{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "machine_shape": "hm",
      "gpuType": "V100"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "wS_st3blv2jp",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "1e438668-f0cc-4edc-cf36-99309a0295af"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Downloading...\n",
            "From: https://drive.google.com/uc?id=1mqCsj6uO4WIbv3FRO9OTKXc8Wd8lYbKa\n",
            "To: /content/Dataset.zip\n",
            "\r  0%|          | 0.00/700M [00:00<?, ?B/s]\r  3%|▎         | 21.0M/700M [00:00<00:03, 208MB/s]\r  6%|▋         | 44.6M/700M [00:00<00:02, 224MB/s]\r 10%|▉         | 67.1M/700M [00:00<00:05, 123MB/s]\r 13%|█▎        | 90.7M/700M [00:00<00:04, 152MB/s]\r 16%|█▌        | 111M/700M [00:00<00:03, 165MB/s] \r 21%|██        | 145M/700M [00:00<00:02, 214MB/s]\r 25%|██▌       | 176M/700M [00:00<00:02, 240MB/s]\r 29%|██▉       | 203M/700M [00:00<00:02, 243MB/s]\r 34%|███▍      | 240M/700M [00:01<00:01, 279MB/s]\r 38%|███▊      | 269M/700M [00:01<00:01, 252MB/s]\r 43%|████▎     | 303M/700M [00:01<00:01, 274MB/s]\r 48%|████▊     | 334M/700M [00:01<00:01, 284MB/s]\r 52%|█████▏    | 364M/700M [00:01<00:01, 213MB/s]\r 57%|█████▋    | 400M/700M [00:01<00:01, 246MB/s]\r 63%|██████▎   | 438M/700M [00:01<00:00, 279MB/s]\r 67%|██████▋   | 470M/700M [00:02<00:00, 259MB/s]\r 73%|███████▎  | 509M/700M [00:02<00:00, 290MB/s]\r 78%|███████▊  | 543M/700M [00:02<00:00, 302MB/s]\r 83%|████████▎ | 580M/700M [00:02<00:00, 322MB/s]\r 89%|████████▊ | 619M/700M [00:02<00:00, 340MB/s]\r 94%|█████████▎| 655M/700M [00:02<00:00, 323MB/s]\r 98%|█████████▊| 688M/700M [00:02<00:00, 322MB/s]\r100%|██████████| 700M/700M [00:02<00:00, 261MB/s]\n",
            "Downloading...\n",
            "From: https://drive.google.com/uc?id=1CzMKzBpTBT_rhk-iMQAGe4_1Hsf5kmnE\n",
            "To: /content/TestDataSet.zip\n",
            "\r  0%|          | 0.00/833M [00:00<?, ?B/s]\r  2%|▏         | 14.2M/833M [00:00<00:05, 141MB/s]\r  7%|▋         | 55.6M/833M [00:00<00:02, 301MB/s]\r 10%|█         | 86.0M/833M [00:00<00:02, 295MB/s]\r 15%|█▌        | 125M/833M [00:00<00:02, 331MB/s] \r 19%|█▉        | 159M/833M [00:00<00:04, 136MB/s]\r 24%|██▍       | 198M/833M [00:01<00:03, 180MB/s]\r 28%|██▊       | 235M/833M [00:01<00:03, 199MB/s]\r 33%|███▎      | 274M/833M [00:01<00:02, 237MB/s]\r 38%|███▊      | 316M/833M [00:01<00:01, 277MB/s]\r 42%|████▏     | 350M/833M [00:01<00:01, 288MB/s]\r 46%|████▋     | 386M/833M [00:01<00:01, 274MB/s]\r 51%|█████     | 422M/833M [00:01<00:01, 294MB/s]\r 56%|█████▌    | 464M/833M [00:01<00:01, 326MB/s]\r 60%|█████▉    | 499M/833M [00:01<00:01, 320MB/s]\r 64%|██████▍   | 537M/833M [00:02<00:00, 322MB/s]\r 69%|██████▉   | 576M/833M [00:02<00:00, 338MB/s]\r 73%|███████▎  | 612M/833M [00:02<00:00, 344MB/s]\r 78%|███████▊  | 650M/833M [00:02<00:00, 355MB/s]\r 82%|████████▏ | 686M/833M [00:02<00:00, 318MB/s]\r 87%|████████▋ | 728M/833M [00:02<00:00, 343MB/s]\r 92%|█████████▏| 763M/833M [00:02<00:00, 275MB/s]\r 95%|█████████▌| 794M/833M [00:02<00:00, 242MB/s]\r 99%|█████████▉| 829M/833M [00:03<00:00, 266MB/s]\r100%|██████████| 833M/833M [00:03<00:00, 271MB/s]\n"
          ]
        }
      ],
      "source": [
        "#@title 데이터셋 불러오기\n",
        "%%bash\n",
        "gdown 1mqCsj6uO4WIbv3FRO9OTKXc8Wd8lYbKa\n",
        "unzip -q Dataset.zip\n",
        "mkdir Dataset\n",
        "mv /content/masked /content/half /content/unmasked /content/Dataset\n",
        "gdown 1CzMKzBpTBT_rhk-iMQAGe4_1Hsf5kmnE\n",
        "unzip -q TestDataSet.zip"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install -qq wandb\n",
        "import wandb\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.optim as optim\n",
        "from torch.utils.data import DataLoader\n",
        "from torchvision import transforms\n",
        "from torchvision.transforms import v2\n",
        "from torchvision.datasets import ImageFolder\n",
        "from torchsummary import summary\n",
        "from sklearn.metrics import accuracy_score, f1_score\n",
        "from imblearn.over_sampling import SMOTE\n",
        "import numpy as np"
      ],
      "metadata": {
        "id": "xvPS6Xwcv6FM"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#@title half를 masked에 합치기\n",
        "import os\n",
        "import shutil\n",
        "half_folder = '/content/Dataset/half'  # \"half\" 폴더의 경로\n",
        "unmasked_folder = '/content/Dataset/masked'  # \"masked\" 폴더의 경로\n",
        "\n",
        "if not os.path.exists(unmasked_folder):\n",
        "    os.makedirs(unmasked_folder)\n",
        "for filename in os.listdir(half_folder):\n",
        "    source_path = os.path.join(half_folder, filename)\n",
        "    destination_path = os.path.join(unmasked_folder, filename)\n",
        "    if os.path.exists(destination_path):\n",
        "        new_filename = filename\n",
        "        name, ext = os.path.splitext(filename)\n",
        "        count = 1\n",
        "        while os.path.exists(os.path.join(unmasked_folder, new_filename)):\n",
        "            new_filename = f\"{name}_{count}{ext}\"\n",
        "            count += 1\n",
        "        destination_path = os.path.join(unmasked_folder, new_filename)\n",
        "    shutil.move(source_path, destination_path)\n",
        "shutil.rmtree(half_folder)\n",
        "print(\"파일 이동 및 폴더 삭제가 완료되었습니다.\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TsbVuJPkZShv",
        "outputId": "2c3c3765-85ba-45a3-c839-f5222ffda61d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "파일 이동 및 폴더 삭제가 완료되었습니다.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# 이미지 데이터 폴더 경로 설정\n",
        "train_folder = '/content/Dataset'\n",
        "valid_folder = '/content/valid'\n",
        "\n",
        "# 이미지 전처리 및 데이터 로드\n",
        "transform = transforms.v2.Compose([\n",
        "    v2.Resize((512, 512)),\n",
        "    v2.RandomVerticalFlip(p=0.5),\n",
        "    #v2.AugMix(),\n",
        "    v2.Compose([v2.ToImage(), v2.ToDtype(torch.float32, scale=True)]),\n",
        "    v2.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])\n",
        "])\n",
        "transform_valid = transforms.v2.Compose([\n",
        "    v2.Resize((512,512)),\n",
        "    v2.Compose([v2.ToImage(), v2.ToDtype(torch.float32, scale=True)]),\n",
        "    v2.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])\n",
        "])"
      ],
      "metadata": {
        "id": "i2QMdOq0wI17"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# 훈련 데이터셋 생성\n",
        "train_dataset = ImageFolder(root=train_folder, transform=transform)\n",
        "# 검증 데이터셋 생성\n",
        "valid_dataset = ImageFolder(root=valid_folder, transform=transform_valid)\n",
        "\n",
        "# DataLoader를 사용하여 데이터 배치로 로드 (배치 크기 변경)\n",
        "batch_size = 256\n",
        "train_loader = DataLoader(train_dataset, num_workers=8, batch_size=batch_size, shuffle=True, pin_memory=True)\n",
        "valid_loader = DataLoader(valid_dataset, num_workers=8, batch_size=batch_size, shuffle=False, pin_memory=True)\n",
        "\n",
        "# 데이터 확인 (예시로 첫 번째 배치만 출력)\n",
        "#for batch in valid_loader:\n",
        "#    X_validation, y_validation = batch\n",
        "#    break"
      ],
      "metadata": {
        "id": "ldA26_nukhkl"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# 데이터 일부만 로드하여 SMOTE 적용\n",
        "num_samples_for_smote = 1000  # 적절한 크기로 조절\n",
        "X_train = []\n",
        "y_train = []\n",
        "\n",
        "for batch in train_loader:\n",
        "    x_batch, y_batch = batch\n",
        "    X_train.append(x_batch.numpy())\n",
        "    y_train.append(y_batch.numpy())\n",
        "    if len(X_train) * batch_size >= num_samples_for_smote:\n",
        "        break\n",
        "\n",
        "X_train = np.concatenate(X_train)\n",
        "y_train = np.concatenate(y_train)\n",
        "\n",
        "# SMOTE 적용\n",
        "smote = SMOTE(sampling_strategy='auto', random_state=42)\n",
        "X_train_resampled, y_train_resampled = smote.fit_resample(X_train.reshape(-1, 224 * 224 * 3), y_train)\n",
        "\n",
        "# 텐서로 변환\n",
        "X_train_resampled = torch.tensor(X_train_resampled.reshape(-1, 3, 224, 224))\n",
        "y_train_resampled = torch.tensor(y_train_resampled)"
      ],
      "metadata": {
        "id": "CmUPTYdYkmDe"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "  # 로지스틱 회귀 모델 정의\n",
        "class LogisticRegressionModel(nn.Module):\n",
        "    def __init__(self, input_size):\n",
        "        super(LogisticRegressionModel, self).__init__()\n",
        "        self.flatten = nn.Flatten()\n",
        "        self.dropout = nn.Dropout(p = 0.5)\n",
        "        self.linear = nn.Linear(input_size, 1)\n",
        "        self.sigmoid = nn.Sigmoid()\n",
        "\n",
        "    def forward(self, x):\n",
        "        x = self.flatten(x)\n",
        "        x = self.linear(x)\n",
        "        x = self.sigmoid(x)\n",
        "        return x"
      ],
      "metadata": {
        "id": "V5c8SK718-3W"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# 모델 초기화\n",
        "input_size = 512 * 512 * 3  # 이미지 크기 및 채널 수에 따라 조정\n",
        "model = LogisticRegressionModel(input_size).cuda()\n",
        "\n",
        "# 손실 함수 및 최적화 알고리즘 설정 (AdamW)\n",
        "criterion = nn.BCELoss()\n",
        "optimizer = optim.AdamW(model.parameters(), lr=0.0001)\n",
        "summary(model, (3,512,512))"
      ],
      "metadata": {
        "id": "LDscXKbR8_4Q",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "6213483c-c024-42a4-a390-210dd0a8ff47"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "----------------------------------------------------------------\n",
            "        Layer (type)               Output Shape         Param #\n",
            "================================================================\n",
            "           Flatten-1               [-1, 786432]               0\n",
            "            Linear-2                    [-1, 1]         786,433\n",
            "           Sigmoid-3                    [-1, 1]               0\n",
            "================================================================\n",
            "Total params: 786,433\n",
            "Trainable params: 786,433\n",
            "Non-trainable params: 0\n",
            "----------------------------------------------------------------\n",
            "Input size (MB): 3.00\n",
            "Forward/backward pass size (MB): 6.00\n",
            "Params size (MB): 3.00\n",
            "Estimated Total Size (MB): 12.00\n",
            "----------------------------------------------------------------\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "wandb.init(project='logistic')\n",
        "wandb.run.name = 'logistic'\n",
        "wandb.run.save()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 157
        },
        "id": "d7rq0P-TY2M3",
        "outputId": "de5ef2d9-26b3-4912-dbfb-3cd240a293d4"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\u001b[34m\u001b[1mwandb\u001b[0m: Currently logged in as: \u001b[33mdtckon1234\u001b[0m (\u001b[33mk-on\u001b[0m). Use \u001b[1m`wandb login --relogin`\u001b[0m to force relogin\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "Tracking run with wandb version 0.16.0"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "Run data is saved locally in <code>/content/wandb/run-20231127_021651-hy0seqbg</code>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "Syncing run <strong><a href='https://wandb.ai/k-on/logistic/runs/hy0seqbg' target=\"_blank\">distinctive-hill-10</a></strong> to <a href='https://wandb.ai/k-on/logistic' target=\"_blank\">Weights & Biases</a> (<a href='https://wandb.me/run' target=\"_blank\">docs</a>)<br/>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              " View project at <a href='https://wandb.ai/k-on/logistic' target=\"_blank\">https://wandb.ai/k-on/logistic</a>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              " View run at <a href='https://wandb.ai/k-on/logistic/runs/hy0seqbg' target=\"_blank\">https://wandb.ai/k-on/logistic/runs/hy0seqbg</a>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\u001b[34m\u001b[1mwandb\u001b[0m: \u001b[33mWARNING\u001b[0m Calling wandb.run.save without any arguments is deprecated.Changes to attributes are automatically persisted.\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "True"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# 학습\n",
        "num_epochs = 50\n",
        "for epoch in range(num_epochs):\n",
        "    model.train()\n",
        "    for i, (inputs, labels) in enumerate(train_loader):\n",
        "        inputs, labels = inputs.cuda(), labels.cuda()\n",
        "        optimizer.zero_grad()\n",
        "        outputs = model(inputs)\n",
        "\n",
        "        loss = criterion(outputs, labels.float().view(-1, 1))\n",
        "        loss.backward()\n",
        "        optimizer.step()\n",
        "        total_steps = int(129600/batch_size)\n",
        "        print(f'Epoch:[{epoch+1}/{num_epochs}] Step:[{i+1}/{total_steps+1}] Step_loss:{loss.item():.5f}')\n",
        "        wandb.log({\"Step_Loss\": loss.item()})\n",
        "    # Print progress\n",
        "    print(f'Epoch [{epoch+1}/{num_epochs}], Loss: {loss.item()}')\n",
        "    # 모델 평가\n",
        "    model.eval()\n",
        "    all_predictions = []\n",
        "    all_labels = []\n",
        "    with torch.no_grad():\n",
        "      for inputs, labels in valid_loader:\n",
        "        inputs, labels = inputs.cuda(), labels.cuda()\n",
        "        outputs = model(inputs).cuda()\n",
        "        predictions = (outputs > 0.5).float()\n",
        "    all_predictions.append(predictions)\n",
        "    all_labels.append(labels)\n",
        "    all_predictions = torch.cat(all_predictions).cpu()\n",
        "    all_labels = torch.cat(all_labels).cpu()\n",
        "    accuracy = accuracy_score(all_labels, all_predictions)\n",
        "    print(f'Validation Accuracy: {accuracy}')\n",
        "    f1 = f1_score(all_labels, all_predictions)\n",
        "    print(f'Validation F1 Score: {f1}')\n",
        "    wandb.log({\"Step_Loss\": loss.item(),\"Validation Acc\":accuracy, \"F1 Score\":f1 })"
      ],
      "metadata": {
        "id": "74dutkal9DtG",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "868d9390-2e6d-459c-cbda-a75d097d92dd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch:[1/50] Step:[1/507] Step_loss:0.92611\n",
            "Epoch:[1/50] Step:[2/507] Step_loss:17.96501\n",
            "Epoch:[1/50] Step:[3/507] Step_loss:28.02427\n",
            "Epoch:[1/50] Step:[4/507] Step_loss:29.39998\n",
            "Epoch:[1/50] Step:[5/507] Step_loss:26.54171\n",
            "Epoch:[1/50] Step:[6/507] Step_loss:32.07275\n",
            "Epoch:[1/50] Step:[7/507] Step_loss:33.72956\n",
            "Epoch:[1/50] Step:[8/507] Step_loss:30.71145\n",
            "Epoch:[1/50] Step:[9/507] Step_loss:30.26371\n",
            "Epoch:[1/50] Step:[10/507] Step_loss:34.73898\n",
            "Epoch:[1/50] Step:[11/507] Step_loss:34.69295\n",
            "Epoch:[1/50] Step:[12/507] Step_loss:34.23483\n",
            "Epoch:[1/50] Step:[13/507] Step_loss:34.02637\n",
            "Epoch:[1/50] Step:[14/507] Step_loss:30.99093\n",
            "Epoch:[1/50] Step:[15/507] Step_loss:32.75654\n",
            "Epoch:[1/50] Step:[16/507] Step_loss:28.81564\n",
            "Epoch:[1/50] Step:[17/507] Step_loss:28.92200\n",
            "Epoch:[1/50] Step:[18/507] Step_loss:30.63656\n",
            "Epoch:[1/50] Step:[19/507] Step_loss:32.38949\n",
            "Epoch:[1/50] Step:[20/507] Step_loss:34.04531\n",
            "Epoch:[1/50] Step:[21/507] Step_loss:29.78942\n",
            "Epoch:[1/50] Step:[22/507] Step_loss:28.24821\n",
            "Epoch:[1/50] Step:[23/507] Step_loss:28.28394\n",
            "Epoch:[1/50] Step:[24/507] Step_loss:32.44166\n",
            "Epoch:[1/50] Step:[25/507] Step_loss:29.94911\n",
            "Epoch:[1/50] Step:[26/507] Step_loss:28.06347\n",
            "Epoch:[1/50] Step:[27/507] Step_loss:37.52262\n",
            "Epoch:[1/50] Step:[28/507] Step_loss:33.68245\n",
            "Epoch:[1/50] Step:[29/507] Step_loss:31.82625\n",
            "Epoch:[1/50] Step:[30/507] Step_loss:32.74332\n",
            "Epoch:[1/50] Step:[31/507] Step_loss:30.91573\n",
            "Epoch:[1/50] Step:[32/507] Step_loss:29.83210\n",
            "Epoch:[1/50] Step:[33/507] Step_loss:33.05833\n",
            "Epoch:[1/50] Step:[34/507] Step_loss:32.35764\n",
            "Epoch:[1/50] Step:[35/507] Step_loss:33.38229\n",
            "Epoch:[1/50] Step:[36/507] Step_loss:31.27427\n",
            "Epoch:[1/50] Step:[37/507] Step_loss:35.55132\n",
            "Epoch:[1/50] Step:[38/507] Step_loss:34.79531\n",
