flair_training_1.5K.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Copy of flair_training_1.5K",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "metadata": {
        "id": "tyJ3ukprrvRF",
        "colab_type": "code",
        "outputId": "e1591765-ef8e-44d3-8a20-a4e1fb266948",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 2264
        }
      },
      "cell_type": "code",
      "source": [
        "!pip install flair\n",
        "!wget -c http:https://gemeos.org/henrique/training_data_1450.csv\n",
        "!wget -c https://github.com/nathanshartmann/portuguese_word_embeddings/raw/master/preprocessing.py\n",
        "\n",
        "!wget -c http:https://gemeos.org/henrique/health_word2vec_300v1.tar.gz -O health_word2vec_300v1.tar.gz #0.84\n",
        "#!wget -c http:https://gemeos.org/henrique/health_fasttext_300v1.tar.gz -O health_fasttext_300v1.tar.gz #0.86\n",
        "!tar -zxvf health_word2vec_300v1.tar.gz"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting flair\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/cc/6a/fea4f9533feea65f8b2ff3d6a9934dba6227df81bbc180915cbd30c5db95/flair-0.4.0.tar.gz (70kB)\n",
            "\u001b[K    100% |████████████████████████████████| 71kB 3.4MB/s \n",
            "\u001b[?25hRequirement already satisfied: torch==1.0.0 in /usr/local/lib/python3.6/dist-packages (from flair) (1.0.0)\n",
            "Collecting gensim==3.4.0 (from flair)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/33/33/df6cb7acdcec5677ed130f4800f67509d24dbec74a03c329fcbf6b0864f0/gensim-3.4.0-cp36-cp36m-manylinux1_x86_64.whl (22.6MB)\n",
            "\u001b[K    100% |████████████████████████████████| 22.6MB 2.4MB/s \n",
            "\u001b[?25hCollecting typing==3.6.4 (from flair)\n",
            "  Downloading https://files.pythonhosted.org/packages/05/2b/2b05bf1d5a9dd450447c9a5df3e118a465e5d3cb12b73b7220a5064a403f/typing-3.6.4-py3-none-any.whl\n",
            "Collecting tqdm==4.26.0 (from flair)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/79/43/19c9fee28110cd47f73e6bc596394337fe9f3e5825b4de402bbf30b3beb5/tqdm-4.26.0-py2.py3-none-any.whl (43kB)\n",
            "\u001b[K    100% |████████████████████████████████| 51kB 22.3MB/s \n",
            "\u001b[?25hCollecting segtok==1.5.7 (from flair)\n",
            "  Downloading https://files.pythonhosted.org/packages/1d/59/6ed78856ab99d2da04084b59e7da797972baa0efecb71546b16d48e49d9b/segtok-1.5.7.tar.gz\n",
            "Collecting matplotlib==3.0.0 (from flair)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/ed/89/dd823436a5f8d5ca9304b51b554863bfd366ca84708d5812f5ee87c923bc/matplotlib-3.0.0-cp36-cp36m-manylinux1_x86_64.whl (12.8MB)\n",
            "\u001b[K    100% |████████████████████████████████| 12.8MB 4.4MB/s \n",
            "\u001b[?25hCollecting mpld3==0.3 (from flair)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/91/95/a52d3a83d0a29ba0d6898f6727e9858fe7a43f6c2ce81a5fe7e05f0f4912/mpld3-0.3.tar.gz (788kB)\n",
            "\u001b[K    100% |████████████████████████████████| 798kB 27.1MB/s \n",
            "\u001b[?25hRequirement already satisfied: sklearn in /usr/local/lib/python3.6/dist-packages (from flair) (0.0)\n",
            "Collecting sqlitedict==1.6.0 (from flair)\n",
            "  Downloading https://files.pythonhosted.org/packages/0f/1c/c757b93147a219cf1e25cef7e1ad9b595b7f802159493c45ce116521caff/sqlitedict-1.6.0.tar.gz\n",
            "Collecting deprecated==1.2.4 (from flair)\n",
            "  Downloading https://files.pythonhosted.org/packages/a9/cb/c1a39ee51e3042df8b284e22c9c440ffad1c25f451bddd4bf9a8dc17cd75/Deprecated-1.2.4-py2.py3-none-any.whl\n",
            "Requirement already satisfied: hyperopt==0.1.1 in /usr/local/lib/python3.6/dist-packages (from flair) (0.1.1)\n",
            "Collecting pytorch-pretrained-bert==0.3.0 (from flair)\n",
            "  Downloading https://files.pythonhosted.org/packages/26/85/de4dd7e018a197280752881adf7b4142886f20155145f641f7c803c0018a/pytorch_pretrained_bert-0.3.0-py3-none-any.whl\n",
            "Requirement already satisfied: numpy>=1.11.3 in /usr/local/lib/python3.6/dist-packages (from gensim==3.4.0->flair) (1.14.6)\n",
            "Requirement already satisfied: smart-open>=1.2.1 in /usr/local/lib/python3.6/dist-packages (from gensim==3.4.0->flair) (1.8.0)\n",
            "Requirement already satisfied: six>=1.5.0 in /usr/local/lib/python3.6/dist-packages (from gensim==3.4.0->flair) (1.11.0)\n",
            "Requirement already satisfied: scipy>=0.18.1 in /usr/local/lib/python3.6/dist-packages (from gensim==3.4.0->flair) (1.1.0)\n",
            "Requirement already satisfied: regex in /usr/local/lib/python3.6/dist-packages (from segtok==1.5.7->flair) (2018.1.10)\n",
            "Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib==3.0.0->flair) (2.3.1)\n",
            "Requirement already satisfied: python-dateutil>=2.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib==3.0.0->flair) (2.5.3)\n",
            "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.6/dist-packages (from matplotlib==3.0.0->flair) (0.10.0)\n",
            "Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib==3.0.0->flair) (1.0.1)\n",
            "Requirement already satisfied: scikit-learn in /usr/local/lib/python3.6/dist-packages (from sklearn->flair) (0.20.2)\n",
            "Requirement already satisfied: wrapt<2,>=1 in /usr/local/lib/python3.6/dist-packages (from deprecated==1.2.4->flair) (1.11.1)\n",
            "Requirement already satisfied: networkx in /usr/local/lib/python3.6/dist-packages (from hyperopt==0.1.1->flair) (2.2)\n",
            "Requirement already satisfied: future in /usr/local/lib/python3.6/dist-packages (from hyperopt==0.1.1->flair) (0.16.0)\n",
            "Requirement already satisfied: pymongo in /usr/local/lib/python3.6/dist-packages (from hyperopt==0.1.1->flair) (3.7.2)\n",
            "Requirement already satisfied: boto3 in /usr/local/lib/python3.6/dist-packages (from pytorch-pretrained-bert==0.3.0->flair) (1.9.89)\n",
            "Requirement already satisfied: requests in /usr/local/lib/python3.6/dist-packages (from pytorch-pretrained-bert==0.3.0->flair) (2.18.4)\n",
            "Requirement already satisfied: boto>=2.32 in /usr/local/lib/python3.6/dist-packages (from smart-open>=1.2.1->gensim==3.4.0->flair) (2.49.0)\n",
            "Requirement already satisfied: bz2file in /usr/local/lib/python3.6/dist-packages (from smart-open>=1.2.1->gensim==3.4.0->flair) (0.98)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.6/dist-packages (from kiwisolver>=1.0.1->matplotlib==3.0.0->flair) (40.8.0)\n",
            "Requirement already satisfied: decorator>=4.3.0 in /usr/local/lib/python3.6/dist-packages (from networkx->hyperopt==0.1.1->flair) (4.3.2)\n",
            "Requirement already satisfied: s3transfer<0.3.0,>=0.2.0 in /usr/local/lib/python3.6/dist-packages (from boto3->pytorch-pretrained-bert==0.3.0->flair) (0.2.0)\n",
            "Requirement already satisfied: botocore<1.13.0,>=1.12.89 in /usr/local/lib/python3.6/dist-packages (from boto3->pytorch-pretrained-bert==0.3.0->flair) (1.12.89)\n",
            "Requirement already satisfied: jmespath<1.0.0,>=0.7.1 in /usr/local/lib/python3.6/dist-packages (from boto3->pytorch-pretrained-bert==0.3.0->flair) (0.9.3)\n",
            "Requirement already satisfied: idna<2.7,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests->pytorch-pretrained-bert==0.3.0->flair) (2.6)\n",
            "Requirement already satisfied: urllib3<1.23,>=1.21.1 in /usr/local/lib/python3.6/dist-packages (from requests->pytorch-pretrained-bert==0.3.0->flair) (1.22)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.6/dist-packages (from requests->pytorch-pretrained-bert==0.3.0->flair) (2018.11.29)\n",
            "Requirement already satisfied: chardet<3.1.0,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests->pytorch-pretrained-bert==0.3.0->flair) (3.0.4)\n",
            "Requirement already satisfied: docutils>=0.10 in /usr/local/lib/python3.6/dist-packages (from botocore<1.13.0,>=1.12.89->boto3->pytorch-pretrained-bert==0.3.0->flair) (0.14)\n",
            "Building wheels for collected packages: flair, segtok, mpld3, sqlitedict\n",
            "  Building wheel for flair (setup.py) ... \u001b[?25ldone\n",
            "\u001b[?25h  Stored in directory: /root/.cache/pip/wheels/54/1c/24/39ec59521bb1a83306cf5224a50f99aad60b55dd3b42d32ed0\n",
            "  Building wheel for segtok (setup.py) ... \u001b[?25ldone\n",
            "\u001b[?25h  Stored in directory: /root/.cache/pip/wheels/15/ee/a8/6112173f1386d33eebedb3f73429cfa41a4c3084556bcee254\n",
            "  Building wheel for mpld3 (setup.py) ... \u001b[?25ldone\n",
            "\u001b[?25h  Stored in directory: /root/.cache/pip/wheels/c0/47/fb/8a64f89aecfe0059830479308ad42d62e898a3e3cefdf6ba28\n",
            "  Building wheel for sqlitedict (setup.py) ... \u001b[?25ldone\n",
            "\u001b[?25h  Stored in directory: /root/.cache/pip/wheels/bd/57/d3/907c3ee02d35e66f674ad0106e61f06eeeb98f6ee66a6cc3fe\n",
            "Successfully built flair segtok mpld3 sqlitedict\n",
            "\u001b[31myellowbrick 0.9.1 has requirement matplotlib!=3.0.0,>=1.5.1, but you'll have matplotlib 3.0.0 which is incompatible.\u001b[0m\n",
            "\u001b[31mthinc 6.12.1 has requirement wrapt<1.11.0,>=1.10.0, but you'll have wrapt 1.11.1 which is incompatible.\u001b[0m\n",
            "\u001b[31mspacy 2.0.18 has requirement numpy>=1.15.0, but you'll have numpy 1.14.6 which is incompatible.\u001b[0m\n",
            "\u001b[31mpymc3 3.6 has requirement joblib<0.13.0, but you'll have joblib 0.13.1 which is incompatible.\u001b[0m\n",
            "\u001b[31mfeaturetools 0.4.1 has requirement pandas>=0.23.0, but you'll have pandas 0.22.0 which is incompatible.\u001b[0m\n",
            "Installing collected packages: gensim, typing, tqdm, segtok, matplotlib, mpld3, sqlitedict, deprecated, pytorch-pretrained-bert, flair\n",
            "  Found existing installation: gensim 3.6.0\n",
            "    Uninstalling gensim-3.6.0:\n",
            "      Successfully uninstalled gensim-3.6.0\n",
            "  Found existing installation: typing 3.6.6\n",
            "    Uninstalling typing-3.6.6:\n",
            "      Successfully uninstalled typing-3.6.6\n",
            "  Found existing installation: tqdm 4.28.1\n",
            "    Uninstalling tqdm-4.28.1:\n",
            "      Successfully uninstalled tqdm-4.28.1\n",
            "  Found existing installation: matplotlib 3.0.2\n",
            "    Uninstalling matplotlib-3.0.2:\n",
            "      Successfully uninstalled matplotlib-3.0.2\n",
            "Successfully installed deprecated-1.2.4 flair-0.4.0 gensim-3.4.0 matplotlib-3.0.0 mpld3-0.3 pytorch-pretrained-bert-0.3.0 segtok-1.5.7 sqlitedict-1.6.0 tqdm-4.26.0 typing-3.6.4\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.colab-display-data+json": {
              "pip_warning": {
                "packages": [
                  "matplotlib",
                  "mpl_toolkits",
                  "typing"
                ]
              }
            }
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "--2019-02-11 08:22:18--  http:https://gemeos.org/henrique/training_data_1450.csv\n",
            "Resolving gemeos.org (gemeos.org)... 187.45.195.133\n",
            "Connecting to gemeos.org (gemeos.org)|187.45.195.133|:80... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 1741760 (1.7M) [text/plain]\n",
            "Saving to: ‘training_data_1450.csv’\n",
            "\n",
            "training_data_1450. 100%[===================>]   1.66M  1.52MB/s    in 1.1s    \n",
            "\n",
            "2019-02-11 08:22:20 (1.52 MB/s) - ‘training_data_1450.csv’ saved [1741760/1741760]\n",
            "\n",
            "--2019-02-11 08:22:21--  https://github.com/nathanshartmann/portuguese_word_embeddings/raw/master/preprocessing.py\n",
            "Resolving github.com (github.com)... 192.30.253.112, 192.30.253.113\n",
            "Connecting to github.com (github.com)|192.30.253.112|:443... connected.\n",
            "HTTP request sent, awaiting response... 302 Found\n",
            "Location: https://raw.githubusercontent.com/nathanshartmann/portuguese_word_embeddings/master/preprocessing.py [following]\n",
            "--2019-02-11 08:22:22--  https://raw.githubusercontent.com/nathanshartmann/portuguese_word_embeddings/master/preprocessing.py\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.0.133, 151.101.64.133, 151.101.128.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.0.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 4510 (4.4K) [text/plain]\n",
            "Saving to: ‘preprocessing.py’\n",
            "\n",
            "preprocessing.py    100%[===================>]   4.40K  --.-KB/s    in 0s      \n",
            "\n",
            "2019-02-11 08:22:22 (108 MB/s) - ‘preprocessing.py’ saved [4510/4510]\n",
            "\n",
            "--2019-02-11 08:22:23--  http:https://gemeos.org/henrique/health_word2vec_300v1.tar.gz\n",
            "Resolving gemeos.org (gemeos.org)... 187.45.195.133\n",
            "Connecting to gemeos.org (gemeos.org)|187.45.195.133|:80... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 179913428 (172M) [application/x-gzip]\n",
            "Saving to: ‘health_word2vec_300v1.tar.gz’\n",
            "\n",
            "health_word2vec_300 100%[===================>] 171.58M  16.4MB/s    in 12s     \n",
            "\n",
            "2019-02-11 08:22:35 (14.8 MB/s) - ‘health_word2vec_300v1.tar.gz’ saved [179913428/179913428]\n",
            "\n",
            "health_word2vec_300.model\n",
            "health_word2vec_300.model.trainables.syn1neg.npy\n",
            "health_word2vec_300.model.wv.vectors.npy\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "JIhHlX-9ue3g",
        "colab_type": "code",
        "outputId": "031f7738-a813-44ca-c264-f3fc8f6c59b9",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "from flair.embeddings import WordEmbeddings\n",
        "import gensim\n",
        "import pandas as pd\n",
        "\n",
        "quedas = pd.read_csv('training_data_1450.csv').drop_duplicates('Evolucao').reset_index()\n",
        "quedas.to_csv('training_data.csv', index=None)\n",
        "\n",
        "custom_embedding = WordEmbeddings('health_word2vec_300.model')\n",
        "\n",
        "quedas.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(1402, 35)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 2
        }
      ]
    },
    {
      "metadata": {
        "id": "gmFZYlxuvb6Y",
        "colab_type": "code",
        "outputId": "3e096fb8-183e-4d5b-dded-d8fb1b82f77e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 105
        }
      },
      "cell_type": "code",
      "source": [
        "import unicodedata\n",
        "import nltk\n",
        "from nltk.corpus import stopwords\n",
        "from flair.data import Sentence, TaggedCorpus\n",
        "\n",
        "nltk.download('punkt')\n",
        "nltk.download('stopwords')\n",
        "\n",
        "import preprocessing\n",
        "\n",
        "def tokenizer2(text):\n",
        "    return_text = []\n",
        "    \n",
        "    for sentence in text:\n",
        "        sentence_clean = preprocessing.clean_text(sentence)\n",
        "        return_text.append(sentence_clean)\n",
        "        \n",
        "    return return_text\n",
        "\n",
        "def remove_accents(input_str):\n",
        "    nfkd_form = unicodedata.normalize('NFKD', input_str)\n",
        "    return u\"\".join([c for c in nfkd_form if not unicodedata.combining(c)])\n",
        "\n",
        "def tokenizer(text):\n",
        "    return_text = []\n",
        "    sw_port = stopwords.words(\"portuguese\")\n",
        "    for sentence in text:\n",
        "        reg_tokenizer = nltk.tokenize.RegexpTokenizer(r'\\w+')\n",
        "        tokens = reg_tokenizer.tokenize(sentence)\n",
        "        return_text.append(' '.join([remove_accents(w.lower()) for w in tokens if w not in sw_port]))\n",
        "        \n",
        "    return return_text\n",
        " \n",
        "  \n",
        "tokens_set = tokenizer(quedas['Evolucao'].values)\n",
        "\n",
        "sentences = []\n",
        "\n",
        "for i, text in enumerate(tokens_set):\n",
        "  sentence = Sentence(text, labels=str(quedas['Target'][i]), use_tokenizer=True)\n",
        "  sentences.append(sentence)\n",
        "  \n",
        "corpus = TaggedCorpus(sentences[:800], sentences[800:900], sentences[900:1070])  "
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[nltk_data] Downloading package punkt to /root/nltk_data...\n",
            "[nltk_data]   Unzipping tokenizers/punkt.zip.\n",
            "[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
            "[nltk_data]   Unzipping corpora/stopwords.zip.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "4drBDe_8EuUG",
        "colab_type": "code",
        "outputId": "40e24c8a-1bc3-4811-8ecd-5fee411f0f95",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 241
        }
      },
      "cell_type": "code",
      "source": [
        "#!rm *s300*\n",
        "#!wget -c http:https://143.107.183.175:22980/download.php?file=embeddings/fasttext/cbow_s300.zip -O cbow_s300.zip #\n",
        "!wget -c http:https://143.107.183.175:22980/download.php?file=embeddings/word2vec/cbow_s300.zip -O cbow_s300.zip #0.77\n",
        "#!wget -c http:https://143.107.183.175:22980/download.php?file=embeddings/glove/glove_s300.zip -O glove_s300.zip\n",
        "#!wget -c http:https://143.107.183.175:22980/download.php?file=embeddings/fasttext/skip_s300.zip -O skip_s300.zip\n",
        "#!wget -c http:https://143.107.183.175:22980/download.php?file=embeddings/word2vec/skip_s300.zip -O skip_s300.zip\n",
        "#!wget -c http:https://143.107.183.175:22980/download.php?file=embeddings/wang2vec/cbow_s300.zip -O cbow_s300.zip #\n",
        "!unzip -o cbow_s300.zip\n",
        "\n",
        "nilc_vectors = gensim.models.KeyedVectors.load_word2vec_format('cbow_s300.txt')\n",
        "nilc_vectors.save('nilc.gensim')\n",
        "\n",
        "nilc_embedding = WordEmbeddings('nilc.gensim')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2019-02-11 08:22:47--  http:https://143.107.183.175:22980/download.php?file=embeddings/word2vec/cbow_s300.zip\n",
            "Connecting to 143.107.183.175:22980... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 929305948 (886M) [application/octet-stream]\n",
            "Saving to: ‘cbow_s300.zip’\n",
            "\n",
            "cbow_s300.zip       100%[===================>] 886.25M  5.22MB/s    in 91s     \n",
            "\n",
            "2019-02-11 08:24:18 (9.75 MB/s) - ‘cbow_s300.zip’ saved [929305948/929305948]\n",
            "\n",
            "Archive:  cbow_s300.zip\n",
            "  inflating: cbow_s300.txt           \n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "Y6v7-dWR2FyW",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "from flair.embeddings import WordEmbeddings, FlairEmbeddings, DocumentLSTMEmbeddings, BertEmbeddings\n",
        "from flair.models import TextClassifier\n",
        "from flair.trainers import ModelTrainer\n",
        "\n",
        "#pt_embeddings = WordEmbeddings('pt')\n",
        "#flair_embedding_forward = FlairEmbeddings('portuguese-forward')\n",
        "#flair_embedding_backward = FlairEmbeddings('portuguese-backward')\n",
        "#bert_embedding = BertEmbeddings('bert-base-multilingual-cased')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "9RnLQ4J7a8-m",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score, roc_auc_score\n",
        "\n",
        "def getScores(y, yPred):\n",
        "    return (accuracy_score(y, yPred), \n",
        "            precision_score(y, yPred), \n",
        "            recall_score(y, yPred),\n",
        "            f1_score(y, yPred, average='micro'),\n",
        "            f1_score(y, yPred, average='macro'),\n",
        "            f1_score(y, yPred, average='binary'),\n",
        "            roc_auc_score(y, yPred))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "7GIMe-dU3tuv",
        "colab_type": "code",
        "outputId": "6b1c0522-bbd5-4d53-b12f-b8558402dcee",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 97770
        }
      },
      "cell_type": "code",
      "source": [
        "from sklearn.model_selection import StratifiedKFold\n",
        "import numpy as np\n",
        "\n",
        "target_set = quedas['Target'].values\n",
        "\n",
        "results = []\n",
        "\n",
        "## TEST\n",
        "kfold_test = StratifiedKFold(n_splits=3)\n",
        "for i, (train, test) in enumerate(kfold_test.split(target_set, target_set)):\n",
        "\n",
        "  ## DEV\n",
        "  kfold_dev = StratifiedKFold(n_splits=10)\n",
        "  for train, dev in kfold_dev.split(target_set[train], target_set[train]):\n",
        "    break\n",
        "\n",
        "  print(len(train), len(dev), len(test))\n",
        "  corpus = TaggedCorpus([sentences[t] for t in train], [sentences[d] for d in dev], [sentences[e] for e in test]) \n",
        "\n",
        "  label_dict = corpus.make_label_dictionary()\n",
        "  \n",
        "  word_embeddings = [custom_embedding,\n",
        "                     nilc_embedding,\n",
        "                     #pt_embeddings,\n",
        "                     #flair_embedding_forward, \n",
        "                     #flair_embedding_backward, \n",
        "                     #bert_embedding,\n",
        "                    ]\n",
        "  \n",
        "  document_embeddings: DocumentLSTMEmbeddings = DocumentLSTMEmbeddings(word_embeddings,\n",
        "                                                                       hidden_size=512,\n",
        "                                                                       reproject_words=True,\n",
        "                                                                       reproject_words_dimension=256,\n",
        "                                                                       )\n",
        "    \n",
        "  classifier = TextClassifier(document_embeddings, label_dictionary=label_dict, multi_label=False)\n",
        "  \n",
        "  trainer = ModelTrainer(classifier, corpus)\n",
        "\n",
        "  trainer.train('quedas_'+str(i)+'/')\n",
        "  \n",
        "  pred_set = []\n",
        "  true_set = []\n",
        "  for e in test:\n",
        "    sentence = sentences[e]\n",
        "    true_set.append(target_set[e])\n",
        "    classifier.predict(sentence)\n",
