543275088072\n",
      "Epoch [12/30], Step [210/313], Loss: 0.003335394198074937\n",
      "Epoch [12/30], Step [220/313], Loss: 0.0010070711141452193\n",
      "Epoch [12/30], Step [230/313], Loss: 0.002416423289105296\n",
      "Epoch [12/30], Step [240/313], Loss: 0.002891355426982045\n",
      "Epoch [12/30], Step [250/313], Loss: 0.0010074833407998085\n",
      "Epoch [12/30], Step [260/313], Loss: 0.005524973385035992\n",
      "Epoch [12/30], Step [270/313], Loss: 0.0015384969301521778\n",
      "Epoch [12/30], Step [280/313], Loss: 0.0021011666394770145\n",
      "Epoch [12/30], Step [290/313], Loss: 0.0017529197502881289\n",
      "Epoch [12/30], Step [300/313], Loss: 0.020917724817991257\n",
      "Epoch [12/30], Step [310/313], Loss: 0.0040904381312429905\n",
      "Epoch [13/30], Step [10/313], Loss: 0.001875840942375362\n",
      "Epoch [13/30], Step [20/313], Loss: 0.0027121736202389\n",
      "Epoch [13/30], Step [30/313], Loss: 0.018022485077381134\n",
      "Epoch [13/30], Step [40/313], Loss: 0.00459736492484808\n",
      "Epoch [13/30], Step [50/313], Loss: 0.0021363261621445417\n",
      "Epoch [13/30], Step [60/313], Loss: 0.0011773793958127499\n",
      "Epoch [13/30], Step [70/313], Loss: 0.004259852226823568\n",
      "Epoch [13/30], Step [80/313], Loss: 0.0015576836885884404\n",
      "Epoch [13/30], Step [90/313], Loss: 0.0012672839220613241\n",
      "Epoch [13/30], Step [100/313], Loss: 0.0027132637333124876\n",
      "Epoch [13/30], Step [110/313], Loss: 0.0009601903147995472\n",
      "Epoch [13/30], Step [120/313], Loss: 0.0010577741777524352\n",
      "Epoch [13/30], Step [130/313], Loss: 0.00391749432310462\n",
      "Epoch [13/30], Step [140/313], Loss: 0.0005563828162848949\n",
      "Epoch [13/30], Step [150/313], Loss: 0.003087173216044903\n",
      "Epoch [13/30], Step [160/313], Loss: 0.0032291340176016092\n",
      "Epoch [13/30], Step [170/313], Loss: 0.000572222750633955\n",
      "Epoch [13/30], Step [180/313], Loss: 0.004583993926644325\n",
      "Epoch [13/30], Step [190/313], Loss: 0.0008514730725437403\n",
      "Epoch [13/30], Step [200/313], Loss: 0.003621084149926901\n",
      "Epoch [13/30], Step [210/313], Loss: 0.0005888264859095216\n",
      "Epoch [13/30], Step [220/313], Loss: 0.002249382669106126\n",
      "Epoch [13/30], Step [230/313], Loss: 0.002420254284515977\n",
      "Epoch [13/30], Step [240/313], Loss: 0.01716754585504532\n",
      "Epoch [13/30], Step [250/313], Loss: 0.004832625389099121\n",
      "Epoch [13/30], Step [260/313], Loss: 0.0027048795018345118\n",
      "Epoch [13/30], Step [270/313], Loss: 0.0006326374132186174\n",
      "Epoch [13/30], Step [280/313], Loss: 0.0016522575169801712\n",
      "Epoch [13/30], Step [290/313], Loss: 0.004547496326267719\n",
      "Epoch [13/30], Step [300/313], Loss: 0.003870235290378332\n",
      "Epoch [13/30], Step [310/313], Loss: 0.001558581250719726\n",
      "Epoch [14/30], Step [10/313], Loss: 0.2516152858734131\n",
      "Epoch [14/30], Step [20/313], Loss: 0.04652177542448044\n",
      "Epoch [14/30], Step [30/313], Loss: 0.005531970877200365\n",
      "Epoch [14/30], Step [40/313], Loss: 0.002488279715180397\n",
      "Epoch [14/30], Step [50/313], Loss: 0.0012278137728571892\n",
      "Epoch [14/30], Step [60/313], Loss: 0.001725470763631165\n",
      "Epoch [14/30], Step [70/313], Loss: 0.002430417574942112\n",
      "Epoch [14/30], Step [80/313], Loss: 0.0007560845697298646\n",
      "Epoch [14/30], Step [90/313], Loss: 0.005811283364892006\n",
      "Epoch [14/30], Step [100/313], Loss: 0.01947508379817009\n",
      "Epoch [14/30], Step [110/313], Loss: 0.004951149225234985\n",
      "Epoch [14/30], Step [120/313], Loss: 0.0016351902158930898\n",
      "Epoch [14/30], Step [130/313], Loss: 0.0010284031741321087\n",
      "Epoch [14/30], Step [140/313], Loss: 0.005473350174725056\n",
      "Epoch [14/30], Step [150/313], Loss: 0.015242570079863071\n",
      "Epoch [14/30], Step [160/313], Loss: 0.0016592468600720167\n",
      "Epoch [14/30], Step [170/313], Loss: 0.0023968135938048363\n",
      "Epoch [14/30], Step [180/313], Loss: 0.005282150581479073\n",
      "Epoch [14/30], Step [190/313], Loss: 0.0031614552717655897\n",
      "Epoch [14/30], Step [200/313], Loss: 0.0035433373413980007\n",
      "Epoch [14/30], Step [210/313], Loss: 0.0021199509501457214\n",
      "Epoch [14/30], Step [220/313], Loss: 0.0010799120645970106\n",
      "Epoch [14/30], Step [230/313], Loss: 0.007874242961406708\n",
      "Epoch [14/30], Step [240/313], Loss: 0.002417711541056633\n",
      "Epoch [14/30], Step [250/313], Loss: 0.0009050248190760612\n",
      "Epoch [14/30], Step [260/313], Loss: 0.002788050100207329\n",
      "Epoch [14/30], Step [270/313], Loss: 0.002715532435104251\n",
      "Epoch [14/30], Step [280/313], Loss: 0.001834469148889184\n",
      "Epoch [14/30], Step [290/313], Loss: 0.014105550944805145\n",
      "Epoch [14/30], Step [300/313], Loss: 0.010613206773996353\n",
      "Epoch [14/30], Step [310/313], Loss: 0.031238727271556854\n",
      "Epoch [15/30], Step [10/313], Loss: 0.0007882853969931602\n",
      "Epoch [15/30], Step [20/313], Loss: 0.013988224789500237\n",
      "Epoch [15/30], Step [30/313], Loss: 0.004829857032746077\n",
      "Epoch [15/30], Step [40/313], Loss: 0.00583559600636363\n",
      "Epoch [15/30], Step [50/313], Loss: 0.013583083637058735\n",
      "Epoch [15/30], Step [60/313], Loss: 0.095360666513443\n",
      "Epoch [15/30], Step [70/313], Loss: 0.01059223897755146\n",
      "Epoch [15/30], Step [80/313], Loss: 0.022290553897619247\n",
      "Epoch [15/30], Step [90/313], Loss: 0.007762265391647816\n",
      "Epoch [15/30], Step [100/313], Loss: 0.0022235785145312548\n",