            "Epoch:[1/50] Step:[39/507] Step_loss:33.01151\n",
            "Epoch:[1/50] Step:[40/507] Step_loss:33.20810\n",
            "Epoch:[1/50] Step:[41/507] Step_loss:31.79251\n",
            "Epoch:[1/50] Step:[42/507] Step_loss:34.28722\n",
            "Epoch:[1/50] Step:[43/507] Step_loss:32.59697\n",
            "Epoch:[1/50] Step:[44/507] Step_loss:24.39297\n",
            "Epoch:[1/50] Step:[45/507] Step_loss:32.61852\n",
            "Epoch:[1/50] Step:[46/507] Step_loss:32.40999\n",
            "Epoch:[1/50] Step:[47/507] Step_loss:29.30087\n",
            "Epoch:[1/50] Step:[48/507] Step_loss:28.62955\n",
            "Epoch:[1/50] Step:[49/507] Step_loss:33.01716\n",
            "Epoch:[1/50] Step:[50/507] Step_loss:26.00561\n",
            "Epoch:[1/50] Step:[51/507] Step_loss:25.43959\n",
            "Epoch:[1/50] Step:[52/507] Step_loss:28.34930\n",
            "Epoch:[1/50] Step:[53/507] Step_loss:28.11903\n",
            "Epoch:[1/50] Step:[54/507] Step_loss:21.28223\n",
            "Epoch:[1/50] Step:[55/507] Step_loss:25.31226\n",
            "Epoch:[1/50] Step:[56/507] Step_loss:27.92940\n",
            "Epoch:[1/50] Step:[57/507] Step_loss:26.51125\n",
            "Epoch:[1/50] Step:[58/507] Step_loss:33.54762\n",
            "Epoch:[1/50] Step:[59/507] Step_loss:28.13580\n",
            "Epoch:[1/50] Step:[60/507] Step_loss:29.33953\n",
            "Epoch:[1/50] Step:[61/507] Step_loss:29.60202\n",
            "Epoch:[1/50] Step:[62/507] Step_loss:26.40533\n",
            "Epoch:[1/50] Step:[63/507] Step_loss:27.86842\n",
            "Epoch:[1/50] Step:[64/507] Step_loss:24.63657\n",
            "Epoch:[1/50] Step:[65/507] Step_loss:23.67387\n",
            "Epoch:[1/50] Step:[66/507] Step_loss:24.68282\n",
            "Epoch:[1/50] Step:[67/507] Step_loss:23.16006\n",
            "Epoch:[1/50] Step:[68/507] Step_loss:20.67647\n",
            "Epoch:[1/50] Step:[69/507] Step_loss:15.56627\n",
            "Epoch:[1/50] Step:[70/507] Step_loss:18.84995\n",
            "Epoch:[1/50] Step:[71/507] Step_loss:19.97908\n",
            "Epoch:[1/50] Step:[72/507] Step_loss:25.70021\n",
            "Epoch:[1/50] Step:[73/507] Step_loss:24.50501\n",
            "Epoch:[1/50] Step:[74/507] Step_loss:31.28442\n",
            "Epoch:[1/50] Step:[75/507] Step_loss:35.20430\n",
            "Epoch:[1/50] Step:[76/507] Step_loss:33.91978\n",
            "Epoch:[1/50] Step:[77/507] Step_loss:29.49848\n",
            "Epoch:[1/50] Step:[78/507] Step_loss:26.32557\n",
            "Epoch:[1/50] Step:[79/507] Step_loss:18.67106\n",
            "Epoch:[1/50] Step:[80/507] Step_loss:9.74836\n",
            "Epoch:[1/50] Step:[81/507] Step_loss:10.40807\n",
            "Epoch:[1/50] Step:[82/507] Step_loss:12.97014\n",
            "Epoch:[1/50] Step:[83/507] Step_loss:20.10498\n",
            "Epoch:[1/50] Step:[84/507] Step_loss:28.29266\n",
            "Epoch:[1/50] Step:[85/507] Step_loss:26.74813\n",
            "Epoch:[1/50] Step:[86/507] Step_loss:17.88354\n",
            "Epoch:[1/50] Step:[87/507] Step_loss:13.72224\n",
            "Epoch:[1/50] Step:[88/507] Step_loss:9.06506\n",
            "Epoch:[1/50] Step:[89/507] Step_loss:8.81968\n",
            "Epoch:[1/50] Step:[90/507] Step_loss:8.61392\n",
            "Epoch:[1/50] Step:[91/507] Step_loss:18.23205\n",
            "Epoch:[1/50] Step:[92/507] Step_loss:21.37077\n",
            "Epoch:[1/50] Step:[93/507] Step_loss:23.52138\n",
            "Epoch:[1/50] Step:[94/507] Step_loss:15.39310\n",
            "Epoch:[1/50] Step:[95/507] Step_loss:12.48295\n",
            "Epoch:[1/50] Step:[96/507] Step_loss:10.13402\n",
            "Epoch:[1/50] Step:[97/507] Step_loss:8.59303\n",
            "Epoch:[1/50] Step:[98/507] Step_loss:5.96533\n",
            "Epoch:[1/50] Step:[99/507] Step_loss:9.25312\n",
            "Epoch:[1/50] Step:[100/507] Step_loss:13.01657\n",
            "Epoch:[1/50] Step:[101/507] Step_loss:19.23004\n",
            "Epoch:[1/50] Step:[102/507] Step_loss:19.19162\n",
            "Epoch:[1/50] Step:[103/507] Step_loss:12.97668\n",
            "Epoch:[1/50] Step:[104/507] Step_loss:9.78840\n",
            "Epoch:[1/50] Step:[105/507] Step_loss:7.51038\n",
            "Epoch:[1/50] Step:[106/507] Step_loss:9.36928\n",
            "Epoch:[1/50] Step:[107/507] Step_loss:4.23162\n",
            "Epoch:[1/50] Step:[108/507] Step_loss:7.70447\n",
            "Epoch:[1/50] Step:[109/507] Step_loss:7.51899\n",
            "Epoch:[1/50] Step:[110/507] Step_loss:11.36247\n",
            "Epoch:[1/50] Step:[111/507] Step_loss:6.20753\n",
            "Epoch:[1/50] Step:[112/507] Step_loss:6.64942\n",
            "Epoch:[1/50] Step:[113/507] Step_loss:5.93941\n",
            "Epoch:[1/50] Step:[114/507] Step_loss:6.62042\n",
            "Epoch:[1/50] Step:[115/507] Step_loss:7.04383\n",
            "Epoch:[1/50] Step:[116/507] Step_loss:6.99548\n",
            "Epoch:[1/50] Step:[117/507] Step_loss:8.56467\n",
            "Epoch:[1/50] Step:[118/507] Step_loss:7.30877\n",
            "Epoch:[1/50] Step:[119/507] Step_loss:6.22988\n",
            "Epoch:[1/50] Step:[120/507] Step_loss:5.97125\n",
            "Epoch:[1/50] Step:[121/507] Step_loss:5.79874\n",
            "Epoch:[1/50] Step:[122/507] Step_loss:5.67384\n",
            "Epoch:[1/50] Step:[123/507] Step_loss:7.26565\n",
            "Epoch:[1/50] Step:[124/507] Step_loss:12.62330\n",
            "Epoch:[1/50] Step:[125/507] Step_loss:11.89493\n",
            "Epoch:[1/50] Step:[126/507] Step_loss:7.91636\n",
            "Epoch:[1/50] Step:[127/507] Step_loss:6.62266\n",
            "Epoch:[1/50] Step:[128/507] Step_loss:5.74803\n",
            "Epoch:[1/50] Step:[129/507] Step_loss:5.97066\n",
            "Epoch:[1/50] Step:[130/507] Step_loss:10.03120\n",
            "Epoch:[1/50] Step:[131/507] Step_loss:9.11483\n",
            "Epoch:[1/50] Step:[132/507] Step_loss:3.93907\n",
            "Epoch:[1/50] Step:[133/507] Step_loss:5.49733\n",
            "Epoch:[1/50] Step:[134/507] Step_loss:4.96593\n",
            "Epoch:[1/50] Step:[135/507] Step_loss:5.51588\n",
            "Epoch:[1/50] Step:[136/507] Step_loss:6.35289\n",
            "Epoch:[1/50] Step:[137/507] Step_loss:6.16147\n",
            "Epoch:[1/50] Step:[138/507] Step_loss:6.15155\n",
            "Epoch:[1/50] Step:[139/507] Step_loss:6.12096\n",
            "Epoch:[1/50] Step:[140/507] Step_loss:5.99878\n",
            "Epoch:[1/50] Step:[141/507] Step_loss:5.39830\n",
            "Epoch:[1/50] Step:[142/507] Step_loss:4.79124\n",
            "Epoch:[1/50] Step:[143/507] Step_loss:6.14693\n",
            "Epoch:[1/50] Step:[144/507] Step_loss:6.38969\n",
            "Epoch:[1/50] Step:[145/507] Step_loss:3.46339\n",
            "Epoch:[1/50] Step:[146/507] Step_loss:5.62094\n",
            "Epoch:[1/50] Step:[147/507] Step_loss:5.04719\n",
            "Epoch:[1/50] Step:[148/507] Step_loss:5.06141\n",
            "Epoch:[1/50] Step:[149/507] Step_loss:6.30240\n",
            "Epoch:[1/50] Step:[150/507] Step_loss:7.53268\n",
            "Epoch:[1/50] Step:[151/507] Step_loss:5.84713\n",
            "Epoch:[1/50] Step:[152/507] Step_loss:4.51679\n",
            "Epoch:[1/50] Step:[153/507] Step_loss:5.79714\n",
            "Epoch:[1/50] Step:[154/507] Step_loss:5.12860\n",
            "Epoch:[1/50] Step:[155/507] Step_loss:5.60337\n",
            "Epoch:[1/50] Step:[156/507] Step_loss:7.00513\n",
            "Epoch:[1/50] Step:[157/507] Step_loss:5.77549\n",
            "Epoch:[1/50] Step:[158/507] Step_loss:7.17724\n",
            "Epoch:[1/50] Step:[159/507] Step_loss:6.69492\n",
            "Epoch:[1/50] Step:[160/507] Step_loss:4.74565\n",
            "Epoch:[1/50] Step:[161/507] Step_loss:5.54885\n",
            "Epoch:[1/50] Step:[162/507] Step_loss:5.17696\n",
            "Epoch:[1/50] Step:[163/507] Step_loss:3.70041\n",
            "Epoch:[1/50] Step:[164/507] Step_loss:4.61016\n",
            "Epoch:[1/50] Step:[165/507] Step_loss:3.95670\n",
            "Epoch:[1/50] Step:[166/507] Step_loss:3.74288\n",
            "Epoch:[1/50] Step:[167/507] Step_loss:3.15144\n",
            "Epoch:[1/50] Step:[168/507] Step_loss:4.76511\n",
            "Epoch:[1/50] Step:[169/507] Step_loss:6.02930\n",
            "Epoch:[1/50] Step:[170/507] Step_loss:3.89361\n",
            "Epoch:[1/50] Step:[171/507] Step_loss:5.61248\n",
            "Epoch:[1/50] Step:[172/507] Step_loss:3.30935\n",
            "Epoch:[1/50] Step:[173/507] Step_loss:2.89106\n",
            "Epoch:[1/50] Step:[174/507] Step_loss:5.17167\n",
            "Epoch:[1/50] Step:[175/507] Step_loss:4.11209\n",
            "Epoch:[1/50] Step:[176/507] Step_loss:3.96491\n",
            "Epoch:[1/50] Step:[177/507] Step_loss:3.26520\n",
            "Epoch:[1/50] Step:[178/507] Step_loss:4.57134\n",
            "Epoch:[1/50] Step:[179/507] Step_loss:3.19261\n",
            "Epoch:[1/50] Step:[180/507] Step_loss:2.93406\n",
            "Epoch:[1/50] Step:[181/507] Step_loss:3.76886\n",
            "Epoch:[1/50] Step:[182/507] Step_loss:5.90026\n",
            "Epoch:[1/50] Step:[183/507] Step_loss:2.74273\n",
            "Epoch:[1/50] Step:[184/507] Step_loss:4.23037\n",
            "Epoch:[1/50] Step:[185/507] Step_loss:3.60169\n",
            "Epoch:[1/50] Step:[186/507] Step_loss:4.76283\n",
            "Epoch:[1/50] Step:[187/507] Step_loss:6.27309\n",
            "Epoch:[1/50] Step:[188/507] Step_loss:3.41997\n",
            "Epoch:[1/50] Step:[189/507] Step_loss:3.94913\n",
            "Epoch:[1/50] Step:[190/507] Step_loss:4.68968\n",
            "Epoch:[1/50] Step:[191/507] Step_loss:4.96434\n",
            "Epoch:[1/50] Step:[192/507] Step_loss:4.70638\n",
            "Epoch:[1/50] Step:[193/507] Step_loss:2.04756\n",
            "Epoch:[1/50] Step:[194/507] Step_loss:4.21817\n",
            "Epoch:[1/50] Step:[195/507] Step_loss:2.83263\n",
            "Epoch:[1/50] Step:[196/507] Step_loss:3.90032\n",
            "Epoch:[1/50] Step:[197/507] Step_loss:6.64316\n",
            "Epoch:[1/50] Step:[198/507] Step_loss:2.57438\n",
            "Epoch:[1/50] Step:[199/507] Step_loss:2.69501\n",
            "Epoch:[1/50] Step:[200/507] Step_loss:4.70036\n",
            "Epoch:[1/50] Step:[201/507] Step_loss:3.24096\n",
            "Epoch:[1/50] Step:[202/507] Step_loss:4.14258\n",
            "Epoch:[1/50] Step:[203/507] Step_loss:2.96619\n",
            "Epoch:[1/50] Step:[204/507] Step_loss:2.21719\n",
            "Epoch:[1/50] Step:[205/507] Step_loss:3.36806\n",
            "Epoch:[1/50] Step:[206/507] Step_loss:3.32886\n",
            "Epoch:[1/50] Step:[207/507] Step_loss:2.97723\n",
            "Epoch:[1/50] Step:[208/507] Step_loss:2.87672\n",
            "Epoch:[1/50] Step:[209/507] Step_loss:3.21946\n",
            "Epoch:[1/50] Step:[210/507] Step_loss:3.38829\n",
            "Epoch:[1/50] Step:[211/507] Step_loss:3.95113\n",
            "Epoch:[1/50] Step:[212/507] Step_loss:5.76735\n",
            "Epoch:[1/50] Step:[213/507] Step_loss:2.04309\n",
            "Epoch:[1/50] Step:[214/507] Step_loss:2.97689\n",
            "Epoch:[1/50] Step:[215/507] Step_loss:2.79256\n",
            "Epoch:[1/50] Step:[216/507] Step_loss:3.19151\n",
            "Epoch:[1/50] Step:[217/507] Step_loss:4.86707\n",
            "Epoch:[1/50] Step:[218/507] Step_loss:2.99453\n",
            "Epoch:[1/50] Step:[219/507] Step_loss:2.94790\n",
            "Epoch:[1/50] Step:[220/507] Step_loss:2.36783\n",
            "Epoch:[1/50] Step:[221/507] Step_loss:3.56491\n",
            "Epoch:[1/50] Step:[222/507] Step_loss:3.98279\n",
            "Epoch:[1/50] Step:[223/507] Step_loss:5.16863\n",
            "Epoch:[1/50] Step:[224/507] Step_loss:2.31808\n",
            "Epoch:[1/50] Step:[225/507] Step_loss:2.92254\n",
            "Epoch:[1/50] Step:[226/507] Step_loss:2.23477\n",
            "Epoch:[1/50] Step:[227/507] Step_loss:5.99232\n",
            "Epoch:[1/50] Step:[228/507] Step_loss:6.89509\n",
            "Epoch:[1/50] Step:[229/507] Step_loss:2.96621\n",
            "Epoch:[1/50] Step:[230/507] Step_loss:1.83197\n",
            "Epoch:[1/50] Step:[231/507] Step_loss:2.68319\n",
            "Epoch:[1/50] Step:[232/507] Step_loss:5.11894\n",
            "Epoch:[1/50] Step:[233/507] Step_loss:7.31092\n",
            "Epoch:[1/50] Step:[234/507] Step_loss:5.41534\n",
            "Epoch:[1/50] Step:[235/507] Step_loss:4.14642\n",
            "Epoch:[1/50] Step:[236/507] Step_loss:3.84945\n",
            "Epoch:[1/50] Step:[237/507] Step_loss:2.38346\n",
            "Epoch:[1/50] Step:[238/507] Step_loss:4.85729\n",
            "Epoch:[1/50] Step:[239/507] Step_loss:6.44695\n",
            "Epoch:[1/50] Step:[240/507] Step_loss:3.13412\n",
            "Epoch:[1/50] Step:[241/507] Step_loss:1.99244\n",
            "Epoch:[1/50] Step:[242/507] Step_loss:4.64823\n",
            "Epoch:[1/50] Step:[243/507] Step_loss:1.80297\n",
            "Epoch:[1/50] Step:[244/507] Step_loss:5.41778\n",
            "Epoch:[1/50] Step:[245/507] Step_loss:4.62681\n",
            "Epoch:[1/50] Step:[246/507] Step_loss:3.76307\n",
            "Epoch:[1/50] Step:[247/507] Step_loss:2.60401\n",
            "Epoch:[1/50] Step:[248/507] Step_loss:1.73992\n",
            "Epoch:[1/50] Step:[249/507] Step_loss:5.79286\n",
            "Epoch:[1/50] Step:[250/507] Step_loss:3.38524\n",
            "Epoch:[1/50] Step:[251/507] Step_loss:2.00778\n",
            "Epoch:[1/50] Step:[252/507] Step_loss:1.25079\n",
            "Epoch:[1/50] Step:[253/507] Step_loss:3.01153\n",
            "Epoch:[1/50] Step:[254/507] Step_loss:3.15938\n",
            "Epoch:[1/50] Step:[255/507] Step_loss:4.31002\n",
            "Epoch:[1/50] Step:[256/507] Step_loss:1.36425\n",
            "Epoch:[1/50] Step:[257/507] Step_loss:3.11848\n",
            "Epoch:[1/50] Step:[258/507] Step_loss:2.28397\n",
            "Epoch:[1/50] Step:[259/507] Step_loss:2.71198\n",
            "Epoch:[1/50] Step:[260/507] Step_loss:2.20367\n",
            "Epoch:[1/50] Step:[261/507] Step_loss:2.84278\n",
            "Epoch:[1/50] Step:[262/507] Step_loss:2.72556\n",
            "Epoch:[1/50] Step:[263/507] Step_loss:1.56079\n",
            "Epoch:[1/50] Step:[264/507] Step_loss:2.50402\n",
            "Epoch:[1/50] Step:[265/507] Step_loss:2.56844\n",
            "Epoch:[1/50] Step:[266/507] Step_loss:4.66357\n",
            "Epoch:[1/50] Step:[267/507] Step_loss:3.73470\n",
            "Epoch:[1/50] Step:[268/507] Step_loss:3.30302\n",
            "Epoch:[1/50] Step:[269/507] Step_loss:3.12708\n",
            "Epoch:[1/50] Step:[270/507] Step_loss:2.43492\n",
            "Epoch:[1/50] Step:[271/507] Step_loss:3.00805\n",
            "Epoch:[1/50] Step:[272/507] Step_loss:1.45751\n",
            "Epoch:[1/50] Step:[273/507] Step_loss:3.29807\n",
            "Epoch:[1/50] Step:[274/507] Step_loss:2.04684\n",
            "Epoch:[1/50] Step:[275/507] Step_loss:3.02976\n",
            "Epoch:[1/50] Step:[276/507] Step_loss:2.45750\n",
            "Epoch:[1/50] Step:[277/507] Step_loss:3.02472\n",
            "Epoch:[1/50] Step:[278/507] Step_loss:1.11688\n",
            "Epoch:[1/50] Step:[279/507] Step_loss:1.58531\n",
            "Epoch:[1/50] Step:[280/507] Step_loss:2.09165\n",
            "Epoch:[1/50] Step:[281/507] Step_loss:3.04142\n",
            "Epoch:[1/50] Step:[282/507] Step_loss:3.46727\n",
            "Epoch:[1/50] Step:[283/507] Step_loss:2.07321\n",
            "Epoch:[1/50] Step:[284/507] Step_loss:1.53825\n",
            "Epoch:[1/50] Step:[285/507] Step_loss:1.08048\n",
            "Epoch:[1/50] Step:[286/507] Step_loss:2.57221\n",
            "Epoch:[1/50] Step:[287/507] Step_loss:2.28836\n",
            "Epoch:[1/50] Step:[288/507] Step_loss:3.11222\n",
            "Epoch:[1/50] Step:[289/507] Step_loss:1.37344\n",
            "Epoch:[1/50] Step:[290/507] Step_loss:3.17950\n",
            "Epoch:[1/50] Step:[291/507] Step_loss:1.46037\n",
            "Epoch:[1/50] Step:[292/507] Step_loss:1.00352\n",
            "Epoch:[1/50] Step:[293/507] Step_loss:3.11486\n",
            "Epoch:[1/50] Step:[294/507] Step_loss:1.64939\n",
            "Epoch:[1/50] Step:[295/507] Step_loss:2.82286\n",
            "Epoch:[1/50] Step:[296/507] Step_loss:1.60020\n",
            "Epoch:[1/50] Step:[297/507] Step_loss:4.50341\n",
            "Epoch:[1/50] Step:[298/507] Step_loss:3.19506\n",
            "Epoch:[1/50] Step:[299/507] Step_loss:3.51352\n",
            "Epoch:[1/50] Step:[300/507] Step_loss:1.97046\n",
            "Epoch:[1/50] Step:[301/507] Step_loss:1.48875\n",
            "Epoch:[1/50] Step:[302/507] Step_loss:1.53113\n",
            "Epoch:[1/50] Step:[303/507] Step_loss:3.14394\n",
            "Epoch:[1/50] Step:[304/507] Step_loss:3.05371\n",
            "Epoch:[1/50] Step:[305/507] Step_loss:2.79744\n",
            "Epoch:[1/50] Step:[306/507] Step_loss:3.07208\n",
            "Epoch:[1/50] Step:[307/507] Step_loss:1.16550\n",