        "    pred_set.append(int(sentence.labels[0].value))\n",
        "\n",
        "  a, p, r, fmi, fma, fb, roc = getScores(true_set,pred_set)\n",
        "  results.append(fb)\n",
        "  print (round(a,3), round(p,3), round(r,3), round(fmi,3), round(fma,3), round(fb,3), round(roc,3))\n",
        "  \n",
        "  #break\n",
        "  \n",
        "print(results)\n",
        "print('Mean: ', np.mean(results), 'Std: ', np.std(results))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "839 95 468\n",
            "2019-02-11 08:28:33,185 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:28:33,187 Evaluation method: MICRO_F1_SCORE\n",
            "2019-02-11 08:28:33,192 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:28:33,879 epoch 1 - iter 0/27 - loss 0.02758908\n",
            "2019-02-11 08:28:34,859 epoch 1 - iter 2/27 - loss 0.05171199\n",
            "2019-02-11 08:28:35,871 epoch 1 - iter 4/27 - loss 0.06803527\n",
            "2019-02-11 08:28:36,820 epoch 1 - iter 6/27 - loss 0.07204746\n",
            "2019-02-11 08:28:37,744 epoch 1 - iter 8/27 - loss 0.07167609\n",
            "2019-02-11 08:28:39,108 epoch 1 - iter 10/27 - loss 0.07094002\n",
            "2019-02-11 08:28:39,978 epoch 1 - iter 12/27 - loss 0.06780434\n",
            "2019-02-11 08:28:40,963 epoch 1 - iter 14/27 - loss 0.06533567\n",
            "2019-02-11 08:28:41,958 epoch 1 - iter 16/27 - loss 0.06449418\n",
            "2019-02-11 08:28:43,534 epoch 1 - iter 18/27 - loss 0.06315144\n",
            "2019-02-11 08:28:44,561 epoch 1 - iter 20/27 - loss 0.06166700\n",
            "2019-02-11 08:28:45,560 epoch 1 - iter 22/27 - loss 0.05971067\n",
            "2019-02-11 08:28:47,103 epoch 1 - iter 24/27 - loss 0.05781486\n",
            "2019-02-11 08:28:47,790 epoch 1 - iter 26/27 - loss 0.05763743\n",
            "2019-02-11 08:28:47,804 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:28:47,805 EPOCH 1 done: loss 0.0576 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:28:48,800 DEV  : loss 0.02195508 - f-score 0.6316 - acc 0.6316\n",
            "2019-02-11 08:28:53,232 TEST : loss 0.02208458 - f-score 0.6474 - acc 0.6474\n",
            "2019-02-11 08:29:12,508 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:29:12,762 epoch 2 - iter 0/27 - loss 0.02169669\n",
            "2019-02-11 08:29:13,495 epoch 2 - iter 2/27 - loss 0.03282100\n",
            "2019-02-11 08:29:14,671 epoch 2 - iter 4/27 - loss 0.03258940\n",
            "2019-02-11 08:29:15,343 epoch 2 - iter 6/27 - loss 0.03140275\n",
            "2019-02-11 08:29:16,175 epoch 2 - iter 8/27 - loss 0.03323443\n",
            "2019-02-11 08:29:17,449 epoch 2 - iter 10/27 - loss 0.03123621\n",
            "2019-02-11 08:29:18,555 epoch 2 - iter 12/27 - loss 0.02966753\n",
            "2019-02-11 08:29:19,317 epoch 2 - iter 14/27 - loss 0.02912895\n",
            "2019-02-11 08:29:20,099 epoch 2 - iter 16/27 - loss 0.02992936\n",
            "2019-02-11 08:29:20,929 epoch 2 - iter 18/27 - loss 0.03005168\n",
            "2019-02-11 08:29:21,666 epoch 2 - iter 20/27 - loss 0.02938979\n",
            "2019-02-11 08:29:22,277 epoch 2 - iter 22/27 - loss 0.02917307\n",
            "2019-02-11 08:29:23,016 epoch 2 - iter 24/27 - loss 0.02892919\n",
            "2019-02-11 08:29:23,511 epoch 2 - iter 26/27 - loss 0.02908288\n",
            "2019-02-11 08:29:23,524 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:29:23,525 EPOCH 2 done: loss 0.0291 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:29:24,159 DEV  : loss 0.02310936 - f-score 0.6105 - acc 0.6105\n",
            "2019-02-11 08:29:28,295 TEST : loss 0.02362916 - f-score 0.6068 - acc 0.6068\n",
            "2019-02-11 08:29:46,638 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:29:47,208 epoch 3 - iter 0/27 - loss 0.02562850\n",
            "2019-02-11 08:29:48,136 epoch 3 - iter 2/27 - loss 0.03181825\n",
            "2019-02-11 08:29:49,402 epoch 3 - iter 4/27 - loss 0.03510916\n",
            "2019-02-11 08:29:50,579 epoch 3 - iter 6/27 - loss 0.03143580\n",
            "2019-02-11 08:29:51,400 epoch 3 - iter 8/27 - loss 0.02978881\n",
            "2019-02-11 08:29:52,256 epoch 3 - iter 10/27 - loss 0.02976570\n",
            "2019-02-11 08:29:52,746 epoch 3 - iter 12/27 - loss 0.02892081\n",
            "2019-02-11 08:29:53,585 epoch 3 - iter 14/27 - loss 0.02895735\n",
            "2019-02-11 08:29:54,317 epoch 3 - iter 16/27 - loss 0.02927843\n",
            "2019-02-11 08:29:55,208 epoch 3 - iter 18/27 - loss 0.02916255\n",
            "2019-02-11 08:29:56,226 epoch 3 - iter 20/27 - loss 0.02875840\n",
            "2019-02-11 08:29:56,916 epoch 3 - iter 22/27 - loss 0.02795560\n",
            "2019-02-11 08:29:57,426 epoch 3 - iter 24/27 - loss 0.02851525\n",
            "2019-02-11 08:29:58,011 epoch 3 - iter 26/27 - loss 0.02894777\n",
            "2019-02-11 08:29:58,024 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:29:58,025 EPOCH 3 done: loss 0.0289 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:29:58,706 DEV  : loss 0.02477501 - f-score 0.5263 - acc 0.5263\n",
            "2019-02-11 08:30:03,202 TEST : loss 0.02459433 - f-score 0.5491 - acc 0.5491\n",
            "2019-02-11 08:30:21,240 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:30:21,640 epoch 4 - iter 0/27 - loss 0.03117821\n",
            "2019-02-11 08:30:22,486 epoch 4 - iter 2/27 - loss 0.03597003\n",
            "2019-02-11 08:30:23,230 epoch 4 - iter 4/27 - loss 0.03057512\n",
            "2019-02-11 08:30:23,962 epoch 4 - iter 6/27 - loss 0.02772910\n",
            "2019-02-11 08:30:24,692 epoch 4 - iter 8/27 - loss 0.02592927\n",
            "2019-02-11 08:30:25,348 epoch 4 - iter 10/27 - loss 0.02638677\n",
            "2019-02-11 08:30:26,373 epoch 4 - iter 12/27 - loss 0.02739184\n",
            "2019-02-11 08:30:26,995 epoch 4 - iter 14/27 - loss 0.02656283\n",
            "2019-02-11 08:30:28,340 epoch 4 - iter 16/27 - loss 0.02609623\n",
            "2019-02-11 08:30:29,552 epoch 4 - iter 18/27 - loss 0.02596024\n",
            "2019-02-11 08:30:30,357 epoch 4 - iter 20/27 - loss 0.02642508\n",
            "2019-02-11 08:30:31,158 epoch 4 - iter 22/27 - loss 0.02590866\n",
            "2019-02-11 08:30:31,904 epoch 4 - iter 24/27 - loss 0.02576583\n",
            "2019-02-11 08:30:32,273 epoch 4 - iter 26/27 - loss 0.02662449\n",
            "2019-02-11 08:30:32,287 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:30:32,289 EPOCH 4 done: loss 0.0266 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:30:32,964 DEV  : loss 0.05489873 - f-score 0.5263 - acc 0.5263\n",
            "2019-02-11 08:30:37,534 TEST : loss 0.05369084 - f-score 0.5385 - acc 0.5385\n",
            "2019-02-11 08:30:55,231 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:30:55,870 epoch 5 - iter 0/27 - loss 0.05171692\n",
            "2019-02-11 08:30:56,687 epoch 5 - iter 2/27 - loss 0.03486621\n",
            "2019-02-11 08:30:57,995 epoch 5 - iter 4/27 - loss 0.02932716\n",
            "2019-02-11 08:30:59,175 epoch 5 - iter 6/27 - loss 0.02583709\n",
            "2019-02-11 08:30:59,916 epoch 5 - iter 8/27 - loss 0.02644474\n",
            "2019-02-11 08:31:00,705 epoch 5 - iter 10/27 - loss 0.02771010\n",
            "2019-02-11 08:31:01,426 epoch 5 - iter 12/27 - loss 0.02716100\n",
            "2019-02-11 08:31:02,134 epoch 5 - iter 14/27 - loss 0.02713007\n",
            "2019-02-11 08:31:02,651 epoch 5 - iter 16/27 - loss 0.02670624\n",
            "2019-02-11 08:31:03,371 epoch 5 - iter 18/27 - loss 0.02657100\n",
            "2019-02-11 08:31:04,120 epoch 5 - iter 20/27 - loss 0.02611885\n",
            "2019-02-11 08:31:04,910 epoch 5 - iter 22/27 - loss 0.02627897\n",
            "2019-02-11 08:31:05,640 epoch 5 - iter 24/27 - loss 0.02613621\n",
            "2019-02-11 08:31:06,178 epoch 5 - iter 26/27 - loss 0.02665392\n",
            "2019-02-11 08:31:06,190 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:31:06,192 EPOCH 5 done: loss 0.0267 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:31:06,855 DEV  : loss 0.02503199 - f-score 0.5263 - acc 0.5263\n",
            "2019-02-11 08:31:11,367 TEST : loss 0.02372567 - f-score 0.5406 - acc 0.5406\n",
            "2019-02-11 08:31:11,370 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:31:11,809 epoch 6 - iter 0/27 - loss 0.02234727\n",
            "2019-02-11 08:31:12,707 epoch 6 - iter 2/27 - loss 0.02158467\n",
            "2019-02-11 08:31:13,302 epoch 6 - iter 4/27 - loss 0.02512250\n",
            "2019-02-11 08:31:14,046 epoch 6 - iter 6/27 - loss 0.02596054\n",
            "2019-02-11 08:31:14,766 epoch 6 - iter 8/27 - loss 0.02560530\n",
            "2019-02-11 08:31:15,589 epoch 6 - iter 10/27 - loss 0.02761441\n",
            "2019-02-11 08:31:16,440 epoch 6 - iter 12/27 - loss 0.02766278\n",
            "2019-02-11 08:31:17,410 epoch 6 - iter 14/27 - loss 0.02739800\n",
            "2019-02-11 08:31:18,261 epoch 6 - iter 16/27 - loss 0.02726635\n",
            "2019-02-11 08:31:18,975 epoch 6 - iter 18/27 - loss 0.02678132\n",
            "2019-02-11 08:31:20,078 epoch 6 - iter 20/27 - loss 0.02602446\n",
            "2019-02-11 08:31:21,351 epoch 6 - iter 22/27 - loss 0.02555145\n",
            "2019-02-11 08:31:22,154 epoch 6 - iter 24/27 - loss 0.02587396\n",
            "2019-02-11 08:31:22,681 epoch 6 - iter 26/27 - loss 0.02684746\n",
            "2019-02-11 08:31:22,693 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:31:22,694 EPOCH 6 done: loss 0.0268 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 08:31:23,348 DEV  : loss 0.02607711 - f-score 0.5368 - acc 0.5368\n",
            "2019-02-11 08:31:27,649 TEST : loss 0.02545453 - f-score 0.5662 - acc 0.5662\n",
            "2019-02-11 08:31:27,652 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:31:28,032 epoch 7 - iter 0/27 - loss 0.02858212\n",
            "2019-02-11 08:31:28,608 epoch 7 - iter 2/27 - loss 0.02763368\n",
            "2019-02-11 08:31:29,339 epoch 7 - iter 4/27 - loss 0.02770934\n",
            "2019-02-11 08:31:30,206 epoch 7 - iter 6/27 - loss 0.02716306\n",
            "2019-02-11 08:31:30,799 epoch 7 - iter 8/27 - loss 0.02663483\n",
            "2019-02-11 08:31:31,462 epoch 7 - iter 10/27 - loss 0.02590699\n",
            "2019-02-11 08:31:32,311 epoch 7 - iter 12/27 - loss 0.02555001\n",
            "2019-02-11 08:31:33,172 epoch 7 - iter 14/27 - loss 0.02502198\n",
            "2019-02-11 08:31:34,326 epoch 7 - iter 16/27 - loss 0.02466707\n",
            "2019-02-11 08:31:35,055 epoch 7 - iter 18/27 - loss 0.02455253\n",
            "2019-02-11 08:31:35,806 epoch 7 - iter 20/27 - loss 0.02641321\n",
            "2019-02-11 08:31:36,389 epoch 7 - iter 22/27 - loss 0.02646287\n",
            "2019-02-11 08:31:37,777 epoch 7 - iter 24/27 - loss 0.02605841\n",
            "2019-02-11 08:31:38,216 epoch 7 - iter 26/27 - loss 0.02736096\n",
            "2019-02-11 08:31:38,229 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:31:38,230 EPOCH 7 done: loss 0.0274 - lr 0.1000 - bad epochs 2\n",
            "2019-02-11 08:31:38,839 DEV  : loss 0.02055985 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 08:31:43,044 TEST : loss 0.02034629 - f-score 0.6688 - acc 0.6688\n",
            "2019-02-11 08:31:43,047 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:31:43,800 epoch 8 - iter 0/27 - loss 0.02439743\n",
            "2019-02-11 08:31:44,551 epoch 8 - iter 2/27 - loss 0.02248557\n",
            "2019-02-11 08:31:45,279 epoch 8 - iter 4/27 - loss 0.02301727\n",
            "2019-02-11 08:31:45,994 epoch 8 - iter 6/27 - loss 0.02364530\n",
            "2019-02-11 08:31:46,804 epoch 8 - iter 8/27 - loss 0.02535035\n",
            "2019-02-11 08:31:47,518 epoch 8 - iter 10/27 - loss 0.02529754\n",
            "2019-02-11 08:31:48,166 epoch 8 - iter 12/27 - loss 0.02490409\n",
            "2019-02-11 08:31:48,730 epoch 8 - iter 14/27 - loss 0.02438060\n",
            "2019-02-11 08:31:49,328 epoch 8 - iter 16/27 - loss 0.02543641\n",
            "2019-02-11 08:31:50,037 epoch 8 - iter 18/27 - loss 0.02536348\n",
            "2019-02-11 08:31:50,663 epoch 8 - iter 20/27 - loss 0.02481730\n",
            "2019-02-11 08:31:51,753 epoch 8 - iter 22/27 - loss 0.02483892\n",
            "2019-02-11 08:31:52,828 epoch 8 - iter 24/27 - loss 0.02462891\n",
            "2019-02-11 08:31:53,217 epoch 8 - iter 26/27 - loss 0.02538518\n",
            "2019-02-11 08:31:53,228 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:31:53,229 EPOCH 8 done: loss 0.0254 - lr 0.1000 - bad epochs 3\n",
            "2019-02-11 08:31:53,836 DEV  : loss 0.02154092 - f-score 0.5368 - acc 0.5368\n",
            "2019-02-11 08:31:57,925 TEST : loss 0.01991938 - f-score 0.6261 - acc 0.6261\n",
            "2019-02-11 08:32:14,972 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:32:15,451 epoch 9 - iter 0/27 - loss 0.02089767\n",
            "2019-02-11 08:32:16,419 epoch 9 - iter 2/27 - loss 0.02457110\n",
            "2019-02-11 08:32:17,156 epoch 9 - iter 4/27 - loss 0.03008956\n",
            "2019-02-11 08:32:18,279 epoch 9 - iter 6/27 - loss 0.02745587\n",
            "2019-02-11 08:32:19,111 epoch 9 - iter 8/27 - loss 0.02889819\n",
            "2019-02-11 08:32:19,863 epoch 9 - iter 10/27 - loss 0.02820899\n",
            "2019-02-11 08:32:20,717 epoch 9 - iter 12/27 - loss 0.02713671\n",
            "2019-02-11 08:32:21,210 epoch 9 - iter 14/27 - loss 0.02662573\n",
            "2019-02-11 08:32:21,952 epoch 9 - iter 16/27 - loss 0.02575762\n",
            "2019-02-11 08:32:22,753 epoch 9 - iter 18/27 - loss 0.02564065\n",
            "2019-02-11 08:32:23,788 epoch 9 - iter 20/27 - loss 0.02624300\n",
            "2019-02-11 08:32:24,439 epoch 9 - iter 22/27 - loss 0.02611433\n",
            "2019-02-11 08:32:25,145 epoch 9 - iter 24/27 - loss 0.02597934\n",
            "2019-02-11 08:32:25,573 epoch 9 - iter 26/27 - loss 0.02761084\n",
            "2019-02-11 08:32:25,585 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:32:25,586 EPOCH 9 done: loss 0.0276 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:32:26,210 DEV  : loss 0.02096766 - f-score 0.6211 - acc 0.6211\n",
            "2019-02-11 08:32:30,417 TEST : loss 0.02026514 - f-score 0.6453 - acc 0.6453\n",
            "2019-02-11 08:32:30,436 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:32:30,750 epoch 10 - iter 0/27 - loss 0.02201578\n",
            "2019-02-11 08:32:31,356 epoch 10 - iter 2/27 - loss 0.02252441\n",
            "2019-02-11 08:32:32,136 epoch 10 - iter 4/27 - loss 0.02498514\n",
            "2019-02-11 08:32:33,243 epoch 10 - iter 6/27 - loss 0.02380140\n",
            "2019-02-11 08:32:33,817 epoch 10 - iter 8/27 - loss 0.02353335\n",
            "2019-02-11 08:32:34,599 epoch 10 - iter 10/27 - loss 0.02379457\n",
            "2019-02-11 08:32:35,502 epoch 10 - iter 12/27 - loss 0.02389190\n",
            "2019-02-11 08:32:36,694 epoch 10 - iter 14/27 - loss 0.02423726\n",
            "2019-02-11 08:32:37,430 epoch 10 - iter 16/27 - loss 0.02389840\n",
            "2019-02-11 08:32:38,277 epoch 10 - iter 18/27 - loss 0.02354682\n",
            "2019-02-11 08:32:39,004 epoch 10 - iter 20/27 - loss 0.02356402\n",
            "2019-02-11 08:32:39,813 epoch 10 - iter 22/27 - loss 0.02428347\n",
            "2019-02-11 08:32:40,599 epoch 10 - iter 24/27 - loss 0.02401206\n",
            "2019-02-11 08:32:40,954 epoch 10 - iter 26/27 - loss 0.02449655\n",
            "2019-02-11 08:32:40,965 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:32:40,967 EPOCH 10 done: loss 0.0245 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 08:32:41,586 DEV  : loss 0.02312580 - f-score 0.6105 - acc 0.6105\n",
            "2019-02-11 08:32:45,779 TEST : loss 0.02110423 - f-score 0.6496 - acc 0.6496\n",
            "2019-02-11 08:33:02,916 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:33:03,389 epoch 11 - iter 0/27 - loss 0.01970967\n",
            "2019-02-11 08:33:04,048 epoch 11 - iter 2/27 - loss 0.02896300\n",
            "2019-02-11 08:33:05,147 epoch 11 - iter 4/27 - loss 0.02568234\n",
            "2019-02-11 08:33:05,920 epoch 11 - iter 6/27 - loss 0.02582404\n",
            "2019-02-11 08:33:06,539 epoch 11 - iter 8/27 - loss 0.02547624\n",
            "2019-02-11 08:33:07,269 epoch 11 - iter 10/27 - loss 0.02588685\n",
            "2019-02-11 08:33:07,918 epoch 11 - iter 12/27 - loss 0.02629650\n",
            "2019-02-11 08:33:08,823 epoch 11 - iter 14/27 - loss 0.02581385\n",
            "2019-02-11 08:33:09,486 epoch 11 - iter 16/27 - loss 0.02522893\n",
            "2019-02-11 08:33:10,352 epoch 11 - iter 18/27 - loss 0.02519027\n",
            "2019-02-11 08:33:11,213 epoch 11 - iter 20/27 - loss 0.02501034\n",
            "2019-02-11 08:33:12,430 epoch 11 - iter 22/27 - loss 0.02521581\n",
            "2019-02-11 08:33:13,208 epoch 11 - iter 24/27 - loss 0.02503725\n",
            "2019-02-11 08:33:13,675 epoch 11 - iter 26/27 - loss 0.02559447\n",
            "2019-02-11 08:33:13,685 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:33:13,686 EPOCH 11 done: loss 0.0256 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:33:14,321 DEV  : loss 0.02383835 - f-score 0.5895 - acc 0.5895\n",
            "2019-02-11 08:33:18,555 TEST : loss 0.02127666 - f-score 0.6325 - acc 0.6325\n",
            "2019-02-11 08:33:18,558 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:33:18,876 epoch 12 - iter 0/27 - loss 0.01740193\n",
            "2019-02-11 08:33:19,504 epoch 12 - iter 2/27 - loss 0.02649861\n",
            "2019-02-11 08:33:20,128 epoch 12 - iter 4/27 - loss 0.02467739\n",
            "2019-02-11 08:33:21,007 epoch 12 - iter 6/27 - loss 0.02439919\n",
            "2019-02-11 08:33:21,692 epoch 12 - iter 8/27 - loss 0.02309933\n",
            "2019-02-11 08:33:22,457 epoch 12 - iter 10/27 - loss 0.02347881\n",
            "2019-02-11 08:33:23,255 epoch 12 - iter 12/27 - loss 0.02407610\n",
            "2019-02-11 08:33:24,180 epoch 12 - iter 14/27 - loss 0.02443281\n",
            "2019-02-11 08:33:24,866 epoch 12 - iter 16/27 - loss 0.02514530\n",
            "2019-02-11 08:33:25,785 epoch 12 - iter 18/27 - loss 0.02521366\n",
            "2019-02-11 08:33:26,514 epoch 12 - iter 20/27 - loss 0.02498465\n",
            "2019-02-11 08:33:27,737 epoch 12 - iter 22/27 - loss 0.02464531\n",
            "2019-02-11 08:33:28,449 epoch 12 - iter 24/27 - loss 0.02488360\n",
            "2019-02-11 08:33:29,067 epoch 12 - iter 26/27 - loss 0.02567811\n",
            "2019-02-11 08:33:29,079 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:33:29,080 EPOCH 12 done: loss 0.0257 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 08:33:29,703 DEV  : loss 0.02414319 - f-score 0.5895 - acc 0.5895\n",
            "2019-02-11 08:33:33,874 TEST : loss 0.02125841 - f-score 0.6346 - acc 0.6346\n",
            "2019-02-11 08:33:33,876 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:33:34,268 epoch 13 - iter 0/27 - loss 0.02408131\n",
            "2019-02-11 08:33:35,326 epoch 13 - iter 2/27 - loss 0.02842312\n",
            "2019-02-11 08:33:36,107 epoch 13 - iter 4/27 - loss 0.02769668\n",
            "2019-02-11 08:33:36,838 epoch 13 - iter 6/27 - loss 0.02605506\n",
            "2019-02-11 08:33:37,487 epoch 13 - iter 8/27 - loss 0.02557441\n",
            "2019-02-11 08:33:38,189 epoch 13 - iter 10/27 - loss 0.02574805\n",
            "2019-02-11 08:33:38,888 epoch 13 - iter 12/27 - loss 0.02527342\n",
            "2019-02-11 08:33:39,495 epoch 13 - iter 14/27 - loss 0.02488652\n",
            "2019-02-11 08:33:40,445 epoch 13 - iter 16/27 - loss 0.02502164\n",
            "2019-02-11 08:33:41,118 epoch 13 - iter 18/27 - loss 0.02465601\n",
            "2019-02-11 08:33:41,715 epoch 13 - iter 20/27 - loss 0.02442677\n",
            "2019-02-11 08:33:42,433 epoch 13 - iter 22/27 - loss 0.02439277\n",
            "2019-02-11 08:33:43,725 epoch 13 - iter 24/27 - loss 0.02408339\n",
            "2019-02-11 08:33:44,162 epoch 13 - iter 26/27 - loss 0.02433574\n",
            "2019-02-11 08:33:44,173 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:33:44,174 EPOCH 13 done: loss 0.0243 - lr 0.1000 - bad epochs 2\n",
            "2019-02-11 08:33:44,790 DEV  : loss 0.03279693 - f-score 0.5474 - acc 0.5474\n",
            "2019-02-11 08:33:49,093 TEST : loss 0.02855659 - f-score 0.5577 - acc 0.5577\n",
            "2019-02-11 08:34:07,309 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:34:08,004 epoch 14 - iter 0/27 - loss 0.02438605\n",
            "2019-02-11 08:34:08,826 epoch 14 - iter 2/27 - loss 0.02902768\n",
            "2019-02-11 08:34:09,692 epoch 14 - iter 4/27 - loss 0.02570243\n",
            "2019-02-11 08:34:10,398 epoch 14 - iter 6/27 - loss 0.02419067\n",
            "2019-02-11 08:34:11,270 epoch 14 - iter 8/27 - loss 0.02538470\n",
            "2019-02-11 08:34:11,939 epoch 14 - iter 10/27 - loss 0.02532147\n",
            "2019-02-11 08:34:12,672 epoch 14 - iter 12/27 - loss 0.02459890\n",
            "2019-02-11 08:34:13,327 epoch 14 - iter 14/27 - loss 0.02482694\n",
            "2019-02-11 08:34:13,962 epoch 14 - iter 16/27 - loss 0.02396347\n",
            "2019-02-11 08:34:14,708 epoch 14 - iter 18/27 - loss 0.02382364\n",
            "2019-02-11 08:34:15,892 epoch 14 - iter 20/27 - loss 0.02434537\n",
            "2019-02-11 08:34:16,621 epoch 14 - iter 22/27 - loss 0.02451994\n",
            "2019-02-11 08:34:17,225 epoch 14 - iter 24/27 - loss 0.02449090\n",
            "2019-02-11 08:34:18,011 epoch 14 - iter 26/27 - loss 0.02509710\n",
            "2019-02-11 08:34:18,022 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:34:18,023 EPOCH 14 done: loss 0.0251 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:34:18,629 DEV  : loss 0.01896110 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 08:34:22,764 TEST : loss 0.01722460 - f-score 0.7244 - acc 0.7244\n",
            "2019-02-11 08:34:22,777 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:34:23,173 epoch 15 - iter 0/27 - loss 0.02352237\n",
            "2019-02-11 08:34:23,870 epoch 15 - iter 2/27 - loss 0.02110863\n",
            "2019-02-11 08:34:24,432 epoch 15 - iter 4/27 - loss 0.02846260\n",
            "2019-02-11 08:34:25,099 epoch 15 - iter 6/27 - loss 0.02583704\n",
            "2019-02-11 08:34:25,902 epoch 15 - iter 8/27 - loss 0.02525085\n",
            "2019-02-11 08:34:26,807 epoch 15 - iter 10/27 - loss 0.02538026\n",
            "2019-02-11 08:34:28,050 epoch 15 - iter 12/27 - loss 0.02442056\n",
            "2019-02-11 08:34:29,159 epoch 15 - iter 14/27 - loss 0.02508696\n",