      "Epoch [15/30], Step [110/313], Loss: 0.011482669040560722\n",
      "Epoch [15/30], Step [120/313], Loss: 0.0013181744143366814\n",
      "Epoch [15/30], Step [130/313], Loss: 0.0026214448735117912\n",
      "Epoch [15/30], Step [140/313], Loss: 0.015565788373351097\n",
      "Epoch [15/30], Step [150/313], Loss: 0.0031943642534315586\n",
      "Epoch [15/30], Step [160/313], Loss: 0.006838412024080753\n",
      "Epoch [15/30], Step [170/313], Loss: 0.0019760210998356342\n",
      "Epoch [15/30], Step [180/313], Loss: 0.0010808113729581237\n",
      "Epoch [15/30], Step [190/313], Loss: 0.001300798961892724\n",
      "Epoch [15/30], Step [200/313], Loss: 0.0013378339353948832\n",
      "Epoch [15/30], Step [210/313], Loss: 0.00201842631213367\n",
      "Epoch [15/30], Step [220/313], Loss: 0.0003606000682339072\n",
      "Epoch [15/30], Step [230/313], Loss: 0.0027857471723109484\n",
      "Epoch [15/30], Step [240/313], Loss: 0.008627374656498432\n",
      "Epoch [15/30], Step [250/313], Loss: 0.008818966336548328\n",
      "Epoch [15/30], Step [260/313], Loss: 0.0009933005785569549\n",
      "Epoch [15/30], Step [270/313], Loss: 0.0022163190878927708\n",
      "Epoch [15/30], Step [280/313], Loss: 0.03508758172392845\n",
      "Epoch [15/30], Step [290/313], Loss: 0.051675666123628616\n",
      "Epoch [15/30], Step [300/313], Loss: 0.021193940192461014\n",
      "Epoch [15/30], Step [310/313], Loss: 0.011664892546832561\n",
      "Epoch [16/30], Step [10/313], Loss: 0.006526912562549114\n",
      "Epoch [16/30], Step [20/313], Loss: 0.0028408984653651714\n",
      "Epoch [16/30], Step [30/313], Loss: 0.021016038954257965\n",
      "Epoch [16/30], Step [40/313], Loss: 0.013479169458150864\n",
      "Epoch [16/30], Step [50/313], Loss: 0.014882362447679043\n",
      "Epoch [16/30], Step [60/313], Loss: 0.0038716942071914673\n",
      "Epoch [16/30], Step [70/313], Loss: 0.004287414252758026\n",
      "Epoch [16/30], Step [80/313], Loss: 0.0035079335793852806\n",
      "Epoch [16/30], Step [90/313], Loss: 0.0015232829609885812\n",
      "Epoch [16/30], Step [100/313], Loss: 0.010040495544672012\n",
      "Epoch [16/30], Step [110/313], Loss: 0.003925842698663473\n",
      "Epoch [16/30], Step [120/313], Loss: 0.01081059593707323\n",
      "Epoch [16/30], Step [130/313], Loss: 0.0009692262392491102\n",
      "Epoch [16/30], Step [140/313], Loss: 0.0012825592420995235\n",
      "Epoch [16/30], Step [150/313], Loss: 0.006270278245210648\n",
      "Epoch [16/30], Step [160/313], Loss: 0.0023394792806357145\n",
      "Epoch [16/30], Step [170/313], Loss: 0.0029960840474814177\n",
      "Epoch [16/30], Step [180/313], Loss: 0.0025097473990172148\n",
      "Epoch [16/30], Step [190/313], Loss: 0.005070335231721401\n",
      "Epoch [16/30], Step [200/313], Loss: 0.0063769882544875145\n",
      "Epoch [16/30], Step [210/313], Loss: 0.0003996636369265616\n",
      "Epoch [16/30], Step [220/313], Loss: 0.015230769291520119\n",
      "Epoch [16/30], Step [230/313], Loss: 0.0026473645120859146\n",
      "Epoch [16/30], Step [240/313], Loss: 0.0016280034324154258\n",
      "Epoch [16/30], Step [250/313], Loss: 0.011121454648673534\n",
      "Epoch [16/30], Step [260/313], Loss: 0.012853581458330154\n",
      "Epoch [16/30], Step [270/313], Loss: 0.001933576539158821\n",
      "Epoch [16/30], Step [280/313], Loss: 0.001379550201818347\n",
      "Epoch [16/30], Step [290/313], Loss: 0.004952574148774147\n",
      "Epoch [16/30], Step [300/313], Loss: 0.0034933655988425016\n",
      "Epoch [16/30], Step [310/313], Loss: 0.0012554738204926252\n",
      "Epoch [17/30], Step [10/313], Loss: 0.0017144621815532446\n",
      "Epoch [17/30], Step [20/313], Loss: 0.0011126295430585742\n",
      "Epoch [17/30], Step [30/313], Loss: 0.0017920206300914288\n",
      "Epoch [17/30], Step [40/313], Loss: 0.0026039707008749247\n",
      "Epoch [17/30], Step [50/313], Loss: 0.0033683294896036386\n",
      "Epoch [17/30], Step [60/313], Loss: 0.0009376167436130345\n",
      "Epoch [17/30], Step [70/313], Loss: 0.0005533398361876607\n",
      "Epoch [17/30], Step [80/313], Loss: 0.0013075119350105524\n",
      "Epoch [17/30], Step [90/313], Loss: 0.0018830334302037954\n",
      "Epoch [17/30], Step [100/313], Loss: 0.0004971680464223027\n",
      "Epoch [17/30], Step [110/313], Loss: 0.0007653090870007873\n",
      "Epoch [17/30], Step [120/313], Loss: 0.0006789776962250471\n",
      "Epoch [17/30], Step [130/313], Loss: 0.0023840083740651608\n",
      "Epoch [17/30], Step [140/313], Loss: 0.019704923033714294\n",
      "Epoch [17/30], Step [150/313], Loss: 0.0015654296148568392\n",
      "Epoch [17/30], Step [160/313], Loss: 0.0013441017363220453\n",
      "Epoch [17/30], Step [170/313], Loss: 0.000715950271114707\n",
      "Epoch [17/30], Step [180/313], Loss: 0.0019750669598579407\n",
      "Epoch [17/30], Step [190/313], Loss: 0.0017121679848060012\n",
      "Epoch [17/30], Step [200/313], Loss: 0.002037780126556754\n",
      "Epoch [17/30], Step [210/313], Loss: 0.0019098478369414806\n",
      "Epoch [17/30], Step [220/313], Loss: 0.0010559840593487024\n",
      "Epoch [17/30], Step [230/313], Loss: 0.0014104522997513413\n",
      "Epoch [17/30], Step [240/313], Loss: 0.00019538257038220763\n",
      "Epoch [17/30], Step [250/313], Loss: 0.0031522931531071663\n",
      "Epoch [17/30], Step [260/313], Loss: 0.0005015234928578138\n",
      "Epoch [17/30], Step [270/313], Loss: 0.001170352683402598\n",
      "Epoch [17/30], Step [280/313], Loss: 0.0017845669062808156\n",
      "Epoch [17/30], Step [290/313], Loss: 0.000590215902775526\n",
      "Epoch [17/30], Step [300/313], Loss: 0.0006593961734324694\n",