            "Epoch:[1/50] Step:[308/507] Step_loss:3.24557\n",
            "Epoch:[1/50] Step:[309/507] Step_loss:2.13058\n",
            "Epoch:[1/50] Step:[310/507] Step_loss:2.18990\n",
            "Epoch:[1/50] Step:[311/507] Step_loss:2.97801\n",
            "Epoch:[1/50] Step:[312/507] Step_loss:1.53431\n",
            "Epoch:[1/50] Step:[313/507] Step_loss:1.96889\n",
            "Epoch:[1/50] Step:[314/507] Step_loss:1.98787\n",
            "Epoch:[1/50] Step:[315/507] Step_loss:1.84116\n",
            "Epoch:[1/50] Step:[316/507] Step_loss:2.87419\n",
            "Epoch:[1/50] Step:[317/507] Step_loss:2.24376\n",
            "Epoch:[1/50] Step:[318/507] Step_loss:2.56879\n",
            "Epoch:[1/50] Step:[319/507] Step_loss:1.45585\n",
            "Epoch:[1/50] Step:[320/507] Step_loss:2.27891\n",
            "Epoch:[1/50] Step:[321/507] Step_loss:1.97838\n",
            "Epoch:[1/50] Step:[322/507] Step_loss:0.85934\n",
            "Epoch:[1/50] Step:[323/507] Step_loss:3.88165\n",
            "Epoch:[1/50] Step:[324/507] Step_loss:1.00591\n",
            "Epoch:[1/50] Step:[325/507] Step_loss:1.54966\n",
            "Epoch:[1/50] Step:[326/507] Step_loss:1.20623\n",
            "Epoch:[1/50] Step:[327/507] Step_loss:2.78212\n",
            "Epoch:[1/50] Step:[328/507] Step_loss:1.26858\n",
            "Epoch:[1/50] Step:[329/507] Step_loss:1.91665\n",
            "Epoch:[1/50] Step:[330/507] Step_loss:2.67017\n",
            "Epoch:[1/50] Step:[331/507] Step_loss:1.71618\n",
            "Epoch:[1/50] Step:[332/507] Step_loss:1.69502\n",
            "Epoch:[1/50] Step:[333/507] Step_loss:1.87905\n",
            "Epoch:[1/50] Step:[334/507] Step_loss:1.15751\n",
            "Epoch:[1/50] Step:[335/507] Step_loss:1.91683\n",
            "Epoch:[1/50] Step:[336/507] Step_loss:2.29406\n",
            "Epoch:[1/50] Step:[337/507] Step_loss:2.90863\n",
            "Epoch:[1/50] Step:[338/507] Step_loss:1.04099\n",
            "Epoch:[1/50] Step:[339/507] Step_loss:2.19316\n",
            "Epoch:[1/50] Step:[340/507] Step_loss:2.09609\n",
            "Epoch:[1/50] Step:[341/507] Step_loss:1.54241\n",
            "Epoch:[1/50] Step:[342/507] Step_loss:2.99688\n",
            "Epoch:[1/50] Step:[343/507] Step_loss:1.33235\n",
            "Epoch:[1/50] Step:[344/507] Step_loss:1.60309\n",
            "Epoch:[1/50] Step:[345/507] Step_loss:2.21865\n",
            "Epoch:[1/50] Step:[346/507] Step_loss:1.77592\n",
            "Epoch:[1/50] Step:[347/507] Step_loss:1.74770\n",
            "Epoch:[1/50] Step:[348/507] Step_loss:2.49574\n",
            "Epoch:[1/50] Step:[349/507] Step_loss:1.88444\n",
            "Epoch:[1/50] Step:[350/507] Step_loss:1.52206\n",
            "Epoch:[1/50] Step:[351/507] Step_loss:0.88929\n",
            "Epoch:[1/50] Step:[352/507] Step_loss:2.49620\n",
            "Epoch:[1/50] Step:[353/507] Step_loss:2.15080\n",
            "Epoch:[1/50] Step:[354/507] Step_loss:1.90408\n",
            "Epoch:[1/50] Step:[355/507] Step_loss:2.53713\n",
            "Epoch:[1/50] Step:[356/507] Step_loss:1.15943\n",
            "Epoch:[1/50] Step:[357/507] Step_loss:1.19485\n",
            "Epoch:[1/50] Step:[358/507] Step_loss:0.85711\n",
            "Epoch:[1/50] Step:[359/507] Step_loss:1.72009\n",
            "Epoch:[1/50] Step:[360/507] Step_loss:1.26768\n",
            "Epoch:[1/50] Step:[361/507] Step_loss:0.83057\n",
            "Epoch:[1/50] Step:[362/507] Step_loss:1.12722\n",
            "Epoch:[1/50] Step:[363/507] Step_loss:1.33289\n",
            "Epoch:[1/50] Step:[364/507] Step_loss:1.43974\n",
            "Epoch:[1/50] Step:[365/507] Step_loss:0.80294\n",
            "Epoch:[1/50] Step:[366/507] Step_loss:2.03684\n",
            "Epoch:[1/50] Step:[367/507] Step_loss:0.72075\n",
            "Epoch:[1/50] Step:[368/507] Step_loss:1.18788\n",
            "Epoch:[1/50] Step:[369/507] Step_loss:1.11456\n",
            "Epoch:[1/50] Step:[370/507] Step_loss:2.11674\n",
            "Epoch:[1/50] Step:[371/507] Step_loss:1.15062\n",
            "Epoch:[1/50] Step:[372/507] Step_loss:0.84876\n",
            "Epoch:[1/50] Step:[373/507] Step_loss:1.57108\n",
            "Epoch:[1/50] Step:[374/507] Step_loss:1.13873\n",
            "Epoch:[1/50] Step:[375/507] Step_loss:2.89710\n",
            "Epoch:[1/50] Step:[376/507] Step_loss:1.39762\n",
            "Epoch:[1/50] Step:[377/507] Step_loss:1.46731\n",
            "Epoch:[1/50] Step:[378/507] Step_loss:1.75947\n",
            "Epoch:[1/50] Step:[379/507] Step_loss:2.41514\n",
            "Epoch:[1/50] Step:[380/507] Step_loss:0.90709\n",
            "Epoch:[1/50] Step:[381/507] Step_loss:1.95711\n",
            "Epoch:[1/50] Step:[382/507] Step_loss:1.40523\n",
            "Epoch:[1/50] Step:[383/507] Step_loss:1.71672\n",
            "Epoch:[1/50] Step:[384/507] Step_loss:1.47623\n",
            "Epoch:[1/50] Step:[385/507] Step_loss:1.17616\n",
            "Epoch:[1/50] Step:[386/507] Step_loss:0.71004\n",
            "Epoch:[1/50] Step:[387/507] Step_loss:1.22165\n",
            "Epoch:[1/50] Step:[388/507] Step_loss:1.06883\n",
            "Epoch:[1/50] Step:[389/507] Step_loss:0.71391\n",
            "Epoch:[1/50] Step:[390/507] Step_loss:1.41523\n",
            "Epoch:[1/50] Step:[391/507] Step_loss:1.65320\n",
            "Epoch:[1/50] Step:[392/507] Step_loss:1.86203\n",
            "Epoch:[1/50] Step:[393/507] Step_loss:1.52043\n",
            "Epoch:[1/50] Step:[394/507] Step_loss:1.62533\n",
            "Epoch:[1/50] Step:[395/507] Step_loss:2.31918\n",
            "Epoch:[1/50] Step:[396/507] Step_loss:1.33171\n",
            "Epoch:[1/50] Step:[397/507] Step_loss:1.80649\n",
            "Epoch:[1/50] Step:[398/507] Step_loss:1.39107\n",
            "Epoch:[1/50] Step:[399/507] Step_loss:1.52419\n",
            "Epoch:[1/50] Step:[400/507] Step_loss:0.60676\n",
            "Epoch:[1/50] Step:[401/507] Step_loss:2.06950\n",
            "Epoch:[1/50] Step:[402/507] Step_loss:1.80053\n",
            "Epoch:[1/50] Step:[403/507] Step_loss:0.97065\n",
            "Epoch:[1/50] Step:[404/507] Step_loss:1.09843\n",
            "Epoch:[1/50] Step:[405/507] Step_loss:0.90966\n",
            "Epoch:[1/50] Step:[406/507] Step_loss:1.67525\n",
            "Epoch:[1/50] Step:[407/507] Step_loss:1.01427\n",
            "Epoch:[1/50] Step:[408/507] Step_loss:1.07018\n",
            "Epoch:[1/50] Step:[409/507] Step_loss:1.26123\n",
            "Epoch:[1/50] Step:[410/507] Step_loss:0.92358\n",
            "Epoch:[1/50] Step:[411/507] Step_loss:0.75123\n",
            "Epoch:[1/50] Step:[412/507] Step_loss:0.80864\n",
            "Epoch:[1/50] Step:[413/507] Step_loss:2.43034\n",
            "Epoch:[1/50] Step:[414/507] Step_loss:0.73967\n",
            "Epoch:[1/50] Step:[415/507] Step_loss:2.69512\n",
            "Epoch:[1/50] Step:[416/507] Step_loss:1.14368\n",
            "Epoch:[1/50] Step:[417/507] Step_loss:1.60990\n",
            "Epoch:[1/50] Step:[418/507] Step_loss:1.82549\n",
            "Epoch:[1/50] Step:[419/507] Step_loss:1.66112\n",
            "Epoch:[1/50] Step:[420/507] Step_loss:1.17843\n",
            "Epoch:[1/50] Step:[421/507] Step_loss:2.43580\n",
            "Epoch:[1/50] Step:[422/507] Step_loss:1.84346\n",
            "Epoch:[1/50] Step:[423/507] Step_loss:1.74972\n",
            "Epoch:[1/50] Step:[424/507] Step_loss:0.95733\n",
            "Epoch:[1/50] Step:[425/507] Step_loss:1.73874\n",
            "Epoch:[1/50] Step:[426/507] Step_loss:1.58781\n",
            "Epoch:[1/50] Step:[427/507] Step_loss:1.26615\n",
            "Epoch:[1/50] Step:[428/507] Step_loss:0.93681\n",
            "Epoch:[1/50] Step:[429/507] Step_loss:1.15098\n",
            "Epoch:[1/50] Step:[430/507] Step_loss:0.92149\n",
            "Epoch:[1/50] Step:[431/507] Step_loss:1.13854\n",
            "Epoch:[1/50] Step:[432/507] Step_loss:1.26096\n",
            "Epoch:[1/50] Step:[433/507] Step_loss:1.45568\n",
            "Epoch:[1/50] Step:[434/507] Step_loss:1.69419\n",
            "Epoch:[1/50] Step:[435/507] Step_loss:1.54674\n",
            "Epoch:[1/50] Step:[436/507] Step_loss:1.00193\n",
            "Epoch:[1/50] Step:[437/507] Step_loss:0.91411\n",
            "Epoch:[1/50] Step:[438/507] Step_loss:0.90873\n",
            "Epoch:[1/50] Step:[439/507] Step_loss:0.88485\n",
            "Epoch:[1/50] Step:[440/507] Step_loss:1.00824\n",
            "Epoch:[1/50] Step:[441/507] Step_loss:0.86211\n",
            "Epoch:[1/50] Step:[442/507] Step_loss:1.15494\n",
            "Epoch:[1/50] Step:[443/507] Step_loss:1.01235\n",
            "Epoch:[1/50] Step:[444/507] Step_loss:1.26224\n",
            "Epoch:[1/50] Step:[445/507] Step_loss:1.44127\n",
            "Epoch:[1/50] Step:[446/507] Step_loss:0.74421\n",
            "Epoch:[1/50] Step:[447/507] Step_loss:0.82201\n",
            "Epoch:[1/50] Step:[448/507] Step_loss:1.57184\n",
            "Epoch:[1/50] Step:[449/507] Step_loss:1.12224\n",
            "Epoch:[1/50] Step:[450/507] Step_loss:0.87346\n",
            "Epoch:[1/50] Step:[451/507] Step_loss:0.87326\n",
            "Epoch:[1/50] Step:[452/507] Step_loss:0.98655\n",
            "Epoch:[1/50] Step:[453/507] Step_loss:0.78332\n",
            "Epoch:[1/50] Step:[454/507] Step_loss:1.49061\n",
            "Epoch:[1/50] Step:[455/507] Step_loss:1.54528\n",
            "Epoch:[1/50] Step:[456/507] Step_loss:1.09332\n",
            "Epoch:[1/50] Step:[457/507] Step_loss:1.47448\n",
            "Epoch:[1/50] Step:[458/507] Step_loss:1.06882\n",
            "Epoch:[1/50] Step:[459/507] Step_loss:0.80238\n",
            "Epoch:[1/50] Step:[460/507] Step_loss:0.74793\n",
            "Epoch:[1/50] Step:[461/507] Step_loss:1.30561\n",
            "Epoch:[1/50] Step:[462/507] Step_loss:1.41186\n",
            "Epoch:[1/50] Step:[463/507] Step_loss:1.23626\n",
            "Epoch:[1/50] Step:[464/507] Step_loss:1.32021\n",
            "Epoch:[1/50] Step:[465/507] Step_loss:0.88194\n",
            "Epoch:[1/50] Step:[466/507] Step_loss:0.99428\n",
            "Epoch:[1/50] Step:[467/507] Step_loss:1.35391\n",
            "Epoch:[1/50] Step:[468/507] Step_loss:1.18079\n",
            "Epoch:[1/50] Step:[469/507] Step_loss:0.84952\n",
            "Epoch:[1/50] Step:[470/507] Step_loss:1.00799\n",
            "Epoch:[1/50] Step:[471/507] Step_loss:1.40627\n",
            "Epoch:[1/50] Step:[472/507] Step_loss:1.00473\n",
            "Epoch:[1/50] Step:[473/507] Step_loss:2.89443\n",
            "Epoch:[1/50] Step:[474/507] Step_loss:0.81322\n",
            "Epoch:[1/50] Step:[475/507] Step_loss:2.80365\n",
            "Epoch:[1/50] Step:[476/507] Step_loss:0.83903\n",
            "Epoch:[1/50] Step:[477/507] Step_loss:2.22877\n",
            "Epoch:[1/50] Step:[478/507] Step_loss:0.95060\n",
            "Epoch:[1/50] Step:[479/507] Step_loss:1.32690\n",
            "Epoch:[1/50] Step:[480/507] Step_loss:1.52121\n",
            "Epoch:[1/50] Step:[481/507] Step_loss:2.67477\n",
            "Epoch:[1/50] Step:[482/507] Step_loss:2.53134\n",
            "Epoch:[1/50] Step:[483/507] Step_loss:2.30193\n",
            "Epoch:[1/50] Step:[484/507] Step_loss:2.74622\n",
            "Epoch:[1/50] Step:[485/507] Step_loss:2.97292\n",
            "Epoch:[1/50] Step:[486/507] Step_loss:1.68141\n",
            "Epoch:[1/50] Step:[487/507] Step_loss:1.91996\n",
            "Epoch:[1/50] Step:[488/507] Step_loss:1.38936\n",
            "Epoch:[1/50] Step:[489/507] Step_loss:1.50205\n",
            "Epoch:[1/50] Step:[490/507] Step_loss:2.10311\n",
            "Epoch:[1/50] Step:[491/507] Step_loss:1.77018\n",
            "Epoch:[1/50] Step:[492/507] Step_loss:1.74421\n",
            "Epoch:[1/50] Step:[493/507] Step_loss:2.21668\n",
            "Epoch:[1/50] Step:[494/507] Step_loss:4.29334\n",
            "Epoch:[1/50] Step:[495/507] Step_loss:1.57612\n",
            "Epoch:[1/50] Step:[496/507] Step_loss:2.44599\n",
            "Epoch:[1/50] Step:[497/507] Step_loss:2.02818\n",
            "Epoch:[1/50] Step:[498/507] Step_loss:1.93965\n",
            "Epoch:[1/50] Step:[499/507] Step_loss:2.66897\n",
            "Epoch:[1/50] Step:[500/507] Step_loss:1.96058\n",
            "Epoch:[1/50] Step:[501/507] Step_loss:1.56164\n",
            "Epoch:[1/50] Step:[502/507] Step_loss:2.57596\n",
            "Epoch:[1/50] Step:[503/507] Step_loss:1.47773\n",
            "Epoch:[1/50] Step:[504/507] Step_loss:1.81897\n",
            "Epoch:[1/50] Step:[505/507] Step_loss:0.89378\n",
            "Epoch:[1/50] Step:[506/507] Step_loss:2.68417\n",
            "Epoch:[1/50] Step:[507/507] Step_loss:0.83558\n",
            "Epoch [1/50], Loss: 0.835576057434082\n",
            "Validation Accuracy: 0.6388888888888888\n",
            "Validation F1 Score: 0.7796610169491525\n",
            "Epoch:[2/50] Step:[1/507] Step_loss:0.80731\n",
            "Epoch:[2/50] Step:[2/507] Step_loss:3.00143\n",
            "Epoch:[2/50] Step:[3/507] Step_loss:1.89791\n",
            "Epoch:[2/50] Step:[4/507] Step_loss:1.97817\n",
            "Epoch:[2/50] Step:[5/507] Step_loss:1.20411\n",
            "Epoch:[2/50] Step:[6/507] Step_loss:2.92139\n",
            "Epoch:[2/50] Step:[7/507] Step_loss:1.73959\n",
            "Epoch:[2/50] Step:[8/507] Step_loss:1.19739\n",
            "Epoch:[2/50] Step:[9/507] Step_loss:1.85242\n",
            "Epoch:[2/50] Step:[10/507] Step_loss:1.33927\n",
            "Epoch:[2/50] Step:[11/507] Step_loss:1.81729\n",
            "Epoch:[2/50] Step:[12/507] Step_loss:1.61638\n",
            "Epoch:[2/50] Step:[13/507] Step_loss:1.84715\n",
            "Epoch:[2/50] Step:[14/507] Step_loss:2.09218\n",
            "Epoch:[2/50] Step:[15/507] Step_loss:0.92745\n",
            "Epoch:[2/50] Step:[16/507] Step_loss:1.30077\n",
            "Epoch:[2/50] Step:[17/507] Step_loss:1.68029\n",
            "Epoch:[2/50] Step:[18/507] Step_loss:1.17470\n",
            "Epoch:[2/50] Step:[19/507] Step_loss:1.09315\n",
            "Epoch:[2/50] Step:[20/507] Step_loss:1.62476\n",
            "Epoch:[2/50] Step:[21/507] Step_loss:1.42844\n",
            "Epoch:[2/50] Step:[22/507] Step_loss:2.92032\n",
            "Epoch:[2/50] Step:[23/507] Step_loss:1.73711\n",
            "Epoch:[2/50] Step:[24/507] Step_loss:1.94157\n",
            "Epoch:[2/50] Step:[25/507] Step_loss:1.87755\n",
            "Epoch:[2/50] Step:[26/507] Step_loss:2.49575\n",
            "Epoch:[2/50] Step:[27/507] Step_loss:0.91843\n",
            "Epoch:[2/50] Step:[28/507] Step_loss:2.44327\n",
            "Epoch:[2/50] Step:[29/507] Step_loss:0.72752\n",
            "Epoch:[2/50] Step:[30/507] Step_loss:1.24884\n",
            "Epoch:[2/50] Step:[31/507] Step_loss:1.91002\n",
            "Epoch:[2/50] Step:[32/507] Step_loss:1.72281\n",
            "Epoch:[2/50] Step:[33/507] Step_loss:1.20823\n",
            "Epoch:[2/50] Step:[34/507] Step_loss:2.28713\n",
            "Epoch:[2/50] Step:[35/507] Step_loss:0.82891\n",
            "Epoch:[2/50] Step:[36/507] Step_loss:1.53356\n",
            "Epoch:[2/50] Step:[37/507] Step_loss:1.36161\n",
            "Epoch:[2/50] Step:[38/507] Step_loss:0.92988\n",
            "Epoch:[2/50] Step:[39/507] Step_loss:1.20160\n",
            "Epoch:[2/50] Step:[40/507] Step_loss:2.01955\n",
            "Epoch:[2/50] Step:[41/507] Step_loss:1.26757\n",
            "Epoch:[2/50] Step:[42/507] Step_loss:1.16253\n",
            "Epoch:[2/50] Step:[43/507] Step_loss:1.64908\n",
            "Epoch:[2/50] Step:[44/507] Step_loss:1.34496\n",
            "Epoch:[2/50] Step:[45/507] Step_loss:1.52965\n",
            "Epoch:[2/50] Step:[46/507] Step_loss:1.84050\n",
            "Epoch:[2/50] Step:[47/507] Step_loss:1.49851\n",
            "Epoch:[2/50] Step:[48/507] Step_loss:1.07116\n",
            "Epoch:[2/50] Step:[49/507] Step_loss:3.83271\n",
            "Epoch:[2/50] Step:[50/507] Step_loss:1.09933\n",
            "Epoch:[2/50] Step:[51/507] Step_loss:2.01188\n",
            "Epoch:[2/50] Step:[52/507] Step_loss:1.74332\n",
            "Epoch:[2/50] Step:[53/507] Step_loss:1.60296\n",
            "Epoch:[2/50] Step:[54/507] Step_loss:1.34406\n",
            "Epoch:[2/50] Step:[55/507] Step_loss:0.99706\n",
            "Epoch:[2/50] Step:[56/507] Step_loss:1.45575\n",
            "Epoch:[2/50] Step:[57/507] Step_loss:1.31064\n",
            "Epoch:[2/50] Step:[58/507] Step_loss:2.49319\n",
            "Epoch:[2/50] Step:[59/507] Step_loss:1.77228\n",
            "Epoch:[2/50] Step:[60/507] Step_loss:1.41876\n",
            "Epoch:[2/50] Step:[61/507] Step_loss:2.28221\n",
            "Epoch:[2/50] Step:[62/507] Step_loss:2.22110\n",
            "Epoch:[2/50] Step:[63/507] Step_loss:2.10574\n",
            "Epoch:[2/50] Step:[64/507] Step_loss:0.74620\n",
            "Epoch:[2/50] Step:[65/507] Step_loss:1.49775\n",
            "Epoch:[2/50] Step:[66/507] Step_loss:1.63482\n",
            "Epoch:[2/50] Step:[67/507] Step_loss:1.98260\n",