            "2019-02-11 08:34:29,702 epoch 15 - iter 16/27 - loss 0.02470296\n",
            "2019-02-11 08:34:30,362 epoch 15 - iter 18/27 - loss 0.02451211\n",
            "2019-02-11 08:34:31,131 epoch 15 - iter 20/27 - loss 0.02507658\n",
            "2019-02-11 08:34:31,850 epoch 15 - iter 22/27 - loss 0.02508158\n",
            "2019-02-11 08:34:32,427 epoch 15 - iter 24/27 - loss 0.02478592\n",
            "2019-02-11 08:34:32,826 epoch 15 - iter 26/27 - loss 0.02473902\n",
            "2019-02-11 08:34:32,837 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:34:32,839 EPOCH 15 done: loss 0.0247 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 08:34:33,445 DEV  : loss 0.02255815 - f-score 0.6211 - acc 0.6211\n",
            "2019-02-11 08:34:37,572 TEST : loss 0.01758791 - f-score 0.7265 - acc 0.7265\n",
            "2019-02-11 08:34:37,574 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:34:37,962 epoch 16 - iter 0/27 - loss 0.02127368\n",
            "2019-02-11 08:34:38,653 epoch 16 - iter 2/27 - loss 0.02047308\n",
            "2019-02-11 08:34:39,494 epoch 16 - iter 4/27 - loss 0.02036262\n",
            "2019-02-11 08:34:40,593 epoch 16 - iter 6/27 - loss 0.02137533\n",
            "2019-02-11 08:34:41,177 epoch 16 - iter 8/27 - loss 0.02165745\n",
            "2019-02-11 08:34:41,874 epoch 16 - iter 10/27 - loss 0.02282342\n",
            "2019-02-11 08:34:42,673 epoch 16 - iter 12/27 - loss 0.02290506\n",
            "2019-02-11 08:34:43,612 epoch 16 - iter 14/27 - loss 0.02386816\n",
            "2019-02-11 08:34:44,349 epoch 16 - iter 16/27 - loss 0.02307137\n",
            "2019-02-11 08:34:44,933 epoch 16 - iter 18/27 - loss 0.02274574\n",
            "2019-02-11 08:34:45,544 epoch 16 - iter 20/27 - loss 0.02274225\n",
            "2019-02-11 08:34:46,353 epoch 16 - iter 22/27 - loss 0.02258112\n",
            "2019-02-11 08:34:47,098 epoch 16 - iter 24/27 - loss 0.02243821\n",
            "2019-02-11 08:34:47,917 epoch 16 - iter 26/27 - loss 0.02288066\n",
            "2019-02-11 08:34:47,929 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:34:47,931 EPOCH 16 done: loss 0.0229 - lr 0.1000 - bad epochs 2\n",
            "2019-02-11 08:34:48,532 DEV  : loss 0.02116361 - f-score 0.6105 - acc 0.6105\n",
            "2019-02-11 08:34:52,798 TEST : loss 0.01824114 - f-score 0.6966 - acc 0.6966\n",
            "2019-02-11 08:35:09,550 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:35:10,232 epoch 17 - iter 0/27 - loss 0.02642301\n",
            "2019-02-11 08:35:10,952 epoch 17 - iter 2/27 - loss 0.02768940\n",
            "2019-02-11 08:35:11,467 epoch 17 - iter 4/27 - loss 0.02885403\n",
            "2019-02-11 08:35:12,587 epoch 17 - iter 6/27 - loss 0.02648319\n",
            "2019-02-11 08:35:13,384 epoch 17 - iter 8/27 - loss 0.02427208\n",
            "2019-02-11 08:35:14,028 epoch 17 - iter 10/27 - loss 0.02536609\n",
            "2019-02-11 08:35:14,705 epoch 17 - iter 12/27 - loss 0.02463743\n",
            "2019-02-11 08:35:15,356 epoch 17 - iter 14/27 - loss 0.02429083\n",
            "2019-02-11 08:35:16,066 epoch 17 - iter 16/27 - loss 0.02410653\n",
            "2019-02-11 08:35:17,203 epoch 17 - iter 18/27 - loss 0.02461598\n",
            "2019-02-11 08:35:18,007 epoch 17 - iter 20/27 - loss 0.02440562\n",
            "2019-02-11 08:35:18,697 epoch 17 - iter 22/27 - loss 0.02398999\n",
            "2019-02-11 08:35:19,476 epoch 17 - iter 24/27 - loss 0.02364935\n",
            "2019-02-11 08:35:20,037 epoch 17 - iter 26/27 - loss 0.02437728\n",
            "2019-02-11 08:35:20,048 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:35:20,049 EPOCH 17 done: loss 0.0244 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:35:20,659 DEV  : loss 0.01904892 - f-score 0.6737 - acc 0.6737\n",
            "2019-02-11 08:35:24,874 TEST : loss 0.01554511 - f-score 0.7735 - acc 0.7735\n",
            "2019-02-11 08:35:24,877 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:35:25,207 epoch 18 - iter 0/27 - loss 0.02409917\n",
            "2019-02-11 08:35:25,986 epoch 18 - iter 2/27 - loss 0.02436421\n",
            "2019-02-11 08:35:26,642 epoch 18 - iter 4/27 - loss 0.02374993\n",
            "2019-02-11 08:35:27,407 epoch 18 - iter 6/27 - loss 0.02196272\n",
            "2019-02-11 08:35:28,151 epoch 18 - iter 8/27 - loss 0.02125201\n",
            "2019-02-11 08:35:28,838 epoch 18 - iter 10/27 - loss 0.01990570\n",
            "2019-02-11 08:35:29,547 epoch 18 - iter 12/27 - loss 0.02030967\n",
            "2019-02-11 08:35:30,358 epoch 18 - iter 14/27 - loss 0.02103757\n",
            "2019-02-11 08:35:31,302 epoch 18 - iter 16/27 - loss 0.02074093\n",
            "2019-02-11 08:35:31,974 epoch 18 - iter 18/27 - loss 0.02072725\n",
            "2019-02-11 08:35:33,219 epoch 18 - iter 20/27 - loss 0.02081416\n",
            "2019-02-11 08:35:33,940 epoch 18 - iter 22/27 - loss 0.02074514\n",
            "2019-02-11 08:35:34,638 epoch 18 - iter 24/27 - loss 0.02080573\n",
            "2019-02-11 08:35:35,027 epoch 18 - iter 26/27 - loss 0.02122848\n",
            "2019-02-11 08:35:35,038 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:35:35,039 EPOCH 18 done: loss 0.0212 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 08:35:35,655 DEV  : loss 0.02101426 - f-score 0.5895 - acc 0.5895\n",
            "2019-02-11 08:35:39,982 TEST : loss 0.01522809 - f-score 0.7607 - acc 0.7607\n",
            "2019-02-11 08:35:57,055 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:35:57,643 epoch 19 - iter 0/27 - loss 0.01847053\n",
            "2019-02-11 08:35:58,476 epoch 19 - iter 2/27 - loss 0.02077107\n",
            "2019-02-11 08:35:59,384 epoch 19 - iter 4/27 - loss 0.01915770\n",
            "2019-02-11 08:36:00,955 epoch 19 - iter 6/27 - loss 0.02001008\n",
            "2019-02-11 08:36:01,714 epoch 19 - iter 8/27 - loss 0.02046037\n",
            "2019-02-11 08:36:02,395 epoch 19 - iter 10/27 - loss 0.01952612\n",
            "2019-02-11 08:36:03,211 epoch 19 - iter 12/27 - loss 0.02004170\n",
            "2019-02-11 08:36:04,090 epoch 19 - iter 14/27 - loss 0.02054333\n",
            "2019-02-11 08:36:04,666 epoch 19 - iter 16/27 - loss 0.02034998\n",
            "2019-02-11 08:36:05,229 epoch 19 - iter 18/27 - loss 0.02085837\n",
            "2019-02-11 08:36:05,972 epoch 19 - iter 20/27 - loss 0.02125340\n",
            "2019-02-11 08:36:06,630 epoch 19 - iter 22/27 - loss 0.02124071\n",
            "2019-02-11 08:36:07,330 epoch 19 - iter 24/27 - loss 0.02097232\n",
            "2019-02-11 08:36:07,873 epoch 19 - iter 26/27 - loss 0.02156733\n",
            "2019-02-11 08:36:07,885 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:36:07,886 EPOCH 19 done: loss 0.0216 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:36:08,505 DEV  : loss 0.02632090 - f-score 0.6526 - acc 0.6526\n",
            "2019-02-11 08:36:12,700 TEST : loss 0.02179886 - f-score 0.7137 - acc 0.7137\n",
            "2019-02-11 08:36:12,703 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:36:13,111 epoch 20 - iter 0/27 - loss 0.03677829\n",
            "2019-02-11 08:36:14,133 epoch 20 - iter 2/27 - loss 0.02453984\n",
            "2019-02-11 08:36:14,943 epoch 20 - iter 4/27 - loss 0.02291812\n",
            "2019-02-11 08:36:15,701 epoch 20 - iter 6/27 - loss 0.02445341\n",
            "2019-02-11 08:36:16,343 epoch 20 - iter 8/27 - loss 0.02241694\n",
            "2019-02-11 08:36:16,900 epoch 20 - iter 10/27 - loss 0.02133206\n",
            "2019-02-11 08:36:17,577 epoch 20 - iter 12/27 - loss 0.02115992\n",
            "2019-02-11 08:36:18,270 epoch 20 - iter 14/27 - loss 0.02056527\n",
            "2019-02-11 08:36:19,416 epoch 20 - iter 16/27 - loss 0.02105148\n",
            "2019-02-11 08:36:20,718 epoch 20 - iter 18/27 - loss 0.02104818\n",
            "2019-02-11 08:36:21,334 epoch 20 - iter 20/27 - loss 0.02061798\n",
            "2019-02-11 08:36:21,939 epoch 20 - iter 22/27 - loss 0.02056721\n",
            "2019-02-11 08:36:22,667 epoch 20 - iter 24/27 - loss 0.02037355\n",
            "2019-02-11 08:36:23,078 epoch 20 - iter 26/27 - loss 0.02053057\n",
            "2019-02-11 08:36:23,089 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:36:23,090 EPOCH 20 done: loss 0.0205 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 08:36:23,715 DEV  : loss 0.03086124 - f-score 0.5895 - acc 0.5895\n",
            "2019-02-11 08:36:27,906 TEST : loss 0.02102771 - f-score 0.6581 - acc 0.6581\n",
            "2019-02-11 08:36:44,938 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:36:45,464 epoch 21 - iter 0/27 - loss 0.02226019\n",
            "2019-02-11 08:36:46,137 epoch 21 - iter 2/27 - loss 0.02155101\n",
            "2019-02-11 08:36:47,014 epoch 21 - iter 4/27 - loss 0.01997963\n",
            "2019-02-11 08:36:48,130 epoch 21 - iter 6/27 - loss 0.02046124\n",
            "2019-02-11 08:36:48,846 epoch 21 - iter 8/27 - loss 0.02047543\n",
            "2019-02-11 08:36:49,987 epoch 21 - iter 10/27 - loss 0.02088790\n",
            "2019-02-11 08:36:50,561 epoch 21 - iter 12/27 - loss 0.01996684\n",
            "2019-02-11 08:36:51,345 epoch 21 - iter 14/27 - loss 0.02027207\n",
            "2019-02-11 08:36:52,297 epoch 21 - iter 16/27 - loss 0.02091458\n",
            "2019-02-11 08:36:52,996 epoch 21 - iter 18/27 - loss 0.02132392\n",
            "2019-02-11 08:36:53,731 epoch 21 - iter 20/27 - loss 0.02116537\n",
            "2019-02-11 08:36:54,464 epoch 21 - iter 22/27 - loss 0.02095819\n",
            "2019-02-11 08:36:55,285 epoch 21 - iter 24/27 - loss 0.02112025\n",
            "2019-02-11 08:36:55,736 epoch 21 - iter 26/27 - loss 0.02152363\n",
            "2019-02-11 08:36:55,750 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:36:55,752 EPOCH 21 done: loss 0.0215 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:36:56,366 DEV  : loss 0.02313410 - f-score 0.6526 - acc 0.6526\n",
            "2019-02-11 08:37:00,584 TEST : loss 0.01735145 - f-score 0.7500 - acc 0.7500\n",
            "2019-02-11 08:37:00,586 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:37:00,985 epoch 22 - iter 0/27 - loss 0.02378900\n",
            "2019-02-11 08:37:01,702 epoch 22 - iter 2/27 - loss 0.02218146\n",
            "2019-02-11 08:37:02,460 epoch 22 - iter 4/27 - loss 0.01919618\n",
            "2019-02-11 08:37:03,009 epoch 22 - iter 6/27 - loss 0.01949174\n",
            "2019-02-11 08:37:04,188 epoch 22 - iter 8/27 - loss 0.02088973\n",
            "2019-02-11 08:37:05,089 epoch 22 - iter 10/27 - loss 0.02178864\n",
            "2019-02-11 08:37:05,967 epoch 22 - iter 12/27 - loss 0.02086039\n",
            "2019-02-11 08:37:06,662 epoch 22 - iter 14/27 - loss 0.02009213\n",
            "2019-02-11 08:37:07,298 epoch 22 - iter 16/27 - loss 0.02031885\n",
            "2019-02-11 08:37:07,969 epoch 22 - iter 18/27 - loss 0.02021477\n",
            "2019-02-11 08:37:08,511 epoch 22 - iter 20/27 - loss 0.02022880\n",
            "2019-02-11 08:37:09,287 epoch 22 - iter 22/27 - loss 0.01995305\n",
            "2019-02-11 08:37:10,621 epoch 22 - iter 24/27 - loss 0.01988180\n",
            "2019-02-11 08:37:11,046 epoch 22 - iter 26/27 - loss 0.02028442\n",
            "2019-02-11 08:37:11,058 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:37:11,059 EPOCH 22 done: loss 0.0203 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 08:37:11,689 DEV  : loss 0.02821774 - f-score 0.6000 - acc 0.6000\n",
            "2019-02-11 08:37:15,906 TEST : loss 0.01729438 - f-score 0.7286 - acc 0.7286\n",
            "2019-02-11 08:37:32,982 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:37:33,528 epoch 23 - iter 0/27 - loss 0.02610574\n",
            "2019-02-11 08:37:34,708 epoch 23 - iter 2/27 - loss 0.02181340\n",
            "2019-02-11 08:37:35,514 epoch 23 - iter 4/27 - loss 0.02138894\n",
            "2019-02-11 08:37:36,372 epoch 23 - iter 6/27 - loss 0.01903968\n",
            "2019-02-11 08:37:37,141 epoch 23 - iter 8/27 - loss 0.01863935\n",
            "2019-02-11 08:37:37,834 epoch 23 - iter 10/27 - loss 0.01961733\n",
            "2019-02-11 08:37:38,407 epoch 23 - iter 12/27 - loss 0.01995254\n",
            "2019-02-11 08:37:39,028 epoch 23 - iter 14/27 - loss 0.01905290\n",
            "2019-02-11 08:37:40,168 epoch 23 - iter 16/27 - loss 0.01883567\n",
            "2019-02-11 08:37:41,094 epoch 23 - iter 18/27 - loss 0.01839564\n",
            "2019-02-11 08:37:41,924 epoch 23 - iter 20/27 - loss 0.01790108\n",
            "2019-02-11 08:37:42,532 epoch 23 - iter 22/27 - loss 0.01750316\n",
            "2019-02-11 08:37:43,143 epoch 23 - iter 24/27 - loss 0.01735942\n",
            "2019-02-11 08:37:43,591 epoch 23 - iter 26/27 - loss 0.01761150\n",
            "2019-02-11 08:37:43,602 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:37:43,603 EPOCH 23 done: loss 0.0176 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:37:44,224 DEV  : loss 0.02914619 - f-score 0.6211 - acc 0.6211\n",
            "2019-02-11 08:37:48,419 TEST : loss 0.01625979 - f-score 0.7585 - acc 0.7585\n",
            "2019-02-11 08:38:05,429 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:38:05,873 epoch 24 - iter 0/27 - loss 0.01952597\n",
            "2019-02-11 08:38:06,627 epoch 24 - iter 2/27 - loss 0.02064057\n",
            "2019-02-11 08:38:07,702 epoch 24 - iter 4/27 - loss 0.02060977\n",
            "2019-02-11 08:38:08,515 epoch 24 - iter 6/27 - loss 0.01914031\n",
            "2019-02-11 08:38:09,302 epoch 24 - iter 8/27 - loss 0.01925794\n",
            "2019-02-11 08:38:10,048 epoch 24 - iter 10/27 - loss 0.02015549\n",
            "2019-02-11 08:38:10,751 epoch 24 - iter 12/27 - loss 0.01979722\n",
            "2019-02-11 08:38:11,745 epoch 24 - iter 14/27 - loss 0.01942802\n",
            "2019-02-11 08:38:12,397 epoch 24 - iter 16/27 - loss 0.01852471\n",
            "2019-02-11 08:38:13,060 epoch 24 - iter 18/27 - loss 0.01864901\n",
            "2019-02-11 08:38:13,983 epoch 24 - iter 20/27 - loss 0.01884308\n",
            "2019-02-11 08:38:15,194 epoch 24 - iter 22/27 - loss 0.01935587\n",
            "2019-02-11 08:38:15,823 epoch 24 - iter 24/27 - loss 0.01912610\n",
            "2019-02-11 08:38:16,210 epoch 24 - iter 26/27 - loss 0.02012138\n",
            "2019-02-11 08:38:16,221 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:38:16,222 EPOCH 24 done: loss 0.0201 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:38:16,824 DEV  : loss 0.03225997 - f-score 0.5789 - acc 0.5789\n",
            "2019-02-11 08:38:20,965 TEST : loss 0.01790547 - f-score 0.7543 - acc 0.7543\n",
            "2019-02-11 08:38:20,968 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:38:21,568 epoch 25 - iter 0/27 - loss 0.02636263\n",
            "2019-02-11 08:38:22,127 epoch 25 - iter 2/27 - loss 0.02047478\n",
            "2019-02-11 08:38:22,794 epoch 25 - iter 4/27 - loss 0.02085354\n",
            "2019-02-11 08:38:23,485 epoch 25 - iter 6/27 - loss 0.02015376\n",
            "2019-02-11 08:38:24,174 epoch 25 - iter 8/27 - loss 0.02108920\n",
            "2019-02-11 08:38:24,845 epoch 25 - iter 10/27 - loss 0.01965240\n",
            "2019-02-11 08:38:25,459 epoch 25 - iter 12/27 - loss 0.01884936\n",
            "2019-02-11 08:38:26,611 epoch 25 - iter 14/27 - loss 0.01871273\n",
            "2019-02-11 08:38:27,671 epoch 25 - iter 16/27 - loss 0.01921794\n",
            "2019-02-11 08:38:28,421 epoch 25 - iter 18/27 - loss 0.01920802\n",
            "2019-02-11 08:38:29,297 epoch 25 - iter 20/27 - loss 0.01869306\n",
            "2019-02-11 08:38:30,022 epoch 25 - iter 22/27 - loss 0.01851624\n",
            "2019-02-11 08:38:30,639 epoch 25 - iter 24/27 - loss 0.01873916\n",
            "2019-02-11 08:38:31,208 epoch 25 - iter 26/27 - loss 0.01850679\n",
            "2019-02-11 08:38:31,219 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:38:31,220 EPOCH 25 done: loss 0.0185 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 08:38:31,834 DEV  : loss 0.02684499 - f-score 0.6316 - acc 0.6316\n",
            "2019-02-11 08:38:35,932 TEST : loss 0.01419734 - f-score 0.7991 - acc 0.7991\n",
            "2019-02-11 08:38:35,934 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:38:36,325 epoch 26 - iter 0/27 - loss 0.00634151\n",
            "2019-02-11 08:38:37,070 epoch 26 - iter 2/27 - loss 0.01536357\n",
            "2019-02-11 08:38:38,060 epoch 26 - iter 4/27 - loss 0.01551134\n",
            "2019-02-11 08:38:38,723 epoch 26 - iter 6/27 - loss 0.01707896\n",
            "2019-02-11 08:38:39,411 epoch 26 - iter 8/27 - loss 0.01677173\n",
            "2019-02-11 08:38:40,007 epoch 26 - iter 10/27 - loss 0.01676307\n",
            "2019-02-11 08:38:40,661 epoch 26 - iter 12/27 - loss 0.01611457\n",
            "2019-02-11 08:38:41,751 epoch 26 - iter 14/27 - loss 0.01667878\n",
            "2019-02-11 08:38:42,333 epoch 26 - iter 16/27 - loss 0.01664089\n",
            "2019-02-11 08:38:43,118 epoch 26 - iter 18/27 - loss 0.01803613\n",
            "2019-02-11 08:38:44,321 epoch 26 - iter 20/27 - loss 0.01863312\n",
            "2019-02-11 08:38:45,133 epoch 26 - iter 22/27 - loss 0.01818976\n",
            "2019-02-11 08:38:45,853 epoch 26 - iter 24/27 - loss 0.01820842\n",
            "2019-02-11 08:38:46,264 epoch 26 - iter 26/27 - loss 0.01878596\n",
            "2019-02-11 08:38:46,275 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:38:46,276 EPOCH 26 done: loss 0.0188 - lr 0.1000 - bad epochs 2\n",
            "2019-02-11 08:38:46,899 DEV  : loss 0.03110595 - f-score 0.5895 - acc 0.5895\n",
            "2019-02-11 08:38:51,015 TEST : loss 0.01684804 - f-score 0.7415 - acc 0.7415\n",
            "2019-02-11 08:38:51,018 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:38:51,243 epoch 27 - iter 0/27 - loss 0.01439744\n",
            "2019-02-11 08:38:51,833 epoch 27 - iter 2/27 - loss 0.01549181\n",
            "2019-02-11 08:38:52,966 epoch 27 - iter 4/27 - loss 0.01444939\n",
            "2019-02-11 08:38:53,634 epoch 27 - iter 6/27 - loss 0.01440898\n",
            "2019-02-11 08:38:54,264 epoch 27 - iter 8/27 - loss 0.01497076\n",
            "2019-02-11 08:38:55,227 epoch 27 - iter 10/27 - loss 0.01547049\n",
            "2019-02-11 08:38:55,930 epoch 27 - iter 12/27 - loss 0.01627861\n",
            "2019-02-11 08:38:56,700 epoch 27 - iter 14/27 - loss 0.01579839\n",
            "2019-02-11 08:38:57,340 epoch 27 - iter 16/27 - loss 0.01641970\n",
            "2019-02-11 08:38:58,602 epoch 27 - iter 18/27 - loss 0.01654923\n",
            "2019-02-11 08:38:59,353 epoch 27 - iter 20/27 - loss 0.01679604\n",
            "2019-02-11 08:39:00,066 epoch 27 - iter 22/27 - loss 0.01678829\n",
            "2019-02-11 08:39:00,789 epoch 27 - iter 24/27 - loss 0.01708430\n",
            "2019-02-11 08:39:01,269 epoch 27 - iter 26/27 - loss 0.01744979\n",
            "2019-02-11 08:39:01,281 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:39:01,282 EPOCH 27 done: loss 0.0174 - lr 0.1000 - bad epochs 3\n",
            "2019-02-11 08:39:01,894 DEV  : loss 0.02955232 - f-score 0.6316 - acc 0.6316\n",
            "2019-02-11 08:39:07,207 TEST : loss 0.01634915 - f-score 0.7735 - acc 0.7735\n",
            "2019-02-11 08:39:24,778 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:39:25,309 epoch 28 - iter 0/27 - loss 0.01483723\n",
            "2019-02-11 08:39:26,039 epoch 28 - iter 2/27 - loss 0.01993426\n",
            "2019-02-11 08:39:26,866 epoch 28 - iter 4/27 - loss 0.01878825\n",
            "2019-02-11 08:39:28,132 epoch 28 - iter 6/27 - loss 0.01845445\n",
            "2019-02-11 08:39:28,697 epoch 28 - iter 8/27 - loss 0.01942559\n",
            "2019-02-11 08:39:29,668 epoch 28 - iter 10/27 - loss 0.02025127\n",
            "2019-02-11 08:39:30,370 epoch 28 - iter 12/27 - loss 0.02023503\n",
            "2019-02-11 08:39:31,196 epoch 28 - iter 14/27 - loss 0.01985691\n",
            "2019-02-11 08:39:31,847 epoch 28 - iter 16/27 - loss 0.01953358\n",
            "2019-02-11 08:39:32,577 epoch 28 - iter 18/27 - loss 0.01870023\n",
            "2019-02-11 08:39:33,344 epoch 28 - iter 20/27 - loss 0.01881659\n",
            "2019-02-11 08:39:34,483 epoch 28 - iter 22/27 - loss 0.01869238\n",
            "2019-02-11 08:39:35,144 epoch 28 - iter 24/27 - loss 0.01801321\n",
            "2019-02-11 08:39:35,600 epoch 28 - iter 26/27 - loss 0.01835907\n",
            "2019-02-11 08:39:35,611 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:39:35,613 EPOCH 28 done: loss 0.0184 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:39:36,232 DEV  : loss 0.02792804 - f-score 0.6421 - acc 0.6421\n",
            "2019-02-11 08:39:40,453 TEST : loss 0.01632869 - f-score 0.7863 - acc 0.7863\n",
            "2019-02-11 08:39:40,455 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:39:40,842 epoch 29 - iter 0/27 - loss 0.02032512\n",
            "2019-02-11 08:39:41,724 epoch 29 - iter 2/27 - loss 0.01843003\n",
            "2019-02-11 08:39:42,680 epoch 29 - iter 4/27 - loss 0.01755381\n",
            "2019-02-11 08:39:43,288 epoch 29 - iter 6/27 - loss 0.01816305\n",
            "2019-02-11 08:39:43,989 epoch 29 - iter 8/27 - loss 0.01750584\n",
            "2019-02-11 08:39:45,114 epoch 29 - iter 10/27 - loss 0.01716941\n",
            "2019-02-11 08:39:45,814 epoch 29 - iter 12/27 - loss 0.01759697\n",
            "2019-02-11 08:39:46,438 epoch 29 - iter 14/27 - loss 0.01706791\n",
            "2019-02-11 08:39:47,170 epoch 29 - iter 16/27 - loss 0.01648718\n",
            "2019-02-11 08:39:47,858 epoch 29 - iter 18/27 - loss 0.01651560\n",
            "2019-02-11 08:39:48,649 epoch 29 - iter 20/27 - loss 0.01652710\n",
            "2019-02-11 08:39:49,413 epoch 29 - iter 22/27 - loss 0.01640913\n",
            "2019-02-11 08:39:50,548 epoch 29 - iter 24/27 - loss 0.01618437\n",
            "2019-02-11 08:39:51,008 epoch 29 - iter 26/27 - loss 0.01650662\n",
            "2019-02-11 08:39:51,021 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:39:51,023 EPOCH 29 done: loss 0.0165 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 08:39:51,634 DEV  : loss 0.02597061 - f-score 0.6842 - acc 0.6842\n",
            "2019-02-11 08:39:55,961 TEST : loss 0.01342553 - f-score 0.8269 - acc 0.8269\n",
            "2019-02-11 08:40:13,061 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:40:13,595 epoch 30 - iter 0/27 - loss 0.02054450\n",
            "2019-02-11 08:40:14,718 epoch 30 - iter 2/27 - loss 0.01797651\n",
            "2019-02-11 08:40:15,412 epoch 30 - iter 4/27 - loss 0.01854666\n",