      "Epoch [17/30], Step [310/313], Loss: 0.0018056248081848025\n",
      "Epoch [18/30], Step [10/313], Loss: 0.0016606042627245188\n",
      "Epoch [18/30], Step [20/313], Loss: 0.0033490490168333054\n",
      "Epoch [18/30], Step [30/313], Loss: 0.006542292423546314\n",
      "Epoch [18/30], Step [40/313], Loss: 0.003088164608925581\n",
      "Epoch [18/30], Step [50/313], Loss: 0.00871807150542736\n",
      "Epoch [18/30], Step [60/313], Loss: 0.0006337630329653621\n",
      "Epoch [18/30], Step [70/313], Loss: 0.006622693967074156\n",
      "Epoch [18/30], Step [80/313], Loss: 0.0006666436092928052\n",
      "Epoch [18/30], Step [90/313], Loss: 0.021386750042438507\n",
      "Epoch [18/30], Step [100/313], Loss: 0.003288431791588664\n",
      "Epoch [18/30], Step [110/313], Loss: 0.002214284148067236\n",
      "Epoch [18/30], Step [120/313], Loss: 0.009589701890945435\n",
      "Epoch [18/30], Step [130/313], Loss: 0.0033355194609612226\n",
      "Epoch [18/30], Step [140/313], Loss: 0.005349620245397091\n",
      "Epoch [18/30], Step [150/313], Loss: 0.0011909925378859043\n",
      "Epoch [18/30], Step [160/313], Loss: 0.011505847796797752\n",
      "Epoch [18/30], Step [170/313], Loss: 0.0011411455925554037\n",
      "Epoch [18/30], Step [180/313], Loss: 0.0022304269950836897\n",
      "Epoch [18/30], Step [190/313], Loss: 0.0022100242786109447\n",
      "Epoch [18/30], Step [200/313], Loss: 0.00042238234891556203\n",
      "Epoch [18/30], Step [210/313], Loss: 0.0008470473112538457\n",
      "Epoch [18/30], Step [220/313], Loss: 0.011797569692134857\n",
      "Epoch [18/30], Step [230/313], Loss: 0.004206890240311623\n",
      "Epoch [18/30], Step [240/313], Loss: 0.0007628645980730653\n",
      "Epoch [18/30], Step [250/313], Loss: 0.0007857754826545715\n",
      "Epoch [18/30], Step [260/313], Loss: 0.000507944670971483\n",
      "Epoch [18/30], Step [270/313], Loss: 0.0005738630425184965\n",
      "Epoch [18/30], Step [280/313], Loss: 0.0004262597649358213\n",
      "Epoch [18/30], Step [290/313], Loss: 0.013170238584280014\n",
      "Epoch [18/30], Step [300/313], Loss: 0.004864540882408619\n",
      "Epoch [18/30], Step [310/313], Loss: 0.0030310451984405518\n",
      "Epoch [19/30], Step [10/313], Loss: 0.001254493952728808\n",
      "Epoch [19/30], Step [20/313], Loss: 0.0034274994395673275\n",
      "Epoch [19/30], Step [30/313], Loss: 0.0009821192361414433\n",
      "Epoch [19/30], Step [40/313], Loss: 0.0006891567027196288\n",
      "Epoch [19/30], Step [50/313], Loss: 0.0019659900572150946\n",
      "Epoch [19/30], Step [60/313], Loss: 0.0006876468542031944\n",
      "Epoch [19/30], Step [70/313], Loss: 0.0017637989949434996\n",
      "Epoch [19/30], Step [80/313], Loss: 0.0004380595055408776\n",
      "Epoch [19/30], Step [90/313], Loss: 0.0005323780933395028\n",
      "Epoch [19/30], Step [100/313], Loss: 0.00048326444812119007\n",
      "Epoch [19/30], Step [110/313], Loss: 0.008078076876699924\n",
      "Epoch [19/30], Step [120/313], Loss: 0.019013915210962296\n",
      "Epoch [19/30], Step [130/313], Loss: 0.05780253931879997\n",
      "Epoch [19/30], Step [140/313], Loss: 0.015481197275221348\n",
      "Epoch [19/30], Step [150/313], Loss: 0.0045767296105623245\n",
      "Epoch [19/30], Step [160/313], Loss: 0.00494813546538353\n",
      "Epoch [19/30], Step [170/313], Loss: 0.029156014323234558\n",
      "Epoch [19/30], Step [180/313], Loss: 0.0028872170951217413\n",
      "Epoch [19/30], Step [190/313], Loss: 0.0037234879564493895\n",
      "Epoch [19/30], Step [200/313], Loss: 0.004458596929907799\n",
      "Epoch [19/30], Step [210/313], Loss: 0.0019266395829617977\n",
      "Epoch [19/30], Step [220/313], Loss: 0.00031762433354742825\n",
      "Epoch [19/30], Step [230/313], Loss: 0.0023059577215462923\n",
      "Epoch [19/30], Step [240/313], Loss: 0.0016672518104314804\n",
      "Epoch [19/30], Step [250/313], Loss: 0.009825422428548336\n",
      "Epoch [19/30], Step [260/313], Loss: 0.0032469879370182753\n",
      "Epoch [19/30], Step [270/313], Loss: 0.001352854771539569\n",
      "Epoch [19/30], Step [280/313], Loss: 0.002961828140541911\n",
      "Epoch [19/30], Step [290/313], Loss: 0.00866740196943283\n",
      "Epoch [19/30], Step [300/313], Loss: 0.001438603620044887\n",
      "Epoch [19/30], Step [310/313], Loss: 0.000530826800968498\n",
      "Epoch [20/30], Step [10/313], Loss: 0.005371480714529753\n",
      "Epoch [20/30], Step [20/313], Loss: 0.0007122871465981007\n",
      "Epoch [20/30], Step [30/313], Loss: 0.001201746636070311\n",
      "Epoch [20/30], Step [40/313], Loss: 0.0014263509074226022\n",
      "Epoch [20/30], Step [50/313], Loss: 0.001528667053207755\n",
      "Epoch [20/30], Step [60/313], Loss: 0.0017440335359424353\n",
      "Epoch [20/30], Step [70/313], Loss: 0.0005949631449766457\n",
      "Epoch [20/30], Step [80/313], Loss: 0.0006567325908690691\n",
      "Epoch [20/30], Step [90/313], Loss: 0.0012388197937980294\n",
      "Epoch [20/30], Step [100/313], Loss: 0.0005266045918688178\n",
      "Epoch [20/30], Step [110/313], Loss: 0.00044082271051593125\n",
      "Epoch [20/30], Step [120/313], Loss: 0.000746274774428457\n",
      "Epoch [20/30], Step [130/313], Loss: 0.0017682238249108195\n",
      "Epoch [20/30], Step [140/313], Loss: 0.0012392783537507057\n",
      "Epoch [20/30], Step [150/313], Loss: 0.00027818180387839675\n",
      "Epoch [20/30], Step [160/313], Loss: 0.0009894219692796469\n",
      "Epoch [20/30], Step [170/313], Loss: 0.00017200762522406876\n",
      "Epoch [20/30], Step [180/313], Loss: 0.0012796163791790605\n",
      "Epoch [20/30], Step [190/313], Loss: 0.0005772769800387323\n",
      "Epoch [20/30], Step [200/313], Loss: 0.0022429819218814373\n",