            "Epoch:[2/50] Step:[68/507] Step_loss:0.98004\n",
            "Epoch:[2/50] Step:[69/507] Step_loss:1.56891\n",
            "Epoch:[2/50] Step:[70/507] Step_loss:1.62377\n",
            "Epoch:[2/50] Step:[71/507] Step_loss:1.08905\n",
            "Epoch:[2/50] Step:[72/507] Step_loss:1.22042\n",
            "Epoch:[2/50] Step:[73/507] Step_loss:1.52020\n",
            "Epoch:[2/50] Step:[74/507] Step_loss:1.57692\n",
            "Epoch:[2/50] Step:[75/507] Step_loss:1.66368\n",
            "Epoch:[2/50] Step:[76/507] Step_loss:1.52605\n",
            "Epoch:[2/50] Step:[77/507] Step_loss:1.23679\n",
            "Epoch:[2/50] Step:[78/507] Step_loss:1.32458\n",
            "Epoch:[2/50] Step:[79/507] Step_loss:1.74764\n",
            "Epoch:[2/50] Step:[80/507] Step_loss:1.10560\n",
            "Epoch:[2/50] Step:[81/507] Step_loss:1.25540\n",
            "Epoch:[2/50] Step:[82/507] Step_loss:0.98920\n",
            "Epoch:[2/50] Step:[83/507] Step_loss:0.97032\n",
            "Epoch:[2/50] Step:[84/507] Step_loss:1.73983\n",
            "Epoch:[2/50] Step:[85/507] Step_loss:1.01191\n",
            "Epoch:[2/50] Step:[86/507] Step_loss:1.03296\n",
            "Epoch:[2/50] Step:[87/507] Step_loss:0.97327\n",
            "Epoch:[2/50] Step:[88/507] Step_loss:1.14630\n",
            "Epoch:[2/50] Step:[89/507] Step_loss:0.87056\n",
            "Epoch:[2/50] Step:[90/507] Step_loss:1.54997\n",
            "Epoch:[2/50] Step:[91/507] Step_loss:1.50637\n",
            "Epoch:[2/50] Step:[92/507] Step_loss:1.03027\n",
            "Epoch:[2/50] Step:[93/507] Step_loss:1.11075\n",
            "Epoch:[2/50] Step:[94/507] Step_loss:1.07877\n",
            "Epoch:[2/50] Step:[95/507] Step_loss:1.42811\n",
            "Epoch:[2/50] Step:[96/507] Step_loss:1.25391\n",
            "Epoch:[2/50] Step:[97/507] Step_loss:1.37703\n",
            "Epoch:[2/50] Step:[98/507] Step_loss:0.69321\n",
            "Epoch:[2/50] Step:[99/507] Step_loss:1.11837\n",
            "Epoch:[2/50] Step:[100/507] Step_loss:1.66315\n",
            "Epoch:[2/50] Step:[101/507] Step_loss:1.58983\n",
            "Epoch:[2/50] Step:[102/507] Step_loss:1.20364\n",
            "Epoch:[2/50] Step:[103/507] Step_loss:1.68731\n",
            "Epoch:[2/50] Step:[104/507] Step_loss:0.97084\n",
            "Epoch:[2/50] Step:[105/507] Step_loss:0.88522\n",
            "Epoch:[2/50] Step:[106/507] Step_loss:1.45807\n",
            "Epoch:[2/50] Step:[107/507] Step_loss:1.27420\n",
            "Epoch:[2/50] Step:[108/507] Step_loss:1.21897\n",
            "Epoch:[2/50] Step:[109/507] Step_loss:1.53852\n",
            "Epoch:[2/50] Step:[110/507] Step_loss:1.70982\n",
            "Epoch:[2/50] Step:[111/507] Step_loss:1.02503\n",
            "Epoch:[2/50] Step:[112/507] Step_loss:1.48950\n",
            "Epoch:[2/50] Step:[113/507] Step_loss:1.69132\n",
            "Epoch:[2/50] Step:[114/507] Step_loss:0.60667\n",
            "Epoch:[2/50] Step:[115/507] Step_loss:1.26023\n",
            "Epoch:[2/50] Step:[116/507] Step_loss:2.02723\n",
            "Epoch:[2/50] Step:[117/507] Step_loss:0.89540\n",
            "Epoch:[2/50] Step:[118/507] Step_loss:0.61320\n",
            "Epoch:[2/50] Step:[119/507] Step_loss:1.79480\n",
            "Epoch:[2/50] Step:[120/507] Step_loss:1.27121\n",
            "Epoch:[2/50] Step:[121/507] Step_loss:0.90080\n",
            "Epoch:[2/50] Step:[122/507] Step_loss:1.06103\n",
            "Epoch:[2/50] Step:[123/507] Step_loss:0.92740\n",
            "Epoch:[2/50] Step:[124/507] Step_loss:1.41279\n",
            "Epoch:[2/50] Step:[125/507] Step_loss:1.31683\n",
            "Epoch:[2/50] Step:[126/507] Step_loss:1.10007\n",
            "Epoch:[2/50] Step:[127/507] Step_loss:0.80650\n",
            "Epoch:[2/50] Step:[128/507] Step_loss:0.94752\n",
            "Epoch:[2/50] Step:[129/507] Step_loss:1.39110\n",
            "Epoch:[2/50] Step:[130/507] Step_loss:0.74575\n",
            "Epoch:[2/50] Step:[131/507] Step_loss:1.33122\n",
            "Epoch:[2/50] Step:[132/507] Step_loss:0.56531\n",
            "Epoch:[2/50] Step:[133/507] Step_loss:0.79229\n",
            "Epoch:[2/50] Step:[134/507] Step_loss:0.79666\n",
            "Epoch:[2/50] Step:[135/507] Step_loss:0.98684\n",
            "Epoch:[2/50] Step:[136/507] Step_loss:0.74012\n",
            "Epoch:[2/50] Step:[137/507] Step_loss:1.44063\n",
            "Epoch:[2/50] Step:[138/507] Step_loss:0.89216\n",
            "Epoch:[2/50] Step:[139/507] Step_loss:0.95554\n",
            "Epoch:[2/50] Step:[140/507] Step_loss:0.90611\n",
            "Epoch:[2/50] Step:[141/507] Step_loss:0.88254\n",
            "Epoch:[2/50] Step:[142/507] Step_loss:1.16141\n",
            "Epoch:[2/50] Step:[143/507] Step_loss:0.86438\n",
            "Epoch:[2/50] Step:[144/507] Step_loss:1.09465\n",
            "Epoch:[2/50] Step:[145/507] Step_loss:0.99398\n",
            "Epoch:[2/50] Step:[146/507] Step_loss:0.86529\n",
            "Epoch:[2/50] Step:[147/507] Step_loss:0.71884\n",
            "Epoch:[2/50] Step:[148/507] Step_loss:0.60127\n",
            "Epoch:[2/50] Step:[149/507] Step_loss:1.02882\n",
            "Epoch:[2/50] Step:[150/507] Step_loss:0.98560\n",
            "Epoch:[2/50] Step:[151/507] Step_loss:0.73204\n",
            "Epoch:[2/50] Step:[152/507] Step_loss:1.00157\n",
            "Epoch:[2/50] Step:[153/507] Step_loss:0.75427\n",
            "Epoch:[2/50] Step:[154/507] Step_loss:0.97000\n",
            "Epoch:[2/50] Step:[155/507] Step_loss:0.58223\n",
            "Epoch:[2/50] Step:[156/507] Step_loss:1.48224\n",
            "Epoch:[2/50] Step:[157/507] Step_loss:0.78306\n",
            "Epoch:[2/50] Step:[158/507] Step_loss:0.62523\n",
            "Epoch:[2/50] Step:[159/507] Step_loss:1.16300\n",
            "Epoch:[2/50] Step:[160/507] Step_loss:1.28267\n",
            "Epoch:[2/50] Step:[161/507] Step_loss:1.70974\n",
            "Epoch:[2/50] Step:[162/507] Step_loss:1.00557\n",
            "Epoch:[2/50] Step:[163/507] Step_loss:0.78198\n",
            "Epoch:[2/50] Step:[164/507] Step_loss:1.85565\n",
            "Epoch:[2/50] Step:[165/507] Step_loss:2.58938\n",
            "Epoch:[2/50] Step:[166/507] Step_loss:0.97621\n",
            "Epoch:[2/50] Step:[167/507] Step_loss:1.05319\n",
            "Epoch:[2/50] Step:[168/507] Step_loss:1.19149\n",
            "Epoch:[2/50] Step:[169/507] Step_loss:0.86920\n",
            "Epoch:[2/50] Step:[170/507] Step_loss:0.84911\n",
            "Epoch:[2/50] Step:[171/507] Step_loss:1.02380\n",
            "Epoch:[2/50] Step:[172/507] Step_loss:1.34127\n",
            "Epoch:[2/50] Step:[173/507] Step_loss:0.89718\n",
            "Epoch:[2/50] Step:[174/507] Step_loss:1.32553\n",
            "Epoch:[2/50] Step:[175/507] Step_loss:0.77175\n",
            "Epoch:[2/50] Step:[176/507] Step_loss:0.79350\n",
            "Epoch:[2/50] Step:[177/507] Step_loss:1.64338\n",
            "Epoch:[2/50] Step:[178/507] Step_loss:1.42816\n",
            "Epoch:[2/50] Step:[179/507] Step_loss:0.66426\n",
            "Epoch:[2/50] Step:[180/507] Step_loss:1.06074\n",
            "Epoch:[2/50] Step:[181/507] Step_loss:1.12909\n",
            "Epoch:[2/50] Step:[182/507] Step_loss:0.57047\n",
            "Epoch:[2/50] Step:[183/507] Step_loss:0.69467\n",
            "Epoch:[2/50] Step:[184/507] Step_loss:1.06841\n",
            "Epoch:[2/50] Step:[185/507] Step_loss:0.72875\n",
            "Epoch:[2/50] Step:[186/507] Step_loss:0.91812\n",
            "Epoch:[2/50] Step:[187/507] Step_loss:0.93238\n",
            "Epoch:[2/50] Step:[188/507] Step_loss:0.66144\n",
            "Epoch:[2/50] Step:[189/507] Step_loss:1.33415\n",
            "Epoch:[2/50] Step:[190/507] Step_loss:0.74704\n",
            "Epoch:[2/50] Step:[191/507] Step_loss:1.41858\n",
            "Epoch:[2/50] Step:[192/507] Step_loss:1.23392\n",
            "Epoch:[2/50] Step:[193/507] Step_loss:1.70091\n",
            "Epoch:[2/50] Step:[194/507] Step_loss:0.69827\n",
            "Epoch:[2/50] Step:[195/507] Step_loss:1.02619\n",
            "Epoch:[2/50] Step:[196/507] Step_loss:1.46821\n",
            "Epoch:[2/50] Step:[197/507] Step_loss:1.48110\n",
            "Epoch:[2/50] Step:[198/507] Step_loss:1.15164\n",
            "Epoch:[2/50] Step:[199/507] Step_loss:1.19386\n",
            "Epoch:[2/50] Step:[200/507] Step_loss:0.65019\n",
            "Epoch:[2/50] Step:[201/507] Step_loss:2.01366\n",
            "Epoch:[2/50] Step:[202/507] Step_loss:1.18499\n",
            "Epoch:[2/50] Step:[203/507] Step_loss:1.40060\n",
            "Epoch:[2/50] Step:[204/507] Step_loss:1.20638\n",
            "Epoch:[2/50] Step:[205/507] Step_loss:2.73679\n",
            "Epoch:[2/50] Step:[206/507] Step_loss:1.00016\n",
            "Epoch:[2/50] Step:[207/507] Step_loss:1.54454\n",
            "Epoch:[2/50] Step:[208/507] Step_loss:1.62813\n",
            "Epoch:[2/50] Step:[209/507] Step_loss:2.00736\n",
            "Epoch:[2/50] Step:[210/507] Step_loss:1.11580\n",
            "Epoch:[2/50] Step:[211/507] Step_loss:1.31366\n",
            "Epoch:[2/50] Step:[212/507] Step_loss:1.18309\n",
            "Epoch:[2/50] Step:[213/507] Step_loss:1.11269\n",
            "Epoch:[2/50] Step:[214/507] Step_loss:1.72435\n",
            "Epoch:[2/50] Step:[215/507] Step_loss:1.76886\n",
            "Epoch:[2/50] Step:[216/507] Step_loss:0.78210\n",
            "Epoch:[2/50] Step:[217/507] Step_loss:0.69414\n",
            "Epoch:[2/50] Step:[218/507] Step_loss:2.34898\n",
            "Epoch:[2/50] Step:[219/507] Step_loss:0.59339\n",
            "Epoch:[2/50] Step:[220/507] Step_loss:1.65857\n",
            "Epoch:[2/50] Step:[221/507] Step_loss:1.08391\n",
            "Epoch:[2/50] Step:[222/507] Step_loss:1.58073\n",
            "Epoch:[2/50] Step:[223/507] Step_loss:0.92962\n",
            "Epoch:[2/50] Step:[224/507] Step_loss:1.27138\n",
            "Epoch:[2/50] Step:[225/507] Step_loss:1.54114\n",
            "Epoch:[2/50] Step:[226/507] Step_loss:1.30728\n",
            "Epoch:[2/50] Step:[227/507] Step_loss:0.88598\n",
            "Epoch:[2/50] Step:[228/507] Step_loss:1.45972\n",
            "Epoch:[2/50] Step:[229/507] Step_loss:1.49722\n",
            "Epoch:[2/50] Step:[230/507] Step_loss:1.60486\n",
            "Epoch:[2/50] Step:[231/507] Step_loss:2.04505\n",
            "Epoch:[2/50] Step:[232/507] Step_loss:1.14872\n",
            "Epoch:[2/50] Step:[233/507] Step_loss:0.76561\n",
            "Epoch:[2/50] Step:[234/507] Step_loss:1.04297\n",
            "Epoch:[2/50] Step:[235/507] Step_loss:1.12234\n",
            "Epoch:[2/50] Step:[236/507] Step_loss:2.66504\n",
            "Epoch:[2/50] Step:[237/507] Step_loss:1.27120\n",
            "Epoch:[2/50] Step:[238/507] Step_loss:1.78318\n",
            "Epoch:[2/50] Step:[239/507] Step_loss:0.67836\n",
            "Epoch:[2/50] Step:[240/507] Step_loss:3.48670\n",
            "Epoch:[2/50] Step:[241/507] Step_loss:1.50600\n",
            "Epoch:[2/50] Step:[242/507] Step_loss:1.13461\n",
            "Epoch:[2/50] Step:[243/507] Step_loss:1.41517\n",
            "Epoch:[2/50] Step:[244/507] Step_loss:0.71582\n",
            "Epoch:[2/50] Step:[245/507] Step_loss:3.45139\n",
            "Epoch:[2/50] Step:[246/507] Step_loss:1.34293\n",
            "Epoch:[2/50] Step:[247/507] Step_loss:1.87054\n",
            "Epoch:[2/50] Step:[248/507] Step_loss:0.87853\n",
            "Epoch:[2/50] Step:[249/507] Step_loss:1.47921\n",
            "Epoch:[2/50] Step:[250/507] Step_loss:1.01285\n",
            "Epoch:[2/50] Step:[251/507] Step_loss:1.66798\n",
            "Epoch:[2/50] Step:[252/507] Step_loss:1.74072\n",
            "Epoch:[2/50] Step:[253/507] Step_loss:1.02474\n",
            "Epoch:[2/50] Step:[254/507] Step_loss:2.48155\n",
            "Epoch:[2/50] Step:[255/507] Step_loss:1.00452\n",
            "Epoch:[2/50] Step:[256/507] Step_loss:1.85191\n",
            "Epoch:[2/50] Step:[257/507] Step_loss:0.55482\n",
            "Epoch:[2/50] Step:[258/507] Step_loss:1.09427\n",
            "Epoch:[2/50] Step:[259/507] Step_loss:1.24811\n",
            "Epoch:[2/50] Step:[260/507] Step_loss:1.26311\n",
            "Epoch:[2/50] Step:[261/507] Step_loss:1.34288\n",
            "Epoch:[2/50] Step:[262/507] Step_loss:1.35192\n",
            "Epoch:[2/50] Step:[263/507] Step_loss:2.51078\n",
            "Epoch:[2/50] Step:[264/507] Step_loss:1.08452\n",
            "Epoch:[2/50] Step:[265/507] Step_loss:1.23766\n",
            "Epoch:[2/50] Step:[266/507] Step_loss:1.46078\n",
            "Epoch:[2/50] Step:[267/507] Step_loss:0.94000\n",
            "Epoch:[2/50] Step:[268/507] Step_loss:1.95412\n",
            "Epoch:[2/50] Step:[269/507] Step_loss:0.87669\n",
            "Epoch:[2/50] Step:[270/507] Step_loss:0.97684\n",
            "Epoch:[2/50] Step:[271/507] Step_loss:0.73182\n",
            "Epoch:[2/50] Step:[272/507] Step_loss:1.10058\n",
            "Epoch:[2/50] Step:[273/507] Step_loss:0.87022\n",
            "Epoch:[2/50] Step:[274/507] Step_loss:0.79762\n",
            "Epoch:[2/50] Step:[275/507] Step_loss:1.47426\n",
            "Epoch:[2/50] Step:[276/507] Step_loss:0.94210\n",
            "Epoch:[2/50] Step:[277/507] Step_loss:0.76615\n",
            "Epoch:[2/50] Step:[278/507] Step_loss:1.07952\n",
            "Epoch:[2/50] Step:[279/507] Step_loss:0.74152\n",
            "Epoch:[2/50] Step:[280/507] Step_loss:0.99619\n",
            "Epoch:[2/50] Step:[281/507] Step_loss:2.50008\n",
            "Epoch:[2/50] Step:[282/507] Step_loss:0.84467\n",
            "Epoch:[2/50] Step:[283/507] Step_loss:0.69144\n",
            "Epoch:[2/50] Step:[284/507] Step_loss:1.78321\n",
            "Epoch:[2/50] Step:[285/507] Step_loss:0.72758\n",
            "Epoch:[2/50] Step:[286/507] Step_loss:1.75827\n",
            "Epoch:[2/50] Step:[287/507] Step_loss:0.87105\n",
            "Epoch:[2/50] Step:[288/507] Step_loss:0.94028\n",
            "Epoch:[2/50] Step:[289/507] Step_loss:1.26360\n",
            "Epoch:[2/50] Step:[290/507] Step_loss:1.02795\n",
            "Epoch:[2/50] Step:[291/507] Step_loss:1.05230\n",
            "Epoch:[2/50] Step:[292/507] Step_loss:1.23303\n",
            "Epoch:[2/50] Step:[293/507] Step_loss:1.30751\n",
            "Epoch:[2/50] Step:[294/507] Step_loss:1.11813\n",
            "Epoch:[2/50] Step:[295/507] Step_loss:0.60946\n",
            "Epoch:[2/50] Step:[296/507] Step_loss:0.88591\n",
            "Epoch:[2/50] Step:[297/507] Step_loss:1.03286\n",
            "Epoch:[2/50] Step:[298/507] Step_loss:0.88084\n",
            "Epoch:[2/50] Step:[299/507] Step_loss:0.92273\n",
            "Epoch:[2/50] Step:[300/507] Step_loss:0.74391\n",
            "Epoch:[2/50] Step:[301/507] Step_loss:2.16656\n",
            "Epoch:[2/50] Step:[302/507] Step_loss:0.55176\n",
            "Epoch:[2/50] Step:[303/507] Step_loss:1.11029\n",
            "Epoch:[2/50] Step:[304/507] Step_loss:0.65945\n",
            "Epoch:[2/50] Step:[305/507] Step_loss:1.32537\n",
            "Epoch:[2/50] Step:[306/507] Step_loss:0.68014\n",
            "Epoch:[2/50] Step:[307/507] Step_loss:0.83418\n",
            "Epoch:[2/50] Step:[308/507] Step_loss:0.68470\n",
            "Epoch:[2/50] Step:[309/507] Step_loss:0.65603\n",
            "Epoch:[2/50] Step:[310/507] Step_loss:1.34200\n",
            "Epoch:[2/50] Step:[311/507] Step_loss:0.56751\n",
            "Epoch:[2/50] Step:[312/507] Step_loss:0.83478\n",
            "Epoch:[2/50] Step:[313/507] Step_loss:1.53469\n",
            "Epoch:[2/50] Step:[314/507] Step_loss:0.78708\n",
            "Epoch:[2/50] Step:[315/507] Step_loss:0.59134\n",
            "Epoch:[2/50] Step:[316/507] Step_loss:0.88091\n",
            "Epoch:[2/50] Step:[317/507] Step_loss:0.61285\n",
            "Epoch:[2/50] Step:[318/507] Step_loss:0.72631\n",
            "Epoch:[2/50] Step:[319/507] Step_loss:0.85362\n",
            "Epoch:[2/50] Step:[320/507] Step_loss:0.73598\n",
            "Epoch:[2/50] Step:[321/507] Step_loss:1.41856\n",
            "Epoch:[2/50] Step:[322/507] Step_loss:0.93619\n",
            "Epoch:[2/50] Step:[323/507] Step_loss:1.34528\n",
            "Epoch:[2/50] Step:[324/507] Step_loss:0.64846\n",
            "Epoch:[2/50] Step:[325/507] Step_loss:1.07563\n",
            "Epoch:[2/50] Step:[326/507] Step_loss:0.63619\n",
            "Epoch:[2/50] Step:[327/507] Step_loss:0.92514\n",
            "Epoch:[2/50] Step:[328/507] Step_loss:1.02270\n",
            "Epoch:[2/50] Step:[329/507] Step_loss:0.74082\n",
            "Epoch:[2/50] Step:[330/507] Step_loss:0.73306\n",
            "Epoch:[2/50] Step:[331/507] Step_loss:1.17971\n",
            "Epoch:[2/50] Step:[332/507] Step_loss:0.77856\n",
            "Epoch:[2/50] Step:[333/507] Step_loss:1.02656\n",
            "Epoch:[2/50] Step:[334/507] Step_loss:0.66911\n",
            "Epoch:[2/50] Step:[335/507] Step_loss:1.30127\n",
            "Epoch:[2/50] Step:[336/507] Step_loss:1.19240\n",