            "2019-02-11 08:40:16,140 epoch 30 - iter 6/27 - loss 0.01933793\n",
            "2019-02-11 08:40:16,803 epoch 30 - iter 8/27 - loss 0.01931521\n",
            "2019-02-11 08:40:17,672 epoch 30 - iter 10/27 - loss 0.01850229\n",
            "2019-02-11 08:40:18,478 epoch 30 - iter 12/27 - loss 0.01822300\n",
            "2019-02-11 08:40:19,188 epoch 30 - iter 14/27 - loss 0.01801475\n",
            "2019-02-11 08:40:19,931 epoch 30 - iter 16/27 - loss 0.01726019\n",
            "2019-02-11 08:40:20,639 epoch 30 - iter 18/27 - loss 0.01686958\n",
            "2019-02-11 08:40:21,429 epoch 30 - iter 20/27 - loss 0.01738990\n",
            "2019-02-11 08:40:22,558 epoch 30 - iter 22/27 - loss 0.01737164\n",
            "2019-02-11 08:40:23,177 epoch 30 - iter 24/27 - loss 0.01725448\n",
            "2019-02-11 08:40:23,818 epoch 30 - iter 26/27 - loss 0.01782283\n",
            "2019-02-11 08:40:23,830 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:40:23,831 EPOCH 30 done: loss 0.0178 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:40:24,457 DEV  : loss 0.02730430 - f-score 0.6526 - acc 0.6526\n",
            "2019-02-11 08:40:28,686 TEST : loss 0.01289316 - f-score 0.8312 - acc 0.8312\n",
            "2019-02-11 08:40:28,689 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:40:29,089 epoch 31 - iter 0/27 - loss 0.01418367\n",
            "2019-02-11 08:40:29,935 epoch 31 - iter 2/27 - loss 0.01344896\n",
            "2019-02-11 08:40:30,475 epoch 31 - iter 4/27 - loss 0.01637304\n",
            "2019-02-11 08:40:31,260 epoch 31 - iter 6/27 - loss 0.01522564\n",
            "2019-02-11 08:40:32,366 epoch 31 - iter 8/27 - loss 0.01454194\n",
            "2019-02-11 08:40:33,176 epoch 31 - iter 10/27 - loss 0.01418922\n",
            "2019-02-11 08:40:33,733 epoch 31 - iter 12/27 - loss 0.01524672\n",
            "2019-02-11 08:40:34,426 epoch 31 - iter 14/27 - loss 0.01613383\n",
            "2019-02-11 08:40:35,672 epoch 31 - iter 16/27 - loss 0.01630838\n",
            "2019-02-11 08:40:36,209 epoch 31 - iter 18/27 - loss 0.01671655\n",
            "2019-02-11 08:40:36,875 epoch 31 - iter 20/27 - loss 0.01693337\n",
            "2019-02-11 08:40:37,585 epoch 31 - iter 22/27 - loss 0.01670319\n",
            "2019-02-11 08:40:38,401 epoch 31 - iter 24/27 - loss 0.01663189\n",
            "2019-02-11 08:40:38,847 epoch 31 - iter 26/27 - loss 0.01679970\n",
            "2019-02-11 08:40:38,859 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:40:38,860 EPOCH 31 done: loss 0.0168 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 08:40:39,481 DEV  : loss 0.02952342 - f-score 0.6105 - acc 0.6105\n",
            "2019-02-11 08:40:43,698 TEST : loss 0.01352234 - f-score 0.8226 - acc 0.8226\n",
            "2019-02-11 08:40:43,701 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:40:44,024 epoch 32 - iter 0/27 - loss 0.02405609\n",
            "2019-02-11 08:40:44,630 epoch 32 - iter 2/27 - loss 0.01711379\n",
            "2019-02-11 08:40:45,398 epoch 32 - iter 4/27 - loss 0.01655019\n",
            "2019-02-11 08:40:46,393 epoch 32 - iter 6/27 - loss 0.01616035\n",
            "2019-02-11 08:40:47,105 epoch 32 - iter 8/27 - loss 0.01779243\n",
            "2019-02-11 08:40:47,849 epoch 32 - iter 10/27 - loss 0.01762548\n",
            "2019-02-11 08:40:48,495 epoch 32 - iter 12/27 - loss 0.01741336\n",
            "2019-02-11 08:40:49,591 epoch 32 - iter 14/27 - loss 0.01742124\n",
            "2019-02-11 08:40:50,181 epoch 32 - iter 16/27 - loss 0.01829785\n",
            "2019-02-11 08:40:50,811 epoch 32 - iter 18/27 - loss 0.01838749\n",
            "2019-02-11 08:40:51,551 epoch 32 - iter 20/27 - loss 0.01826860\n",
            "2019-02-11 08:40:52,751 epoch 32 - iter 22/27 - loss 0.01834107\n",
            "2019-02-11 08:40:53,727 epoch 32 - iter 24/27 - loss 0.01780888\n",
            "2019-02-11 08:40:54,205 epoch 32 - iter 26/27 - loss 0.01847181\n",
            "2019-02-11 08:40:54,216 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:40:54,217 EPOCH 32 done: loss 0.0185 - lr 0.1000 - bad epochs 2\n",
            "2019-02-11 08:40:54,834 DEV  : loss 0.03268221 - f-score 0.6526 - acc 0.6526\n",
            "2019-02-11 08:40:59,154 TEST : loss 0.01470269 - f-score 0.8205 - acc 0.8205\n",
            "2019-02-11 08:40:59,157 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:40:59,536 epoch 33 - iter 0/27 - loss 0.01864474\n",
            "2019-02-11 08:41:00,301 epoch 33 - iter 2/27 - loss 0.01697562\n",
            "2019-02-11 08:41:01,060 epoch 33 - iter 4/27 - loss 0.01644827\n",
            "2019-02-11 08:41:02,300 epoch 33 - iter 6/27 - loss 0.01749557\n",
            "2019-02-11 08:41:03,262 epoch 33 - iter 8/27 - loss 0.01708237\n",
            "2019-02-11 08:41:04,063 epoch 33 - iter 10/27 - loss 0.01620170\n",
            "2019-02-11 08:41:04,829 epoch 33 - iter 12/27 - loss 0.01616814\n",
            "2019-02-11 08:41:05,625 epoch 33 - iter 14/27 - loss 0.01525625\n",
            "2019-02-11 08:41:06,345 epoch 33 - iter 16/27 - loss 0.01510504\n",
            "2019-02-11 08:41:06,942 epoch 33 - iter 18/27 - loss 0.01506896\n",
            "2019-02-11 08:41:08,009 epoch 33 - iter 20/27 - loss 0.01504630\n",
            "2019-02-11 08:41:08,702 epoch 33 - iter 22/27 - loss 0.01481521\n",
            "2019-02-11 08:41:09,382 epoch 33 - iter 24/27 - loss 0.01489693\n",
            "2019-02-11 08:41:09,803 epoch 33 - iter 26/27 - loss 0.01579599\n",
            "2019-02-11 08:41:09,814 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:41:09,816 EPOCH 33 done: loss 0.0158 - lr 0.1000 - bad epochs 3\n",
            "2019-02-11 08:41:10,440 DEV  : loss 0.02953634 - f-score 0.6105 - acc 0.6105\n",
            "2019-02-11 08:41:14,669 TEST : loss 0.01297598 - f-score 0.8397 - acc 0.8397\n",
            "2019-02-11 08:41:31,647 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:41:32,097 epoch 34 - iter 0/27 - loss 0.01817691\n",
            "2019-02-11 08:41:32,948 epoch 34 - iter 2/27 - loss 0.01485226\n",
            "2019-02-11 08:41:33,653 epoch 34 - iter 4/27 - loss 0.01755368\n",
            "2019-02-11 08:41:34,584 epoch 34 - iter 6/27 - loss 0.01582557\n",
            "2019-02-11 08:41:35,302 epoch 34 - iter 8/27 - loss 0.01578662\n",
            "2019-02-11 08:41:35,989 epoch 34 - iter 10/27 - loss 0.01612036\n",
            "2019-02-11 08:41:36,690 epoch 34 - iter 12/27 - loss 0.01628080\n",
            "2019-02-11 08:41:37,364 epoch 34 - iter 14/27 - loss 0.01644147\n",
            "2019-02-11 08:41:38,283 epoch 34 - iter 16/27 - loss 0.01714405\n",
            "2019-02-11 08:41:39,532 epoch 34 - iter 18/27 - loss 0.01705512\n",
            "2019-02-11 08:41:40,243 epoch 34 - iter 20/27 - loss 0.01682893\n",
            "2019-02-11 08:41:41,307 epoch 34 - iter 22/27 - loss 0.01671233\n",
            "2019-02-11 08:41:41,876 epoch 34 - iter 24/27 - loss 0.01669055\n",
            "2019-02-11 08:41:42,363 epoch 34 - iter 26/27 - loss 0.01733553\n",
            "2019-02-11 08:41:42,374 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:41:42,376 EPOCH 34 done: loss 0.0173 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:41:43,006 DEV  : loss 0.03175106 - f-score 0.6316 - acc 0.6316\n",
            "2019-02-11 08:41:47,369 TEST : loss 0.01307287 - f-score 0.8462 - acc 0.8462\n",
            "2019-02-11 08:41:47,372 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:41:47,745 epoch 35 - iter 0/27 - loss 0.01383549\n",
            "2019-02-11 08:41:48,508 epoch 35 - iter 2/27 - loss 0.01597268\n",
            "2019-02-11 08:41:49,433 epoch 35 - iter 4/27 - loss 0.01737890\n",
            "2019-02-11 08:41:50,065 epoch 35 - iter 6/27 - loss 0.01652548\n",
            "2019-02-11 08:41:50,997 epoch 35 - iter 8/27 - loss 0.01485549\n",
            "2019-02-11 08:41:51,632 epoch 35 - iter 10/27 - loss 0.01419149\n",
            "2019-02-11 08:41:52,464 epoch 35 - iter 12/27 - loss 0.01379385\n",
            "2019-02-11 08:41:53,072 epoch 35 - iter 14/27 - loss 0.01350136\n",
            "2019-02-11 08:41:54,230 epoch 35 - iter 16/27 - loss 0.01417122\n",
            "2019-02-11 08:41:55,236 epoch 35 - iter 18/27 - loss 0.01389900\n",
            "2019-02-11 08:41:55,986 epoch 35 - iter 20/27 - loss 0.01390218\n",
            "2019-02-11 08:41:56,672 epoch 35 - iter 22/27 - loss 0.01423104\n",
            "2019-02-11 08:41:57,347 epoch 35 - iter 24/27 - loss 0.01456400\n",
            "2019-02-11 08:41:57,839 epoch 35 - iter 26/27 - loss 0.01552548\n",
            "2019-02-11 08:41:57,851 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:41:57,852 EPOCH 35 done: loss 0.0155 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 08:41:58,479 DEV  : loss 0.03033316 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 08:42:02,688 TEST : loss 0.01285495 - f-score 0.8419 - acc 0.8419\n",
            "2019-02-11 08:42:19,763 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:42:20,293 epoch 36 - iter 0/27 - loss 0.01409601\n",
            "2019-02-11 08:42:20,890 epoch 36 - iter 2/27 - loss 0.01279273\n",
            "2019-02-11 08:42:21,616 epoch 36 - iter 4/27 - loss 0.01242012\n",
            "2019-02-11 08:42:22,985 epoch 36 - iter 6/27 - loss 0.01422239\n",
            "2019-02-11 08:42:23,720 epoch 36 - iter 8/27 - loss 0.01398134\n",
            "2019-02-11 08:42:24,428 epoch 36 - iter 10/27 - loss 0.01312555\n",
            "2019-02-11 08:42:25,162 epoch 36 - iter 12/27 - loss 0.01219946\n",
            "2019-02-11 08:42:25,907 epoch 36 - iter 14/27 - loss 0.01333432\n",
            "2019-02-11 08:42:26,586 epoch 36 - iter 16/27 - loss 0.01339244\n",
            "2019-02-11 08:42:27,169 epoch 36 - iter 18/27 - loss 0.01395490\n",
            "2019-02-11 08:42:27,898 epoch 36 - iter 20/27 - loss 0.01404556\n",
            "2019-02-11 08:42:28,719 epoch 36 - iter 22/27 - loss 0.01422351\n",
            "2019-02-11 08:42:29,459 epoch 36 - iter 24/27 - loss 0.01425080\n",
            "2019-02-11 08:42:30,375 epoch 36 - iter 26/27 - loss 0.01429501\n",
            "2019-02-11 08:42:30,386 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:42:30,387 EPOCH 36 done: loss 0.0143 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:42:31,009 DEV  : loss 0.03245191 - f-score 0.6105 - acc 0.6105\n",
            "2019-02-11 08:42:35,371 TEST : loss 0.01217363 - f-score 0.8504 - acc 0.8504\n",
            "2019-02-11 08:42:52,496 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:42:52,952 epoch 37 - iter 0/27 - loss 0.00889235\n",
            "2019-02-11 08:42:53,810 epoch 37 - iter 2/27 - loss 0.01356457\n",
            "2019-02-11 08:42:54,583 epoch 37 - iter 4/27 - loss 0.01350213\n",
            "2019-02-11 08:42:55,574 epoch 37 - iter 6/27 - loss 0.01360017\n",
            "2019-02-11 08:42:56,722 epoch 37 - iter 8/27 - loss 0.01425627\n",
            "2019-02-11 08:42:57,469 epoch 37 - iter 10/27 - loss 0.01285735\n",
            "2019-02-11 08:42:58,143 epoch 37 - iter 12/27 - loss 0.01254897\n",
            "2019-02-11 08:42:58,932 epoch 37 - iter 14/27 - loss 0.01270010\n",
            "2019-02-11 08:43:00,155 epoch 37 - iter 16/27 - loss 0.01360850\n",
            "2019-02-11 08:43:00,919 epoch 37 - iter 18/27 - loss 0.01341382\n",
            "2019-02-11 08:43:01,666 epoch 37 - iter 20/27 - loss 0.01381061\n",
            "2019-02-11 08:43:02,329 epoch 37 - iter 22/27 - loss 0.01375974\n",
            "2019-02-11 08:43:03,132 epoch 37 - iter 24/27 - loss 0.01381419\n",
            "2019-02-11 08:43:03,508 epoch 37 - iter 26/27 - loss 0.01371822\n",
            "2019-02-11 08:43:03,520 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:43:03,521 EPOCH 37 done: loss 0.0137 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:43:04,143 DEV  : loss 0.03397957 - f-score 0.6211 - acc 0.6211\n",
            "2019-02-11 08:43:08,383 TEST : loss 0.01259359 - f-score 0.8419 - acc 0.8419\n",
            "2019-02-11 08:43:26,122 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:43:26,549 epoch 38 - iter 0/27 - loss 0.01428149\n",
            "2019-02-11 08:43:27,396 epoch 38 - iter 2/27 - loss 0.01819350\n",
            "2019-02-11 08:43:28,200 epoch 38 - iter 4/27 - loss 0.01442196\n",
            "2019-02-11 08:43:28,889 epoch 38 - iter 6/27 - loss 0.01482709\n",
            "2019-02-11 08:43:29,531 epoch 38 - iter 8/27 - loss 0.01552913\n",
            "2019-02-11 08:43:30,469 epoch 38 - iter 10/27 - loss 0.01546063\n",
            "2019-02-11 08:43:31,194 epoch 38 - iter 12/27 - loss 0.01457444\n",
            "2019-02-11 08:43:31,856 epoch 38 - iter 14/27 - loss 0.01365977\n",
            "2019-02-11 08:43:32,528 epoch 38 - iter 16/27 - loss 0.01414200\n",
            "2019-02-11 08:43:33,094 epoch 38 - iter 18/27 - loss 0.01395453\n",
            "2019-02-11 08:43:33,833 epoch 38 - iter 20/27 - loss 0.01429471\n",
            "2019-02-11 08:43:34,902 epoch 38 - iter 22/27 - loss 0.01391742\n",
            "2019-02-11 08:43:36,347 epoch 38 - iter 24/27 - loss 0.01408701\n",
            "2019-02-11 08:43:36,983 epoch 38 - iter 26/27 - loss 0.01423901\n",
            "2019-02-11 08:43:36,995 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:43:36,996 EPOCH 38 done: loss 0.0142 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:43:37,637 DEV  : loss 0.03876059 - f-score 0.6105 - acc 0.6105\n",
            "2019-02-11 08:43:41,885 TEST : loss 0.01506322 - f-score 0.8312 - acc 0.8312\n",
            "2019-02-11 08:43:41,888 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:43:42,235 epoch 39 - iter 0/27 - loss 0.01312309\n",
            "2019-02-11 08:43:43,005 epoch 39 - iter 2/27 - loss 0.01081070\n",
            "2019-02-11 08:43:43,581 epoch 39 - iter 4/27 - loss 0.01091053\n",
            "2019-02-11 08:43:44,321 epoch 39 - iter 6/27 - loss 0.01215287\n",
            "2019-02-11 08:43:45,448 epoch 39 - iter 8/27 - loss 0.01149892\n",
            "2019-02-11 08:43:46,411 epoch 39 - iter 10/27 - loss 0.01138261\n",
            "2019-02-11 08:43:47,118 epoch 39 - iter 12/27 - loss 0.01165954\n",
            "2019-02-11 08:43:48,020 epoch 39 - iter 14/27 - loss 0.01167470\n",
            "2019-02-11 08:43:48,699 epoch 39 - iter 16/27 - loss 0.01150556\n",
            "2019-02-11 08:43:49,490 epoch 39 - iter 18/27 - loss 0.01155973\n",
            "2019-02-11 08:43:50,063 epoch 39 - iter 20/27 - loss 0.01133281\n",
            "2019-02-11 08:43:51,220 epoch 39 - iter 22/27 - loss 0.01163000\n",
            "2019-02-11 08:43:51,926 epoch 39 - iter 24/27 - loss 0.01205386\n",
            "2019-02-11 08:43:52,402 epoch 39 - iter 26/27 - loss 0.01253442\n",
            "2019-02-11 08:43:52,414 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:43:52,415 EPOCH 39 done: loss 0.0125 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 08:43:53,057 DEV  : loss 0.03442365 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 08:43:57,281 TEST : loss 0.01324078 - f-score 0.8547 - acc 0.8547\n",
            "2019-02-11 08:44:14,519 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:44:15,026 epoch 40 - iter 0/27 - loss 0.01149067\n",
            "2019-02-11 08:44:15,732 epoch 40 - iter 2/27 - loss 0.00805476\n",
            "2019-02-11 08:44:17,086 epoch 40 - iter 4/27 - loss 0.00941675\n",
            "2019-02-11 08:44:18,012 epoch 40 - iter 6/27 - loss 0.00913842\n",
            "2019-02-11 08:44:18,640 epoch 40 - iter 8/27 - loss 0.00989353\n",
            "2019-02-11 08:44:19,602 epoch 40 - iter 10/27 - loss 0.00991386\n",
            "2019-02-11 08:44:20,376 epoch 40 - iter 12/27 - loss 0.01032740\n",
            "2019-02-11 08:44:21,135 epoch 40 - iter 14/27 - loss 0.01089619\n",
            "2019-02-11 08:44:22,295 epoch 40 - iter 16/27 - loss 0.01132218\n",
            "2019-02-11 08:44:23,013 epoch 40 - iter 18/27 - loss 0.01185437\n",
            "2019-02-11 08:44:23,813 epoch 40 - iter 20/27 - loss 0.01167406\n",
            "2019-02-11 08:44:24,614 epoch 40 - iter 22/27 - loss 0.01194287\n",
            "2019-02-11 08:44:25,300 epoch 40 - iter 24/27 - loss 0.01203078\n",
            "2019-02-11 08:44:25,687 epoch 40 - iter 26/27 - loss 0.01211251\n",
            "2019-02-11 08:44:25,698 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:44:25,699 EPOCH 40 done: loss 0.0121 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:44:26,324 DEV  : loss 0.03965748 - f-score 0.6421 - acc 0.6421\n",
            "2019-02-11 08:44:30,513 TEST : loss 0.01214373 - f-score 0.8761 - acc 0.8761\n",
            "2019-02-11 08:44:47,539 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:44:48,379 epoch 41 - iter 0/27 - loss 0.01428907\n",
            "2019-02-11 08:44:49,119 epoch 41 - iter 2/27 - loss 0.01224801\n",
            "2019-02-11 08:44:49,761 epoch 41 - iter 4/27 - loss 0.01063992\n",
            "2019-02-11 08:44:50,468 epoch 41 - iter 6/27 - loss 0.01015361\n",
            "2019-02-11 08:44:51,208 epoch 41 - iter 8/27 - loss 0.01006309\n",
            "2019-02-11 08:44:51,892 epoch 41 - iter 10/27 - loss 0.01051795\n",
            "2019-02-11 08:44:52,659 epoch 41 - iter 12/27 - loss 0.01010207\n",
            "2019-02-11 08:44:53,527 epoch 41 - iter 14/27 - loss 0.01004284\n",
            "2019-02-11 08:44:54,234 epoch 41 - iter 16/27 - loss 0.01010579\n",
            "2019-02-11 08:44:55,520 epoch 41 - iter 18/27 - loss 0.01066410\n",
            "2019-02-11 08:44:56,256 epoch 41 - iter 20/27 - loss 0.01033031\n",
            "2019-02-11 08:44:56,856 epoch 41 - iter 22/27 - loss 0.01048351\n",
            "2019-02-11 08:44:57,768 epoch 41 - iter 24/27 - loss 0.01077822\n",
            "2019-02-11 08:44:58,163 epoch 41 - iter 26/27 - loss 0.01123588\n",
            "2019-02-11 08:44:58,174 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:44:58,175 EPOCH 41 done: loss 0.0112 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:44:58,798 DEV  : loss 0.04118387 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 08:45:03,115 TEST : loss 0.01605155 - f-score 0.8248 - acc 0.8248\n",
            "2019-02-11 08:45:20,056 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:45:20,435 epoch 42 - iter 0/27 - loss 0.01452532\n",
            "2019-02-11 08:45:21,268 epoch 42 - iter 2/27 - loss 0.01048823\n",
            "2019-02-11 08:45:22,187 epoch 42 - iter 4/27 - loss 0.01025186\n",
            "2019-02-11 08:45:23,830 epoch 42 - iter 6/27 - loss 0.01012727\n",
            "2019-02-11 08:45:24,353 epoch 42 - iter 8/27 - loss 0.00959399\n",
            "2019-02-11 08:45:25,047 epoch 42 - iter 10/27 - loss 0.01067772\n",
            "2019-02-11 08:45:25,703 epoch 42 - iter 12/27 - loss 0.01150157\n",
            "2019-02-11 08:45:26,411 epoch 42 - iter 14/27 - loss 0.01089293\n",
            "2019-02-11 08:45:27,179 epoch 42 - iter 16/27 - loss 0.01079286\n",
            "2019-02-11 08:45:27,814 epoch 42 - iter 18/27 - loss 0.01114921\n",
            "2019-02-11 08:45:28,700 epoch 42 - iter 20/27 - loss 0.01084125\n",
            "2019-02-11 08:45:29,479 epoch 42 - iter 22/27 - loss 0.01131979\n",
            "2019-02-11 08:45:30,211 epoch 42 - iter 24/27 - loss 0.01132774\n",
            "2019-02-11 08:45:30,549 epoch 42 - iter 26/27 - loss 0.01292898\n",
            "2019-02-11 08:45:30,560 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:45:30,561 EPOCH 42 done: loss 0.0129 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:45:31,182 DEV  : loss 0.03857610 - f-score 0.6842 - acc 0.6842\n",
            "2019-02-11 08:45:35,440 TEST : loss 0.01151184 - f-score 0.8974 - acc 0.8974\n",
            "2019-02-11 08:45:35,443 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:45:35,747 epoch 43 - iter 0/27 - loss 0.00856316\n",
            "2019-02-11 08:45:36,470 epoch 43 - iter 2/27 - loss 0.00967531\n",
            "2019-02-11 08:45:37,147 epoch 43 - iter 4/27 - loss 0.01001146\n",
            "2019-02-11 08:45:37,913 epoch 43 - iter 6/27 - loss 0.00949011\n",
            "2019-02-11 08:45:38,853 epoch 43 - iter 8/27 - loss 0.01123326\n",
            "2019-02-11 08:45:39,735 epoch 43 - iter 10/27 - loss 0.01212422\n",
            "2019-02-11 08:45:40,824 epoch 43 - iter 12/27 - loss 0.01177136\n",
            "2019-02-11 08:45:41,441 epoch 43 - iter 14/27 - loss 0.01132392\n",
            "2019-02-11 08:45:42,362 epoch 43 - iter 16/27 - loss 0.01097795\n",
            "2019-02-11 08:45:43,501 epoch 43 - iter 18/27 - loss 0.01114259\n",
            "2019-02-11 08:45:44,221 epoch 43 - iter 20/27 - loss 0.01101344\n",
            "2019-02-11 08:45:44,929 epoch 43 - iter 22/27 - loss 0.01049241\n",
            "2019-02-11 08:45:45,634 epoch 43 - iter 24/27 - loss 0.01084872\n",
            "2019-02-11 08:45:46,074 epoch 43 - iter 26/27 - loss 0.01117891\n",
            "2019-02-11 08:45:46,087 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:45:46,088 EPOCH 43 done: loss 0.0112 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 08:45:46,701 DEV  : loss 0.03747991 - f-score 0.6842 - acc 0.6842\n",
            "2019-02-11 08:45:51,058 TEST : loss 0.01067526 - f-score 0.8996 - acc 0.8996\n",
            "2019-02-11 08:46:08,044 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:46:08,594 epoch 44 - iter 0/27 - loss 0.01078594\n",
            "2019-02-11 08:46:09,466 epoch 44 - iter 2/27 - loss 0.01069824\n",
            "2019-02-11 08:46:10,715 epoch 44 - iter 4/27 - loss 0.01155571\n",
            "2019-02-11 08:46:11,430 epoch 44 - iter 6/27 - loss 0.01185631\n",
            "2019-02-11 08:46:12,294 epoch 44 - iter 8/27 - loss 0.01117536\n",
            "2019-02-11 08:46:12,979 epoch 44 - iter 10/27 - loss 0.01028560\n",
            "2019-02-11 08:46:13,648 epoch 44 - iter 12/27 - loss 0.01075572\n",
            "2019-02-11 08:46:14,395 epoch 44 - iter 14/27 - loss 0.01033621\n",
            "2019-02-11 08:46:15,004 epoch 44 - iter 16/27 - loss 0.01057948\n",
            "2019-02-11 08:46:15,660 epoch 44 - iter 18/27 - loss 0.01070681\n",
            "2019-02-11 08:46:16,542 epoch 44 - iter 20/27 - loss 0.01091116\n",
            "2019-02-11 08:46:17,514 epoch 44 - iter 22/27 - loss 0.01083771\n",
            "2019-02-11 08:46:18,115 epoch 44 - iter 24/27 - loss 0.01128952\n",
            "2019-02-11 08:46:18,929 epoch 44 - iter 26/27 - loss 0.01127410\n",
            "2019-02-11 08:46:18,939 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:46:18,940 EPOCH 44 done: loss 0.0113 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:46:19,553 DEV  : loss 0.04065067 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 08:46:23,709 TEST : loss 0.01114967 - f-score 0.9103 - acc 0.9103\n",