      "Epoch [20/30], Step [210/313], Loss: 0.001829092507250607\n",
      "Epoch [20/30], Step [220/313], Loss: 0.000908552436158061\n",
      "Epoch [20/30], Step [230/313], Loss: 0.00038267701165750623\n",
      "Epoch [20/30], Step [240/313], Loss: 0.0014600128633901477\n",
      "Epoch [20/30], Step [250/313], Loss: 0.0013315326068550348\n",
      "Epoch [20/30], Step [260/313], Loss: 0.0006692414171993732\n",
      "Epoch [20/30], Step [270/313], Loss: 0.0004063521628268063\n",
      "Epoch [20/30], Step [280/313], Loss: 0.00013922536163590848\n",
      "Epoch [20/30], Step [290/313], Loss: 0.00041004965896718204\n",
      "Epoch [20/30], Step [300/313], Loss: 0.0001425666268914938\n",
      "Epoch [20/30], Step [310/313], Loss: 0.0007141928654164076\n",
      "Epoch [21/30], Step [10/313], Loss: 0.0011647704523056746\n",
      "Epoch [21/30], Step [20/313], Loss: 0.00023136789968702942\n",
      "Epoch [21/30], Step [30/313], Loss: 0.00038470758590847254\n",
      "Epoch [21/30], Step [40/313], Loss: 0.0004143323749303818\n",
      "Epoch [21/30], Step [50/313], Loss: 0.0024774246849119663\n",
      "Epoch [21/30], Step [60/313], Loss: 0.0007118038483895361\n",
      "Epoch [21/30], Step [70/313], Loss: 0.0004520104848779738\n",
      "Epoch [21/30], Step [80/313], Loss: 0.0011037831427529454\n",
      "Epoch [21/30], Step [90/313], Loss: 0.0005970967467874289\n",
      "Epoch [21/30], Step [100/313], Loss: 0.0005332420114427805\n",
      "Epoch [21/30], Step [110/313], Loss: 0.0010461658239364624\n",
      "Epoch [21/30], Step [120/313], Loss: 0.00020448080613277853\n",
      "Epoch [21/30], Step [130/313], Loss: 0.0017234896076843143\n",
      "Epoch [21/30], Step [140/313], Loss: 0.00013685511657968163\n",
      "Epoch [21/30], Step [150/313], Loss: 0.0005309266853146255\n",
      "Epoch [21/30], Step [160/313], Loss: 0.0028906099032610655\n",
      "Epoch [21/30], Step [170/313], Loss: 0.0007555973716080189\n",
      "Epoch [21/30], Step [180/313], Loss: 0.0007338951108977199\n",
      "Epoch [21/30], Step [190/313], Loss: 0.00033776185591705143\n",
      "Epoch [21/30], Step [200/313], Loss: 0.0008473204798065126\n",
      "Epoch [21/30], Step [210/313], Loss: 0.0004939733189530671\n",
      "Epoch [21/30], Step [220/313], Loss: 0.00019807566422969103\n",
      "Epoch [21/30], Step [230/313], Loss: 0.00012332717597018927\n",
      "Epoch [21/30], Step [240/313], Loss: 0.0006525202188640833\n",
      "Epoch [21/30], Step [250/313], Loss: 0.0005330569110810757\n",
      "Epoch [21/30], Step [260/313], Loss: 0.00016516222967766225\n",
      "Epoch [21/30], Step [270/313], Loss: 0.00049886095803231\n",
      "Epoch [21/30], Step [280/313], Loss: 0.0001876106980489567\n",
      "Epoch [21/30], Step [290/313], Loss: 0.0002126150211552158\n",
      "Epoch [21/30], Step [300/313], Loss: 0.0005267770029604435\n",
      "Epoch [21/30], Step [310/313], Loss: 0.0002836457279045135\n",
      "Epoch [22/30], Step [10/313], Loss: 0.0002447457518428564\n",
      "Epoch [22/30], Step [20/313], Loss: 0.00018368795281276107\n",
      "Epoch [22/30], Step [30/313], Loss: 0.0003225059772375971\n",
      "Epoch [22/30], Step [40/313], Loss: 0.00023749686079099774\n",
      "Epoch [22/30], Step [50/313], Loss: 0.00019359287398401648\n",
      "Epoch [22/30], Step [60/313], Loss: 0.00014390196884050965\n",
      "Epoch [22/30], Step [70/313], Loss: 0.000128074039821513\n",
      "Epoch [22/30], Step [80/313], Loss: 0.00022903189528733492\n",
      "Epoch [22/30], Step [90/313], Loss: 0.00019406055798754096\n",
      "Epoch [22/30], Step [100/313], Loss: 0.00019305181922391057\n",
      "Epoch [22/30], Step [110/313], Loss: 0.0002473010972607881\n",
      "Epoch [22/30], Step [120/313], Loss: 0.00013240182306617498\n",
      "Epoch [22/30], Step [130/313], Loss: 0.000236942243645899\n",
      "Epoch [22/30], Step [140/313], Loss: 0.00010556984489085153\n",
      "Epoch [22/30], Step [150/313], Loss: 0.00012330705067142844\n",
      "Epoch [22/30], Step [160/313], Loss: 0.00020070935715921223\n",
      "Epoch [22/30], Step [170/313], Loss: 0.0004950646543875337\n",
      "Epoch [22/30], Step [180/313], Loss: 0.00017297938757110387\n",
      "Epoch [22/30], Step [190/313], Loss: 0.00011375757458154112\n",
      "Epoch [22/30], Step [200/313], Loss: 0.0003758098464459181\n",
      "Epoch [22/30], Step [210/313], Loss: 0.0002200877934228629\n",
      "Epoch [22/30], Step [220/313], Loss: 0.0007748354692012072\n",
      "Epoch [22/30], Step [230/313], Loss: 0.00017129929619841278\n",
      "Epoch [22/30], Step [240/313], Loss: 0.00016400324238929898\n",
      "Epoch [22/30], Step [250/313], Loss: 0.00022208652808330953\n",
      "Epoch [22/30], Step [260/313], Loss: 0.0003124619834125042\n",
      "Epoch [22/30], Step [270/313], Loss: 0.000196434382814914\n",
      "Epoch [22/30], Step [280/313], Loss: 0.00048679535393603146\n",
      "Epoch [22/30], Step [290/313], Loss: 0.0002672692353371531\n",
      "Epoch [22/30], Step [300/313], Loss: 0.0002463902928866446\n",
      "Epoch [22/30], Step [310/313], Loss: 0.0002462558331899345\n",
      "Epoch [23/30], Step [10/313], Loss: 0.00016009327373467386\n",
      "Epoch [23/30], Step [20/313], Loss: 0.0003350909682922065\n",
      "Epoch [23/30], Step [30/313], Loss: 0.0002294974692631513\n",
      "Epoch [23/30], Step [40/313], Loss: 0.00037524697836488485\n",
      "Epoch [23/30], Step [50/313], Loss: 0.0003077170695178211\n",
      "Epoch [23/30], Step [60/313], Loss: 0.0005297524621710181\n",
      "Epoch [23/30], Step [70/313], Loss: 0.00012013882223982364\n",
      "Epoch [23/30], Step [80/313], Loss: 0.0003665246767923236\n",