            "Epoch:[2/50] Step:[337/507] Step_loss:1.56614\n",
            "Epoch:[2/50] Step:[338/507] Step_loss:1.28924\n",
            "Epoch:[2/50] Step:[339/507] Step_loss:0.63192\n",
            "Epoch:[2/50] Step:[340/507] Step_loss:1.77486\n",
            "Epoch:[2/50] Step:[341/507] Step_loss:0.74892\n",
            "Epoch:[2/50] Step:[342/507] Step_loss:1.13245\n",
            "Epoch:[2/50] Step:[343/507] Step_loss:1.06279\n",
            "Epoch:[2/50] Step:[344/507] Step_loss:1.02469\n",
            "Epoch:[2/50] Step:[345/507] Step_loss:0.97290\n",
            "Epoch:[2/50] Step:[346/507] Step_loss:1.23597\n",
            "Epoch:[2/50] Step:[347/507] Step_loss:1.21992\n",
            "Epoch:[2/50] Step:[348/507] Step_loss:0.62781\n",
            "Epoch:[2/50] Step:[349/507] Step_loss:1.70141\n",
            "Epoch:[2/50] Step:[350/507] Step_loss:1.11245\n",
            "Epoch:[2/50] Step:[351/507] Step_loss:0.67987\n",
            "Epoch:[2/50] Step:[352/507] Step_loss:1.02830\n",
            "Epoch:[2/50] Step:[353/507] Step_loss:0.84895\n",
            "Epoch:[2/50] Step:[354/507] Step_loss:0.79920\n",
            "Epoch:[2/50] Step:[355/507] Step_loss:0.59439\n",
            "Epoch:[2/50] Step:[356/507] Step_loss:0.83393\n",
            "Epoch:[2/50] Step:[357/507] Step_loss:0.83445\n",
            "Epoch:[2/50] Step:[358/507] Step_loss:1.49566\n",
            "Epoch:[2/50] Step:[359/507] Step_loss:0.97054\n",
            "Epoch:[2/50] Step:[360/507] Step_loss:0.72477\n",
            "Epoch:[2/50] Step:[361/507] Step_loss:3.28220\n",
            "Epoch:[2/50] Step:[362/507] Step_loss:0.80032\n",
            "Epoch:[2/50] Step:[363/507] Step_loss:0.97850\n",
            "Epoch:[2/50] Step:[364/507] Step_loss:0.89150\n",
            "Epoch:[2/50] Step:[365/507] Step_loss:0.83714\n",
            "Epoch:[2/50] Step:[366/507] Step_loss:1.06333\n",
            "Epoch:[2/50] Step:[367/507] Step_loss:1.43856\n",
            "Epoch:[2/50] Step:[368/507] Step_loss:1.29512\n",
            "Epoch:[2/50] Step:[369/507] Step_loss:0.75861\n",
            "Epoch:[2/50] Step:[370/507] Step_loss:1.10799\n",
            "Epoch:[2/50] Step:[371/507] Step_loss:0.97220\n",
            "Epoch:[2/50] Step:[372/507] Step_loss:0.91618\n",
            "Epoch:[2/50] Step:[373/507] Step_loss:0.89536\n",
            "Epoch:[2/50] Step:[374/507] Step_loss:1.39383\n",
            "Epoch:[2/50] Step:[375/507] Step_loss:1.07760\n",
            "Epoch:[2/50] Step:[376/507] Step_loss:2.21005\n",
            "Epoch:[2/50] Step:[377/507] Step_loss:1.07423\n",
            "Epoch:[2/50] Step:[378/507] Step_loss:2.44670\n",
            "Epoch:[2/50] Step:[379/507] Step_loss:1.07439\n",
            "Epoch:[2/50] Step:[380/507] Step_loss:1.36134\n",
            "Epoch:[2/50] Step:[381/507] Step_loss:1.91968\n",
            "Epoch:[2/50] Step:[382/507] Step_loss:1.50280\n",
            "Epoch:[2/50] Step:[383/507] Step_loss:2.87094\n",
            "Epoch:[2/50] Step:[384/507] Step_loss:1.05074\n",
            "Epoch:[2/50] Step:[385/507] Step_loss:2.66221\n",
            "Epoch:[2/50] Step:[386/507] Step_loss:1.20271\n",
            "Epoch:[2/50] Step:[387/507] Step_loss:2.02995\n",
            "Epoch:[2/50] Step:[388/507] Step_loss:1.34333\n",
            "Epoch:[2/50] Step:[389/507] Step_loss:2.30311\n",
            "Epoch:[2/50] Step:[390/507] Step_loss:1.72860\n",
            "Epoch:[2/50] Step:[391/507] Step_loss:1.28172\n",
            "Epoch:[2/50] Step:[392/507] Step_loss:1.30915\n",
            "Epoch:[2/50] Step:[393/507] Step_loss:1.53796\n",
            "Epoch:[2/50] Step:[394/507] Step_loss:0.74845\n",
            "Epoch:[2/50] Step:[395/507] Step_loss:1.49914\n",
            "Epoch:[2/50] Step:[396/507] Step_loss:1.55317\n",
            "Epoch:[2/50] Step:[397/507] Step_loss:0.93917\n",
            "Epoch:[2/50] Step:[398/507] Step_loss:1.42414\n",
            "Epoch:[2/50] Step:[399/507] Step_loss:0.76963\n",
            "Epoch:[2/50] Step:[400/507] Step_loss:1.31952\n",
            "Epoch:[2/50] Step:[401/507] Step_loss:1.13492\n",
            "Epoch:[2/50] Step:[402/507] Step_loss:0.87789\n",
            "Epoch:[2/50] Step:[403/507] Step_loss:0.89928\n",
            "Epoch:[2/50] Step:[404/507] Step_loss:1.41511\n",
            "Epoch:[2/50] Step:[405/507] Step_loss:0.84958\n",
            "Epoch:[2/50] Step:[406/507] Step_loss:1.25596\n",
            "Epoch:[2/50] Step:[407/507] Step_loss:1.79370\n",
            "Epoch:[2/50] Step:[408/507] Step_loss:1.23886\n",
            "Epoch:[2/50] Step:[409/507] Step_loss:2.55764\n",
            "Epoch:[2/50] Step:[410/507] Step_loss:0.81915\n",
            "Epoch:[2/50] Step:[411/507] Step_loss:1.02662\n",
            "Epoch:[2/50] Step:[412/507] Step_loss:1.05714\n",
            "Epoch:[2/50] Step:[413/507] Step_loss:1.88943\n",
            "Epoch:[2/50] Step:[414/507] Step_loss:1.71558\n",
            "Epoch:[2/50] Step:[415/507] Step_loss:0.87286\n",
            "Epoch:[2/50] Step:[416/507] Step_loss:1.11158\n",
            "Epoch:[2/50] Step:[417/507] Step_loss:1.20309\n",
            "Epoch:[2/50] Step:[418/507] Step_loss:1.66493\n",
            "Epoch:[2/50] Step:[419/507] Step_loss:1.00770\n",
            "Epoch:[2/50] Step:[420/507] Step_loss:1.31944\n",
            "Epoch:[2/50] Step:[421/507] Step_loss:1.83764\n",
            "Epoch:[2/50] Step:[422/507] Step_loss:0.66302\n",
            "Epoch:[2/50] Step:[423/507] Step_loss:4.19271\n",
            "Epoch:[2/50] Step:[424/507] Step_loss:1.63603\n",
            "Epoch:[2/50] Step:[425/507] Step_loss:2.03297\n",
            "Epoch:[2/50] Step:[426/507] Step_loss:0.98718\n",
            "Epoch:[2/50] Step:[427/507] Step_loss:0.79079\n",
            "Epoch:[2/50] Step:[428/507] Step_loss:2.13097\n",
            "Epoch:[2/50] Step:[429/507] Step_loss:1.15403\n",
            "Epoch:[2/50] Step:[430/507] Step_loss:1.22999\n",
            "Epoch:[2/50] Step:[431/507] Step_loss:2.52245\n",
            "Epoch:[2/50] Step:[432/507] Step_loss:0.88035\n",
            "Epoch:[2/50] Step:[433/507] Step_loss:2.75817\n",
            "Epoch:[2/50] Step:[434/507] Step_loss:1.07774\n",
            "Epoch:[2/50] Step:[435/507] Step_loss:0.93123\n",
            "Epoch:[2/50] Step:[436/507] Step_loss:2.92149\n",
            "Epoch:[2/50] Step:[437/507] Step_loss:1.23748\n",
            "Epoch:[2/50] Step:[438/507] Step_loss:1.09497\n",
            "Epoch:[2/50] Step:[439/507] Step_loss:1.08852\n",
            "Epoch:[2/50] Step:[440/507] Step_loss:0.72120\n",
            "Epoch:[2/50] Step:[441/507] Step_loss:0.80999\n",
            "Epoch:[2/50] Step:[442/507] Step_loss:0.74362\n",
            "Epoch:[2/50] Step:[443/507] Step_loss:1.75155\n",
            "Epoch:[2/50] Step:[444/507] Step_loss:1.60020\n",
            "Epoch:[2/50] Step:[445/507] Step_loss:1.55721\n",
            "Epoch:[2/50] Step:[446/507] Step_loss:0.87045\n",
            "Epoch:[2/50] Step:[447/507] Step_loss:1.16343\n",
            "Epoch:[2/50] Step:[448/507] Step_loss:0.87320\n",
            "Epoch:[2/50] Step:[449/507] Step_loss:0.74354\n",
            "Epoch:[2/50] Step:[450/507] Step_loss:1.28667\n",
            "Epoch:[2/50] Step:[451/507] Step_loss:0.73503\n",
            "Epoch:[2/50] Step:[452/507] Step_loss:2.28178\n",
            "Epoch:[2/50] Step:[453/507] Step_loss:0.69395\n",
            "Epoch:[2/50] Step:[454/507] Step_loss:0.71925\n",
            "Epoch:[2/50] Step:[455/507] Step_loss:1.05454\n",
            "Epoch:[2/50] Step:[456/507] Step_loss:1.15030\n",
            "Epoch:[2/50] Step:[457/507] Step_loss:0.90394\n",
            "Epoch:[2/50] Step:[458/507] Step_loss:0.93938\n",
            "Epoch:[2/50] Step:[459/507] Step_loss:1.11034\n",
            "Epoch:[2/50] Step:[460/507] Step_loss:0.85856\n",
            "Epoch:[2/50] Step:[461/507] Step_loss:0.86326\n",
            "Epoch:[2/50] Step:[462/507] Step_loss:1.14530\n",
            "Epoch:[2/50] Step:[463/507] Step_loss:1.30824\n",
            "Epoch:[2/50] Step:[464/507] Step_loss:1.27763\n",
            "Epoch:[2/50] Step:[465/507] Step_loss:1.56347\n",
            "Epoch:[2/50] Step:[466/507] Step_loss:1.14349\n",
            "Epoch:[2/50] Step:[467/507] Step_loss:1.06893\n",
            "Epoch:[2/50] Step:[468/507] Step_loss:1.20318\n",
            "Epoch:[2/50] Step:[469/507] Step_loss:1.04536\n",
            "Epoch:[2/50] Step:[470/507] Step_loss:0.69828\n",
            "Epoch:[2/50] Step:[471/507] Step_loss:0.85258\n",
            "Epoch:[2/50] Step:[472/507] Step_loss:1.44118\n",
            "Epoch:[2/50] Step:[473/507] Step_loss:0.91289\n",
            "Epoch:[2/50] Step:[474/507] Step_loss:1.51814\n",
            "Epoch:[2/50] Step:[475/507] Step_loss:1.17503\n",
            "Epoch:[2/50] Step:[476/507] Step_loss:1.16701\n",
            "Epoch:[2/50] Step:[477/507] Step_loss:2.18699\n",
            "Epoch:[2/50] Step:[478/507] Step_loss:1.72992\n",
            "Epoch:[2/50] Step:[479/507] Step_loss:0.82489\n",
            "Epoch:[2/50] Step:[480/507] Step_loss:1.52565\n",
            "Epoch:[2/50] Step:[481/507] Step_loss:1.75263\n",
            "Epoch:[2/50] Step:[482/507] Step_loss:1.06265\n",
            "Epoch:[2/50] Step:[483/507] Step_loss:1.32161\n",
            "Epoch:[2/50] Step:[484/507] Step_loss:1.49199\n",
            "Epoch:[2/50] Step:[485/507] Step_loss:1.48106\n",
            "Epoch:[2/50] Step:[486/507] Step_loss:1.02796\n",
            "Epoch:[2/50] Step:[487/507] Step_loss:0.85753\n",
            "Epoch:[2/50] Step:[488/507] Step_loss:0.85934\n",
            "Epoch:[2/50] Step:[489/507] Step_loss:1.29633\n",
            "Epoch:[2/50] Step:[490/507] Step_loss:1.76230\n",
            "Epoch:[2/50] Step:[491/507] Step_loss:1.44758\n",
            "Epoch:[2/50] Step:[492/507] Step_loss:0.81904\n",
            "Epoch:[2/50] Step:[493/507] Step_loss:1.07435\n",
            "Epoch:[2/50] Step:[494/507] Step_loss:0.75258\n",
            "Epoch:[2/50] Step:[495/507] Step_loss:0.78359\n",
            "Epoch:[2/50] Step:[496/507] Step_loss:0.91705\n",
            "Epoch:[2/50] Step:[497/507] Step_loss:0.94225\n",
            "Epoch:[2/50] Step:[498/507] Step_loss:1.26407\n",
            "Epoch:[2/50] Step:[499/507] Step_loss:0.81127\n",
            "Epoch:[2/50] Step:[500/507] Step_loss:1.15880\n",
            "Epoch:[2/50] Step:[501/507] Step_loss:1.77391\n",
            "Epoch:[2/50] Step:[502/507] Step_loss:1.29002\n",
            "Epoch:[2/50] Step:[503/507] Step_loss:0.91688\n",
            "Epoch:[2/50] Step:[504/507] Step_loss:1.42233\n",
            "Epoch:[2/50] Step:[505/507] Step_loss:0.86659\n",
            "Epoch:[2/50] Step:[506/507] Step_loss:0.86733\n",
            "Epoch:[2/50] Step:[507/507] Step_loss:1.27705\n",
            "Epoch [2/50], Loss: 1.2770473957061768\n",
            "Validation Accuracy: 0.7916666666666666\n",
            "Validation F1 Score: 0.8837209302325582\n",
            "Epoch:[3/50] Step:[1/507] Step_loss:0.97031\n",
            "Epoch:[3/50] Step:[2/507] Step_loss:1.24351\n",
            "Epoch:[3/50] Step:[3/507] Step_loss:1.17213\n",
            "Epoch:[3/50] Step:[4/507] Step_loss:0.75657\n",
            "Epoch:[3/50] Step:[5/507] Step_loss:0.97895\n",
            "Epoch:[3/50] Step:[6/507] Step_loss:0.92513\n",
            "Epoch:[3/50] Step:[7/507] Step_loss:0.85826\n",
            "Epoch:[3/50] Step:[8/507] Step_loss:0.89523\n",
            "Epoch:[3/50] Step:[9/507] Step_loss:0.75828\n",
            "Epoch:[3/50] Step:[10/507] Step_loss:0.92346\n",
            "Epoch:[3/50] Step:[11/507] Step_loss:1.02386\n",
            "Epoch:[3/50] Step:[12/507] Step_loss:1.14697\n",
            "Epoch:[3/50] Step:[13/507] Step_loss:0.80997\n",
            "Epoch:[3/50] Step:[14/507] Step_loss:0.82402\n",
            "Epoch:[3/50] Step:[15/507] Step_loss:0.68802\n",
            "Epoch:[3/50] Step:[16/507] Step_loss:0.74192\n",
            "Epoch:[3/50] Step:[17/507] Step_loss:0.74590\n",
            "Epoch:[3/50] Step:[18/507] Step_loss:0.65004\n",
            "Epoch:[3/50] Step:[19/507] Step_loss:0.67302\n",
            "Epoch:[3/50] Step:[20/507] Step_loss:0.63654\n",
            "Epoch:[3/50] Step:[21/507] Step_loss:0.84976\n",
            "Epoch:[3/50] Step:[22/507] Step_loss:0.54716\n",
            "Epoch:[3/50] Step:[23/507] Step_loss:0.94014\n",
            "Epoch:[3/50] Step:[24/507] Step_loss:0.60989\n",
            "Epoch:[3/50] Step:[25/507] Step_loss:0.57690\n",
            "Epoch:[3/50] Step:[26/507] Step_loss:1.34547\n",
            "Epoch:[3/50] Step:[27/507] Step_loss:0.83299\n",
            "Epoch:[3/50] Step:[28/507] Step_loss:1.27992\n",
            "Epoch:[3/50] Step:[29/507] Step_loss:0.61227\n",
            "Epoch:[3/50] Step:[30/507] Step_loss:0.68960\n",
            "Epoch:[3/50] Step:[31/507] Step_loss:1.03534\n",
            "Epoch:[3/50] Step:[32/507] Step_loss:0.73475\n",
            "Epoch:[3/50] Step:[33/507] Step_loss:0.78283\n",
            "Epoch:[3/50] Step:[34/507] Step_loss:1.19413\n",
            "Epoch:[3/50] Step:[35/507] Step_loss:1.05840\n",
            "Epoch:[3/50] Step:[36/507] Step_loss:1.14257\n",
            "Epoch:[3/50] Step:[37/507] Step_loss:0.94033\n",
            "Epoch:[3/50] Step:[38/507] Step_loss:0.99710\n",
            "Epoch:[3/50] Step:[39/507] Step_loss:0.58397\n",
            "Epoch:[3/50] Step:[40/507] Step_loss:0.86409\n",
            "Epoch:[3/50] Step:[41/507] Step_loss:0.81236\n",
            "Epoch:[3/50] Step:[42/507] Step_loss:0.51131\n",
            "Epoch:[3/50] Step:[43/507] Step_loss:0.76465\n",
            "Epoch:[3/50] Step:[44/507] Step_loss:1.02839\n",
            "Epoch:[3/50] Step:[45/507] Step_loss:1.08698\n",
            "Epoch:[3/50] Step:[46/507] Step_loss:1.13875\n",
            "Epoch:[3/50] Step:[47/507] Step_loss:0.81204\n",
            "Epoch:[3/50] Step:[48/507] Step_loss:1.04449\n",
            "Epoch:[3/50] Step:[49/507] Step_loss:0.48409\n",
            "Epoch:[3/50] Step:[50/507] Step_loss:1.05669\n",
            "Epoch:[3/50] Step:[51/507] Step_loss:0.79000\n",
            "Epoch:[3/50] Step:[52/507] Step_loss:0.98820\n",
            "Epoch:[3/50] Step:[53/507] Step_loss:0.82114\n",
            "Epoch:[3/50] Step:[54/507] Step_loss:0.73856\n",
            "Epoch:[3/50] Step:[55/507] Step_loss:0.66449\n",
            "Epoch:[3/50] Step:[56/507] Step_loss:1.05201\n",
            "Epoch:[3/50] Step:[57/507] Step_loss:0.79889\n",
            "Epoch:[3/50] Step:[58/507] Step_loss:0.71615\n",
            "Epoch:[3/50] Step:[59/507] Step_loss:0.75254\n",
            "Epoch:[3/50] Step:[60/507] Step_loss:0.61195\n",
            "Epoch:[3/50] Step:[61/507] Step_loss:0.69675\n",
            "Epoch:[3/50] Step:[62/507] Step_loss:0.71338\n",
            "Epoch:[3/50] Step:[63/507] Step_loss:0.50279\n",
            "Epoch:[3/50] Step:[64/507] Step_loss:0.70442\n",
            "Epoch:[3/50] Step:[65/507] Step_loss:0.95332\n",
            "Epoch:[3/50] Step:[66/507] Step_loss:0.70634\n",
            "Epoch:[3/50] Step:[67/507] Step_loss:0.77429\n",
            "Epoch:[3/50] Step:[68/507] Step_loss:0.73150\n",
            "Epoch:[3/50] Step:[69/507] Step_loss:0.76781\n",
            "Epoch:[3/50] Step:[70/507] Step_loss:1.14045\n",
            "Epoch:[3/50] Step:[71/507] Step_loss:0.61496\n",
            "Epoch:[3/50] Step:[72/507] Step_loss:0.62359\n",
            "Epoch:[3/50] Step:[73/507] Step_loss:0.89659\n",
            "Epoch:[3/50] Step:[74/507] Step_loss:0.66964\n",
            "Epoch:[3/50] Step:[75/507] Step_loss:1.30391\n",
            "Epoch:[3/50] Step:[76/507] Step_loss:0.71554\n",
            "Epoch:[3/50] Step:[77/507] Step_loss:1.19400\n",
            "Epoch:[3/50] Step:[78/507] Step_loss:0.85775\n",
            "Epoch:[3/50] Step:[79/507] Step_loss:1.54381\n",
            "Epoch:[3/50] Step:[80/507] Step_loss:0.54566\n",
            "Epoch:[3/50] Step:[81/507] Step_loss:1.10638\n",
            "Epoch:[3/50] Step:[82/507] Step_loss:1.14171\n",
            "Epoch:[3/50] Step:[83/507] Step_loss:1.09940\n",
            "Epoch:[3/50] Step:[84/507] Step_loss:0.74004\n",
            "Epoch:[3/50] Step:[85/507] Step_loss:1.01610\n",
            "Epoch:[3/50] Step:[86/507] Step_loss:0.70532\n",
            "Epoch:[3/50] Step:[87/507] Step_loss:1.05757\n",
            "Epoch:[3/50] Step:[88/507] Step_loss:0.95952\n",
            "Epoch:[3/50] Step:[89/507] Step_loss:1.05206\n",
            "Epoch:[3/50] Step:[90/507] Step_loss:0.76493\n",
            "Epoch:[3/50] Step:[91/507] Step_loss:0.82366\n",
            "Epoch:[3/50] Step:[92/507] Step_loss:0.86865\n",
            "Epoch:[3/50] Step:[93/507] Step_loss:0.74024\n",
            "Epoch:[3/50] Step:[94/507] Step_loss:0.71529\n",
            "Epoch:[3/50] Step:[95/507] Step_loss:0.59778\n",
            "Epoch:[3/50] Step:[96/507] Step_loss:0.69420\n",
            "Epoch:[3/50] Step:[97/507] Step_loss:0.68898\n",