            "2019-02-11 08:46:23,712 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:46:24,448 epoch 45 - iter 0/27 - loss 0.01555083\n",
            "2019-02-11 08:46:25,104 epoch 45 - iter 2/27 - loss 0.01030761\n",
            "2019-02-11 08:46:26,321 epoch 45 - iter 4/27 - loss 0.01108437\n",
            "2019-02-11 08:46:27,181 epoch 45 - iter 6/27 - loss 0.01161584\n",
            "2019-02-11 08:46:27,782 epoch 45 - iter 8/27 - loss 0.01100436\n",
            "2019-02-11 08:46:28,573 epoch 45 - iter 10/27 - loss 0.01047350\n",
            "2019-02-11 08:46:29,299 epoch 45 - iter 12/27 - loss 0.01121996\n",
            "2019-02-11 08:46:29,975 epoch 45 - iter 14/27 - loss 0.01067804\n",
            "2019-02-11 08:46:30,658 epoch 45 - iter 16/27 - loss 0.01052310\n",
            "2019-02-11 08:46:31,169 epoch 45 - iter 18/27 - loss 0.01037194\n",
            "2019-02-11 08:46:32,085 epoch 45 - iter 20/27 - loss 0.01006422\n",
            "2019-02-11 08:46:32,970 epoch 45 - iter 22/27 - loss 0.00990857\n",
            "2019-02-11 08:46:33,668 epoch 45 - iter 24/27 - loss 0.00987327\n",
            "2019-02-11 08:46:34,114 epoch 45 - iter 26/27 - loss 0.01005851\n",
            "2019-02-11 08:46:34,125 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:46:34,126 EPOCH 45 done: loss 0.0101 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 08:46:34,745 DEV  : loss 0.04541897 - f-score 0.6421 - acc 0.6421\n",
            "2019-02-11 08:46:38,951 TEST : loss 0.01244524 - f-score 0.8953 - acc 0.8953\n",
            "2019-02-11 08:46:56,067 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:46:56,442 epoch 46 - iter 0/27 - loss 0.00516776\n",
            "2019-02-11 08:46:57,258 epoch 46 - iter 2/27 - loss 0.01245715\n",
            "2019-02-11 08:46:57,870 epoch 46 - iter 4/27 - loss 0.01049049\n",
            "2019-02-11 08:46:58,572 epoch 46 - iter 6/27 - loss 0.01013108\n",
            "2019-02-11 08:46:59,476 epoch 46 - iter 8/27 - loss 0.01007964\n",
            "2019-02-11 08:47:00,246 epoch 46 - iter 10/27 - loss 0.00975772\n",
            "2019-02-11 08:47:01,010 epoch 46 - iter 12/27 - loss 0.01010106\n",
            "2019-02-11 08:47:01,709 epoch 46 - iter 14/27 - loss 0.01010429\n",
            "2019-02-11 08:47:02,605 epoch 46 - iter 16/27 - loss 0.00998057\n",
            "2019-02-11 08:47:03,646 epoch 46 - iter 18/27 - loss 0.00978559\n",
            "2019-02-11 08:47:04,287 epoch 46 - iter 20/27 - loss 0.00996622\n",
            "2019-02-11 08:47:05,203 epoch 46 - iter 22/27 - loss 0.00979929\n",
            "2019-02-11 08:47:06,326 epoch 46 - iter 24/27 - loss 0.00985280\n",
            "2019-02-11 08:47:06,770 epoch 46 - iter 26/27 - loss 0.01017106\n",
            "2019-02-11 08:47:06,781 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:47:06,782 EPOCH 46 done: loss 0.0102 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:47:07,400 DEV  : loss 0.04750267 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 08:47:11,638 TEST : loss 0.01326465 - f-score 0.8974 - acc 0.8974\n",
            "2019-02-11 08:47:11,641 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:47:12,038 epoch 47 - iter 0/27 - loss 0.00615819\n",
            "2019-02-11 08:47:12,927 epoch 47 - iter 2/27 - loss 0.00733590\n",
            "2019-02-11 08:47:13,729 epoch 47 - iter 4/27 - loss 0.00960543\n",
            "2019-02-11 08:47:14,400 epoch 47 - iter 6/27 - loss 0.01028685\n",
            "2019-02-11 08:47:14,962 epoch 47 - iter 8/27 - loss 0.00975497\n",
            "2019-02-11 08:47:15,939 epoch 47 - iter 10/27 - loss 0.00904009\n",
            "2019-02-11 08:47:16,601 epoch 47 - iter 12/27 - loss 0.00891958\n",
            "2019-02-11 08:47:17,371 epoch 47 - iter 14/27 - loss 0.00952634\n",
            "2019-02-11 08:47:18,009 epoch 47 - iter 16/27 - loss 0.00964925\n",
            "2019-02-11 08:47:18,886 epoch 47 - iter 18/27 - loss 0.00970531\n",
            "2019-02-11 08:47:19,566 epoch 47 - iter 20/27 - loss 0.00971480\n",
            "2019-02-11 08:47:20,167 epoch 47 - iter 22/27 - loss 0.00945737\n",
            "2019-02-11 08:47:21,729 epoch 47 - iter 24/27 - loss 0.00994029\n",
            "2019-02-11 08:47:22,202 epoch 47 - iter 26/27 - loss 0.01024387\n",
            "2019-02-11 08:47:22,213 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:47:22,214 EPOCH 47 done: loss 0.0102 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 08:47:22,832 DEV  : loss 0.04514008 - f-score 0.6737 - acc 0.6737\n",
            "2019-02-11 08:47:27,047 TEST : loss 0.01224392 - f-score 0.8953 - acc 0.8953\n",
            "2019-02-11 08:47:27,050 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:47:27,865 epoch 48 - iter 0/27 - loss 0.01031348\n",
            "2019-02-11 08:47:28,642 epoch 48 - iter 2/27 - loss 0.00674206\n",
            "2019-02-11 08:47:29,432 epoch 48 - iter 4/27 - loss 0.00663087\n",
            "2019-02-11 08:47:30,202 epoch 48 - iter 6/27 - loss 0.00854626\n",
            "2019-02-11 08:47:31,038 epoch 48 - iter 8/27 - loss 0.00937862\n",
            "2019-02-11 08:47:31,757 epoch 48 - iter 10/27 - loss 0.00880357\n",
            "2019-02-11 08:47:32,479 epoch 48 - iter 12/27 - loss 0.00813157\n",
            "2019-02-11 08:47:33,205 epoch 48 - iter 14/27 - loss 0.00795050\n",
            "2019-02-11 08:47:33,713 epoch 48 - iter 16/27 - loss 0.00796154\n",
            "2019-02-11 08:47:34,343 epoch 48 - iter 18/27 - loss 0.00808345\n",
            "2019-02-11 08:47:35,050 epoch 48 - iter 20/27 - loss 0.00872718\n",
            "2019-02-11 08:47:35,517 epoch 48 - iter 22/27 - loss 0.00929796\n",
            "2019-02-11 08:47:36,734 epoch 48 - iter 24/27 - loss 0.00920781\n",
            "2019-02-11 08:47:37,444 epoch 48 - iter 26/27 - loss 0.00983010\n",
            "2019-02-11 08:47:37,455 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:47:37,456 EPOCH 48 done: loss 0.0098 - lr 0.1000 - bad epochs 2\n",
            "2019-02-11 08:47:38,078 DEV  : loss 0.04840868 - f-score 0.6526 - acc 0.6526\n",
            "2019-02-11 08:47:42,365 TEST : loss 0.01181997 - f-score 0.9081 - acc 0.9081\n",
            "2019-02-11 08:47:59,401 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:47:59,826 epoch 49 - iter 0/27 - loss 0.00265940\n",
            "2019-02-11 08:48:00,576 epoch 49 - iter 2/27 - loss 0.00851744\n",
            "2019-02-11 08:48:01,500 epoch 49 - iter 4/27 - loss 0.00694692\n",
            "2019-02-11 08:48:02,264 epoch 49 - iter 6/27 - loss 0.00640360\n",
            "2019-02-11 08:48:02,901 epoch 49 - iter 8/27 - loss 0.00703178\n",
            "2019-02-11 08:48:03,551 epoch 49 - iter 10/27 - loss 0.00767926\n",
            "2019-02-11 08:48:04,212 epoch 49 - iter 12/27 - loss 0.00776136\n",
            "2019-02-11 08:48:05,275 epoch 49 - iter 14/27 - loss 0.00712086\n",
            "2019-02-11 08:48:05,923 epoch 49 - iter 16/27 - loss 0.00689527\n",
            "2019-02-11 08:48:06,841 epoch 49 - iter 18/27 - loss 0.00686804\n",
            "2019-02-11 08:48:07,564 epoch 49 - iter 20/27 - loss 0.00710167\n",
            "2019-02-11 08:48:08,680 epoch 49 - iter 22/27 - loss 0.00746021\n",
            "2019-02-11 08:48:09,406 epoch 49 - iter 24/27 - loss 0.00777340\n",
            "2019-02-11 08:48:09,894 epoch 49 - iter 26/27 - loss 0.00918619\n",
            "2019-02-11 08:48:09,905 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:48:09,906 EPOCH 49 done: loss 0.0092 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:48:10,533 DEV  : loss 0.05139158 - f-score 0.6421 - acc 0.6421\n",
            "2019-02-11 08:48:15,054 TEST : loss 0.01225091 - f-score 0.9038 - acc 0.9038\n",
            "2019-02-11 08:48:32,669 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:48:33,187 epoch 50 - iter 0/27 - loss 0.00264030\n",
            "2019-02-11 08:48:34,276 epoch 50 - iter 2/27 - loss 0.00356856\n",
            "2019-02-11 08:48:35,046 epoch 50 - iter 4/27 - loss 0.00870115\n",
            "2019-02-11 08:48:35,780 epoch 50 - iter 6/27 - loss 0.00773640\n",
            "2019-02-11 08:48:36,572 epoch 50 - iter 8/27 - loss 0.00790860\n",
            "2019-02-11 08:48:37,104 epoch 50 - iter 10/27 - loss 0.00768458\n",
            "2019-02-11 08:48:37,824 epoch 50 - iter 12/27 - loss 0.00760111\n",
            "2019-02-11 08:48:38,512 epoch 50 - iter 14/27 - loss 0.00818750\n",
            "2019-02-11 08:48:39,161 epoch 50 - iter 16/27 - loss 0.00842185\n",
            "2019-02-11 08:48:40,140 epoch 50 - iter 18/27 - loss 0.00868913\n",
            "2019-02-11 08:48:40,725 epoch 50 - iter 20/27 - loss 0.00852247\n",
            "2019-02-11 08:48:41,287 epoch 50 - iter 22/27 - loss 0.00836756\n",
            "2019-02-11 08:48:42,849 epoch 50 - iter 24/27 - loss 0.00886658\n",
            "2019-02-11 08:48:43,305 epoch 50 - iter 26/27 - loss 0.00920763\n",
            "2019-02-11 08:48:43,316 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:48:43,317 EPOCH 50 done: loss 0.0092 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:48:43,936 DEV  : loss 0.05025981 - f-score 0.6105 - acc 0.6105\n",
            "2019-02-11 08:48:48,132 TEST : loss 0.01248743 - f-score 0.8953 - acc 0.8953\n",
            "2019-02-11 08:48:48,134 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:48:48,453 epoch 51 - iter 0/27 - loss 0.00770123\n",
            "2019-02-11 08:48:49,125 epoch 51 - iter 2/27 - loss 0.00443875\n",
            "2019-02-11 08:48:49,942 epoch 51 - iter 4/27 - loss 0.00798738\n",
            "2019-02-11 08:48:50,688 epoch 51 - iter 6/27 - loss 0.00903333\n",
            "2019-02-11 08:48:51,297 epoch 51 - iter 8/27 - loss 0.00876331\n",
            "2019-02-11 08:48:52,148 epoch 51 - iter 10/27 - loss 0.00849017\n",
            "2019-02-11 08:48:53,411 epoch 51 - iter 12/27 - loss 0.00802371\n",
            "2019-02-11 08:48:54,075 epoch 51 - iter 14/27 - loss 0.00817084\n",
            "2019-02-11 08:48:54,800 epoch 51 - iter 16/27 - loss 0.00785784\n",
            "2019-02-11 08:48:55,725 epoch 51 - iter 18/27 - loss 0.00888859\n",
            "2019-02-11 08:48:56,386 epoch 51 - iter 20/27 - loss 0.00922683\n",
            "2019-02-11 08:48:57,144 epoch 51 - iter 22/27 - loss 0.00882086\n",
            "2019-02-11 08:48:57,914 epoch 51 - iter 24/27 - loss 0.00891496\n",
            "2019-02-11 08:48:58,301 epoch 51 - iter 26/27 - loss 0.00885003\n",
            "2019-02-11 08:48:58,312 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:48:58,313 EPOCH 51 done: loss 0.0089 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 08:48:58,934 DEV  : loss 0.04670784 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 08:49:03,266 TEST : loss 0.01107916 - f-score 0.9103 - acc 0.9103\n",
            "2019-02-11 08:49:20,351 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:49:21,166 epoch 52 - iter 0/27 - loss 0.00632595\n",
            "2019-02-11 08:49:21,963 epoch 52 - iter 2/27 - loss 0.00840907\n",
            "2019-02-11 08:49:22,756 epoch 52 - iter 4/27 - loss 0.00953549\n",
            "2019-02-11 08:49:23,410 epoch 52 - iter 6/27 - loss 0.00956967\n",
            "2019-02-11 08:49:24,169 epoch 52 - iter 8/27 - loss 0.00815492\n",
            "2019-02-11 08:49:25,325 epoch 52 - iter 10/27 - loss 0.00839532\n",
            "2019-02-11 08:49:26,102 epoch 52 - iter 12/27 - loss 0.00787967\n",
            "2019-02-11 08:49:27,214 epoch 52 - iter 14/27 - loss 0.00789703\n",
            "2019-02-11 08:49:27,902 epoch 52 - iter 16/27 - loss 0.00811093\n",
            "2019-02-11 08:49:28,669 epoch 52 - iter 18/27 - loss 0.00835465\n",
            "2019-02-11 08:49:29,369 epoch 52 - iter 20/27 - loss 0.00872925\n",
            "2019-02-11 08:49:30,025 epoch 52 - iter 22/27 - loss 0.00894702\n",
            "2019-02-11 08:49:30,797 epoch 52 - iter 24/27 - loss 0.00882433\n",
            "2019-02-11 08:49:31,734 epoch 52 - iter 26/27 - loss 0.00877875\n",
            "2019-02-11 08:49:31,746 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:49:31,749 EPOCH 52 done: loss 0.0088 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:49:32,437 DEV  : loss 0.05919128 - f-score 0.6105 - acc 0.6105\n",
            "2019-02-11 08:49:36,667 TEST : loss 0.01519723 - f-score 0.8846 - acc 0.8846\n",
            "2019-02-11 08:49:53,784 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:49:54,301 epoch 53 - iter 0/27 - loss 0.00689258\n",
            "2019-02-11 08:49:55,112 epoch 53 - iter 2/27 - loss 0.00474738\n",
            "2019-02-11 08:49:55,700 epoch 53 - iter 4/27 - loss 0.00518384\n",
            "2019-02-11 08:49:56,820 epoch 53 - iter 6/27 - loss 0.00656420\n",
            "2019-02-11 08:49:57,635 epoch 53 - iter 8/27 - loss 0.00660316\n",
            "2019-02-11 08:49:58,549 epoch 53 - iter 10/27 - loss 0.00672599\n",
            "2019-02-11 08:49:59,246 epoch 53 - iter 12/27 - loss 0.00633491\n",
            "2019-02-11 08:50:00,425 epoch 53 - iter 14/27 - loss 0.00637066\n",
            "2019-02-11 08:50:01,088 epoch 53 - iter 16/27 - loss 0.00649070\n",
            "2019-02-11 08:50:01,685 epoch 53 - iter 18/27 - loss 0.00617121\n",
            "2019-02-11 08:50:02,452 epoch 53 - iter 20/27 - loss 0.00590125\n",
            "2019-02-11 08:50:03,295 epoch 53 - iter 22/27 - loss 0.00591314\n",
            "2019-02-11 08:50:04,059 epoch 53 - iter 24/27 - loss 0.00621011\n",
            "2019-02-11 08:50:04,451 epoch 53 - iter 26/27 - loss 0.00629952\n",
            "2019-02-11 08:50:04,462 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:50:04,463 EPOCH 53 done: loss 0.0063 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:50:05,091 DEV  : loss 0.05108314 - f-score 0.6842 - acc 0.6842\n",
            "2019-02-11 08:50:09,334 TEST : loss 0.01232350 - f-score 0.9038 - acc 0.9038\n",
            "2019-02-11 08:50:26,407 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:50:26,881 epoch 54 - iter 0/27 - loss 0.00495604\n",
            "2019-02-11 08:50:27,625 epoch 54 - iter 2/27 - loss 0.00722873\n",
            "2019-02-11 08:50:28,324 epoch 54 - iter 4/27 - loss 0.00629485\n",
            "2019-02-11 08:50:29,142 epoch 54 - iter 6/27 - loss 0.00701598\n",
            "2019-02-11 08:50:30,235 epoch 54 - iter 8/27 - loss 0.00720217\n",
            "2019-02-11 08:50:30,839 epoch 54 - iter 10/27 - loss 0.00650567\n",
            "2019-02-11 08:50:31,531 epoch 54 - iter 12/27 - loss 0.00677217\n",
            "2019-02-11 08:50:32,284 epoch 54 - iter 14/27 - loss 0.00751222\n",
            "2019-02-11 08:50:32,962 epoch 54 - iter 16/27 - loss 0.00769463\n",
            "2019-02-11 08:50:33,657 epoch 54 - iter 18/27 - loss 0.00751003\n",
            "2019-02-11 08:50:34,422 epoch 54 - iter 20/27 - loss 0.00710887\n",
            "2019-02-11 08:50:35,514 epoch 54 - iter 22/27 - loss 0.00713798\n",
            "2019-02-11 08:50:36,353 epoch 54 - iter 24/27 - loss 0.00726598\n",
            "2019-02-11 08:50:36,826 epoch 54 - iter 26/27 - loss 0.00706627\n",
            "2019-02-11 08:50:36,837 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:50:36,838 EPOCH 54 done: loss 0.0071 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:50:37,471 DEV  : loss 0.06132209 - f-score 0.6737 - acc 0.6737\n",
            "2019-02-11 08:50:41,940 TEST : loss 0.01418503 - f-score 0.9145 - acc 0.9145\n",
            "2019-02-11 08:50:41,943 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:50:42,294 epoch 55 - iter 0/27 - loss 0.00280388\n",
            "2019-02-11 08:50:42,940 epoch 55 - iter 2/27 - loss 0.00524023\n",
            "2019-02-11 08:50:43,632 epoch 55 - iter 4/27 - loss 0.00546418\n",
            "2019-02-11 08:50:44,343 epoch 55 - iter 6/27 - loss 0.00616400\n",
            "2019-02-11 08:50:45,127 epoch 55 - iter 8/27 - loss 0.00658931\n",
            "2019-02-11 08:50:45,635 epoch 55 - iter 10/27 - loss 0.00620477\n",
            "2019-02-11 08:50:46,423 epoch 55 - iter 12/27 - loss 0.00604906\n",
            "2019-02-11 08:50:47,313 epoch 55 - iter 14/27 - loss 0.00578984\n",
            "2019-02-11 08:50:48,053 epoch 55 - iter 16/27 - loss 0.00580260\n",
            "2019-02-11 08:50:48,754 epoch 55 - iter 18/27 - loss 0.00560629\n",
            "2019-02-11 08:50:49,997 epoch 55 - iter 20/27 - loss 0.00545291\n",
            "2019-02-11 08:50:51,343 epoch 55 - iter 22/27 - loss 0.00582488\n",
            "2019-02-11 08:50:52,009 epoch 55 - iter 24/27 - loss 0.00595449\n",
            "2019-02-11 08:50:52,601 epoch 55 - iter 26/27 - loss 0.00625441\n",
            "2019-02-11 08:50:52,614 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:50:52,615 EPOCH 55 done: loss 0.0063 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 08:50:53,237 DEV  : loss 0.05295098 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 08:50:57,472 TEST : loss 0.01273241 - f-score 0.9103 - acc 0.9103\n",
            "2019-02-11 08:51:14,564 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:51:15,064 epoch 56 - iter 0/27 - loss 0.00982987\n",
            "2019-02-11 08:51:15,950 epoch 56 - iter 2/27 - loss 0.00857864\n",
            "2019-02-11 08:51:16,564 epoch 56 - iter 4/27 - loss 0.00793397\n",
            "2019-02-11 08:51:17,608 epoch 56 - iter 6/27 - loss 0.00707221\n",
            "2019-02-11 08:51:18,213 epoch 56 - iter 8/27 - loss 0.00694733\n",
            "2019-02-11 08:51:18,837 epoch 56 - iter 10/27 - loss 0.00725789\n",
            "2019-02-11 08:51:19,574 epoch 56 - iter 12/27 - loss 0.00663112\n",
            "2019-02-11 08:51:20,332 epoch 56 - iter 14/27 - loss 0.00643215\n",
            "2019-02-11 08:51:21,264 epoch 56 - iter 16/27 - loss 0.00644479\n",
            "2019-02-11 08:51:21,995 epoch 56 - iter 18/27 - loss 0.00607789\n",
            "2019-02-11 08:51:22,763 epoch 56 - iter 20/27 - loss 0.00633823\n",
            "2019-02-11 08:51:23,397 epoch 56 - iter 22/27 - loss 0.00610750\n",
            "2019-02-11 08:51:24,620 epoch 56 - iter 24/27 - loss 0.00593043\n",
            "2019-02-11 08:51:25,065 epoch 56 - iter 26/27 - loss 0.00630836\n",
            "2019-02-11 08:51:25,076 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:51:25,077 EPOCH 56 done: loss 0.0063 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 08:51:25,709 DEV  : loss 0.06824962 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 08:51:30,166 TEST : loss 0.01658166 - f-score 0.9017 - acc 0.9017\n",
            "2019-02-11 08:51:30,169 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:51:30,612 epoch 57 - iter 0/27 - loss 0.00568863\n",
            "2019-02-11 08:51:31,308 epoch 57 - iter 2/27 - loss 0.00714530\n",
            "2019-02-11 08:51:32,259 epoch 57 - iter 4/27 - loss 0.00809626\n",
            "2019-02-11 08:51:33,416 epoch 57 - iter 6/27 - loss 0.00647682\n",
            "2019-02-11 08:51:34,151 epoch 57 - iter 8/27 - loss 0.00753163\n",
            "2019-02-11 08:51:35,198 epoch 57 - iter 10/27 - loss 0.00818343\n",
            "2019-02-11 08:51:35,856 epoch 57 - iter 12/27 - loss 0.00762754\n",
            "2019-02-11 08:51:36,680 epoch 57 - iter 14/27 - loss 0.00779497\n",
            "2019-02-11 08:51:37,584 epoch 57 - iter 16/27 - loss 0.00785941\n",
            "2019-02-11 08:51:38,307 epoch 57 - iter 18/27 - loss 0.00739280\n",
            "2019-02-11 08:51:39,028 epoch 57 - iter 20/27 - loss 0.00766026\n",
            "2019-02-11 08:51:39,681 epoch 57 - iter 22/27 - loss 0.00764098\n",
            "2019-02-11 08:51:40,474 epoch 57 - iter 24/27 - loss 0.00768665\n",
            "2019-02-11 08:51:40,980 epoch 57 - iter 26/27 - loss 0.00763016\n",
            "2019-02-11 08:51:40,991 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:51:40,992 EPOCH 57 done: loss 0.0076 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 08:51:41,611 DEV  : loss 0.06360523 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 08:51:45,832 TEST : loss 0.01471141 - f-score 0.9167 - acc 0.9167\n",
            "2019-02-11 08:51:45,834 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:51:46,167 epoch 58 - iter 0/27 - loss 0.00895465\n",
            "2019-02-11 08:51:46,889 epoch 58 - iter 2/27 - loss 0.00444765\n",
            "2019-02-11 08:51:47,796 epoch 58 - iter 4/27 - loss 0.00553481\n",
            "2019-02-11 08:51:48,642 epoch 58 - iter 6/27 - loss 0.00469378\n",
            "2019-02-11 08:51:49,223 epoch 58 - iter 8/27 - loss 0.00431161\n",
            "2019-02-11 08:51:49,856 epoch 58 - iter 10/27 - loss 0.00404268\n",
            "2019-02-11 08:51:50,745 epoch 58 - iter 12/27 - loss 0.00437798\n",
            "2019-02-11 08:51:51,888 epoch 58 - iter 14/27 - loss 0.00546526\n",
            "2019-02-11 08:51:52,528 epoch 58 - iter 16/27 - loss 0.00563087\n",
            "2019-02-11 08:51:53,223 epoch 58 - iter 18/27 - loss 0.00579507\n",
            "2019-02-11 08:51:53,925 epoch 58 - iter 20/27 - loss 0.00602575\n",
            "2019-02-11 08:51:55,146 epoch 58 - iter 22/27 - loss 0.00595393\n",
            "2019-02-11 08:51:55,854 epoch 58 - iter 24/27 - loss 0.00615429\n",
            "2019-02-11 08:51:56,380 epoch 58 - iter 26/27 - loss 0.00639861\n",
            "2019-02-11 08:51:56,391 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:51:56,392 EPOCH 58 done: loss 0.0064 - lr 0.1000 - bad epochs 2\n",
            "2019-02-11 08:51:57,009 DEV  : loss 0.06037766 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 08:52:01,319 TEST : loss 0.01342751 - f-score 0.9188 - acc 0.9188\n",
            "2019-02-11 08:52:01,322 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:52:01,674 epoch 59 - iter 0/27 - loss 0.00705194\n",
            "2019-02-11 08:52:02,321 epoch 59 - iter 2/27 - loss 0.00780894\n",
            "2019-02-11 08:52:03,092 epoch 59 - iter 4/27 - loss 0.00541924\n",
            "2019-02-11 08:52:03,849 epoch 59 - iter 6/27 - loss 0.00613005\n",
            "2019-02-11 08:52:04,855 epoch 59 - iter 8/27 - loss 0.00613899\n",
            "2019-02-11 08:52:05,590 epoch 59 - iter 10/27 - loss 0.00635852\n",
            "2019-02-11 08:52:06,271 epoch 59 - iter 12/27 - loss 0.00656226\n",
            "2019-02-11 08:52:07,503 epoch 59 - iter 14/27 - loss 0.00611253\n",
            "2019-02-11 08:52:08,522 epoch 59 - iter 16/27 - loss 0.00634274\n",
            "2019-02-11 08:52:09,186 epoch 59 - iter 18/27 - loss 0.00658916\n",
            "2019-02-11 08:52:09,992 epoch 59 - iter 20/27 - loss 0.00676508\n",
            "2019-02-11 08:52:10,873 epoch 59 - iter 22/27 - loss 0.00745052\n",
            "2019-02-11 08:52:11,508 epoch 59 - iter 24/27 - loss 0.00759057\n",
            "2019-02-11 08:52:12,008 epoch 59 - iter 26/27 - loss 0.00741815\n",