      "Epoch [23/30], Step [90/313], Loss: 0.00022694630024489015\n",
      "Epoch [23/30], Step [100/313], Loss: 0.0002858669322449714\n",
      "Epoch [23/30], Step [110/313], Loss: 0.00011371784785296768\n",
      "Epoch [23/30], Step [120/313], Loss: 0.00015076945419423282\n",
      "Epoch [23/30], Step [130/313], Loss: 0.0002902383857872337\n",
      "Epoch [23/30], Step [140/313], Loss: 0.00011911360343219712\n",
      "Epoch [23/30], Step [150/313], Loss: 0.00037829717621207237\n",
      "Epoch [23/30], Step [160/313], Loss: 0.00015438963600900024\n",
      "Epoch [23/30], Step [170/313], Loss: 0.00012768275337293744\n",
      "Epoch [23/30], Step [180/313], Loss: 0.00034360343124717474\n",
      "Epoch [23/30], Step [190/313], Loss: 0.0001749175862642005\n",
      "Epoch [23/30], Step [200/313], Loss: 0.0002723862126003951\n",
      "Epoch [23/30], Step [210/313], Loss: 0.00021414345246739686\n",
      "Epoch [23/30], Step [220/313], Loss: 0.0003410138306207955\n",
      "Epoch [23/30], Step [230/313], Loss: 0.0005942313000559807\n",
      "Epoch [23/30], Step [240/313], Loss: 0.0007419027388095856\n",
      "Epoch [23/30], Step [250/313], Loss: 0.0003381345886737108\n",
      "Epoch [23/30], Step [260/313], Loss: 7.38035814720206e-05\n",
      "Epoch [23/30], Step [270/313], Loss: 0.0004884094232693315\n",
      "Epoch [23/30], Step [280/313], Loss: 0.0003010043583344668\n",
      "Epoch [23/30], Step [290/313], Loss: 0.00019072818395216018\n",
      "Epoch [23/30], Step [300/313], Loss: 0.00012543739285320044\n",
      "Epoch [23/30], Step [310/313], Loss: 0.00012212924775667489\n",
      "Epoch [24/30], Step [10/313], Loss: 0.00016213319031521678\n",
      "Epoch [24/30], Step [20/313], Loss: 0.00013910123379901052\n",
      "Epoch [24/30], Step [30/313], Loss: 8.172998059308156e-05\n",
      "Epoch [24/30], Step [40/313], Loss: 0.00017099911929108202\n",
      "Epoch [24/30], Step [50/313], Loss: 0.00039874116191640496\n",
      "Epoch [24/30], Step [60/313], Loss: 0.00026752345729619265\n",
      "Epoch [24/30], Step [70/313], Loss: 0.00011552961223060265\n",
      "Epoch [24/30], Step [80/313], Loss: 0.00039506342727690935\n",
      "Epoch [24/30], Step [90/313], Loss: 0.00012879897258244455\n",
      "Epoch [24/30], Step [100/313], Loss: 0.00042037590174004436\n",
      "Epoch [24/30], Step [110/313], Loss: 0.00018058187561109662\n",
      "Epoch [24/30], Step [120/313], Loss: 0.00012769080058205873\n",
      "Epoch [24/30], Step [130/313], Loss: 4.92384351673536e-05\n",
      "Epoch [24/30], Step [140/313], Loss: 0.00012083313049515709\n",
      "Epoch [24/30], Step [150/313], Loss: 0.00011982108117081225\n",
      "Epoch [24/30], Step [160/313], Loss: 0.000251474033575505\n",
      "Epoch [24/30], Step [170/313], Loss: 0.00014766740787308663\n",
      "Epoch [24/30], Step [180/313], Loss: 0.0005246911896392703\n",
      "Epoch [24/30], Step [190/313], Loss: 0.00020834202587138861\n",
      "Epoch [24/30], Step [200/313], Loss: 0.00016901319031603634\n",
      "Epoch [24/30], Step [210/313], Loss: 0.0002340827923035249\n",
      "Epoch [24/30], Step [220/313], Loss: 0.00039661783375777304\n",
      "Epoch [24/30], Step [230/313], Loss: 0.00015992889530025423\n",
      "Epoch [24/30], Step [240/313], Loss: 0.00019921000057365745\n",
      "Epoch [24/30], Step [250/313], Loss: 0.00018162481137551367\n",
      "Epoch [24/30], Step [260/313], Loss: 0.00039732790901325643\n",
      "Epoch [24/30], Step [270/313], Loss: 0.0001326756610069424\n",
      "Epoch [24/30], Step [280/313], Loss: 0.0002700376498978585\n",
      "Epoch [24/30], Step [290/313], Loss: 0.00013055687304586172\n",
      "Epoch [24/30], Step [300/313], Loss: 0.0002958517870865762\n",
      "Epoch [24/30], Step [310/313], Loss: 0.0002053928910754621\n",
      "Epoch [25/30], Step [10/313], Loss: 0.0002579743741080165\n",
      "Epoch [25/30], Step [20/313], Loss: 0.00022284929582383484\n",
      "Epoch [25/30], Step [30/313], Loss: 0.00013915618183091283\n",
      "Epoch [25/30], Step [40/313], Loss: 0.0001804184284992516\n",
      "Epoch [25/30], Step [50/313], Loss: 0.0002643334155436605\n",
      "Epoch [25/30], Step [60/313], Loss: 0.00011104947770945728\n",
      "Epoch [25/30], Step [70/313], Loss: 0.0003758222737815231\n",
      "Epoch [25/30], Step [80/313], Loss: 0.00016793677059467882\n",
      "Epoch [25/30], Step [90/313], Loss: 0.0002299774350831285\n",
      "Epoch [25/30], Step [100/313], Loss: 0.0001309892104472965\n",
      "Epoch [25/30], Step [110/313], Loss: 0.0001602810516487807\n",
      "Epoch [25/30], Step [120/313], Loss: 0.00015889189671725035\n",
      "Epoch [25/30], Step [130/313], Loss: 0.00017264645430259407\n",
      "Epoch [25/30], Step [140/313], Loss: 0.0001243817387148738\n",
      "Epoch [25/30], Step [150/313], Loss: 0.00022889982210472226\n",
      "Epoch [25/30], Step [160/313], Loss: 0.00022483165957964957\n",
      "Epoch [25/30], Step [170/313], Loss: 9.339069947600365e-05\n",
      "Epoch [25/30], Step [180/313], Loss: 7.09763917257078e-05\n",
      "Epoch [25/30], Step [190/313], Loss: 5.484528446686454e-05\n",
      "Epoch [25/30], Step [200/313], Loss: 7.304152677534148e-05\n",
      "Epoch [25/30], Step [210/313], Loss: 0.00017945520812645555\n",
      "Epoch [25/30], Step [220/313], Loss: 0.0001910389109980315\n",
      "Epoch [25/30], Step [230/313], Loss: 0.0001743126194924116\n",
      "Epoch [25/30], Step [240/313], Loss: 0.00013277243124321103\n",
      "Epoch [25/30], Step [250/313], Loss: 0.0005532772047445178\n",
      "Epoch [25/30], Step [260/313], Loss: 0.00019947992404922843\n",
      "Epoch [25/30], Step [270/313], Loss: 0.0001913194719236344\n",
      "Epoch [25/30], Step [280/313], Loss: 0.00014543073484674096\n",