            "Epoch:[3/50] Step:[98/507] Step_loss:1.32953\n",
            "Epoch:[3/50] Step:[99/507] Step_loss:0.66986\n",
            "Epoch:[3/50] Step:[100/507] Step_loss:1.01572\n",
            "Epoch:[3/50] Step:[101/507] Step_loss:0.79854\n",
            "Epoch:[3/50] Step:[102/507] Step_loss:0.80218\n",
            "Epoch:[3/50] Step:[103/507] Step_loss:0.66796\n",
            "Epoch:[3/50] Step:[104/507] Step_loss:0.95887\n",
            "Epoch:[3/50] Step:[105/507] Step_loss:0.60517\n",
            "Epoch:[3/50] Step:[106/507] Step_loss:0.75290\n",
            "Epoch:[3/50] Step:[107/507] Step_loss:1.20406\n",
            "Epoch:[3/50] Step:[108/507] Step_loss:0.79422\n",
            "Epoch:[3/50] Step:[109/507] Step_loss:0.64818\n",
            "Epoch:[3/50] Step:[110/507] Step_loss:0.70651\n",
            "Epoch:[3/50] Step:[111/507] Step_loss:0.69956\n",
            "Epoch:[3/50] Step:[112/507] Step_loss:0.64185\n",
            "Epoch:[3/50] Step:[113/507] Step_loss:0.57158\n",
            "Epoch:[3/50] Step:[114/507] Step_loss:0.71702\n",
            "Epoch:[3/50] Step:[115/507] Step_loss:0.81660\n",
            "Epoch:[3/50] Step:[116/507] Step_loss:0.65813\n",
            "Epoch:[3/50] Step:[117/507] Step_loss:0.68678\n",
            "Epoch:[3/50] Step:[118/507] Step_loss:0.71637\n",
            "Epoch:[3/50] Step:[119/507] Step_loss:0.76153\n",
            "Epoch:[3/50] Step:[120/507] Step_loss:0.70396\n",
            "Epoch:[3/50] Step:[121/507] Step_loss:0.75643\n",
            "Epoch:[3/50] Step:[122/507] Step_loss:0.72386\n",
            "Epoch:[3/50] Step:[123/507] Step_loss:0.91976\n",
            "Epoch:[3/50] Step:[124/507] Step_loss:0.81337\n",
            "Epoch:[3/50] Step:[125/507] Step_loss:0.74683\n",
            "Epoch:[3/50] Step:[126/507] Step_loss:0.92301\n",
            "Epoch:[3/50] Step:[127/507] Step_loss:0.80325\n",
            "Epoch:[3/50] Step:[128/507] Step_loss:0.87950\n",
            "Epoch:[3/50] Step:[129/507] Step_loss:0.67294\n",
            "Epoch:[3/50] Step:[130/507] Step_loss:0.88784\n",
            "Epoch:[3/50] Step:[131/507] Step_loss:0.86496\n",
            "Epoch:[3/50] Step:[132/507] Step_loss:0.71385\n",
            "Epoch:[3/50] Step:[133/507] Step_loss:1.15573\n",
            "Epoch:[3/50] Step:[134/507] Step_loss:0.85745\n",
            "Epoch:[3/50] Step:[135/507] Step_loss:0.75867\n",
            "Epoch:[3/50] Step:[136/507] Step_loss:0.90571\n",
            "Epoch:[3/50] Step:[137/507] Step_loss:0.65841\n",
            "Epoch:[3/50] Step:[138/507] Step_loss:0.66632\n",
            "Epoch:[3/50] Step:[139/507] Step_loss:1.64018\n",
            "Epoch:[3/50] Step:[140/507] Step_loss:0.82646\n",
            "Epoch:[3/50] Step:[141/507] Step_loss:0.59025\n",
            "Epoch:[3/50] Step:[142/507] Step_loss:0.89731\n",
            "Epoch:[3/50] Step:[143/507] Step_loss:0.69081\n",
            "Epoch:[3/50] Step:[144/507] Step_loss:0.85364\n",
            "Epoch:[3/50] Step:[145/507] Step_loss:0.79013\n",
            "Epoch:[3/50] Step:[146/507] Step_loss:0.69906\n",
            "Epoch:[3/50] Step:[147/507] Step_loss:0.85919\n",
            "Epoch:[3/50] Step:[148/507] Step_loss:0.98737\n",
            "Epoch:[3/50] Step:[149/507] Step_loss:1.10206\n",
            "Epoch:[3/50] Step:[150/507] Step_loss:0.88975\n",
            "Epoch:[3/50] Step:[151/507] Step_loss:1.09517\n",
            "Epoch:[3/50] Step:[152/507] Step_loss:0.89288\n",
            "Epoch:[3/50] Step:[153/507] Step_loss:1.64655\n",
            "Epoch:[3/50] Step:[154/507] Step_loss:0.63715\n",
            "Epoch:[3/50] Step:[155/507] Step_loss:0.75533\n",
            "Epoch:[3/50] Step:[156/507] Step_loss:0.88690\n",
            "Epoch:[3/50] Step:[157/507] Step_loss:0.82200\n",
            "Epoch:[3/50] Step:[158/507] Step_loss:1.01460\n",
            "Epoch:[3/50] Step:[159/507] Step_loss:1.10762\n",
            "Epoch:[3/50] Step:[160/507] Step_loss:0.85170\n",
            "Epoch:[3/50] Step:[161/507] Step_loss:1.07621\n",
            "Epoch:[3/50] Step:[162/507] Step_loss:0.80115\n",
            "Epoch:[3/50] Step:[163/507] Step_loss:0.84243\n",
            "Epoch:[3/50] Step:[164/507] Step_loss:0.73897\n",
            "Epoch:[3/50] Step:[165/507] Step_loss:0.59377\n",
            "Epoch:[3/50] Step:[166/507] Step_loss:0.61287\n",
            "Epoch:[3/50] Step:[167/507] Step_loss:1.03073\n",
            "Epoch:[3/50] Step:[168/507] Step_loss:0.65977\n",
            "Epoch:[3/50] Step:[169/507] Step_loss:0.98535\n",
            "Epoch:[3/50] Step:[170/507] Step_loss:0.80115\n",
            "Epoch:[3/50] Step:[171/507] Step_loss:1.02420\n",
            "Epoch:[3/50] Step:[172/507] Step_loss:0.74361\n",
            "Epoch:[3/50] Step:[173/507] Step_loss:1.12068\n",
            "Epoch:[3/50] Step:[174/507] Step_loss:0.74264\n",
            "Epoch:[3/50] Step:[175/507] Step_loss:0.80571\n",
            "Epoch:[3/50] Step:[176/507] Step_loss:0.74964\n",
            "Epoch:[3/50] Step:[177/507] Step_loss:0.76780\n",
            "Epoch:[3/50] Step:[178/507] Step_loss:0.56256\n",
            "Epoch:[3/50] Step:[179/507] Step_loss:0.74556\n",
            "Epoch:[3/50] Step:[180/507] Step_loss:0.65185\n",
            "Epoch:[3/50] Step:[181/507] Step_loss:0.96296\n",
            "Epoch:[3/50] Step:[182/507] Step_loss:1.23929\n",
            "Epoch:[3/50] Step:[183/507] Step_loss:1.15558\n",
            "Epoch:[3/50] Step:[184/507] Step_loss:0.54136\n",
            "Epoch:[3/50] Step:[185/507] Step_loss:0.73346\n",
            "Epoch:[3/50] Step:[186/507] Step_loss:0.72016\n",
            "Epoch:[3/50] Step:[187/507] Step_loss:0.63718\n",
            "Epoch:[3/50] Step:[188/507] Step_loss:0.75927\n",
            "Epoch:[3/50] Step:[189/507] Step_loss:0.68432\n",
            "Epoch:[3/50] Step:[190/507] Step_loss:0.55152\n",
            "Epoch:[3/50] Step:[191/507] Step_loss:0.84164\n",
            "Epoch:[3/50] Step:[192/507] Step_loss:0.50072\n",
            "Epoch:[3/50] Step:[193/507] Step_loss:0.62446\n",
            "Epoch:[3/50] Step:[194/507] Step_loss:0.46979\n",
            "Epoch:[3/50] Step:[195/507] Step_loss:0.53457\n",
            "Epoch:[3/50] Step:[196/507] Step_loss:0.77163\n",
            "Epoch:[3/50] Step:[197/507] Step_loss:0.51869\n",
            "Epoch:[3/50] Step:[198/507] Step_loss:0.62193\n",
            "Epoch:[3/50] Step:[199/507] Step_loss:0.78306\n",
            "Epoch:[3/50] Step:[200/507] Step_loss:1.03227\n",
            "Epoch:[3/50] Step:[201/507] Step_loss:0.70874\n",
            "Epoch:[3/50] Step:[202/507] Step_loss:0.98389\n",
            "Epoch:[3/50] Step:[203/507] Step_loss:0.75218\n",
            "Epoch:[3/50] Step:[204/507] Step_loss:1.05627\n",
            "Epoch:[3/50] Step:[205/507] Step_loss:0.71714\n",
            "Epoch:[3/50] Step:[206/507] Step_loss:1.14116\n",
            "Epoch:[3/50] Step:[207/507] Step_loss:1.26843\n",
            "Epoch:[3/50] Step:[208/507] Step_loss:1.09749\n",
            "Epoch:[3/50] Step:[209/507] Step_loss:1.43812\n",
            "Epoch:[3/50] Step:[210/507] Step_loss:1.01777\n",
            "Epoch:[3/50] Step:[211/507] Step_loss:2.81762\n",
            "Epoch:[3/50] Step:[212/507] Step_loss:1.80752\n",
            "Epoch:[3/50] Step:[213/507] Step_loss:1.60547\n",
            "Epoch:[3/50] Step:[214/507] Step_loss:1.07807\n",
            "Epoch:[3/50] Step:[215/507] Step_loss:1.07162\n",
            "Epoch:[3/50] Step:[216/507] Step_loss:1.25656\n",
            "Epoch:[3/50] Step:[217/507] Step_loss:0.96910\n",
            "Epoch:[3/50] Step:[218/507] Step_loss:0.77319\n",
            "Epoch:[3/50] Step:[219/507] Step_loss:0.77941\n",
            "Epoch:[3/50] Step:[220/507] Step_loss:0.71991\n",
            "Epoch:[3/50] Step:[221/507] Step_loss:0.91064\n",
            "Epoch:[3/50] Step:[222/507] Step_loss:0.74012\n",
            "Epoch:[3/50] Step:[223/507] Step_loss:0.87111\n",
            "Epoch:[3/50] Step:[224/507] Step_loss:0.87622\n",
            "Epoch:[3/50] Step:[225/507] Step_loss:0.79752\n",
            "Epoch:[3/50] Step:[226/507] Step_loss:0.78631\n",
            "Epoch:[3/50] Step:[227/507] Step_loss:1.30261\n",
            "Epoch:[3/50] Step:[228/507] Step_loss:0.92275\n",
            "Epoch:[3/50] Step:[229/507] Step_loss:0.67749\n",
            "Epoch:[3/50] Step:[230/507] Step_loss:0.59783\n",
            "Epoch:[3/50] Step:[231/507] Step_loss:1.04945\n",
            "Epoch:[3/50] Step:[232/507] Step_loss:0.82217\n",
            "Epoch:[3/50] Step:[233/507] Step_loss:0.71375\n",
            "Epoch:[3/50] Step:[234/507] Step_loss:1.12569\n",
            "Epoch:[3/50] Step:[235/507] Step_loss:0.61346\n",
            "Epoch:[3/50] Step:[236/507] Step_loss:0.69922\n",
            "Epoch:[3/50] Step:[237/507] Step_loss:0.91911\n",
            "Epoch:[3/50] Step:[238/507] Step_loss:0.68004\n",
            "Epoch:[3/50] Step:[239/507] Step_loss:0.81468\n",
            "Epoch:[3/50] Step:[240/507] Step_loss:0.54116\n",
            "Epoch:[3/50] Step:[241/507] Step_loss:1.00639\n",
            "Epoch:[3/50] Step:[242/507] Step_loss:0.52509\n",
            "Epoch:[3/50] Step:[243/507] Step_loss:1.14976\n",
            "Epoch:[3/50] Step:[244/507] Step_loss:0.56500\n",
            "Epoch:[3/50] Step:[245/507] Step_loss:0.73608\n",
            "Epoch:[3/50] Step:[246/507] Step_loss:1.03897\n",
            "Epoch:[3/50] Step:[247/507] Step_loss:0.79922\n",
            "Epoch:[3/50] Step:[248/507] Step_loss:1.15671\n",
            "Epoch:[3/50] Step:[249/507] Step_loss:0.63155\n",
            "Epoch:[3/50] Step:[250/507] Step_loss:0.65303\n",
            "Epoch:[3/50] Step:[251/507] Step_loss:0.67858\n",
            "Epoch:[3/50] Step:[252/507] Step_loss:0.87173\n",
            "Epoch:[3/50] Step:[253/507] Step_loss:0.53563\n",
            "Epoch:[3/50] Step:[254/507] Step_loss:1.27241\n",
            "Epoch:[3/50] Step:[255/507] Step_loss:0.74017\n",
            "Epoch:[3/50] Step:[256/507] Step_loss:1.03963\n",
            "Epoch:[3/50] Step:[257/507] Step_loss:0.66121\n",
            "Epoch:[3/50] Step:[258/507] Step_loss:1.15934\n",
            "Epoch:[3/50] Step:[259/507] Step_loss:0.51704\n",
            "Epoch:[3/50] Step:[260/507] Step_loss:0.76657\n",
            "Epoch:[3/50] Step:[261/507] Step_loss:0.53493\n",
            "Epoch:[3/50] Step:[262/507] Step_loss:0.63179\n",
            "Epoch:[3/50] Step:[263/507] Step_loss:0.73182\n",
            "Epoch:[3/50] Step:[264/507] Step_loss:0.78250\n",
            "Epoch:[3/50] Step:[265/507] Step_loss:0.74953\n",
            "Epoch:[3/50] Step:[266/507] Step_loss:0.63531\n",
            "Epoch:[3/50] Step:[267/507] Step_loss:0.54008\n",
            "Epoch:[3/50] Step:[268/507] Step_loss:0.85962\n",
            "Epoch:[3/50] Step:[269/507] Step_loss:1.08961\n",
            "Epoch:[3/50] Step:[270/507] Step_loss:0.60792\n",
            "Epoch:[3/50] Step:[271/507] Step_loss:1.64637\n",
            "Epoch:[3/50] Step:[272/507] Step_loss:0.64654\n",
            "Epoch:[3/50] Step:[273/507] Step_loss:1.51209\n",
            "Epoch:[3/50] Step:[274/507] Step_loss:0.89547\n",
            "Epoch:[3/50] Step:[275/507] Step_loss:0.74065\n",
            "Epoch:[3/50] Step:[276/507] Step_loss:0.76208\n",
            "Epoch:[3/50] Step:[277/507] Step_loss:0.90229\n",
            "Epoch:[3/50] Step:[278/507] Step_loss:0.79155\n",
            "Epoch:[3/50] Step:[279/507] Step_loss:1.03248\n",
            "Epoch:[3/50] Step:[280/507] Step_loss:1.14262\n",
            "Epoch:[3/50] Step:[281/507] Step_loss:1.14676\n",
            "Epoch:[3/50] Step:[282/507] Step_loss:1.68562\n",
            "Epoch:[3/50] Step:[283/507] Step_loss:0.84255\n",
            "Epoch:[3/50] Step:[284/507] Step_loss:1.52913\n",
            "Epoch:[3/50] Step:[285/507] Step_loss:0.66810\n",
            "Epoch:[3/50] Step:[286/507] Step_loss:0.98149\n",
            "Epoch:[3/50] Step:[287/507] Step_loss:1.05809\n",
            "Epoch:[3/50] Step:[288/507] Step_loss:0.77235\n",
            "Epoch:[3/50] Step:[289/507] Step_loss:1.79510\n",
            "Epoch:[3/50] Step:[290/507] Step_loss:0.45631\n",
            "Epoch:[3/50] Step:[291/507] Step_loss:0.88220\n",
            "Epoch:[3/50] Step:[292/507] Step_loss:0.82328\n",
            "Epoch:[3/50] Step:[293/507] Step_loss:1.37500\n",
            "Epoch:[3/50] Step:[294/507] Step_loss:0.75582\n",
            "Epoch:[3/50] Step:[295/507] Step_loss:0.67409\n",
            "Epoch:[3/50] Step:[296/507] Step_loss:1.17173\n",
            "Epoch:[3/50] Step:[297/507] Step_loss:1.15418\n",
            "Epoch:[3/50] Step:[298/507] Step_loss:0.83873\n",
            "Epoch:[3/50] Step:[299/507] Step_loss:0.87849\n",
            "Epoch:[3/50] Step:[300/507] Step_loss:0.59186\n",
            "Epoch:[3/50] Step:[301/507] Step_loss:0.82950\n",
            "Epoch:[3/50] Step:[302/507] Step_loss:0.82255\n",
            "Epoch:[3/50] Step:[303/507] Step_loss:0.55379\n",
            "Epoch:[3/50] Step:[304/507] Step_loss:1.02425\n",
            "Epoch:[3/50] Step:[305/507] Step_loss:0.76430\n",
            "Epoch:[3/50] Step:[306/507] Step_loss:0.65747\n",
            "Epoch:[3/50] Step:[307/507] Step_loss:1.14609\n",
            "Epoch:[3/50] Step:[308/507] Step_loss:1.00427\n",
            "Epoch:[3/50] Step:[309/507] Step_loss:0.73650\n",
            "Epoch:[3/50] Step:[310/507] Step_loss:0.94238\n",
            "Epoch:[3/50] Step:[311/507] Step_loss:1.24639\n",
            "Epoch:[3/50] Step:[312/507] Step_loss:0.80158\n",
            "Epoch:[3/50] Step:[313/507] Step_loss:0.60998\n",
            "Epoch:[3/50] Step:[314/507] Step_loss:0.76922\n",
            "Epoch:[3/50] Step:[315/507] Step_loss:0.68647\n",
            "Epoch:[3/50] Step:[316/507] Step_loss:0.89140\n",
            "Epoch:[3/50] Step:[317/507] Step_loss:0.77715\n",
            "Epoch:[3/50] Step:[318/507] Step_loss:0.77581\n",
            "Epoch:[3/50] Step:[319/507] Step_loss:0.63237\n",
            "Epoch:[3/50] Step:[320/507] Step_loss:0.87749\n",
            "Epoch:[3/50] Step:[321/507] Step_loss:1.03051\n",
            "Epoch:[3/50] Step:[322/507] Step_loss:1.20154\n",
            "Epoch:[3/50] Step:[323/507] Step_loss:0.89206\n",
            "Epoch:[3/50] Step:[324/507] Step_loss:0.64792\n",
            "Epoch:[3/50] Step:[325/507] Step_loss:0.48061\n",
            "Epoch:[3/50] Step:[326/507] Step_loss:0.73421\n",
            "Epoch:[3/50] Step:[327/507] Step_loss:0.77633\n",
            "Epoch:[3/50] Step:[328/507] Step_loss:0.56027\n",
            "Epoch:[3/50] Step:[329/507] Step_loss:1.14413\n",
            "Epoch:[3/50] Step:[330/507] Step_loss:0.66623\n",
            "Epoch:[3/50] Step:[331/507] Step_loss:1.60273\n",
            "Epoch:[3/50] Step:[332/507] Step_loss:0.98643\n",
            "Epoch:[3/50] Step:[333/507] Step_loss:0.93950\n",
            "Epoch:[3/50] Step:[334/507] Step_loss:1.18937\n",
            "Epoch:[3/50] Step:[335/507] Step_loss:0.63425\n",
            "Epoch:[3/50] Step:[336/507] Step_loss:1.24959\n",
            "Epoch:[3/50] Step:[337/507] Step_loss:0.53942\n",
            "Epoch:[3/50] Step:[338/507] Step_loss:1.00379\n",
            "Epoch:[3/50] Step:[339/507] Step_loss:0.65171\n",
            "Epoch:[3/50] Step:[340/507] Step_loss:0.54821\n",
            "Epoch:[3/50] Step:[341/507] Step_loss:0.94855\n",
            "Epoch:[3/50] Step:[342/507] Step_loss:0.72406\n",
            "Epoch:[3/50] Step:[343/507] Step_loss:0.74798\n",
            "Epoch:[3/50] Step:[344/507] Step_loss:0.93030\n",
            "Epoch:[3/50] Step:[345/507] Step_loss:1.00172\n",
            "Epoch:[3/50] Step:[346/507] Step_loss:0.98377\n",
            "Epoch:[3/50] Step:[347/507] Step_loss:0.59929\n",
            "Epoch:[3/50] Step:[348/507] Step_loss:0.78432\n",
            "Epoch:[3/50] Step:[349/507] Step_loss:0.51220\n",
            "Epoch:[3/50] Step:[350/507] Step_loss:0.78088\n",
            "Epoch:[3/50] Step:[351/507] Step_loss:0.81628\n",
            "Epoch:[3/50] Step:[352/507] Step_loss:0.72019\n",
            "Epoch:[3/50] Step:[353/507] Step_loss:0.74101\n",
            "Epoch:[3/50] Step:[354/507] Step_loss:0.80227\n",
            "Epoch:[3/50] Step:[355/507] Step_loss:1.17732\n",
            "Epoch:[3/50] Step:[356/507] Step_loss:1.11290\n",
            "Epoch:[3/50] Step:[357/507] Step_loss:2.10343\n",
            "Epoch:[3/50] Step:[358/507] Step_loss:0.69751\n",
            "Epoch:[3/50] Step:[359/507] Step_loss:1.02842\n",
            "Epoch:[3/50] Step:[360/507] Step_loss:0.85140\n",
            "Epoch:[3/50] Step:[361/507] Step_loss:1.74704\n",
            "Epoch:[3/50] Step:[362/507] Step_loss:0.74524\n",
            "Epoch:[3/50] Step:[363/507] Step_loss:0.58708\n",
            "Epoch:[3/50] Step:[364/507] Step_loss:0.83500\n",
            "Epoch:[3/50] Step:[365/507] Step_loss:0.57474\n",