            "2019-02-11 08:52:12,019 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:52:12,021 EPOCH 59 done: loss 0.0074 - lr 0.1000 - bad epochs 3\n",
            "2019-02-11 08:52:12,650 DEV  : loss 0.05939491 - f-score 0.6737 - acc 0.6737\n",
            "2019-02-11 08:52:16,877 TEST : loss 0.01354903 - f-score 0.9167 - acc 0.9167\n",
            "Epoch    58: reducing learning rate of group 0 to 5.0000e-02.\n",
            "2019-02-11 08:52:16,881 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:52:17,288 epoch 60 - iter 0/27 - loss 0.00299971\n",
            "2019-02-11 08:52:17,856 epoch 60 - iter 2/27 - loss 0.00468965\n",
            "2019-02-11 08:52:18,592 epoch 60 - iter 4/27 - loss 0.00394316\n",
            "2019-02-11 08:52:19,097 epoch 60 - iter 6/27 - loss 0.00494734\n",
            "2019-02-11 08:52:20,027 epoch 60 - iter 8/27 - loss 0.00456913\n",
            "2019-02-11 08:52:20,667 epoch 60 - iter 10/27 - loss 0.00507710\n",
            "2019-02-11 08:52:21,248 epoch 60 - iter 12/27 - loss 0.00516739\n",
            "2019-02-11 08:52:22,521 epoch 60 - iter 14/27 - loss 0.00569411\n",
            "2019-02-11 08:52:23,163 epoch 60 - iter 16/27 - loss 0.00535097\n",
            "2019-02-11 08:52:23,869 epoch 60 - iter 18/27 - loss 0.00549695\n",
            "2019-02-11 08:52:24,755 epoch 60 - iter 20/27 - loss 0.00554920\n",
            "2019-02-11 08:52:25,453 epoch 60 - iter 22/27 - loss 0.00524124\n",
            "2019-02-11 08:52:26,086 epoch 60 - iter 24/27 - loss 0.00518018\n",
            "2019-02-11 08:52:26,493 epoch 60 - iter 26/27 - loss 0.00557008\n",
            "2019-02-11 08:52:26,504 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:52:26,505 EPOCH 60 done: loss 0.0056 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 08:52:27,142 DEV  : loss 0.05770594 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 08:52:31,533 TEST : loss 0.01265318 - f-score 0.9231 - acc 0.9231\n",
            "2019-02-11 08:52:48,729 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:52:49,139 epoch 61 - iter 0/27 - loss 0.00563622\n",
            "2019-02-11 08:52:50,363 epoch 61 - iter 2/27 - loss 0.00345331\n",
            "2019-02-11 08:52:51,291 epoch 61 - iter 4/27 - loss 0.00351895\n",
            "2019-02-11 08:52:52,068 epoch 61 - iter 6/27 - loss 0.00358622\n",
            "2019-02-11 08:52:53,222 epoch 61 - iter 8/27 - loss 0.00326467\n",
            "2019-02-11 08:52:54,054 epoch 61 - iter 10/27 - loss 0.00475896\n",
            "2019-02-11 08:52:54,667 epoch 61 - iter 12/27 - loss 0.00605072\n",
            "2019-02-11 08:52:55,325 epoch 61 - iter 14/27 - loss 0.00626357\n",
            "2019-02-11 08:52:55,965 epoch 61 - iter 16/27 - loss 0.00589835\n",
            "2019-02-11 08:52:56,769 epoch 61 - iter 18/27 - loss 0.00572015\n",
            "2019-02-11 08:52:57,472 epoch 61 - iter 20/27 - loss 0.00619260\n",
            "2019-02-11 08:52:58,351 epoch 61 - iter 22/27 - loss 0.00600392\n",
            "2019-02-11 08:52:59,123 epoch 61 - iter 24/27 - loss 0.00627644\n",
            "2019-02-11 08:52:59,418 epoch 61 - iter 26/27 - loss 0.00636338\n",
            "2019-02-11 08:52:59,429 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:52:59,431 EPOCH 61 done: loss 0.0064 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 08:53:00,055 DEV  : loss 0.05445006 - f-score 0.6842 - acc 0.6842\n",
            "2019-02-11 08:53:04,282 TEST : loss 0.01262615 - f-score 0.9209 - acc 0.9209\n",
            "2019-02-11 08:53:04,284 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:53:04,620 epoch 62 - iter 0/27 - loss 0.01550227\n",
            "2019-02-11 08:53:05,224 epoch 62 - iter 2/27 - loss 0.00761805\n",
            "2019-02-11 08:53:06,026 epoch 62 - iter 4/27 - loss 0.00712080\n",
            "2019-02-11 08:53:06,702 epoch 62 - iter 6/27 - loss 0.00695225\n",
            "2019-02-11 08:53:07,311 epoch 62 - iter 8/27 - loss 0.00705415\n",
            "2019-02-11 08:53:08,188 epoch 62 - iter 10/27 - loss 0.00738294\n",
            "2019-02-11 08:53:08,916 epoch 62 - iter 12/27 - loss 0.00699883\n",
            "2019-02-11 08:53:09,683 epoch 62 - iter 14/27 - loss 0.00638620\n",
            "2019-02-11 08:53:10,365 epoch 62 - iter 16/27 - loss 0.00623789\n",
            "2019-02-11 08:53:11,007 epoch 62 - iter 18/27 - loss 0.00611668\n",
            "2019-02-11 08:53:11,707 epoch 62 - iter 20/27 - loss 0.00624589\n",
            "2019-02-11 08:53:12,405 epoch 62 - iter 22/27 - loss 0.00619601\n",
            "2019-02-11 08:53:13,631 epoch 62 - iter 24/27 - loss 0.00622328\n",
            "2019-02-11 08:53:14,032 epoch 62 - iter 26/27 - loss 0.00632898\n",
            "2019-02-11 08:53:14,043 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:53:14,044 EPOCH 62 done: loss 0.0063 - lr 0.0500 - bad epochs 1\n",
            "2019-02-11 08:53:14,659 DEV  : loss 0.06094483 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 08:53:18,901 TEST : loss 0.01302259 - f-score 0.9295 - acc 0.9295\n",
            "2019-02-11 08:53:18,904 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:53:19,410 epoch 63 - iter 0/27 - loss 0.00280983\n",
            "2019-02-11 08:53:20,058 epoch 63 - iter 2/27 - loss 0.00340797\n",
            "2019-02-11 08:53:20,832 epoch 63 - iter 4/27 - loss 0.00391058\n",
            "2019-02-11 08:53:21,603 epoch 63 - iter 6/27 - loss 0.00522508\n",
            "2019-02-11 08:53:22,395 epoch 63 - iter 8/27 - loss 0.00568553\n",
            "2019-02-11 08:53:23,078 epoch 63 - iter 10/27 - loss 0.00504237\n",
            "2019-02-11 08:53:23,843 epoch 63 - iter 12/27 - loss 0.00537423\n",
            "2019-02-11 08:53:25,502 epoch 63 - iter 14/27 - loss 0.00508044\n",
            "2019-02-11 08:53:26,106 epoch 63 - iter 16/27 - loss 0.00535447\n",
            "2019-02-11 08:53:26,858 epoch 63 - iter 18/27 - loss 0.00574867\n",
            "2019-02-11 08:53:27,773 epoch 63 - iter 20/27 - loss 0.00611350\n",
            "2019-02-11 08:53:28,363 epoch 63 - iter 22/27 - loss 0.00590065\n",
            "2019-02-11 08:53:29,301 epoch 63 - iter 24/27 - loss 0.00620622\n",
            "2019-02-11 08:53:29,771 epoch 63 - iter 26/27 - loss 0.00602750\n",
            "2019-02-11 08:53:29,782 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:53:29,783 EPOCH 63 done: loss 0.0060 - lr 0.0500 - bad epochs 2\n",
            "2019-02-11 08:53:30,403 DEV  : loss 0.06201892 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 08:53:34,623 TEST : loss 0.01323146 - f-score 0.9209 - acc 0.9209\n",
            "2019-02-11 08:53:34,626 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:53:34,903 epoch 64 - iter 0/27 - loss 0.01317579\n",
            "2019-02-11 08:53:35,915 epoch 64 - iter 2/27 - loss 0.00749881\n",
            "2019-02-11 08:53:36,680 epoch 64 - iter 4/27 - loss 0.00617153\n",
            "2019-02-11 08:53:37,532 epoch 64 - iter 6/27 - loss 0.00495980\n",
            "2019-02-11 08:53:38,205 epoch 64 - iter 8/27 - loss 0.00446974\n",
            "2019-02-11 08:53:39,087 epoch 64 - iter 10/27 - loss 0.00445620\n",
            "2019-02-11 08:53:39,728 epoch 64 - iter 12/27 - loss 0.00416118\n",
            "2019-02-11 08:53:40,840 epoch 64 - iter 14/27 - loss 0.00435083\n",
            "2019-02-11 08:53:42,063 epoch 64 - iter 16/27 - loss 0.00400880\n",
            "2019-02-11 08:53:42,631 epoch 64 - iter 18/27 - loss 0.00458120\n",
            "2019-02-11 08:53:43,258 epoch 64 - iter 20/27 - loss 0.00437421\n",
            "2019-02-11 08:53:44,020 epoch 64 - iter 22/27 - loss 0.00446736\n",
            "2019-02-11 08:53:44,643 epoch 64 - iter 24/27 - loss 0.00470464\n",
            "2019-02-11 08:53:45,000 epoch 64 - iter 26/27 - loss 0.00455867\n",
            "2019-02-11 08:53:45,013 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:53:45,014 EPOCH 64 done: loss 0.0046 - lr 0.0500 - bad epochs 3\n",
            "2019-02-11 08:53:45,624 DEV  : loss 0.06179952 - f-score 0.6737 - acc 0.6737\n",
            "2019-02-11 08:53:49,791 TEST : loss 0.01303606 - f-score 0.9231 - acc 0.9231\n",
            "2019-02-11 08:54:06,801 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:54:07,143 epoch 65 - iter 0/27 - loss 0.00834293\n",
            "2019-02-11 08:54:07,837 epoch 65 - iter 2/27 - loss 0.00521292\n",
            "2019-02-11 08:54:08,593 epoch 65 - iter 4/27 - loss 0.00387841\n",
            "2019-02-11 08:54:09,388 epoch 65 - iter 6/27 - loss 0.00448100\n",
            "2019-02-11 08:54:10,043 epoch 65 - iter 8/27 - loss 0.00413362\n",
            "2019-02-11 08:54:10,726 epoch 65 - iter 10/27 - loss 0.00466076\n",
            "2019-02-11 08:54:11,478 epoch 65 - iter 12/27 - loss 0.00468480\n",
            "2019-02-11 08:54:12,271 epoch 65 - iter 14/27 - loss 0.00422657\n",
            "2019-02-11 08:54:13,467 epoch 65 - iter 16/27 - loss 0.00407319\n",
            "2019-02-11 08:54:14,067 epoch 65 - iter 18/27 - loss 0.00391064\n",
            "2019-02-11 08:54:15,251 epoch 65 - iter 20/27 - loss 0.00383750\n",
            "2019-02-11 08:54:16,197 epoch 65 - iter 22/27 - loss 0.00370843\n",
            "2019-02-11 08:54:16,966 epoch 65 - iter 24/27 - loss 0.00354869\n",
            "2019-02-11 08:54:17,445 epoch 65 - iter 26/27 - loss 0.00377595\n",
            "2019-02-11 08:54:17,458 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:54:17,459 EPOCH 65 done: loss 0.0038 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 08:54:18,084 DEV  : loss 0.05927072 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 08:54:22,523 TEST : loss 0.01258486 - f-score 0.9209 - acc 0.9209\n",
            "2019-02-11 08:54:39,704 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:54:40,058 epoch 66 - iter 0/27 - loss 0.00167640\n",
            "2019-02-11 08:54:40,925 epoch 66 - iter 2/27 - loss 0.00190987\n",
            "2019-02-11 08:54:42,232 epoch 66 - iter 4/27 - loss 0.00421709\n",
            "2019-02-11 08:54:43,104 epoch 66 - iter 6/27 - loss 0.00381254\n",
            "2019-02-11 08:54:43,896 epoch 66 - iter 8/27 - loss 0.00369546\n",
            "2019-02-11 08:54:44,775 epoch 66 - iter 10/27 - loss 0.00393584\n",
            "2019-02-11 08:54:45,431 epoch 66 - iter 12/27 - loss 0.00401885\n",
            "2019-02-11 08:54:46,069 epoch 66 - iter 14/27 - loss 0.00405700\n",
            "2019-02-11 08:54:46,751 epoch 66 - iter 16/27 - loss 0.00427639\n",
            "2019-02-11 08:54:47,646 epoch 66 - iter 18/27 - loss 0.00399603\n",
            "2019-02-11 08:54:48,300 epoch 66 - iter 20/27 - loss 0.00397953\n",
            "2019-02-11 08:54:48,932 epoch 66 - iter 22/27 - loss 0.00444610\n",
            "2019-02-11 08:54:49,651 epoch 66 - iter 24/27 - loss 0.00466557\n",
            "2019-02-11 08:54:50,484 epoch 66 - iter 26/27 - loss 0.00476499\n",
            "2019-02-11 08:54:50,494 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:54:50,495 EPOCH 66 done: loss 0.0048 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 08:54:51,113 DEV  : loss 0.05912002 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 08:54:55,349 TEST : loss 0.01280678 - f-score 0.9274 - acc 0.9274\n",
            "2019-02-11 08:54:55,352 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:54:55,736 epoch 67 - iter 0/27 - loss 0.00157106\n",
            "2019-02-11 08:54:56,477 epoch 67 - iter 2/27 - loss 0.00204385\n",
            "2019-02-11 08:54:57,725 epoch 67 - iter 4/27 - loss 0.00386319\n",
            "2019-02-11 08:54:58,344 epoch 67 - iter 6/27 - loss 0.00336785\n",
            "2019-02-11 08:54:59,104 epoch 67 - iter 8/27 - loss 0.00367633\n",
            "2019-02-11 08:54:59,864 epoch 67 - iter 10/27 - loss 0.00379487\n",
            "2019-02-11 08:55:00,792 epoch 67 - iter 12/27 - loss 0.00418750\n",
            "2019-02-11 08:55:01,496 epoch 67 - iter 14/27 - loss 0.00400281\n",
            "2019-02-11 08:55:02,131 epoch 67 - iter 16/27 - loss 0.00400277\n",
            "2019-02-11 08:55:02,759 epoch 67 - iter 18/27 - loss 0.00402872\n",
            "2019-02-11 08:55:03,684 epoch 67 - iter 20/27 - loss 0.00378472\n",
            "2019-02-11 08:55:04,264 epoch 67 - iter 22/27 - loss 0.00396531\n",
            "2019-02-11 08:55:05,139 epoch 67 - iter 24/27 - loss 0.00390155\n",
            "2019-02-11 08:55:05,617 epoch 67 - iter 26/27 - loss 0.00397505\n",
            "2019-02-11 08:55:05,628 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:55:05,630 EPOCH 67 done: loss 0.0040 - lr 0.0500 - bad epochs 1\n",
            "2019-02-11 08:55:06,246 DEV  : loss 0.06598371 - f-score 0.6737 - acc 0.6737\n",
            "2019-02-11 08:55:10,502 TEST : loss 0.01401980 - f-score 0.9231 - acc 0.9231\n",
            "2019-02-11 08:55:10,505 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:55:10,794 epoch 68 - iter 0/27 - loss 0.00425926\n",
            "2019-02-11 08:55:11,459 epoch 68 - iter 2/27 - loss 0.00276149\n",
            "2019-02-11 08:55:12,206 epoch 68 - iter 4/27 - loss 0.00467097\n",
            "2019-02-11 08:55:12,877 epoch 68 - iter 6/27 - loss 0.00482904\n",
            "2019-02-11 08:55:14,040 epoch 68 - iter 8/27 - loss 0.00559256\n",
            "2019-02-11 08:55:14,935 epoch 68 - iter 10/27 - loss 0.00512456\n",
            "2019-02-11 08:55:15,530 epoch 68 - iter 12/27 - loss 0.00465571\n",
            "2019-02-11 08:55:16,269 epoch 68 - iter 14/27 - loss 0.00477523\n",
            "2019-02-11 08:55:16,910 epoch 68 - iter 16/27 - loss 0.00435934\n",
            "2019-02-11 08:55:17,688 epoch 68 - iter 18/27 - loss 0.00437984\n",
            "2019-02-11 08:55:18,826 epoch 68 - iter 20/27 - loss 0.00448160\n",
            "2019-02-11 08:55:19,491 epoch 68 - iter 22/27 - loss 0.00469678\n",
            "2019-02-11 08:55:20,547 epoch 68 - iter 24/27 - loss 0.00458698\n",
            "2019-02-11 08:55:21,007 epoch 68 - iter 26/27 - loss 0.00458257\n",
            "2019-02-11 08:55:21,019 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:55:21,020 EPOCH 68 done: loss 0.0046 - lr 0.0500 - bad epochs 2\n",
            "2019-02-11 08:55:21,635 DEV  : loss 0.06616296 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 08:55:25,910 TEST : loss 0.01392341 - f-score 0.9231 - acc 0.9231\n",
            "2019-02-11 08:55:25,916 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:55:26,450 epoch 69 - iter 0/27 - loss 0.00943048\n",
            "2019-02-11 08:55:27,475 epoch 69 - iter 2/27 - loss 0.00433960\n",
            "2019-02-11 08:55:28,366 epoch 69 - iter 4/27 - loss 0.00314424\n",
            "2019-02-11 08:55:28,886 epoch 69 - iter 6/27 - loss 0.00362298\n",
            "2019-02-11 08:55:30,061 epoch 69 - iter 8/27 - loss 0.00304877\n",
            "2019-02-11 08:55:30,815 epoch 69 - iter 10/27 - loss 0.00320555\n",
            "2019-02-11 08:55:31,394 epoch 69 - iter 12/27 - loss 0.00363037\n",
            "2019-02-11 08:55:32,124 epoch 69 - iter 14/27 - loss 0.00377478\n",
            "2019-02-11 08:55:33,211 epoch 69 - iter 16/27 - loss 0.00350766\n",
            "2019-02-11 08:55:33,955 epoch 69 - iter 18/27 - loss 0.00336100\n",
            "2019-02-11 08:55:34,721 epoch 69 - iter 20/27 - loss 0.00398942\n",
            "2019-02-11 08:55:35,429 epoch 69 - iter 22/27 - loss 0.00375218\n",
            "2019-02-11 08:55:36,069 epoch 69 - iter 24/27 - loss 0.00388736\n",
            "2019-02-11 08:55:36,576 epoch 69 - iter 26/27 - loss 0.00377165\n",
            "2019-02-11 08:55:36,589 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:55:36,590 EPOCH 69 done: loss 0.0038 - lr 0.0500 - bad epochs 3\n",
            "2019-02-11 08:55:37,226 DEV  : loss 0.06466721 - f-score 0.6737 - acc 0.6737\n",
            "2019-02-11 08:55:41,471 TEST : loss 0.01353693 - f-score 0.9252 - acc 0.9252\n",
            "2019-02-11 08:55:58,600 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:55:58,970 epoch 70 - iter 0/27 - loss 0.00127694\n",
            "2019-02-11 08:56:00,138 epoch 70 - iter 2/27 - loss 0.00397725\n",
            "2019-02-11 08:56:01,041 epoch 70 - iter 4/27 - loss 0.00458903\n",
            "2019-02-11 08:56:02,146 epoch 70 - iter 6/27 - loss 0.00441080\n",
            "2019-02-11 08:56:02,707 epoch 70 - iter 8/27 - loss 0.00460681\n",
            "2019-02-11 08:56:03,394 epoch 70 - iter 10/27 - loss 0.00425584\n",
            "2019-02-11 08:56:04,158 epoch 70 - iter 12/27 - loss 0.00431663\n",
            "2019-02-11 08:56:04,955 epoch 70 - iter 14/27 - loss 0.00443878\n",
            "2019-02-11 08:56:05,837 epoch 70 - iter 16/27 - loss 0.00432353\n",
            "2019-02-11 08:56:06,717 epoch 70 - iter 18/27 - loss 0.00441558\n",
            "2019-02-11 08:56:07,273 epoch 70 - iter 20/27 - loss 0.00479978\n",
            "2019-02-11 08:56:08,020 epoch 70 - iter 22/27 - loss 0.00449274\n",
            "2019-02-11 08:56:08,653 epoch 70 - iter 24/27 - loss 0.00440103\n",
            "2019-02-11 08:56:09,059 epoch 70 - iter 26/27 - loss 0.00442686\n",
            "2019-02-11 08:56:09,070 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:56:09,071 EPOCH 70 done: loss 0.0044 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 08:56:09,696 DEV  : loss 0.06451924 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 08:56:14,138 TEST : loss 0.01354504 - f-score 0.9274 - acc 0.9274\n",
            "2019-02-11 08:56:14,141 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:56:14,567 epoch 71 - iter 0/27 - loss 0.00093855\n",
            "2019-02-11 08:56:15,380 epoch 71 - iter 2/27 - loss 0.00265582\n",
            "2019-02-11 08:56:16,129 epoch 71 - iter 4/27 - loss 0.00556721\n",
            "2019-02-11 08:56:16,760 epoch 71 - iter 6/27 - loss 0.00489579\n",
            "2019-02-11 08:56:17,465 epoch 71 - iter 8/27 - loss 0.00418610\n",
            "2019-02-11 08:56:18,351 epoch 71 - iter 10/27 - loss 0.00465449\n",
            "2019-02-11 08:56:19,139 epoch 71 - iter 12/27 - loss 0.00435208\n",
            "2019-02-11 08:56:20,245 epoch 71 - iter 14/27 - loss 0.00440541\n",
            "2019-02-11 08:56:20,874 epoch 71 - iter 16/27 - loss 0.00428193\n",
            "2019-02-11 08:56:21,642 epoch 71 - iter 18/27 - loss 0.00390197\n",
            "2019-02-11 08:56:22,357 epoch 71 - iter 20/27 - loss 0.00402453\n",
            "2019-02-11 08:56:23,101 epoch 71 - iter 22/27 - loss 0.00396770\n",
            "2019-02-11 08:56:24,278 epoch 71 - iter 24/27 - loss 0.00385210\n",
            "2019-02-11 08:56:24,775 epoch 71 - iter 26/27 - loss 0.00398050\n",
            "2019-02-11 08:56:24,787 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:56:24,788 EPOCH 71 done: loss 0.0040 - lr 0.0500 - bad epochs 1\n",
            "2019-02-11 08:56:25,409 DEV  : loss 0.06264196 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 08:56:29,642 TEST : loss 0.01335459 - f-score 0.9295 - acc 0.9295\n",
            "2019-02-11 08:56:29,644 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:56:30,028 epoch 72 - iter 0/27 - loss 0.00072049\n",
            "2019-02-11 08:56:31,009 epoch 72 - iter 2/27 - loss 0.00164445\n",
            "2019-02-11 08:56:31,790 epoch 72 - iter 4/27 - loss 0.00192272\n",
            "2019-02-11 08:56:32,886 epoch 72 - iter 6/27 - loss 0.00178483\n",
            "2019-02-11 08:56:33,939 epoch 72 - iter 8/27 - loss 0.00209790\n",
            "2019-02-11 08:56:34,779 epoch 72 - iter 10/27 - loss 0.00257888\n",
            "2019-02-11 08:56:35,662 epoch 72 - iter 12/27 - loss 0.00284508\n",
            "2019-02-11 08:56:36,262 epoch 72 - iter 14/27 - loss 0.00308122\n",
            "2019-02-11 08:56:37,030 epoch 72 - iter 16/27 - loss 0.00299343\n",
            "2019-02-11 08:56:37,735 epoch 72 - iter 18/27 - loss 0.00274849\n",
            "2019-02-11 08:56:38,442 epoch 72 - iter 20/27 - loss 0.00269242\n",
            "2019-02-11 08:56:39,158 epoch 72 - iter 22/27 - loss 0.00260017\n",
            "2019-02-11 08:56:39,804 epoch 72 - iter 24/27 - loss 0.00260010\n",
            "2019-02-11 08:56:40,210 epoch 72 - iter 26/27 - loss 0.00290269\n",
            "2019-02-11 08:56:40,220 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:56:40,221 EPOCH 72 done: loss 0.0029 - lr 0.0500 - bad epochs 2\n",
            "2019-02-11 08:56:40,838 DEV  : loss 0.06177498 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 08:56:45,082 TEST : loss 0.01285097 - f-score 0.9295 - acc 0.9295\n",
            "2019-02-11 08:57:02,310 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:57:02,814 epoch 73 - iter 0/27 - loss 0.00131742\n",
            "2019-02-11 08:57:03,783 epoch 73 - iter 2/27 - loss 0.00256321\n",
            "2019-02-11 08:57:04,905 epoch 73 - iter 4/27 - loss 0.00403563\n",
            "2019-02-11 08:57:05,825 epoch 73 - iter 6/27 - loss 0.00357670\n",
            "2019-02-11 08:57:06,521 epoch 73 - iter 8/27 - loss 0.00368427\n",
            "2019-02-11 08:57:07,443 epoch 73 - iter 10/27 - loss 0.00362738\n",
            "2019-02-11 08:57:08,025 epoch 73 - iter 12/27 - loss 0.00387854\n",
            "2019-02-11 08:57:09,226 epoch 73 - iter 14/27 - loss 0.00382230\n",
            "2019-02-11 08:57:09,838 epoch 73 - iter 16/27 - loss 0.00380123\n",
            "2019-02-11 08:57:10,533 epoch 73 - iter 18/27 - loss 0.00409926\n",
            "2019-02-11 08:57:11,209 epoch 73 - iter 20/27 - loss 0.00403766\n",
            "2019-02-11 08:57:11,709 epoch 73 - iter 22/27 - loss 0.00396242\n",
            "2019-02-11 08:57:12,242 epoch 73 - iter 24/27 - loss 0.00387727\n",
            "2019-02-11 08:57:12,692 epoch 73 - iter 26/27 - loss 0.00373517\n",
            "2019-02-11 08:57:12,703 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:57:12,705 EPOCH 73 done: loss 0.0037 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 08:57:13,319 DEV  : loss 0.06198196 - f-score 0.6842 - acc 0.6842\n",
            "2019-02-11 08:57:17,538 TEST : loss 0.01319911 - f-score 0.9274 - acc 0.9274\n",
            "2019-02-11 08:57:17,541 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:57:18,376 epoch 74 - iter 0/27 - loss 0.00529572\n",
            "2019-02-11 08:57:19,069 epoch 74 - iter 2/27 - loss 0.00280970\n",
            "2019-02-11 08:57:19,769 epoch 74 - iter 4/27 - loss 0.00289471\n",
            "2019-02-11 08:57:20,401 epoch 74 - iter 6/27 - loss 0.00282234\n",
            "2019-02-11 08:57:21,251 epoch 74 - iter 8/27 - loss 0.00361749\n",
            "2019-02-11 08:57:22,379 epoch 74 - iter 10/27 - loss 0.00411713\n",
            "2019-02-11 08:57:23,050 epoch 74 - iter 12/27 - loss 0.00411697\n",
            "2019-02-11 08:57:24,053 epoch 74 - iter 14/27 - loss 0.00394563\n",
            "2019-02-11 08:57:24,665 epoch 74 - iter 16/27 - loss 0.00402088\n",