      "Epoch [25/30], Step [290/313], Loss: 0.0003930213279090822\n",
      "Epoch [25/30], Step [300/313], Loss: 0.0001419164618710056\n",
      "Epoch [25/30], Step [310/313], Loss: 0.00031463702907785773\n",
      "Epoch [26/30], Step [10/313], Loss: 0.00018158796592615545\n",
      "Epoch [26/30], Step [20/313], Loss: 0.00012348995369393378\n",
      "Epoch [26/30], Step [30/313], Loss: 9.08758447621949e-05\n",
      "Epoch [26/30], Step [40/313], Loss: 0.00012536939175333828\n",
      "Epoch [26/30], Step [50/313], Loss: 0.0002248562523163855\n",
      "Epoch [26/30], Step [60/313], Loss: 0.000271312368568033\n",
      "Epoch [26/30], Step [70/313], Loss: 0.00022110163990873843\n",
      "Epoch [26/30], Step [80/313], Loss: 0.00011236309364903718\n",
      "Epoch [26/30], Step [90/313], Loss: 0.00010498733900021762\n",
      "Epoch [26/30], Step [100/313], Loss: 8.768475527176633e-05\n",
      "Epoch [26/30], Step [110/313], Loss: 7.730208744760603e-05\n",
      "Epoch [26/30], Step [120/313], Loss: 0.00016638760280329734\n",
      "Epoch [26/30], Step [130/313], Loss: 8.632912795292214e-05\n",
      "Epoch [26/30], Step [140/313], Loss: 0.0001165088833658956\n",
      "Epoch [26/30], Step [150/313], Loss: 9.561722254147753e-05\n",
      "Epoch [26/30], Step [160/313], Loss: 0.0001320533046964556\n",
      "Epoch [26/30], Step [170/313], Loss: 9.743856935529038e-05\n",
      "Epoch [26/30], Step [180/313], Loss: 9.091354149859399e-05\n",
      "Epoch [26/30], Step [190/313], Loss: 0.00012134532153140754\n",
      "Epoch [26/30], Step [200/313], Loss: 0.0002863219706341624\n",
      "Epoch [26/30], Step [210/313], Loss: 0.00013128944556228817\n",
      "Epoch [26/30], Step [220/313], Loss: 0.00011348674888722599\n",
      "Epoch [26/30], Step [230/313], Loss: 0.00012739645899273455\n",
      "Epoch [26/30], Step [240/313], Loss: 0.00014854344772174954\n",
      "Epoch [26/30], Step [250/313], Loss: 3.899114017258398e-05\n",
      "Epoch [26/30], Step [260/313], Loss: 8.389001595787704e-05\n",
      "Epoch [26/30], Step [270/313], Loss: 0.00024807034060359\n",
      "Epoch [26/30], Step [280/313], Loss: 0.00016189554298762232\n",
      "Epoch [26/30], Step [290/313], Loss: 0.00019510663696564734\n",
      "Epoch [26/30], Step [300/313], Loss: 0.00011158972483826801\n",
      "Epoch [26/30], Step [310/313], Loss: 0.0004967094864696264\n",
      "Epoch [27/30], Step [10/313], Loss: 6.516097346320748e-05\n",
      "Epoch [27/30], Step [20/313], Loss: 0.00017919355013873428\n",
      "Epoch [27/30], Step [30/313], Loss: 0.00022241237456910312\n",
      "Epoch [27/30], Step [40/313], Loss: 0.00013042920909356326\n",
      "Epoch [27/30], Step [50/313], Loss: 0.00024675196618773043\n",
      "Epoch [27/30], Step [60/313], Loss: 0.0002604337641969323\n",
      "Epoch [27/30], Step [70/313], Loss: 0.00012914914987049997\n",
      "Epoch [27/30], Step [80/313], Loss: 0.00012568583770189434\n",
      "Epoch [27/30], Step [90/313], Loss: 0.00014956574887037277\n",
      "Epoch [27/30], Step [100/313], Loss: 0.0001592835469637066\n",
      "Epoch [27/30], Step [110/313], Loss: 7.966168050188571e-05\n",
      "Epoch [27/30], Step [120/313], Loss: 0.0001344399352092296\n",
      "Epoch [27/30], Step [130/313], Loss: 7.845433719921857e-05\n",
      "Epoch [27/30], Step [140/313], Loss: 0.00031740072881802917\n",
      "Epoch [27/30], Step [150/313], Loss: 0.00020111181947868317\n",
      "Epoch [27/30], Step [160/313], Loss: 0.0004287299234420061\n",
      "Epoch [27/30], Step [170/313], Loss: 9.751872858032584e-05\n",
      "Epoch [27/30], Step [180/313], Loss: 0.00011318959150230512\n",
      "Epoch [27/30], Step [190/313], Loss: 9.180662163998932e-05\n",
      "Epoch [27/30], Step [200/313], Loss: 0.00042160297743976116\n",
      "Epoch [27/30], Step [210/313], Loss: 0.00024436708190478384\n",
      "Epoch [27/30], Step [220/313], Loss: 0.00016953417798504233\n",
      "Epoch [27/30], Step [230/313], Loss: 0.0004634796059690416\n",
      "Epoch [27/30], Step [240/313], Loss: 0.00010025351366493851\n",
      "Epoch [27/30], Step [250/313], Loss: 0.00011824716057162732\n",
      "Epoch [27/30], Step [260/313], Loss: 0.00017474309424869716\n",
      "Epoch [27/30], Step [270/313], Loss: 8.47724950290285e-05\n",
      "Epoch [27/30], Step [280/313], Loss: 0.0004299496067687869\n",
      "Epoch [27/30], Step [290/313], Loss: 0.0005149826174601912\n",
      "Epoch [27/30], Step [300/313], Loss: 7.750885561108589e-05\n",
      "Epoch [27/30], Step [310/313], Loss: 0.00044366400106810033\n",
      "Epoch [28/30], Step [10/313], Loss: 0.00016165633860509843\n",
      "Epoch [28/30], Step [20/313], Loss: 0.00029645103495568037\n",
      "Epoch [28/30], Step [30/313], Loss: 0.0002005613496294245\n",
      "Epoch [28/30], Step [40/313], Loss: 0.00011585660104174167\n",
      "Epoch [28/30], Step [50/313], Loss: 0.0004118744400329888\n",
      "Epoch [28/30], Step [60/313], Loss: 8.740820339880884e-05\n",
      "Epoch [28/30], Step [70/313], Loss: 0.00010404703789390624\n",
      "Epoch [28/30], Step [80/313], Loss: 0.0002554557577241212\n",
      "Epoch [28/30], Step [90/313], Loss: 0.0003534916031640023\n",
      "Epoch [28/30], Step [100/313], Loss: 0.0001471613795729354\n",
      "Epoch [28/30], Step [110/313], Loss: 0.00015550028183497488\n",
      "Epoch [28/30], Step [120/313], Loss: 0.00016861782933119684\n",
      "Epoch [28/30], Step [130/313], Loss: 0.0001360625319648534\n",
      "Epoch [28/30], Step [140/313], Loss: 8.016111678443849e-05\n",
      "Epoch [28/30], Step [150/313], Loss: 0.00011115481902379543\n",
      "Epoch [28/30], Step [160/313], Loss: 5.0299382564844564e-05\n",