            "Epoch:[3/50] Step:[366/507] Step_loss:0.76251\n",
            "Epoch:[3/50] Step:[367/507] Step_loss:0.53862\n",
            "Epoch:[3/50] Step:[368/507] Step_loss:0.88113\n",
            "Epoch:[3/50] Step:[369/507] Step_loss:0.71332\n",
            "Epoch:[3/50] Step:[370/507] Step_loss:0.96068\n",
            "Epoch:[3/50] Step:[371/507] Step_loss:0.49091\n",
            "Epoch:[3/50] Step:[372/507] Step_loss:1.09918\n",
            "Epoch:[3/50] Step:[373/507] Step_loss:1.08498\n",
            "Epoch:[3/50] Step:[374/507] Step_loss:0.90588\n",
            "Epoch:[3/50] Step:[375/507] Step_loss:1.58259\n",
            "Epoch:[3/50] Step:[376/507] Step_loss:0.77562\n",
            "Epoch:[3/50] Step:[377/507] Step_loss:0.91792\n",
            "Epoch:[3/50] Step:[378/507] Step_loss:0.94602\n",
            "Epoch:[3/50] Step:[379/507] Step_loss:0.79474\n",
            "Epoch:[3/50] Step:[380/507] Step_loss:0.93143\n",
            "Epoch:[3/50] Step:[381/507] Step_loss:0.61653\n",
            "Epoch:[3/50] Step:[382/507] Step_loss:0.85078\n",
            "Epoch:[3/50] Step:[383/507] Step_loss:1.50729\n",
            "Epoch:[3/50] Step:[384/507] Step_loss:1.13981\n",
            "Epoch:[3/50] Step:[385/507] Step_loss:0.55458\n",
            "Epoch:[3/50] Step:[386/507] Step_loss:0.97304\n",
            "Epoch:[3/50] Step:[387/507] Step_loss:0.66434\n",
            "Epoch:[3/50] Step:[388/507] Step_loss:1.41043\n",
            "Epoch:[3/50] Step:[389/507] Step_loss:0.50744\n",
            "Epoch:[3/50] Step:[390/507] Step_loss:1.02449\n",
            "Epoch:[3/50] Step:[391/507] Step_loss:0.63883\n",
            "Epoch:[3/50] Step:[392/507] Step_loss:1.18554\n",
            "Epoch:[3/50] Step:[393/507] Step_loss:0.59068\n",
            "Epoch:[3/50] Step:[394/507] Step_loss:0.79559\n",
            "Epoch:[3/50] Step:[395/507] Step_loss:0.53924\n",
            "Epoch:[3/50] Step:[396/507] Step_loss:0.88893\n",
            "Epoch:[3/50] Step:[397/507] Step_loss:0.56646\n",
            "Epoch:[3/50] Step:[398/507] Step_loss:0.79699\n",
            "Epoch:[3/50] Step:[399/507] Step_loss:0.75471\n",
            "Epoch:[3/50] Step:[400/507] Step_loss:0.99086\n",
            "Epoch:[3/50] Step:[401/507] Step_loss:0.81406\n",
            "Epoch:[3/50] Step:[402/507] Step_loss:0.50749\n",
            "Epoch:[3/50] Step:[403/507] Step_loss:0.54094\n",
            "Epoch:[3/50] Step:[404/507] Step_loss:0.61082\n",
            "Epoch:[3/50] Step:[405/507] Step_loss:0.67834\n",
            "Epoch:[3/50] Step:[406/507] Step_loss:0.86989\n",
            "Epoch:[3/50] Step:[407/507] Step_loss:0.56475\n",
            "Epoch:[3/50] Step:[408/507] Step_loss:1.33733\n",
            "Epoch:[3/50] Step:[409/507] Step_loss:0.53532\n",
            "Epoch:[3/50] Step:[410/507] Step_loss:0.70975\n",
            "Epoch:[3/50] Step:[411/507] Step_loss:0.92952\n",
            "Epoch:[3/50] Step:[412/507] Step_loss:0.86274\n",
            "Epoch:[3/50] Step:[413/507] Step_loss:1.42122\n",
            "Epoch:[3/50] Step:[414/507] Step_loss:0.75327\n",
            "Epoch:[3/50] Step:[415/507] Step_loss:0.65964\n",
            "Epoch:[3/50] Step:[416/507] Step_loss:0.86337\n",
            "Epoch:[3/50] Step:[417/507] Step_loss:0.76657\n",
            "Epoch:[3/50] Step:[418/507] Step_loss:0.88510\n",
            "Epoch:[3/50] Step:[419/507] Step_loss:0.77916\n",
            "Epoch:[3/50] Step:[420/507] Step_loss:0.58571\n",
            "Epoch:[3/50] Step:[421/507] Step_loss:0.64261\n",
            "Epoch:[3/50] Step:[422/507] Step_loss:0.84351\n",
            "Epoch:[3/50] Step:[423/507] Step_loss:0.72609\n",
            "Epoch:[3/50] Step:[424/507] Step_loss:0.80468\n",
            "Epoch:[3/50] Step:[425/507] Step_loss:0.61485\n",
            "Epoch:[3/50] Step:[426/507] Step_loss:1.24846\n",
            "Epoch:[3/50] Step:[427/507] Step_loss:1.08024\n",
            "Epoch:[3/50] Step:[428/507] Step_loss:0.72921\n",
            "Epoch:[3/50] Step:[429/507] Step_loss:0.87012\n",
            "Epoch:[3/50] Step:[430/507] Step_loss:1.25042\n",
            "Epoch:[3/50] Step:[431/507] Step_loss:0.91461\n",
            "Epoch:[3/50] Step:[432/507] Step_loss:1.08649\n",
            "Epoch:[3/50] Step:[433/507] Step_loss:0.79140\n",
            "Epoch:[3/50] Step:[434/507] Step_loss:0.99613\n",
            "Epoch:[3/50] Step:[435/507] Step_loss:0.67516\n",
            "Epoch:[3/50] Step:[436/507] Step_loss:0.74877\n",
            "Epoch:[3/50] Step:[437/507] Step_loss:0.72396\n",
            "Epoch:[3/50] Step:[438/507] Step_loss:1.07678\n",
            "Epoch:[3/50] Step:[439/507] Step_loss:1.15130\n",
            "Epoch:[3/50] Step:[440/507] Step_loss:0.75437\n",
            "Epoch:[3/50] Step:[441/507] Step_loss:0.99135\n",
            "Epoch:[3/50] Step:[442/507] Step_loss:0.66522\n",
            "Epoch:[3/50] Step:[443/507] Step_loss:0.97715\n",
            "Epoch:[3/50] Step:[444/507] Step_loss:1.29466\n",
            "Epoch:[3/50] Step:[445/507] Step_loss:1.23512\n",
            "Epoch:[3/50] Step:[446/507] Step_loss:1.34782\n",
            "Epoch:[3/50] Step:[447/507] Step_loss:2.00570\n",
            "Epoch:[3/50] Step:[448/507] Step_loss:1.02710\n",
            "Epoch:[3/50] Step:[449/507] Step_loss:1.98304\n",
            "Epoch:[3/50] Step:[450/507] Step_loss:1.51917\n",
            "Epoch:[3/50] Step:[451/507] Step_loss:1.50713\n",
            "Epoch:[3/50] Step:[452/507] Step_loss:1.22801\n",
            "Epoch:[3/50] Step:[453/507] Step_loss:1.61343\n",
            "Epoch:[3/50] Step:[454/507] Step_loss:0.64264\n",
            "Epoch:[3/50] Step:[455/507] Step_loss:1.09268\n",
            "Epoch:[3/50] Step:[456/507] Step_loss:1.20647\n",
            "Epoch:[3/50] Step:[457/507] Step_loss:0.72851\n",
            "Epoch:[3/50] Step:[458/507] Step_loss:1.07233\n",
            "Epoch:[3/50] Step:[459/507] Step_loss:1.67077\n",
            "Epoch:[3/50] Step:[460/507] Step_loss:0.71135\n",
            "Epoch:[3/50] Step:[461/507] Step_loss:1.26523\n",
            "Epoch:[3/50] Step:[462/507] Step_loss:1.38369\n",
            "Epoch:[3/50] Step:[463/507] Step_loss:1.83809\n",
            "Epoch:[3/50] Step:[464/507] Step_loss:0.59496\n",
            "Epoch:[3/50] Step:[465/507] Step_loss:0.97512\n",
            "Epoch:[3/50] Step:[466/507] Step_loss:0.57551\n",
            "Epoch:[3/50] Step:[467/507] Step_loss:0.82386\n",
            "Epoch:[3/50] Step:[468/507] Step_loss:0.94913\n",
            "Epoch:[3/50] Step:[469/507] Step_loss:0.93353\n",
            "Epoch:[3/50] Step:[470/507] Step_loss:1.51372\n",
            "Epoch:[3/50] Step:[471/507] Step_loss:0.99599\n",
            "Epoch:[3/50] Step:[472/507] Step_loss:0.75868\n",
            "Epoch:[3/50] Step:[473/507] Step_loss:1.05109\n",
            "Epoch:[3/50] Step:[474/507] Step_loss:0.99136\n",
            "Epoch:[3/50] Step:[475/507] Step_loss:0.69810\n",
            "Epoch:[3/50] Step:[476/507] Step_loss:0.83122\n",
            "Epoch:[3/50] Step:[477/507] Step_loss:0.62133\n",
            "Epoch:[3/50] Step:[478/507] Step_loss:0.89092\n",
            "Epoch:[3/50] Step:[479/507] Step_loss:1.08065\n",
            "Epoch:[3/50] Step:[480/507] Step_loss:0.77499\n",
            "Epoch:[3/50] Step:[481/507] Step_loss:0.63889\n",
            "Epoch:[3/50] Step:[482/507] Step_loss:1.24828\n",
            "Epoch:[3/50] Step:[483/507] Step_loss:0.68686\n",
            "Epoch:[3/50] Step:[484/507] Step_loss:1.13382\n",
            "Epoch:[3/50] Step:[485/507] Step_loss:0.54806\n",
            "Epoch:[3/50] Step:[486/507] Step_loss:1.20363\n",
            "Epoch:[3/50] Step:[487/507] Step_loss:0.95946\n",
            "Epoch:[3/50] Step:[488/507] Step_loss:1.38076\n",
            "Epoch:[3/50] Step:[489/507] Step_loss:0.66425\n",
            "Epoch:[3/50] Step:[490/507] Step_loss:0.73047\n",
            "Epoch:[3/50] Step:[491/507] Step_loss:1.25504\n",
            "Epoch:[3/50] Step:[492/507] Step_loss:0.89657\n",
            "Epoch:[3/50] Step:[493/507] Step_loss:0.66175\n",
            "Epoch:[3/50] Step:[494/507] Step_loss:0.80383\n",
            "Epoch:[3/50] Step:[495/507] Step_loss:0.68474\n",
            "Epoch:[3/50] Step:[496/507] Step_loss:0.73757\n",
            "Epoch:[3/50] Step:[497/507] Step_loss:0.84063\n",
            "Epoch:[3/50] Step:[498/507] Step_loss:1.21627\n",
            "Epoch:[3/50] Step:[499/507] Step_loss:0.95109\n",
            "Epoch:[3/50] Step:[500/507] Step_loss:0.66685\n",
            "Epoch:[3/50] Step:[501/507] Step_loss:1.36728\n",
            "Epoch:[3/50] Step:[502/507] Step_loss:0.80638\n",
            "Epoch:[3/50] Step:[503/507] Step_loss:1.12307\n",
            "Epoch:[3/50] Step:[504/507] Step_loss:0.72435\n",
            "Epoch:[3/50] Step:[505/507] Step_loss:0.82099\n",
            "Epoch:[3/50] Step:[506/507] Step_loss:0.57446\n",
            "Epoch:[3/50] Step:[507/507] Step_loss:1.04337\n",
            "Epoch [3/50], Loss: 1.0433744192123413\n",
            "Validation Accuracy: 0.875\n",
            "Validation F1 Score: 0.9333333333333333\n",
            "Epoch:[4/50] Step:[1/507] Step_loss:0.81674\n",
            "Epoch:[4/50] Step:[2/507] Step_loss:1.07032\n",
            "Epoch:[4/50] Step:[3/507] Step_loss:0.49901\n",
            "Epoch:[4/50] Step:[4/507] Step_loss:0.96361\n",
            "Epoch:[4/50] Step:[5/507] Step_loss:0.75629\n",
            "Epoch:[4/50] Step:[6/507] Step_loss:0.59347\n",
            "Epoch:[4/50] Step:[7/507] Step_loss:0.98469\n",
            "Epoch:[4/50] Step:[8/507] Step_loss:0.61722\n",
            "Epoch:[4/50] Step:[9/507] Step_loss:0.51170\n",
            "Epoch:[4/50] Step:[10/507] Step_loss:0.73004\n",
            "Epoch:[4/50] Step:[11/507] Step_loss:0.79349\n",
            "Epoch:[4/50] Step:[12/507] Step_loss:0.65644\n",
            "Epoch:[4/50] Step:[13/507] Step_loss:0.67312\n",
            "Epoch:[4/50] Step:[14/507] Step_loss:0.44125\n",
            "Epoch:[4/50] Step:[15/507] Step_loss:0.71271\n",
            "Epoch:[4/50] Step:[16/507] Step_loss:0.63779\n",
            "Epoch:[4/50] Step:[17/507] Step_loss:0.64527\n",
            "Epoch:[4/50] Step:[18/507] Step_loss:0.58493\n",
            "Epoch:[4/50] Step:[19/507] Step_loss:0.59278\n",
            "Epoch:[4/50] Step:[20/507] Step_loss:0.87463\n",
            "Epoch:[4/50] Step:[21/507] Step_loss:0.52960\n",
            "Epoch:[4/50] Step:[22/507] Step_loss:0.71828\n",
            "Epoch:[4/50] Step:[23/507] Step_loss:0.51665\n",
            "Epoch:[4/50] Step:[24/507] Step_loss:0.34889\n",
            "Epoch:[4/50] Step:[25/507] Step_loss:0.44234\n",
            "Epoch:[4/50] Step:[26/507] Step_loss:0.62936\n",
            "Epoch:[4/50] Step:[27/507] Step_loss:0.71983\n",
            "Epoch:[4/50] Step:[28/507] Step_loss:0.57779\n",
            "Epoch:[4/50] Step:[29/507] Step_loss:0.47055\n",
            "Epoch:[4/50] Step:[30/507] Step_loss:0.64189\n",
            "Epoch:[4/50] Step:[31/507] Step_loss:0.54564\n",
            "Epoch:[4/50] Step:[32/507] Step_loss:1.30211\n",
            "Epoch:[4/50] Step:[33/507] Step_loss:0.67002\n",
            "Epoch:[4/50] Step:[34/507] Step_loss:0.74379\n",
            "Epoch:[4/50] Step:[35/507] Step_loss:0.53003\n",
            "Epoch:[4/50] Step:[36/507] Step_loss:0.73956\n",
            "Epoch:[4/50] Step:[37/507] Step_loss:1.00729\n",
            "Epoch:[4/50] Step:[38/507] Step_loss:0.64367\n",
            "Epoch:[4/50] Step:[39/507] Step_loss:0.57587\n",
            "Epoch:[4/50] Step:[40/507] Step_loss:0.66184\n",
            "Epoch:[4/50] Step:[41/507] Step_loss:1.06159\n",
            "Epoch:[4/50] Step:[42/507] Step_loss:0.79084\n",
            "Epoch:[4/50] Step:[43/507] Step_loss:1.00791\n",
            "Epoch:[4/50] Step:[44/507] Step_loss:1.20467\n",
            "Epoch:[4/50] Step:[45/507] Step_loss:0.74263\n",
            "Epoch:[4/50] Step:[46/507] Step_loss:1.17951\n",
            "Epoch:[4/50] Step:[47/507] Step_loss:0.64724\n",
            "Epoch:[4/50] Step:[48/507] Step_loss:0.90665\n",
            "Epoch:[4/50] Step:[49/507] Step_loss:0.90008\n",
            "Epoch:[4/50] Step:[50/507] Step_loss:0.92868\n",
            "Epoch:[4/50] Step:[51/507] Step_loss:0.61113\n",
            "Epoch:[4/50] Step:[52/507] Step_loss:1.42930\n",
            "Epoch:[4/50] Step:[53/507] Step_loss:1.03763\n",
            "Epoch:[4/50] Step:[54/507] Step_loss:0.81156\n",
            "Epoch:[4/50] Step:[55/507] Step_loss:0.73158\n",
            "Epoch:[4/50] Step:[56/507] Step_loss:0.82135\n",
            "Epoch:[4/50] Step:[57/507] Step_loss:0.49101\n",
            "Epoch:[4/50] Step:[58/507] Step_loss:0.60278\n",
            "Epoch:[4/50] Step:[59/507] Step_loss:0.73126\n",
            "Epoch:[4/50] Step:[60/507] Step_loss:0.76719\n",
            "Epoch:[4/50] Step:[61/507] Step_loss:1.42232\n",
            "Epoch:[4/50] Step:[62/507] Step_loss:0.71760\n",
            "Epoch:[4/50] Step:[63/507] Step_loss:0.58899\n",
            "Epoch:[4/50] Step:[64/507] Step_loss:0.81850\n",
            "Epoch:[4/50] Step:[65/507] Step_loss:0.64434\n",
            "Epoch:[4/50] Step:[66/507] Step_loss:1.06624\n",
            "Epoch:[4/50] Step:[67/507] Step_loss:0.86384\n",
            "Epoch:[4/50] Step:[68/507] Step_loss:0.66147\n",
            "Epoch:[4/50] Step:[69/507] Step_loss:0.66179\n",
            "Epoch:[4/50] Step:[70/507] Step_loss:0.79945\n",
            "Epoch:[4/50] Step:[71/507] Step_loss:0.96254\n",
            "Epoch:[4/50] Step:[72/507] Step_loss:0.75589\n",
            "Epoch:[4/50] Step:[73/507] Step_loss:0.83529\n",
            "Epoch:[4/50] Step:[74/507] Step_loss:0.66241\n",
            "Epoch:[4/50] Step:[75/507] Step_loss:0.56878\n",
            "Epoch:[4/50] Step:[76/507] Step_loss:0.73045\n",
            "Epoch:[4/50] Step:[77/507] Step_loss:0.54575\n",
            "Epoch:[4/50] Step:[78/507] Step_loss:0.95794\n",
            "Epoch:[4/50] Step:[79/507] Step_loss:0.78840\n",
            "Epoch:[4/50] Step:[80/507] Step_loss:1.28096\n",
            "Epoch:[4/50] Step:[81/507] Step_loss:1.15647\n",
            "Epoch:[4/50] Step:[82/507] Step_loss:1.15548\n",
            "Epoch:[4/50] Step:[83/507] Step_loss:0.92367\n",
            "Epoch:[4/50] Step:[84/507] Step_loss:1.01607\n",
            "Epoch:[4/50] Step:[85/507] Step_loss:1.47520\n",
            "Epoch:[4/50] Step:[86/507] Step_loss:0.80348\n",
            "Epoch:[4/50] Step:[87/507] Step_loss:1.06421\n",
            "Epoch:[4/50] Step:[88/507] Step_loss:1.02677\n",
            "Epoch:[4/50] Step:[89/507] Step_loss:0.92067\n",
            "Epoch:[4/50] Step:[90/507] Step_loss:0.79148\n",
            "Epoch:[4/50] Step:[91/507] Step_loss:0.79846\n",
            "Epoch:[4/50] Step:[92/507] Step_loss:0.76147\n",
            "Epoch:[4/50] Step:[93/507] Step_loss:0.98476\n",
            "Epoch:[4/50] Step:[94/507] Step_loss:0.75246\n",
            "Epoch:[4/50] Step:[95/507] Step_loss:0.83992\n",
            "Epoch:[4/50] Step:[96/507] Step_loss:0.78298\n",
            "Epoch:[4/50] Step:[97/507] Step_loss:0.73207\n",
            "Epoch:[4/50] Step:[98/507] Step_loss:1.10520\n",
            "Epoch:[4/50] Step:[99/507] Step_loss:0.68400\n",
            "Epoch:[4/50] Step:[100/507] Step_loss:1.69592\n",
            "Epoch:[4/50] Step:[101/507] Step_loss:0.67468\n",
            "Epoch:[4/50] Step:[102/507] Step_loss:1.02714\n",
            "Epoch:[4/50] Step:[103/507] Step_loss:0.89241\n",
            "Epoch:[4/50] Step:[104/507] Step_loss:0.98147\n",
            "Epoch:[4/50] Step:[105/507] Step_loss:0.87172\n",
            "Epoch:[4/50] Step:[106/507] Step_loss:1.23156\n",
            "Epoch:[4/50] Step:[107/507] Step_loss:0.65174\n",
            "Epoch:[4/50] Step:[108/507] Step_loss:0.68857\n",
            "Epoch:[4/50] Step:[109/507] Step_loss:0.60651\n",
            "Epoch:[4/50] Step:[110/507] Step_loss:1.13312\n",
            "Epoch:[4/50] Step:[111/507] Step_loss:1.02846\n",
            "Epoch:[4/50] Step:[112/507] Step_loss:0.98692\n",
            "Epoch:[4/50] Step:[113/507] Step_loss:0.66731\n",
            "Epoch:[4/50] Step:[114/507] Step_loss:0.69300\n",
            "Epoch:[4/50] Step:[115/507] Step_loss:0.67716\n",
            "Epoch:[4/50] Step:[116/507] Step_loss:0.81728\n",
            "Epoch:[4/50] Step:[117/507] Step_loss:0.73794\n",
            "Epoch:[4/50] Step:[118/507] Step_loss:0.74714\n",
            "Epoch:[4/50] Step:[119/507] Step_loss:0.68377\n",
            "Epoch:[4/50] Step:[120/507] Step_loss:0.96777\n",
            "Epoch:[4/50] Step:[121/507] Step_loss:0.95824\n",
            "Epoch:[4/50] Step:[122/507] Step_loss:0.69539\n",
            "Epoch:[4/50] Step:[123/507] Step_loss:0.85473\n",
            "Epoch:[4/50] Step:[124/507] Step_loss:0.71898\n",
            "Epoch:[4/50] Step:[125/507] Step_loss:0.63720\n",
            "Epoch:[4/50] Step:[126/507] Step_loss:0.69805\n",