            "2019-02-11 08:57:25,498 epoch 74 - iter 18/27 - loss 0.00370564\n",
            "2019-02-11 08:57:26,184 epoch 74 - iter 20/27 - loss 0.00386583\n",
            "2019-02-11 08:57:26,932 epoch 74 - iter 22/27 - loss 0.00369167\n",
            "2019-02-11 08:57:27,630 epoch 74 - iter 24/27 - loss 0.00363980\n",
            "2019-02-11 08:57:28,061 epoch 74 - iter 26/27 - loss 0.00359705\n",
            "2019-02-11 08:57:28,073 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:57:28,074 EPOCH 74 done: loss 0.0036 - lr 0.0500 - bad epochs 1\n",
            "2019-02-11 08:57:28,710 DEV  : loss 0.06363355 - f-score 0.6737 - acc 0.6737\n",
            "2019-02-11 08:57:32,944 TEST : loss 0.01324778 - f-score 0.9295 - acc 0.9295\n",
            "2019-02-11 08:57:32,947 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:57:33,348 epoch 75 - iter 0/27 - loss 0.00194567\n",
            "2019-02-11 08:57:34,161 epoch 75 - iter 2/27 - loss 0.00358064\n",
            "2019-02-11 08:57:34,956 epoch 75 - iter 4/27 - loss 0.00260433\n",
            "2019-02-11 08:57:36,071 epoch 75 - iter 6/27 - loss 0.00264287\n",
            "2019-02-11 08:57:36,833 epoch 75 - iter 8/27 - loss 0.00294227\n",
            "2019-02-11 08:57:37,391 epoch 75 - iter 10/27 - loss 0.00357252\n",
            "2019-02-11 08:57:38,238 epoch 75 - iter 12/27 - loss 0.00325248\n",
            "2019-02-11 08:57:39,051 epoch 75 - iter 14/27 - loss 0.00304516\n",
            "2019-02-11 08:57:39,820 epoch 75 - iter 16/27 - loss 0.00368529\n",
            "2019-02-11 08:57:40,800 epoch 75 - iter 18/27 - loss 0.00346630\n",
            "2019-02-11 08:57:41,368 epoch 75 - iter 20/27 - loss 0.00345562\n",
            "2019-02-11 08:57:42,497 epoch 75 - iter 22/27 - loss 0.00325533\n",
            "2019-02-11 08:57:43,018 epoch 75 - iter 24/27 - loss 0.00339055\n",
            "2019-02-11 08:57:43,454 epoch 75 - iter 26/27 - loss 0.00351091\n",
            "2019-02-11 08:57:43,464 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:57:43,465 EPOCH 75 done: loss 0.0035 - lr 0.0500 - bad epochs 2\n",
            "2019-02-11 08:57:44,092 DEV  : loss 0.05924071 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 08:57:48,318 TEST : loss 0.01234182 - f-score 0.9359 - acc 0.9359\n",
            "2019-02-11 08:57:48,321 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:57:48,726 epoch 76 - iter 0/27 - loss 0.00379341\n",
            "2019-02-11 08:57:49,904 epoch 76 - iter 2/27 - loss 0.00258738\n",
            "2019-02-11 08:57:50,663 epoch 76 - iter 4/27 - loss 0.00309673\n",
            "2019-02-11 08:57:51,368 epoch 76 - iter 6/27 - loss 0.00352100\n",
            "2019-02-11 08:57:52,445 epoch 76 - iter 8/27 - loss 0.00302556\n",
            "2019-02-11 08:57:53,254 epoch 76 - iter 10/27 - loss 0.00287315\n",
            "2019-02-11 08:57:53,977 epoch 76 - iter 12/27 - loss 0.00270266\n",
            "2019-02-11 08:57:54,616 epoch 76 - iter 14/27 - loss 0.00286109\n",
            "2019-02-11 08:57:55,442 epoch 76 - iter 16/27 - loss 0.00342297\n",
            "2019-02-11 08:57:56,487 epoch 76 - iter 18/27 - loss 0.00336467\n",
            "2019-02-11 08:57:57,230 epoch 76 - iter 20/27 - loss 0.00361783\n",
            "2019-02-11 08:57:57,876 epoch 76 - iter 22/27 - loss 0.00349900\n",
            "2019-02-11 08:57:58,465 epoch 76 - iter 24/27 - loss 0.00331222\n",
            "2019-02-11 08:57:58,916 epoch 76 - iter 26/27 - loss 0.00340371\n",
            "2019-02-11 08:57:58,927 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:57:58,928 EPOCH 76 done: loss 0.0034 - lr 0.0500 - bad epochs 3\n",
            "2019-02-11 08:57:59,547 DEV  : loss 0.06210781 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 08:58:03,760 TEST : loss 0.01278018 - f-score 0.9338 - acc 0.9338\n",
            "Epoch    75: reducing learning rate of group 0 to 2.5000e-02.\n",
            "2019-02-11 08:58:03,763 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:58:04,060 epoch 77 - iter 0/27 - loss 0.01205421\n",
            "2019-02-11 08:58:04,859 epoch 77 - iter 2/27 - loss 0.00562420\n",
            "2019-02-11 08:58:05,639 epoch 77 - iter 4/27 - loss 0.00454919\n",
            "2019-02-11 08:58:06,292 epoch 77 - iter 6/27 - loss 0.00476402\n",
            "2019-02-11 08:58:06,887 epoch 77 - iter 8/27 - loss 0.00503689\n",
            "2019-02-11 08:58:08,038 epoch 77 - iter 10/27 - loss 0.00470741\n",
            "2019-02-11 08:58:08,951 epoch 77 - iter 12/27 - loss 0.00428276\n",
            "2019-02-11 08:58:09,582 epoch 77 - iter 14/27 - loss 0.00397917\n",
            "2019-02-11 08:58:10,332 epoch 77 - iter 16/27 - loss 0.00387634\n",
            "2019-02-11 08:58:11,003 epoch 77 - iter 18/27 - loss 0.00357214\n",
            "2019-02-11 08:58:11,868 epoch 77 - iter 20/27 - loss 0.00357925\n",
            "2019-02-11 08:58:12,774 epoch 77 - iter 22/27 - loss 0.00350720\n",
            "2019-02-11 08:58:13,324 epoch 77 - iter 24/27 - loss 0.00357802\n",
            "2019-02-11 08:58:14,264 epoch 77 - iter 26/27 - loss 0.00363069\n",
            "2019-02-11 08:58:14,277 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:58:14,278 EPOCH 77 done: loss 0.0036 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 08:58:14,901 DEV  : loss 0.06404477 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 08:58:19,099 TEST : loss 0.01312898 - f-score 0.9359 - acc 0.9359\n",
            "2019-02-11 08:58:19,101 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:58:19,490 epoch 78 - iter 0/27 - loss 0.00182629\n",
            "2019-02-11 08:58:20,641 epoch 78 - iter 2/27 - loss 0.00441022\n",
            "2019-02-11 08:58:21,352 epoch 78 - iter 4/27 - loss 0.00288797\n",
            "2019-02-11 08:58:22,102 epoch 78 - iter 6/27 - loss 0.00305314\n",
            "2019-02-11 08:58:22,823 epoch 78 - iter 8/27 - loss 0.00254672\n",
            "2019-02-11 08:58:23,378 epoch 78 - iter 10/27 - loss 0.00240273\n",
            "2019-02-11 08:58:24,025 epoch 78 - iter 12/27 - loss 0.00297090\n",
            "2019-02-11 08:58:24,789 epoch 78 - iter 14/27 - loss 0.00306877\n",
            "2019-02-11 08:58:25,472 epoch 78 - iter 16/27 - loss 0.00286740\n",
            "2019-02-11 08:58:26,268 epoch 78 - iter 18/27 - loss 0.00298700\n",
            "2019-02-11 08:58:27,446 epoch 78 - iter 20/27 - loss 0.00300374\n",
            "2019-02-11 08:58:28,704 epoch 78 - iter 22/27 - loss 0.00316796\n",
            "2019-02-11 08:58:29,495 epoch 78 - iter 24/27 - loss 0.00309844\n",
            "2019-02-11 08:58:30,045 epoch 78 - iter 26/27 - loss 0.00298800\n",
            "2019-02-11 08:58:30,057 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:58:30,059 EPOCH 78 done: loss 0.0030 - lr 0.0250 - bad epochs 1\n",
            "2019-02-11 08:58:30,696 DEV  : loss 0.06335162 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 08:58:34,902 TEST : loss 0.01301631 - f-score 0.9359 - acc 0.9359\n",
            "2019-02-11 08:58:34,904 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:58:35,212 epoch 79 - iter 0/27 - loss 0.00264302\n",
            "2019-02-11 08:58:35,816 epoch 79 - iter 2/27 - loss 0.00252417\n",
            "2019-02-11 08:58:36,572 epoch 79 - iter 4/27 - loss 0.00247998\n",
            "2019-02-11 08:58:37,221 epoch 79 - iter 6/27 - loss 0.00213520\n",
            "2019-02-11 08:58:37,854 epoch 79 - iter 8/27 - loss 0.00190010\n",
            "2019-02-11 08:58:38,933 epoch 79 - iter 10/27 - loss 0.00212139\n",
            "2019-02-11 08:58:39,620 epoch 79 - iter 12/27 - loss 0.00227995\n",
            "2019-02-11 08:58:40,335 epoch 79 - iter 14/27 - loss 0.00206678\n",
            "2019-02-11 08:58:40,957 epoch 79 - iter 16/27 - loss 0.00198681\n",
            "2019-02-11 08:58:41,669 epoch 79 - iter 18/27 - loss 0.00231620\n",
            "2019-02-11 08:58:42,592 epoch 79 - iter 20/27 - loss 0.00235848\n",
            "2019-02-11 08:58:43,829 epoch 79 - iter 22/27 - loss 0.00254393\n",
            "2019-02-11 08:58:44,555 epoch 79 - iter 24/27 - loss 0.00258341\n",
            "2019-02-11 08:58:45,126 epoch 79 - iter 26/27 - loss 0.00257749\n",
            "2019-02-11 08:58:45,138 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:58:45,139 EPOCH 79 done: loss 0.0026 - lr 0.0250 - bad epochs 2\n",
            "2019-02-11 08:58:45,760 DEV  : loss 0.06512622 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 08:58:50,038 TEST : loss 0.01337490 - f-score 0.9338 - acc 0.9338\n",
            "2019-02-11 08:59:07,198 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:59:07,870 epoch 80 - iter 0/27 - loss 0.00033048\n",
            "2019-02-11 08:59:08,630 epoch 80 - iter 2/27 - loss 0.00233718\n",
            "2019-02-11 08:59:09,325 epoch 80 - iter 4/27 - loss 0.00291304\n",
            "2019-02-11 08:59:10,072 epoch 80 - iter 6/27 - loss 0.00275379\n",
            "2019-02-11 08:59:10,918 epoch 80 - iter 8/27 - loss 0.00311855\n",
            "2019-02-11 08:59:11,787 epoch 80 - iter 10/27 - loss 0.00285603\n",
            "2019-02-11 08:59:12,612 epoch 80 - iter 12/27 - loss 0.00318724\n",
            "2019-02-11 08:59:13,311 epoch 80 - iter 14/27 - loss 0.00382240\n",
            "2019-02-11 08:59:13,865 epoch 80 - iter 16/27 - loss 0.00360104\n",
            "2019-02-11 08:59:14,421 epoch 80 - iter 18/27 - loss 0.00391057\n",
            "2019-02-11 08:59:15,670 epoch 80 - iter 20/27 - loss 0.00366209\n",
            "2019-02-11 08:59:16,406 epoch 80 - iter 22/27 - loss 0.00368004\n",
            "2019-02-11 08:59:17,047 epoch 80 - iter 24/27 - loss 0.00369048\n",
            "2019-02-11 08:59:17,921 epoch 80 - iter 26/27 - loss 0.00359945\n",
            "2019-02-11 08:59:17,932 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:59:17,933 EPOCH 80 done: loss 0.0036 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 08:59:18,568 DEV  : loss 0.06433582 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 08:59:22,870 TEST : loss 0.01316260 - f-score 0.9338 - acc 0.9338\n",
            "2019-02-11 08:59:22,872 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:59:23,174 epoch 81 - iter 0/27 - loss 0.00575286\n",
            "2019-02-11 08:59:23,873 epoch 81 - iter 2/27 - loss 0.00291112\n",
            "2019-02-11 08:59:24,559 epoch 81 - iter 4/27 - loss 0.00233877\n",
            "2019-02-11 08:59:25,346 epoch 81 - iter 6/27 - loss 0.00218395\n",
            "2019-02-11 08:59:26,721 epoch 81 - iter 8/27 - loss 0.00234110\n",
            "2019-02-11 08:59:27,690 epoch 81 - iter 10/27 - loss 0.00273085\n",
            "2019-02-11 08:59:28,425 epoch 81 - iter 12/27 - loss 0.00247605\n",
            "2019-02-11 08:59:29,129 epoch 81 - iter 14/27 - loss 0.00243256\n",
            "2019-02-11 08:59:29,934 epoch 81 - iter 16/27 - loss 0.00285968\n",
            "2019-02-11 08:59:30,802 epoch 81 - iter 18/27 - loss 0.00286412\n",
            "2019-02-11 08:59:31,462 epoch 81 - iter 20/27 - loss 0.00274538\n",
            "2019-02-11 08:59:32,100 epoch 81 - iter 22/27 - loss 0.00305134\n",
            "2019-02-11 08:59:32,818 epoch 81 - iter 24/27 - loss 0.00282385\n",
            "2019-02-11 08:59:33,360 epoch 81 - iter 26/27 - loss 0.00274259\n",
            "2019-02-11 08:59:33,371 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:59:33,372 EPOCH 81 done: loss 0.0027 - lr 0.0250 - bad epochs 1\n",
            "2019-02-11 08:59:34,004 DEV  : loss 0.06290106 - f-score 0.7158 - acc 0.7158\n",
            "2019-02-11 08:59:38,378 TEST : loss 0.01291363 - f-score 0.9380 - acc 0.9380\n",
            "2019-02-11 08:59:38,381 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:59:39,263 epoch 82 - iter 0/27 - loss 0.00091571\n",
            "2019-02-11 08:59:39,964 epoch 82 - iter 2/27 - loss 0.00440727\n",
            "2019-02-11 08:59:40,710 epoch 82 - iter 4/27 - loss 0.00316066\n",
            "2019-02-11 08:59:41,458 epoch 82 - iter 6/27 - loss 0.00389236\n",
            "2019-02-11 08:59:42,449 epoch 82 - iter 8/27 - loss 0.00361243\n",
            "2019-02-11 08:59:43,133 epoch 82 - iter 10/27 - loss 0.00304079\n",
            "2019-02-11 08:59:43,997 epoch 82 - iter 12/27 - loss 0.00330599\n",
            "2019-02-11 08:59:44,768 epoch 82 - iter 14/27 - loss 0.00373288\n",
            "2019-02-11 08:59:45,587 epoch 82 - iter 16/27 - loss 0.00390960\n",
            "2019-02-11 08:59:46,434 epoch 82 - iter 18/27 - loss 0.00373183\n",
            "2019-02-11 08:59:47,230 epoch 82 - iter 20/27 - loss 0.00362931\n",
            "2019-02-11 08:59:47,898 epoch 82 - iter 22/27 - loss 0.00357281\n",
            "2019-02-11 08:59:49,109 epoch 82 - iter 24/27 - loss 0.00352241\n",
            "2019-02-11 08:59:49,725 epoch 82 - iter 26/27 - loss 0.00368912\n",
            "2019-02-11 08:59:49,737 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:59:49,738 EPOCH 82 done: loss 0.0037 - lr 0.0250 - bad epochs 2\n",
            "2019-02-11 08:59:50,417 DEV  : loss 0.06578878 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 08:59:54,873 TEST : loss 0.01355862 - f-score 0.9338 - acc 0.9338\n",
            "2019-02-11 08:59:54,876 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 08:59:55,477 epoch 83 - iter 0/27 - loss 0.00099531\n",
            "2019-02-11 08:59:56,230 epoch 83 - iter 2/27 - loss 0.00271390\n",
            "2019-02-11 08:59:56,815 epoch 83 - iter 4/27 - loss 0.00291637\n",
            "2019-02-11 08:59:57,405 epoch 83 - iter 6/27 - loss 0.00328611\n",
            "2019-02-11 08:59:58,201 epoch 83 - iter 8/27 - loss 0.00312084\n",
            "2019-02-11 08:59:59,443 epoch 83 - iter 10/27 - loss 0.00334465\n",
            "2019-02-11 09:00:00,686 epoch 83 - iter 12/27 - loss 0.00304865\n",
            "2019-02-11 09:00:01,395 epoch 83 - iter 14/27 - loss 0.00295793\n",
            "2019-02-11 09:00:02,276 epoch 83 - iter 16/27 - loss 0.00330500\n",
            "2019-02-11 09:00:02,963 epoch 83 - iter 18/27 - loss 0.00337161\n",
            "2019-02-11 09:00:03,782 epoch 83 - iter 20/27 - loss 0.00324064\n",
            "2019-02-11 09:00:04,411 epoch 83 - iter 22/27 - loss 0.00309113\n",
            "2019-02-11 09:00:05,090 epoch 83 - iter 24/27 - loss 0.00289699\n",
            "2019-02-11 09:00:05,547 epoch 83 - iter 26/27 - loss 0.00288367\n",
            "2019-02-11 09:00:05,559 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:00:05,560 EPOCH 83 done: loss 0.0029 - lr 0.0250 - bad epochs 3\n",
            "2019-02-11 09:00:06,168 DEV  : loss 0.06590847 - f-score 0.7158 - acc 0.7158\n",
            "2019-02-11 09:00:10,393 TEST : loss 0.01351557 - f-score 0.9359 - acc 0.9359\n",
            "Epoch    82: reducing learning rate of group 0 to 1.2500e-02.\n",
            "2019-02-11 09:00:10,396 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:00:11,142 epoch 84 - iter 0/27 - loss 0.00069957\n",
            "2019-02-11 09:00:11,815 epoch 84 - iter 2/27 - loss 0.00415786\n",
            "2019-02-11 09:00:13,020 epoch 84 - iter 4/27 - loss 0.00482154\n",
            "2019-02-11 09:00:13,730 epoch 84 - iter 6/27 - loss 0.00480589\n",
            "2019-02-11 09:00:14,215 epoch 84 - iter 8/27 - loss 0.00471512\n",
            "2019-02-11 09:00:15,019 epoch 84 - iter 10/27 - loss 0.00490524\n",
            "2019-02-11 09:00:15,888 epoch 84 - iter 12/27 - loss 0.00429712\n",
            "2019-02-11 09:00:16,573 epoch 84 - iter 14/27 - loss 0.00407400\n",
            "2019-02-11 09:00:17,174 epoch 84 - iter 16/27 - loss 0.00393455\n",
            "2019-02-11 09:00:17,917 epoch 84 - iter 18/27 - loss 0.00393683\n",
            "2019-02-11 09:00:18,591 epoch 84 - iter 20/27 - loss 0.00371716\n",
            "2019-02-11 09:00:19,532 epoch 84 - iter 22/27 - loss 0.00350552\n",
            "2019-02-11 09:00:20,216 epoch 84 - iter 24/27 - loss 0.00338968\n",
            "2019-02-11 09:00:20,808 epoch 84 - iter 26/27 - loss 0.00347512\n",
            "2019-02-11 09:00:20,820 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:00:20,821 EPOCH 84 done: loss 0.0035 - lr 0.0125 - bad epochs 0\n",
            "2019-02-11 09:00:21,435 DEV  : loss 0.06639402 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 09:00:25,843 TEST : loss 0.01357066 - f-score 0.9359 - acc 0.9359\n",
            "2019-02-11 09:00:25,845 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:00:26,293 epoch 85 - iter 0/27 - loss 0.00370610\n",
            "2019-02-11 09:00:26,921 epoch 85 - iter 2/27 - loss 0.00242920\n",
            "2019-02-11 09:00:27,845 epoch 85 - iter 4/27 - loss 0.00379058\n",
            "2019-02-11 09:00:28,654 epoch 85 - iter 6/27 - loss 0.00388957\n",
            "2019-02-11 09:00:29,412 epoch 85 - iter 8/27 - loss 0.00402216\n",
            "2019-02-11 09:00:30,021 epoch 85 - iter 10/27 - loss 0.00334265\n",
            "2019-02-11 09:00:30,831 epoch 85 - iter 12/27 - loss 0.00372657\n",
            "2019-02-11 09:00:32,140 epoch 85 - iter 14/27 - loss 0.00362069\n",
            "2019-02-11 09:00:32,810 epoch 85 - iter 16/27 - loss 0.00350777\n",
            "2019-02-11 09:00:33,566 epoch 85 - iter 18/27 - loss 0.00326540\n",
            "2019-02-11 09:00:34,421 epoch 85 - iter 20/27 - loss 0.00334440\n",
            "2019-02-11 09:00:35,550 epoch 85 - iter 22/27 - loss 0.00313576\n",
            "2019-02-11 09:00:36,066 epoch 85 - iter 24/27 - loss 0.00300162\n",
            "2019-02-11 09:00:36,531 epoch 85 - iter 26/27 - loss 0.00326351\n",
            "2019-02-11 09:00:36,542 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:00:36,544 EPOCH 85 done: loss 0.0033 - lr 0.0125 - bad epochs 1\n",
            "2019-02-11 09:00:37,172 DEV  : loss 0.06677678 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 09:00:41,436 TEST : loss 0.01367030 - f-score 0.9338 - acc 0.9338\n",
            "2019-02-11 09:00:41,439 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:00:41,846 epoch 86 - iter 0/27 - loss 0.00364327\n",
            "2019-02-11 09:00:42,442 epoch 86 - iter 2/27 - loss 0.00258373\n",
            "2019-02-11 09:00:43,182 epoch 86 - iter 4/27 - loss 0.00311103\n",
            "2019-02-11 09:00:44,098 epoch 86 - iter 6/27 - loss 0.00245175\n",
            "2019-02-11 09:00:44,786 epoch 86 - iter 8/27 - loss 0.00302468\n",
            "2019-02-11 09:00:45,534 epoch 86 - iter 10/27 - loss 0.00268501\n",
            "2019-02-11 09:00:46,311 epoch 86 - iter 12/27 - loss 0.00257118\n",
            "2019-02-11 09:00:47,384 epoch 86 - iter 14/27 - loss 0.00253162\n",
            "2019-02-11 09:00:48,160 epoch 86 - iter 16/27 - loss 0.00244183\n",
            "2019-02-11 09:00:49,297 epoch 86 - iter 18/27 - loss 0.00247876\n",
            "2019-02-11 09:00:49,978 epoch 86 - iter 20/27 - loss 0.00248051\n",
            "2019-02-11 09:00:50,672 epoch 86 - iter 22/27 - loss 0.00241217\n",
            "2019-02-11 09:00:51,532 epoch 86 - iter 24/27 - loss 0.00308685\n",
            "2019-02-11 09:00:51,994 epoch 86 - iter 26/27 - loss 0.00324913\n",
            "2019-02-11 09:00:52,005 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:00:52,006 EPOCH 86 done: loss 0.0032 - lr 0.0125 - bad epochs 2\n",
            "2019-02-11 09:00:52,621 DEV  : loss 0.06594871 - f-score 0.7263 - acc 0.7263\n",
            "2019-02-11 09:00:56,830 TEST : loss 0.01349195 - f-score 0.9402 - acc 0.9402\n",
            "2019-02-11 09:00:56,833 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:00:57,109 epoch 87 - iter 0/27 - loss 0.00865449\n",
            "2019-02-11 09:00:57,824 epoch 87 - iter 2/27 - loss 0.00598404\n",
            "2019-02-11 09:00:58,592 epoch 87 - iter 4/27 - loss 0.00415733\n",
            "2019-02-11 09:00:59,291 epoch 87 - iter 6/27 - loss 0.00308834\n",
            "2019-02-11 09:01:00,083 epoch 87 - iter 8/27 - loss 0.00260806\n",
            "2019-02-11 09:01:00,898 epoch 87 - iter 10/27 - loss 0.00231955\n",
            "2019-02-11 09:01:01,614 epoch 87 - iter 12/27 - loss 0.00201843\n",
            "2019-02-11 09:01:02,756 epoch 87 - iter 14/27 - loss 0.00234925\n",
            "2019-02-11 09:01:03,728 epoch 87 - iter 16/27 - loss 0.00217630\n",
            "2019-02-11 09:01:04,565 epoch 87 - iter 18/27 - loss 0.00208406\n",
            "2019-02-11 09:01:05,245 epoch 87 - iter 20/27 - loss 0.00198769\n",
            "2019-02-11 09:01:06,426 epoch 87 - iter 22/27 - loss 0.00207482\n",
            "2019-02-11 09:01:07,153 epoch 87 - iter 24/27 - loss 0.00204084\n",
            "2019-02-11 09:01:07,560 epoch 87 - iter 26/27 - loss 0.00203184\n",
            "2019-02-11 09:01:07,572 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:01:07,573 EPOCH 87 done: loss 0.0020 - lr 0.0125 - bad epochs 3\n",
            "2019-02-11 09:01:08,193 DEV  : loss 0.06688053 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 09:01:12,428 TEST : loss 0.01361159 - f-score 0.9402 - acc 0.9402\n",
            "2019-02-11 09:01:29,628 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:01:30,096 epoch 88 - iter 0/27 - loss 0.00344812\n",
            "2019-02-11 09:01:31,027 epoch 88 - iter 2/27 - loss 0.00386316\n",
            "2019-02-11 09:01:31,714 epoch 88 - iter 4/27 - loss 0.00348529\n",
            "2019-02-11 09:01:32,502 epoch 88 - iter 6/27 - loss 0.00339972\n",
            "2019-02-11 09:01:33,249 epoch 88 - iter 8/27 - loss 0.00297062\n",
            "2019-02-11 09:01:34,850 epoch 88 - iter 10/27 - loss 0.00261768\n",
            "2019-02-11 09:01:35,746 epoch 88 - iter 12/27 - loss 0.00240822\n",
            "2019-02-11 09:01:36,425 epoch 88 - iter 14/27 - loss 0.00231334\n",
            "2019-02-11 09:01:37,125 epoch 88 - iter 16/27 - loss 0.00232674\n",
            "2019-02-11 09:01:37,638 epoch 88 - iter 18/27 - loss 0.00235311\n",
            "2019-02-11 09:01:38,321 epoch 88 - iter 20/27 - loss 0.00234321\n",
            "2019-02-11 09:01:39,056 epoch 88 - iter 22/27 - loss 0.00223359\n",
            "2019-02-11 09:01:39,778 epoch 88 - iter 24/27 - loss 0.00213159\n",
            "2019-02-11 09:01:40,385 epoch 88 - iter 26/27 - loss 0.00256596\n",
            "2019-02-11 09:01:40,396 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:01:40,397 EPOCH 88 done: loss 0.0026 - lr 0.0125 - bad epochs 0\n",
            "2019-02-11 09:01:41,022 DEV  : loss 0.06821528 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 09:01:45,286 TEST : loss 0.01390958 - f-score 0.9359 - acc 0.9359\n",
            "2019-02-11 09:01:45,289 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:01:45,695 epoch 89 - iter 0/27 - loss 0.00181545\n",
            "2019-02-11 09:01:46,587 epoch 89 - iter 2/27 - loss 0.00266651\n",
            "2019-02-11 09:01:47,549 epoch 89 - iter 4/27 - loss 0.00214420\n",