      "Epoch [28/30], Step [170/313], Loss: 0.00010297572589479387\n",
      "Epoch [28/30], Step [180/313], Loss: 6.190899148350582e-05\n",
      "Epoch [28/30], Step [190/313], Loss: 0.00019496987806633115\n",
      "Epoch [28/30], Step [200/313], Loss: 0.00022777346021030098\n",
      "Epoch [28/30], Step [210/313], Loss: 7.571547757834196e-05\n",
      "Epoch [28/30], Step [220/313], Loss: 0.00013180996757000685\n",
      "Epoch [28/30], Step [230/313], Loss: 0.00013008690439164639\n",
      "Epoch [28/30], Step [240/313], Loss: 0.00013048744585830718\n",
      "Epoch [28/30], Step [250/313], Loss: 0.00011961615382460877\n",
      "Epoch [28/30], Step [260/313], Loss: 0.00013957777991890907\n",
      "Epoch [28/30], Step [270/313], Loss: 0.00015584161155857146\n",
      "Epoch [28/30], Step [280/313], Loss: 0.00011676736176013947\n",
      "Epoch [28/30], Step [290/313], Loss: 0.0002440307871438563\n",
      "Epoch [28/30], Step [300/313], Loss: 0.0002041063562501222\n",
      "Epoch [28/30], Step [310/313], Loss: 0.00015123675984796137\n",
      "Epoch [29/30], Step [10/313], Loss: 0.00028932600980624557\n",
      "Epoch [29/30], Step [20/313], Loss: 0.0001362524344585836\n",
      "Epoch [29/30], Step [30/313], Loss: 0.00010456603195052594\n",
      "Epoch [29/30], Step [40/313], Loss: 0.00011090822226833552\n",
      "Epoch [29/30], Step [50/313], Loss: 5.9278034314047545e-05\n",
      "Epoch [29/30], Step [60/313], Loss: 0.00011361252109054476\n",
      "Epoch [29/30], Step [70/313], Loss: 9.503476758254692e-05\n",
      "Epoch [29/30], Step [80/313], Loss: 0.00018169538816437125\n",
      "Epoch [29/30], Step [90/313], Loss: 0.00014310520782601088\n",
      "Epoch [29/30], Step [100/313], Loss: 0.00023160157434176654\n",
      "Epoch [29/30], Step [110/313], Loss: 6.583188223885372e-05\n",
      "Epoch [29/30], Step [120/313], Loss: 0.0001353699917672202\n",
      "Epoch [29/30], Step [130/313], Loss: 0.00019985990365967155\n",
      "Epoch [29/30], Step [140/313], Loss: 0.000125678998301737\n",
      "Epoch [29/30], Step [150/313], Loss: 0.00017119359108619392\n",
      "Epoch [29/30], Step [160/313], Loss: 9.75754446699284e-05\n",
      "Epoch [29/30], Step [170/313], Loss: 5.728251198888756e-05\n",
      "Epoch [29/30], Step [180/313], Loss: 3.581598139135167e-05\n",
      "Epoch [29/30], Step [190/313], Loss: 0.00018339420785196126\n",
      "Epoch [29/30], Step [200/313], Loss: 8.654267730889842e-05\n",
      "Epoch [29/30], Step [210/313], Loss: 0.00027474621310830116\n",
      "Epoch [29/30], Step [220/313], Loss: 0.00016241456614807248\n",
      "Epoch [29/30], Step [230/313], Loss: 0.00016204202256631106\n",
      "Epoch [29/30], Step [240/313], Loss: 8.128324407152832e-05\n",
      "Epoch [29/30], Step [250/313], Loss: 0.0001485565589973703\n",
      "Epoch [29/30], Step [260/313], Loss: 8.37646221043542e-05\n",
      "Epoch [29/30], Step [270/313], Loss: 8.502549462718889e-05\n",
      "Epoch [29/30], Step [280/313], Loss: 0.00011628565698629245\n",
      "Epoch [29/30], Step [290/313], Loss: 5.507275272975676e-05\n",
      "Epoch [29/30], Step [300/313], Loss: 0.0002013868506764993\n",
      "Epoch [29/30], Step [310/313], Loss: 5.394205800257623e-05\n",
      "Epoch [30/30], Step [10/313], Loss: 0.00011707730300258845\n",
      "Epoch [30/30], Step [20/313], Loss: 0.0004546225245576352\n",
      "Epoch [30/30], Step [30/313], Loss: 0.00015238906780723482\n",
      "Epoch [30/30], Step [40/313], Loss: 0.00012500447337515652\n",
      "Epoch [30/30], Step [50/313], Loss: 7.135064515750855e-05\n",
      "Epoch [30/30], Step [60/313], Loss: 0.00017336467863060534\n",
      "Epoch [30/30], Step [70/313], Loss: 0.0002169442013837397\n",
      "Epoch [30/30], Step [80/313], Loss: 0.00012330233585089445\n",
      "Epoch [30/30], Step [90/313], Loss: 0.00013411714462563396\n",
      "Epoch [30/30], Step [100/313], Loss: 9.271818271372467e-05\n",
      "Epoch [30/30], Step [110/313], Loss: 5.7056407968048006e-05\n",
      "Epoch [30/30], Step [120/313], Loss: 0.00011945128062507138\n",
      "Epoch [30/30], Step [130/313], Loss: 0.00011538956459844485\n",
      "Epoch [30/30], Step [140/313], Loss: 6.51854497846216e-05\n",
      "Epoch [30/30], Step [150/313], Loss: 0.00012915520346723497\n",
      "Epoch [30/30], Step [160/313], Loss: 0.00012483104364946485\n",
      "Epoch [30/30], Step [170/313], Loss: 5.291881461744197e-05\n",
      "Epoch [30/30], Step [180/313], Loss: 5.738634354202077e-05\n",
      "Epoch [30/30], Step [190/313], Loss: 0.00011820476356660947\n",
      "Epoch [30/30], Step [200/313], Loss: 8.8020387920551e-05\n",
      "Epoch [30/30], Step [210/313], Loss: 0.0004486521356739104\n",
      "Epoch [30/30], Step [220/313], Loss: 0.00018997379811480641\n",
      "Epoch [30/30], Step [230/313], Loss: 0.00010257336543872952\n",
      "Epoch [30/30], Step [240/313], Loss: 0.00013778411084786057\n",
      "Epoch [30/30], Step [250/313], Loss: 9.600450721336529e-05\n",
      "Epoch [30/30], Step [260/313], Loss: 8.463053382001817e-05\n",
      "Epoch [30/30], Step [270/313], Loss: 0.00011948757310165092\n",
      "Epoch [30/30], Step [280/313], Loss: 0.00010125889093615115\n",
      "Epoch [30/30], Step [290/313], Loss: 7.959624053910375e-05\n",
      "Epoch [30/30], Step [300/313], Loss: 3.7804180465172976e-05\n",
      "Epoch [30/30], Step [310/313], Loss: 9.356289956485853e-05\n"
     ]
    }
   ],
   "source": [
    "def encode_labels(labels):\n",
    "\n",
    "    encoded = []\n",
    "    lengths = []\n",
    "    for label in labels:\n",
    "        encoded_label = [char_to_int[ch] for ch in label]\n",
    "        encoded.extend(encoded_label)\n",
    "        lengths.append(len(encoded_label))\n",