            "Epoch:[4/50] Step:[127/507] Step_loss:0.73623\n",
            "Epoch:[4/50] Step:[128/507] Step_loss:0.76392\n",
            "Epoch:[4/50] Step:[129/507] Step_loss:0.61637\n",
            "Epoch:[4/50] Step:[130/507] Step_loss:0.68499\n",
            "Epoch:[4/50] Step:[131/507] Step_loss:0.79923\n",
            "Epoch:[4/50] Step:[132/507] Step_loss:0.93252\n",
            "Epoch:[4/50] Step:[133/507] Step_loss:0.81469\n",
            "Epoch:[4/50] Step:[134/507] Step_loss:1.02115\n",
            "Epoch:[4/50] Step:[135/507] Step_loss:1.17865\n",
            "Epoch:[4/50] Step:[136/507] Step_loss:0.85171\n",
            "Epoch:[4/50] Step:[137/507] Step_loss:0.56894\n",
            "Epoch:[4/50] Step:[138/507] Step_loss:0.93126\n",
            "Epoch:[4/50] Step:[139/507] Step_loss:0.79074\n",
            "Epoch:[4/50] Step:[140/507] Step_loss:0.84113\n",
            "Epoch:[4/50] Step:[141/507] Step_loss:0.90489\n",
            "Epoch:[4/50] Step:[142/507] Step_loss:0.71938\n",
            "Epoch:[4/50] Step:[143/507] Step_loss:1.36338\n",
            "Epoch:[4/50] Step:[144/507] Step_loss:1.29431\n",
            "Epoch:[4/50] Step:[145/507] Step_loss:1.05395\n",
            "Epoch:[4/50] Step:[146/507] Step_loss:0.69127\n",
            "Epoch:[4/50] Step:[147/507] Step_loss:0.90810\n",
            "Epoch:[4/50] Step:[148/507] Step_loss:0.86423\n",
            "Epoch:[4/50] Step:[149/507] Step_loss:1.33001\n",
            "Epoch:[4/50] Step:[150/507] Step_loss:0.51408\n",
            "Epoch:[4/50] Step:[151/507] Step_loss:1.05500\n",
            "Epoch:[4/50] Step:[152/507] Step_loss:0.65721\n",
            "Epoch:[4/50] Step:[153/507] Step_loss:0.76921\n",
            "Epoch:[4/50] Step:[154/507] Step_loss:0.62676\n",
            "Epoch:[4/50] Step:[155/507] Step_loss:0.62067\n",
            "Epoch:[4/50] Step:[156/507] Step_loss:1.42502\n",
            "Epoch:[4/50] Step:[157/507] Step_loss:1.04080\n",
            "Epoch:[4/50] Step:[158/507] Step_loss:1.83110\n",
            "Epoch:[4/50] Step:[159/507] Step_loss:0.88233\n",
            "Epoch:[4/50] Step:[160/507] Step_loss:1.39010\n",
            "Epoch:[4/50] Step:[161/507] Step_loss:1.20302\n",
            "Epoch:[4/50] Step:[162/507] Step_loss:0.70526\n",
            "Epoch:[4/50] Step:[163/507] Step_loss:1.10852\n",
            "Epoch:[4/50] Step:[164/507] Step_loss:0.99267\n",
            "Epoch:[4/50] Step:[165/507] Step_loss:0.83834\n",
            "Epoch:[4/50] Step:[166/507] Step_loss:1.13523\n",
            "Epoch:[4/50] Step:[167/507] Step_loss:0.53275\n",
            "Epoch:[4/50] Step:[168/507] Step_loss:0.82988\n",
            "Epoch:[4/50] Step:[169/507] Step_loss:1.09459\n",
            "Epoch:[4/50] Step:[170/507] Step_loss:0.88019\n",
            "Epoch:[4/50] Step:[171/507] Step_loss:0.69895\n",
            "Epoch:[4/50] Step:[172/507] Step_loss:1.08467\n",
            "Epoch:[4/50] Step:[173/507] Step_loss:0.94101\n",
            "Epoch:[4/50] Step:[174/507] Step_loss:0.63393\n",
            "Epoch:[4/50] Step:[175/507] Step_loss:0.93176\n",
            "Epoch:[4/50] Step:[176/507] Step_loss:0.74315\n",
            "Epoch:[4/50] Step:[177/507] Step_loss:0.87035\n",
            "Epoch:[4/50] Step:[178/507] Step_loss:0.63771\n",
            "Epoch:[4/50] Step:[179/507] Step_loss:1.04986\n",
            "Epoch:[4/50] Step:[180/507] Step_loss:0.78790\n",
            "Epoch:[4/50] Step:[181/507] Step_loss:1.03096\n",
            "Epoch:[4/50] Step:[182/507] Step_loss:0.55661\n",
            "Epoch:[4/50] Step:[183/507] Step_loss:1.05424\n",
            "Epoch:[4/50] Step:[184/507] Step_loss:0.68027\n",
            "Epoch:[4/50] Step:[185/507] Step_loss:0.62522\n",
            "Epoch:[4/50] Step:[186/507] Step_loss:0.64197\n",
            "Epoch:[4/50] Step:[187/507] Step_loss:0.97753\n",
            "Epoch:[4/50] Step:[188/507] Step_loss:1.14457\n",
            "Epoch:[4/50] Step:[189/507] Step_loss:1.08657\n",
            "Epoch:[4/50] Step:[190/507] Step_loss:1.80112\n",
            "Epoch:[4/50] Step:[191/507] Step_loss:0.61635\n",
            "Epoch:[4/50] Step:[192/507] Step_loss:1.46636\n",
            "Epoch:[4/50] Step:[193/507] Step_loss:0.84997\n",
            "Epoch:[4/50] Step:[194/507] Step_loss:2.00392\n",
            "Epoch:[4/50] Step:[195/507] Step_loss:0.89818\n",
            "Epoch:[4/50] Step:[196/507] Step_loss:0.85126\n",
            "Epoch:[4/50] Step:[197/507] Step_loss:0.77186\n",
            "Epoch:[4/50] Step:[198/507] Step_loss:0.75654\n",
            "Epoch:[4/50] Step:[199/507] Step_loss:0.72636\n",
            "Epoch:[4/50] Step:[200/507] Step_loss:0.82093\n",
            "Epoch:[4/50] Step:[201/507] Step_loss:0.73518\n",
            "Epoch:[4/50] Step:[202/507] Step_loss:1.21435\n",
            "Epoch:[4/50] Step:[203/507] Step_loss:0.69390\n",
            "Epoch:[4/50] Step:[204/507] Step_loss:0.81204\n",
            "Epoch:[4/50] Step:[205/507] Step_loss:0.70899\n",
            "Epoch:[4/50] Step:[206/507] Step_loss:0.81486\n",
            "Epoch:[4/50] Step:[207/507] Step_loss:0.67783\n",
            "Epoch:[4/50] Step:[208/507] Step_loss:0.66961\n",
            "Epoch:[4/50] Step:[209/507] Step_loss:1.02166\n",
            "Epoch:[4/50] Step:[210/507] Step_loss:1.24091\n",
            "Epoch:[4/50] Step:[211/507] Step_loss:0.88621\n",
            "Epoch:[4/50] Step:[212/507] Step_loss:0.75104\n",
            "Epoch:[4/50] Step:[213/507] Step_loss:0.55717\n",
            "Epoch:[4/50] Step:[214/507] Step_loss:0.75919\n",
            "Epoch:[4/50] Step:[215/507] Step_loss:0.61751\n",
            "Epoch:[4/50] Step:[216/507] Step_loss:0.89065\n",
            "Epoch:[4/50] Step:[217/507] Step_loss:0.94440\n",
            "Epoch:[4/50] Step:[218/507] Step_loss:1.28386\n",
            "Epoch:[4/50] Step:[219/507] Step_loss:0.62201\n",
            "Epoch:[4/50] Step:[220/507] Step_loss:0.62382\n",
            "Epoch:[4/50] Step:[221/507] Step_loss:0.69483\n",
            "Epoch:[4/50] Step:[222/507] Step_loss:0.72613\n",
            "Epoch:[4/50] Step:[223/507] Step_loss:0.79483\n",
            "Epoch:[4/50] Step:[224/507] Step_loss:0.74241\n",
            "Epoch:[4/50] Step:[225/507] Step_loss:1.24331\n",
            "Epoch:[4/50] Step:[226/507] Step_loss:0.80169\n",
            "Epoch:[4/50] Step:[227/507] Step_loss:0.53676\n",
            "Epoch:[4/50] Step:[228/507] Step_loss:0.43649\n",
            "Epoch:[4/50] Step:[229/507] Step_loss:0.66876\n",
            "Epoch:[4/50] Step:[230/507] Step_loss:0.55474\n",
            "Epoch:[4/50] Step:[231/507] Step_loss:0.66179\n",
            "Epoch:[4/50] Step:[232/507] Step_loss:0.44688\n",
            "Epoch:[4/50] Step:[233/507] Step_loss:0.54335\n",
            "Epoch:[4/50] Step:[234/507] Step_loss:0.53258\n",
            "Epoch:[4/50] Step:[235/507] Step_loss:0.73207\n",
            "Epoch:[4/50] Step:[236/507] Step_loss:0.67635\n",
            "Epoch:[4/50] Step:[237/507] Step_loss:0.52315\n",
            "Epoch:[4/50] Step:[238/507] Step_loss:0.52517\n",
            "Epoch:[4/50] Step:[239/507] Step_loss:0.68870\n",
            "Epoch:[4/50] Step:[240/507] Step_loss:1.10624\n",
            "Epoch:[4/50] Step:[241/507] Step_loss:0.82708\n",
            "Epoch:[4/50] Step:[242/507] Step_loss:0.57780\n",
            "Epoch:[4/50] Step:[243/507] Step_loss:0.58563\n",
            "Epoch:[4/50] Step:[244/507] Step_loss:0.59547\n",
            "Epoch:[4/50] Step:[245/507] Step_loss:0.66157\n",
            "Epoch:[4/50] Step:[246/507] Step_loss:0.45493\n",
            "Epoch:[4/50] Step:[247/507] Step_loss:0.83660\n",
            "Epoch:[4/50] Step:[248/507] Step_loss:0.52356\n",
            "Epoch:[4/50] Step:[249/507] Step_loss:1.10433\n",
            "Epoch:[4/50] Step:[250/507] Step_loss:0.63271\n",
            "Epoch:[4/50] Step:[251/507] Step_loss:0.65885\n",
            "Epoch:[4/50] Step:[252/507] Step_loss:0.66741\n",
            "Epoch:[4/50] Step:[253/507] Step_loss:0.97988\n",
            "Epoch:[4/50] Step:[254/507] Step_loss:1.28592\n",
            "Epoch:[4/50] Step:[255/507] Step_loss:0.59036\n",
            "Epoch:[4/50] Step:[256/507] Step_loss:1.06557\n",
            "Epoch:[4/50] Step:[257/507] Step_loss:0.67068\n",
            "Epoch:[4/50] Step:[258/507] Step_loss:1.08008\n",
            "Epoch:[4/50] Step:[259/507] Step_loss:0.71096\n",
            "Epoch:[4/50] Step:[260/507] Step_loss:0.76722\n",
            "Epoch:[4/50] Step:[261/507] Step_loss:0.61461\n",
            "Epoch:[4/50] Step:[262/507] Step_loss:0.96911\n",
            "Epoch:[4/50] Step:[263/507] Step_loss:0.71336\n",
            "Epoch:[4/50] Step:[264/507] Step_loss:1.06679\n",
            "Epoch:[4/50] Step:[265/507] Step_loss:0.97514\n",
            "Epoch:[4/50] Step:[266/507] Step_loss:1.79183\n",
            "Epoch:[4/50] Step:[267/507] Step_loss:0.52358\n",
            "Epoch:[4/50] Step:[268/507] Step_loss:1.37989\n",
            "Epoch:[4/50] Step:[269/507] Step_loss:1.16080\n",
            "Epoch:[4/50] Step:[270/507] Step_loss:0.80900\n",
            "Epoch:[4/50] Step:[271/507] Step_loss:1.41804\n",
            "Epoch:[4/50] Step:[272/507] Step_loss:1.11183\n",
            "Epoch:[4/50] Step:[273/507] Step_loss:0.84621\n",
            "Epoch:[4/50] Step:[274/507] Step_loss:0.83818\n",
            "Epoch:[4/50] Step:[275/507] Step_loss:0.62631\n",
            "Epoch:[4/50] Step:[276/507] Step_loss:1.79825\n",
            "Epoch:[4/50] Step:[277/507] Step_loss:0.88582\n",
            "Epoch:[4/50] Step:[278/507] Step_loss:0.81109\n",
            "Epoch:[4/50] Step:[279/507] Step_loss:1.10938\n",
            "Epoch:[4/50] Step:[280/507] Step_loss:1.11158\n",
            "Epoch:[4/50] Step:[281/507] Step_loss:0.74432\n",
            "Epoch:[4/50] Step:[282/507] Step_loss:1.57350\n",
            "Epoch:[4/50] Step:[283/507] Step_loss:1.52433\n",
            "Epoch:[4/50] Step:[284/507] Step_loss:0.73970\n",
            "Epoch:[4/50] Step:[285/507] Step_loss:1.67387\n",
            "Epoch:[4/50] Step:[286/507] Step_loss:0.70665\n",
            "Epoch:[4/50] Step:[287/507] Step_loss:1.12695\n",
            "Epoch:[4/50] Step:[288/507] Step_loss:1.86299\n",
            "Epoch:[4/50] Step:[289/507] Step_loss:0.74069\n",
            "Epoch:[4/50] Step:[290/507] Step_loss:1.49595\n",
            "Epoch:[4/50] Step:[291/507] Step_loss:1.25997\n",
            "Epoch:[4/50] Step:[292/507] Step_loss:0.72727\n",
            "Epoch:[4/50] Step:[293/507] Step_loss:0.84994\n",
            "Epoch:[4/50] Step:[294/507] Step_loss:1.41080\n",
            "Epoch:[4/50] Step:[295/507] Step_loss:0.72050\n",
            "Epoch:[4/50] Step:[296/507] Step_loss:0.78448\n",
            "Epoch:[4/50] Step:[297/507] Step_loss:0.89851\n",
            "Epoch:[4/50] Step:[298/507] Step_loss:0.73023\n",
            "Epoch:[4/50] Step:[299/507] Step_loss:0.54939\n",
            "Epoch:[4/50] Step:[300/507] Step_loss:1.01449\n",
            "Epoch:[4/50] Step:[301/507] Step_loss:1.15813\n",
            "Epoch:[4/50] Step:[302/507] Step_loss:0.60062\n",
            "Epoch:[4/50] Step:[303/507] Step_loss:0.83173\n",
            "Epoch:[4/50] Step:[304/507] Step_loss:0.86099\n",
            "Epoch:[4/50] Step:[305/507] Step_loss:1.13961\n",
            "Epoch:[4/50] Step:[306/507] Step_loss:1.51390\n",
            "Epoch:[4/50] Step:[307/507] Step_loss:0.75600\n",
            "Epoch:[4/50] Step:[308/507] Step_loss:0.88366\n",
            "Epoch:[4/50] Step:[309/507] Step_loss:1.33846\n",
            "Epoch:[4/50] Step:[310/507] Step_loss:0.68129\n",
            "Epoch:[4/50] Step:[311/507] Step_loss:1.31049\n",
            "Epoch:[4/50] Step:[312/507] Step_loss:0.83526\n",
            "Epoch:[4/50] Step:[313/507] Step_loss:0.73763\n",
            "Epoch:[4/50] Step:[314/507] Step_loss:0.66205\n",
            "Epoch:[4/50] Step:[315/507] Step_loss:1.47210\n",
            "Epoch:[4/50] Step:[316/507] Step_loss:0.68710\n",
            "Epoch:[4/50] Step:[317/507] Step_loss:0.67970\n",
            "Epoch:[4/50] Step:[318/507] Step_loss:1.28927\n",
            "Epoch:[4/50] Step:[319/507] Step_loss:0.60418\n",
            "Epoch:[4/50] Step:[320/507] Step_loss:0.75684\n",
            "Epoch:[4/50] Step:[321/507] Step_loss:0.63652\n",
            "Epoch:[4/50] Step:[322/507] Step_loss:0.70192\n",
            "Epoch:[4/50] Step:[323/507] Step_loss:0.74748\n",
            "Epoch:[4/50] Step:[324/507] Step_loss:0.95021\n",
            "Epoch:[4/50] Step:[325/507] Step_loss:0.54101\n",
            "Epoch:[4/50] Step:[326/507] Step_loss:0.53657\n",
            "Epoch:[4/50] Step:[327/507] Step_loss:1.32077\n",
            "Epoch:[4/50] Step:[328/507] Step_loss:0.80177\n",
            "Epoch:[4/50] Step:[329/507] Step_loss:0.75356\n",
            "Epoch:[4/50] Step:[330/507] Step_loss:0.62236\n",
            "Epoch:[4/50] Step:[331/507] Step_loss:1.01292\n",
            "Epoch:[4/50] Step:[332/507] Step_loss:0.83800\n",
            "Epoch:[4/50] Step:[333/507] Step_loss:1.00383\n",
            "Epoch:[4/50] Step:[334/507] Step_loss:1.08777\n",
            "Epoch:[4/50] Step:[335/507] Step_loss:1.07842\n",
            "Epoch:[4/50] Step:[336/507] Step_loss:0.62094\n",
            "Epoch:[4/50] Step:[337/507] Step_loss:0.57270\n",
            "Epoch:[4/50] Step:[338/507] Step_loss:0.83843\n",
            "Epoch:[4/50] Step:[339/507] Step_loss:0.66649\n",
            "Epoch:[4/50] Step:[340/507] Step_loss:0.83521\n",
            "Epoch:[4/50] Step:[341/507] Step_loss:0.54726\n",
            "Epoch:[4/50] Step:[342/507] Step_loss:0.88284\n",
            "Epoch:[4/50] Step:[343/507] Step_loss:1.00489\n",
            "Epoch:[4/50] Step:[344/507] Step_loss:1.24507\n",
            "Epoch:[4/50] Step:[345/507] Step_loss:0.59604\n",
            "Epoch:[4/50] Step:[346/507] Step_loss:0.61470\n",
            "Epoch:[4/50] Step:[347/507] Step_loss:0.96760\n",
            "Epoch:[4/50] Step:[348/507] Step_loss:0.99460\n",
            "Epoch:[4/50] Step:[349/507] Step_loss:0.90839\n",
            "Epoch:[4/50] Step:[350/507] Step_loss:0.54773\n",
            "Epoch:[4/50] Step:[351/507] Step_loss:0.59107\n",
            "Epoch:[4/50] Step:[352/507] Step_loss:0.66067\n",
            "Epoch:[4/50] Step:[353/507] Step_loss:0.61856\n",
            "Epoch:[4/50] Step:[354/507] Step_loss:0.78078\n",
            "Epoch:[4/50] Step:[355/507] Step_loss:0.84348\n",
            "Epoch:[4/50] Step:[356/507] Step_loss:1.04091\n",
            "Epoch:[4/50] Step:[357/507] Step_loss:1.24198\n",
            "Epoch:[4/50] Step:[358/507] Step_loss:1.24034\n",
            "Epoch:[4/50] Step:[359/507] Step_loss:0.57928\n",
            "Epoch:[4/50] Step:[360/507] Step_loss:1.09554\n",
            "Epoch:[4/50] Step:[361/507] Step_loss:1.43374\n",
            "Epoch:[4/50] Step:[362/507] Step_loss:1.01070\n",
            "Epoch:[4/50] Step:[363/507] Step_loss:0.97136\n",
            "Epoch:[4/50] Step:[364/507] Step_loss:1.13584\n",
            "Epoch:[4/50] Step:[365/507] Step_loss:1.05252\n",
            "Epoch:[4/50] Step:[366/507] Step_loss:0.67280\n",
            "Epoch:[4/50] Step:[367/507] Step_loss:0.66295\n",
            "Epoch:[4/50] Step:[368/507] Step_loss:0.57201\n",
            "Epoch:[4/50] Step:[369/507] Step_loss:0.90276\n",
            "Epoch:[4/50] Step:[370/507] Step_loss:0.99978\n",
            "Epoch:[4/50] Step:[371/507] Step_loss:0.59021\n",
            "Epoch:[4/50] Step:[372/507] Step_loss:0.57353\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# 모델 평가\n",
        "model.eval()\n",
        "all_predictions = []\n",
        "all_labels = []\n",
        "with torch.no_grad():\n",
        "    for inputs, labels in valid_loader:\n",
        "        inputs, labels = inputs.cuda(), labels.cuda()\n",
        "        outputs = model(inputs).cuda()\n",
        "        predictions = (outputs > 0.5).float()\n",
        "        all_predictions.append(predictions)\n",
        "        all_labels.append(labels)\n",
        "\n",
        "# 전체 데이터에 대한 예측과 레이블을 하나로 합침\n",
        "all_predictions = torch.cat(all_predictions).cpu()\n",
        "all_labels = torch.cat(all_labels).cpu()"
      ],
      "metadata": {
        "id": "RLWn-yczZ4RY"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# 정확도 계산\n",
        "accuracy = accuracy_score(all_labels, all_predictions)\n",
        "print(f'Validation Accuracy: {accuracy}')\n",
        "\n",
        "# F1 스코어 계산\n",
        "f1 = f1_score(all_labels, all_predictions)\n",
        "print(f'Validation F1 Score: {f1}')"
      ],
      "metadata": {
        "id": "H86915ck9E4a",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e45d2686-7660-4690-8c47-a19a881ed42c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Validation Accuracy: 0.8342592592592593\n",
            "Validation F1 Score: 0.76177801437317\n"
          ]
        }
      ]
    }
  ]
}