    "    return torch.tensor(encoded), lengths\n",
    "\n",
    "print(len(char_to_int))\n",
    "\n",
    "# 训练模型\n",
    "\n",
    "for epoch in range(num_epochs):\n",
    "    for i, (images, labels) in enumerate(train_loader):\n",
    "\n",
    "        images = images.to(device)\n",
    "        \n",
    "        optimizer.zero_grad()\n",
    "\n",
    "        outputs = crnn(images)\n",
    "        outputs = outputs.log_softmax(2)\n",
    "        \n",
    "        # Encode labels into format suitable for CTC\n",
    "        encoded_labels, label_lengths = encode_labels(labels)\n",
    "        \n",
    "        #print(encoded_labels)\n",
    "        #print(\"---------\")\n",
    "        \n",
    "        encoded_labels = encoded_labels.to(device)\n",
    "        output_lengths = torch.full((outputs.size(1),), outputs.size(0), dtype=torch.long)\n",
    "\n",
    "        loss = criterion(outputs, encoded_labels, output_lengths, torch.tensor(label_lengths))\n",
    "        #print(f\"loss:{loss}\")\n",
    "        loss.backward()\n",
    "\n",
    "        optimizer.step()\n",
    "\n",
    "        if (i+1) % 10 == 0:\n",
    "            print(f'Epoch [{epoch+1}/{num_epochs}], Step [{i+1}/{len(train_loader)}], Loss: {loss.item()}')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c8df0c41-78ce-4730-8519-885efca23651",
   "metadata": {},
   "source": [
    "## 保存训练权重"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "42bdb6ec-30d9-420f-80c3-075c3c1dce31",
   "metadata": {},
   "outputs": [],
   "source": [
    "PATH = './crnn_str_single_simple.pth'\n",
    "torch.save(crnn.state_dict(), PATH)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5d4b415b-e296-4dba-98f0-fcddcad989e3",
   "metadata": {},
   "source": [
    "## 测试评估"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "c5c69981-934d-41ee-a9a2-1a9c9d132f7f",
   "metadata": {},
   "outputs": [],
   "source": [
    "#device = \"cpu\"\n",
    "PATH = './crnn_str_single_simple.pth'\n",
    "net = CRNN(num_classes=num_classes).to(device)\n",
    "#net = CRNN(config).to(device)\n",
    "\n",
    "if device == \"cpu\":\n",
    "    net.load_state_dict(torch.load(PATH, map_location='cpu'))\n",
    "else:\n",
    "    net.load_state_dict(torch.load(PATH))\n",
    "\n",
    "net.eval()\n",
    "\n",
    "\n",
    "test_dataset = CustomDataset(root_dir='/home/han/Documents/SRP/dataset/testdata/', transform=transform)\n",
    "test_loader = torch.utils.data.DataLoader(test_dataset, batch_size=1, shuffle=True, num_workers=4)  #此处batch_size必须为1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "639c62e8-26a7-441b-8f67-a8ccbd05bde0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 定义字符到整数的逆映射\n",
    "int_to_char = {i: c for c, i in char_to_int.items()}\n",
    "\n",
    "\n",
    "def decode_labels(preds):\n",
    "    decoded_text = \"\"\n",
    "    prevalue = 0\n",
    "    for pred in preds:\n",
    "        value = pred.item()\n",
    "        if value == 0:\n",
    "            prevalue = value\n",
    "            continue\n",
    "        if value == prevalue:\n",
    "            continue\n",
    "        if value in int_to_char:\n",
    "            decoded_text += int_to_char[value]\n",
    "        prevalue = value\n",
    "    return decoded_text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "7a649c69-78fe-4850-9d7d-9e27022762de",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "misarrangement\n",
      "misarangement\n",
      "food-productive\n",
      "fod-productive\n",
      "patternmaker\n",
      "paternmaker\n",
      "tyrosinuria\n",
      "tyrosinurila\n",
      "test procees:100/1000\n",
      "Skimos\n",
      "skimos\n",
      "Aristomachus\n",
      "aristomachus\n",
      "phytophylogeny\n",
      "phytoplylagry\n",
      "psychotherapies\n",
      "nsxcrhotherapies\n",
      "modernists\n",
      "modernsts\n",
      "Methodistically\n",
      "methodistically\n",
      "test procees:200/1000\n",
      "test procees:300/1000\n",
      "test procees:400/1000\n",
      "test procees:500/1000\n",
      "test procees:600/1000\n",
      "test procees:700/1000\n",
      "test procees:800/1000\n",
      "test procees:900/1000\n",
      "test procees:1000/1000\n",
      "Test Accuracy: 93.50%\n"
     ]
    }
   ],
   "source": [
    "\n",
    "# 对测试数据进行评估\n",
    "test_accuracy = 0.0\n",
    "total_samples = 0\n",
    "\n",
    "i=0\n",
    "error_count = 0\n",
    "with torch.no_grad():\n",
    "    for images, labels in test_loader:\n",
    "        images = images.to(device)\n",
    "        outputs = net(images)\n",
    "\n",
    "        \n",
    "        _, predicted = torch.max(outputs,2)  # 第2个维度是类别维度)\n",
    "\n",
    "        #print(predicted)\n",
    "        if(i<=4):\n",
    "            images_show(images,[decode_labels(predicted)])\n",
    "        if(decode_labels(predicted) ==  labels[0]):\n",
    "            test_accuracy +=1\n",
    "        else:\n",
    "            if(error_count < 10):\n",
    "                print(labels[0])\n",
    "                print(decode_labels(predicted))\n",
    "                error_count+=1\n",
    "        total_samples += len(labels)\n",
    "        i+=1\n",
    "        if i==5:\n",
    "            #break\n",
    "            continue\n",
    "        if((i+1)%100 == 0):\n",
    "            print(f\"test procees:{i+1}/{len(test_loader)}\")\n",
    "\n",
    "test_accuracy /= total_samples\n",
    "print(f\"Test Accuracy: {test_accuracy * 100:.2f}%\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1e239433-1890-4929-9d1b-d641f1b0264b",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.11.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}