            "2019-02-11 09:01:48,398 epoch 89 - iter 6/27 - loss 0.00237396\n",
            "2019-02-11 09:01:49,159 epoch 89 - iter 8/27 - loss 0.00243380\n",
            "2019-02-11 09:01:49,810 epoch 89 - iter 10/27 - loss 0.00228897\n",
            "2019-02-11 09:01:50,860 epoch 89 - iter 12/27 - loss 0.00243527\n",
            "2019-02-11 09:01:51,476 epoch 89 - iter 14/27 - loss 0.00222240\n",
            "2019-02-11 09:01:52,191 epoch 89 - iter 16/27 - loss 0.00221066\n",
            "2019-02-11 09:01:52,953 epoch 89 - iter 18/27 - loss 0.00206442\n",
            "2019-02-11 09:01:53,640 epoch 89 - iter 20/27 - loss 0.00202775\n",
            "2019-02-11 09:01:54,488 epoch 89 - iter 22/27 - loss 0.00206438\n",
            "2019-02-11 09:01:55,634 epoch 89 - iter 24/27 - loss 0.00196013\n",
            "2019-02-11 09:01:56,060 epoch 89 - iter 26/27 - loss 0.00202671\n",
            "2019-02-11 09:01:56,072 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:01:56,073 EPOCH 89 done: loss 0.0020 - lr 0.0125 - bad epochs 1\n",
            "2019-02-11 09:01:56,691 DEV  : loss 0.06794561 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 09:02:00,919 TEST : loss 0.01387568 - f-score 0.9338 - acc 0.9338\n",
            "2019-02-11 09:02:18,126 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:02:18,640 epoch 90 - iter 0/27 - loss 0.00374507\n",
            "2019-02-11 09:02:19,418 epoch 90 - iter 2/27 - loss 0.00352298\n",
            "2019-02-11 09:02:20,230 epoch 90 - iter 4/27 - loss 0.00286375\n",
            "2019-02-11 09:02:21,093 epoch 90 - iter 6/27 - loss 0.00219284\n",
            "2019-02-11 09:02:21,836 epoch 90 - iter 8/27 - loss 0.00232949\n",
            "2019-02-11 09:02:22,561 epoch 90 - iter 10/27 - loss 0.00253277\n",
            "2019-02-11 09:02:23,900 epoch 90 - iter 12/27 - loss 0.00290917\n",
            "2019-02-11 09:02:24,696 epoch 90 - iter 14/27 - loss 0.00271211\n",
            "2019-02-11 09:02:25,703 epoch 90 - iter 16/27 - loss 0.00261605\n",
            "2019-02-11 09:02:26,377 epoch 90 - iter 18/27 - loss 0.00249141\n",
            "2019-02-11 09:02:27,083 epoch 90 - iter 20/27 - loss 0.00244122\n",
            "2019-02-11 09:02:27,850 epoch 90 - iter 22/27 - loss 0.00233831\n",
            "2019-02-11 09:02:28,409 epoch 90 - iter 24/27 - loss 0.00255342\n",
            "2019-02-11 09:02:28,930 epoch 90 - iter 26/27 - loss 0.00254486\n",
            "2019-02-11 09:02:28,941 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:02:28,942 EPOCH 90 done: loss 0.0025 - lr 0.0125 - bad epochs 0\n",
            "2019-02-11 09:02:29,568 DEV  : loss 0.06708584 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 09:02:33,882 TEST : loss 0.01369556 - f-score 0.9338 - acc 0.9338\n",
            "2019-02-11 09:02:33,885 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:02:34,184 epoch 91 - iter 0/27 - loss 0.00270159\n",
            "2019-02-11 09:02:34,911 epoch 91 - iter 2/27 - loss 0.00384749\n",
            "2019-02-11 09:02:35,524 epoch 91 - iter 4/27 - loss 0.00402304\n",
            "2019-02-11 09:02:36,213 epoch 91 - iter 6/27 - loss 0.00408542\n",
            "2019-02-11 09:02:37,485 epoch 91 - iter 8/27 - loss 0.00330274\n",
            "2019-02-11 09:02:38,498 epoch 91 - iter 10/27 - loss 0.00309570\n",
            "2019-02-11 09:02:39,090 epoch 91 - iter 12/27 - loss 0.00276894\n",
            "2019-02-11 09:02:39,923 epoch 91 - iter 14/27 - loss 0.00283486\n",
            "2019-02-11 09:02:40,719 epoch 91 - iter 16/27 - loss 0.00297499\n",
            "2019-02-11 09:02:41,415 epoch 91 - iter 18/27 - loss 0.00315461\n",
            "2019-02-11 09:02:42,010 epoch 91 - iter 20/27 - loss 0.00314006\n",
            "2019-02-11 09:02:42,598 epoch 91 - iter 22/27 - loss 0.00309292\n",
            "2019-02-11 09:02:43,375 epoch 91 - iter 24/27 - loss 0.00305887\n",
            "2019-02-11 09:02:44,305 epoch 91 - iter 26/27 - loss 0.00343220\n",
            "2019-02-11 09:02:44,317 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:02:44,318 EPOCH 91 done: loss 0.0034 - lr 0.0125 - bad epochs 1\n",
            "2019-02-11 09:02:44,950 DEV  : loss 0.06719822 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 09:02:49,280 TEST : loss 0.01370438 - f-score 0.9359 - acc 0.9359\n",
            "2019-02-11 09:02:49,283 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:02:49,563 epoch 92 - iter 0/27 - loss 0.00245391\n",
            "2019-02-11 09:02:50,866 epoch 92 - iter 2/27 - loss 0.00249273\n",
            "2019-02-11 09:02:51,671 epoch 92 - iter 4/27 - loss 0.00224552\n",
            "2019-02-11 09:02:52,374 epoch 92 - iter 6/27 - loss 0.00203020\n",
            "2019-02-11 09:02:52,923 epoch 92 - iter 8/27 - loss 0.00172865\n",
            "2019-02-11 09:02:53,675 epoch 92 - iter 10/27 - loss 0.00170629\n",
            "2019-02-11 09:02:54,657 epoch 92 - iter 12/27 - loss 0.00177702\n",
            "2019-02-11 09:02:55,352 epoch 92 - iter 14/27 - loss 0.00171895\n",
            "2019-02-11 09:02:56,058 epoch 92 - iter 16/27 - loss 0.00175028\n",
            "2019-02-11 09:02:56,710 epoch 92 - iter 18/27 - loss 0.00193748\n",
            "2019-02-11 09:02:57,373 epoch 92 - iter 20/27 - loss 0.00212787\n",
            "2019-02-11 09:02:58,108 epoch 92 - iter 22/27 - loss 0.00230302\n",
            "2019-02-11 09:02:58,753 epoch 92 - iter 24/27 - loss 0.00220544\n",
            "2019-02-11 09:02:59,775 epoch 92 - iter 26/27 - loss 0.00224810\n",
            "2019-02-11 09:02:59,788 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:02:59,789 EPOCH 92 done: loss 0.0022 - lr 0.0125 - bad epochs 2\n",
            "2019-02-11 09:03:00,414 DEV  : loss 0.06671824 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 09:03:04,719 TEST : loss 0.01362423 - f-score 0.9380 - acc 0.9380\n",
            "2019-02-11 09:03:04,721 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:03:05,134 epoch 93 - iter 0/27 - loss 0.00211303\n",
            "2019-02-11 09:03:06,171 epoch 93 - iter 2/27 - loss 0.00180406\n",
            "2019-02-11 09:03:07,449 epoch 93 - iter 4/27 - loss 0.00222054\n",
            "2019-02-11 09:03:08,071 epoch 93 - iter 6/27 - loss 0.00191495\n",
            "2019-02-11 09:03:08,725 epoch 93 - iter 8/27 - loss 0.00220340\n",
            "2019-02-11 09:03:09,574 epoch 93 - iter 10/27 - loss 0.00204152\n",
            "2019-02-11 09:03:10,222 epoch 93 - iter 12/27 - loss 0.00244825\n",
            "2019-02-11 09:03:10,855 epoch 93 - iter 14/27 - loss 0.00222119\n",
            "2019-02-11 09:03:11,542 epoch 93 - iter 16/27 - loss 0.00204943\n",
            "2019-02-11 09:03:12,321 epoch 93 - iter 18/27 - loss 0.00187250\n",
            "2019-02-11 09:03:13,318 epoch 93 - iter 20/27 - loss 0.00212465\n",
            "2019-02-11 09:03:13,950 epoch 93 - iter 22/27 - loss 0.00199092\n",
            "2019-02-11 09:03:14,691 epoch 93 - iter 24/27 - loss 0.00230097\n",
            "2019-02-11 09:03:15,128 epoch 93 - iter 26/27 - loss 0.00238060\n",
            "2019-02-11 09:03:15,140 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:03:15,141 EPOCH 93 done: loss 0.0024 - lr 0.0125 - bad epochs 3\n",
            "2019-02-11 09:03:15,770 DEV  : loss 0.06841853 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 09:03:20,096 TEST : loss 0.01397503 - f-score 0.9316 - acc 0.9316\n",
            "Epoch    92: reducing learning rate of group 0 to 6.2500e-03.\n",
            "2019-02-11 09:03:20,099 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:03:20,507 epoch 94 - iter 0/27 - loss 0.00066582\n",
            "2019-02-11 09:03:21,206 epoch 94 - iter 2/27 - loss 0.00187245\n",
            "2019-02-11 09:03:22,017 epoch 94 - iter 4/27 - loss 0.00347639\n",
            "2019-02-11 09:03:23,469 epoch 94 - iter 6/27 - loss 0.00285241\n",
            "2019-02-11 09:03:24,393 epoch 94 - iter 8/27 - loss 0.00229194\n",
            "2019-02-11 09:03:25,089 epoch 94 - iter 10/27 - loss 0.00249115\n",
            "2019-02-11 09:03:25,649 epoch 94 - iter 12/27 - loss 0.00226044\n",
            "2019-02-11 09:03:26,523 epoch 94 - iter 14/27 - loss 0.00232511\n",
            "2019-02-11 09:03:27,200 epoch 94 - iter 16/27 - loss 0.00233825\n",
            "2019-02-11 09:03:27,802 epoch 94 - iter 18/27 - loss 0.00217548\n",
            "2019-02-11 09:03:28,519 epoch 94 - iter 20/27 - loss 0.00203058\n",
            "2019-02-11 09:03:29,566 epoch 94 - iter 22/27 - loss 0.00205019\n",
            "2019-02-11 09:03:30,246 epoch 94 - iter 24/27 - loss 0.00209812\n",
            "2019-02-11 09:03:30,748 epoch 94 - iter 26/27 - loss 0.00219351\n",
            "2019-02-11 09:03:30,761 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:03:30,762 EPOCH 94 done: loss 0.0022 - lr 0.0063 - bad epochs 0\n",
            "2019-02-11 09:03:31,451 DEV  : loss 0.06894438 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 09:03:36,058 TEST : loss 0.01409057 - f-score 0.9338 - acc 0.9338\n",
            "2019-02-11 09:03:36,061 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:03:36,404 epoch 95 - iter 0/27 - loss 0.00171957\n",
            "2019-02-11 09:03:37,639 epoch 95 - iter 2/27 - loss 0.00550379\n",
            "2019-02-11 09:03:38,397 epoch 95 - iter 4/27 - loss 0.00445672\n",
            "2019-02-11 09:03:39,098 epoch 95 - iter 6/27 - loss 0.00383145\n",
            "2019-02-11 09:03:39,862 epoch 95 - iter 8/27 - loss 0.00325116\n",
            "2019-02-11 09:03:40,848 epoch 95 - iter 10/27 - loss 0.00284425\n",
            "2019-02-11 09:03:42,013 epoch 95 - iter 12/27 - loss 0.00269778\n",
            "2019-02-11 09:03:42,723 epoch 95 - iter 14/27 - loss 0.00256076\n",
            "2019-02-11 09:03:43,631 epoch 95 - iter 16/27 - loss 0.00238708\n",
            "2019-02-11 09:03:44,377 epoch 95 - iter 18/27 - loss 0.00258136\n",
            "2019-02-11 09:03:45,219 epoch 95 - iter 20/27 - loss 0.00263756\n",
            "2019-02-11 09:03:45,754 epoch 95 - iter 22/27 - loss 0.00293866\n",
            "2019-02-11 09:03:46,422 epoch 95 - iter 24/27 - loss 0.00304320\n",
            "2019-02-11 09:03:46,852 epoch 95 - iter 26/27 - loss 0.00332339\n",
            "2019-02-11 09:03:46,863 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:03:46,864 EPOCH 95 done: loss 0.0033 - lr 0.0063 - bad epochs 1\n",
            "2019-02-11 09:03:47,491 DEV  : loss 0.06834665 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 09:03:51,815 TEST : loss 0.01397350 - f-score 0.9338 - acc 0.9338\n",
            "2019-02-11 09:03:51,818 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:03:52,147 epoch 96 - iter 0/27 - loss 0.00015204\n",
            "2019-02-11 09:03:52,855 epoch 96 - iter 2/27 - loss 0.00074835\n",
            "2019-02-11 09:03:53,543 epoch 96 - iter 4/27 - loss 0.00175340\n",
            "2019-02-11 09:03:54,349 epoch 96 - iter 6/27 - loss 0.00183548\n",
            "2019-02-11 09:03:55,065 epoch 96 - iter 8/27 - loss 0.00272111\n",
            "2019-02-11 09:03:56,226 epoch 96 - iter 10/27 - loss 0.00234023\n",
            "2019-02-11 09:03:56,650 epoch 96 - iter 12/27 - loss 0.00291075\n",
            "2019-02-11 09:03:57,567 epoch 96 - iter 14/27 - loss 0.00270145\n",
            "2019-02-11 09:03:58,503 epoch 96 - iter 16/27 - loss 0.00259088\n",
            "2019-02-11 09:03:59,309 epoch 96 - iter 18/27 - loss 0.00270871\n",
            "2019-02-11 09:04:00,088 epoch 96 - iter 20/27 - loss 0.00264427\n",
            "2019-02-11 09:04:01,166 epoch 96 - iter 22/27 - loss 0.00251171\n",
            "2019-02-11 09:04:01,992 epoch 96 - iter 24/27 - loss 0.00260543\n",
            "2019-02-11 09:04:02,444 epoch 96 - iter 26/27 - loss 0.00298583\n",
            "2019-02-11 09:04:02,455 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:04:02,456 EPOCH 96 done: loss 0.0030 - lr 0.0063 - bad epochs 2\n",
            "2019-02-11 09:04:03,091 DEV  : loss 0.06724418 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 09:04:07,406 TEST : loss 0.01374423 - f-score 0.9380 - acc 0.9380\n",
            "2019-02-11 09:04:07,409 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:04:07,772 epoch 97 - iter 0/27 - loss 0.00430180\n",
            "2019-02-11 09:04:08,706 epoch 97 - iter 2/27 - loss 0.00266114\n",
            "2019-02-11 09:04:09,454 epoch 97 - iter 4/27 - loss 0.00326384\n",
            "2019-02-11 09:04:10,176 epoch 97 - iter 6/27 - loss 0.00275268\n",
            "2019-02-11 09:04:10,898 epoch 97 - iter 8/27 - loss 0.00265520\n",
            "2019-02-11 09:04:11,621 epoch 97 - iter 10/27 - loss 0.00332206\n",
            "2019-02-11 09:04:12,798 epoch 97 - iter 12/27 - loss 0.00314588\n",
            "2019-02-11 09:04:13,694 epoch 97 - iter 14/27 - loss 0.00286135\n",
            "2019-02-11 09:04:14,426 epoch 97 - iter 16/27 - loss 0.00283886\n",
            "2019-02-11 09:04:15,534 epoch 97 - iter 18/27 - loss 0.00313078\n",
            "2019-02-11 09:04:16,190 epoch 97 - iter 20/27 - loss 0.00309823\n",
            "2019-02-11 09:04:16,857 epoch 97 - iter 22/27 - loss 0.00304196\n",
            "2019-02-11 09:04:17,515 epoch 97 - iter 24/27 - loss 0.00307302\n",
            "2019-02-11 09:04:18,033 epoch 97 - iter 26/27 - loss 0.00308156\n",
            "2019-02-11 09:04:18,047 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:04:18,048 EPOCH 97 done: loss 0.0031 - lr 0.0063 - bad epochs 3\n",
            "2019-02-11 09:04:18,672 DEV  : loss 0.06744277 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 09:04:23,086 TEST : loss 0.01378584 - f-score 0.9380 - acc 0.9380\n",
            "Epoch    96: reducing learning rate of group 0 to 3.1250e-03.\n",
            "2019-02-11 09:04:23,090 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:04:23,467 epoch 98 - iter 0/27 - loss 0.00157466\n",
            "2019-02-11 09:04:24,130 epoch 98 - iter 2/27 - loss 0.00285178\n",
            "2019-02-11 09:04:24,978 epoch 98 - iter 4/27 - loss 0.00222226\n",
            "2019-02-11 09:04:25,769 epoch 98 - iter 6/27 - loss 0.00195478\n",
            "2019-02-11 09:04:26,837 epoch 98 - iter 8/27 - loss 0.00184494\n",
            "2019-02-11 09:04:27,685 epoch 98 - iter 10/27 - loss 0.00192846\n",
            "2019-02-11 09:04:28,844 epoch 98 - iter 12/27 - loss 0.00192210\n",
            "2019-02-11 09:04:29,430 epoch 98 - iter 14/27 - loss 0.00194459\n",
            "2019-02-11 09:04:30,213 epoch 98 - iter 16/27 - loss 0.00224190\n",
            "2019-02-11 09:04:30,966 epoch 98 - iter 18/27 - loss 0.00232292\n",
            "2019-02-11 09:04:31,722 epoch 98 - iter 20/27 - loss 0.00238281\n",
            "2019-02-11 09:04:32,642 epoch 98 - iter 22/27 - loss 0.00228780\n",
            "2019-02-11 09:04:33,231 epoch 98 - iter 24/27 - loss 0.00218456\n",
            "2019-02-11 09:04:33,621 epoch 98 - iter 26/27 - loss 0.00226618\n",
            "2019-02-11 09:04:33,632 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:04:33,634 EPOCH 98 done: loss 0.0023 - lr 0.0031 - bad epochs 0\n",
            "2019-02-11 09:04:34,256 DEV  : loss 0.06761997 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 09:04:38,539 TEST : loss 0.01382346 - f-score 0.9359 - acc 0.9359\n",
            "2019-02-11 09:04:38,542 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:04:39,143 epoch 99 - iter 0/27 - loss 0.00695707\n",
            "2019-02-11 09:04:40,294 epoch 99 - iter 2/27 - loss 0.00435998\n",
            "2019-02-11 09:04:41,150 epoch 99 - iter 4/27 - loss 0.00432896\n",
            "2019-02-11 09:04:41,894 epoch 99 - iter 6/27 - loss 0.00402090\n",
            "2019-02-11 09:04:42,637 epoch 99 - iter 8/27 - loss 0.00322257\n",
            "2019-02-11 09:04:43,301 epoch 99 - iter 10/27 - loss 0.00366061\n",
            "2019-02-11 09:04:44,431 epoch 99 - iter 12/27 - loss 0.00341180\n",
            "2019-02-11 09:04:45,174 epoch 99 - iter 14/27 - loss 0.00302788\n",
            "2019-02-11 09:04:45,879 epoch 99 - iter 16/27 - loss 0.00271840\n",
            "2019-02-11 09:04:46,558 epoch 99 - iter 18/27 - loss 0.00273873\n",
            "2019-02-11 09:04:47,248 epoch 99 - iter 20/27 - loss 0.00262249\n",
            "2019-02-11 09:04:47,979 epoch 99 - iter 22/27 - loss 0.00252999\n",
            "2019-02-11 09:04:48,500 epoch 99 - iter 24/27 - loss 0.00245025\n",
            "2019-02-11 09:04:49,032 epoch 99 - iter 26/27 - loss 0.00279678\n",
            "2019-02-11 09:04:49,044 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:04:49,046 EPOCH 99 done: loss 0.0028 - lr 0.0031 - bad epochs 1\n",
            "2019-02-11 09:04:49,676 DEV  : loss 0.06755654 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 09:04:54,827 TEST : loss 0.01381321 - f-score 0.9359 - acc 0.9359\n",
            "2019-02-11 09:04:54,830 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:04:55,168 epoch 100 - iter 0/27 - loss 0.00439151\n",
            "2019-02-11 09:04:55,972 epoch 100 - iter 2/27 - loss 0.00199476\n",
            "2019-02-11 09:04:57,199 epoch 100 - iter 4/27 - loss 0.00155565\n",
            "2019-02-11 09:04:58,094 epoch 100 - iter 6/27 - loss 0.00149910\n",
            "2019-02-11 09:04:58,749 epoch 100 - iter 8/27 - loss 0.00225149\n",
            "2019-02-11 09:04:59,303 epoch 100 - iter 10/27 - loss 0.00242141\n",
            "2019-02-11 09:05:00,560 epoch 100 - iter 12/27 - loss 0.00308854\n",
            "2019-02-11 09:05:01,484 epoch 100 - iter 14/27 - loss 0.00314100\n",
            "2019-02-11 09:05:02,120 epoch 100 - iter 16/27 - loss 0.00287295\n",
            "2019-02-11 09:05:02,935 epoch 100 - iter 18/27 - loss 0.00285684\n",
            "2019-02-11 09:05:03,638 epoch 100 - iter 20/27 - loss 0.00273982\n",
            "2019-02-11 09:05:04,271 epoch 100 - iter 22/27 - loss 0.00255443\n",
            "2019-02-11 09:05:05,021 epoch 100 - iter 24/27 - loss 0.00243900\n",
            "2019-02-11 09:05:05,441 epoch 100 - iter 26/27 - loss 0.00233858\n",
            "2019-02-11 09:05:05,452 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:05:05,453 EPOCH 100 done: loss 0.0023 - lr 0.0031 - bad epochs 2\n",
            "2019-02-11 09:05:06,076 DEV  : loss 0.06756706 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 09:05:10,423 TEST : loss 0.01383181 - f-score 0.9359 - acc 0.9359\n",
            "2019-02-11 09:05:27,380 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:05:27,382 Testing using best model ...\n",
            "2019-02-11 09:05:43,613 MICRO_AVG: acc 0.9338 - f1-score 0.9338\n",
            "2019-02-11 09:05:43,614 MARCO_AVG: acc 0.9331 - f1-score 0.9357\n",
            "2019-02-11 09:05:43,618 0          tp: 195 - fp: 10 - fn: 21 - tn: 195 - precision: 0.9512 - recall: 0.9028 - accuracy: 0.9264 - f1-score: 0.9264\n",
            "2019-02-11 09:05:43,621 1          tp: 242 - fp: 21 - fn: 10 - tn: 242 - precision: 0.9202 - recall: 0.9603 - accuracy: 0.9398 - f1-score: 0.9398\n",
            "2019-02-11 09:05:43,624 ----------------------------------------------------------------------------------------------------\n",
            "0.936 0.924 0.96 0.936 0.935 0.942 0.934\n",
            "839 95 468\n",
            "2019-02-11 09:05:48,330 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:05:48,334 Evaluation method: MICRO_F1_SCORE\n",
            "2019-02-11 09:05:48,336 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:05:48,776 epoch 1 - iter 0/27 - loss 0.02496621\n",
            "2019-02-11 09:05:49,592 epoch 1 - iter 2/27 - loss 0.07010529\n",
            "2019-02-11 09:05:50,863 epoch 1 - iter 4/27 - loss 0.06114644\n",
            "2019-02-11 09:05:51,571 epoch 1 - iter 6/27 - loss 0.06317506\n",
            "2019-02-11 09:05:52,353 epoch 1 - iter 8/27 - loss 0.06515024\n",
            "2019-02-11 09:05:52,880 epoch 1 - iter 10/27 - loss 0.06557029\n",
            "2019-02-11 09:05:53,574 epoch 1 - iter 12/27 - loss 0.06402648\n",
            "2019-02-11 09:05:54,318 epoch 1 - iter 14/27 - loss 0.06265694\n",
            "2019-02-11 09:05:54,983 epoch 1 - iter 16/27 - loss 0.06378586\n",
            "2019-02-11 09:05:55,705 epoch 1 - iter 18/27 - loss 0.06141897\n",
            "2019-02-11 09:05:56,624 epoch 1 - iter 20/27 - loss 0.05939550\n",
            "2019-02-11 09:05:57,703 epoch 1 - iter 22/27 - loss 0.05747131\n",
            "2019-02-11 09:05:58,423 epoch 1 - iter 24/27 - loss 0.05603466\n",
            "2019-02-11 09:05:59,102 epoch 1 - iter 26/27 - loss 0.05718038\n",
            "2019-02-11 09:05:59,114 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:05:59,115 EPOCH 1 done: loss 0.0572 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:05:59,764 DEV  : loss 0.02654217 - f-score 0.3368 - acc 0.3368\n",
            "2019-02-11 09:06:03,692 TEST : loss 0.02416610 - f-score 0.4594 - acc 0.4594\n",
            "2019-02-11 09:06:21,388 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:06:21,900 epoch 2 - iter 0/27 - loss 0.02811057\n",
            "2019-02-11 09:06:22,630 epoch 2 - iter 2/27 - loss 0.03485349\n",
            "2019-02-11 09:06:23,350 epoch 2 - iter 4/27 - loss 0.03053838\n",
            "2019-02-11 09:06:24,942 epoch 2 - iter 6/27 - loss 0.03363122\n",
            "2019-02-11 09:06:25,664 epoch 2 - iter 8/27 - loss 0.03330232\n",
            "2019-02-11 09:06:26,612 epoch 2 - iter 10/27 - loss 0.03208366\n",
            "2019-02-11 09:06:27,252 epoch 2 - iter 12/27 - loss 0.03078087\n",
            "2019-02-11 09:06:27,871 epoch 2 - iter 14/27 - loss 0.03191982\n",
            "2019-02-11 09:06:28,882 epoch 2 - iter 16/27 - loss 0.03095900\n",
            "2019-02-11 09:06:29,815 epoch 2 - iter 18/27 - loss 0.03175867\n",
            "2019-02-11 09:06:30,531 epoch 2 - iter 20/27 - loss 0.03181662\n",
            "2019-02-11 09:06:31,187 epoch 2 - iter 22/27 - loss 0.03173433\n",
            "2019-02-11 09:06:31,980 epoch 2 - iter 24/27 - loss 0.03153351\n",
            "2019-02-11 09:06:32,375 epoch 2 - iter 26/27 - loss 0.03173006\n",
            "2019-02-11 09:06:32,389 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:06:32,390 EPOCH 2 done: loss 0.0317 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:06:33,033 DEV  : loss 0.02054595 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 09:06:36,739 TEST : loss 0.02533036 - f-score 0.5385 - acc 0.5385\n",
            "2019-02-11 09:06:54,022 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:06:55,074 epoch 3 - iter 0/27 - loss 0.02720103\n",
            "2019-02-11 09:06:55,851 epoch 3 - iter 2/27 - loss 0.03590266\n",
            "2019-02-11 09:06:56,726 epoch 3 - iter 4/27 - loss 0.03336892\n",
            "2019-02-11 09:06:57,436 epoch 3 - iter 6/27 - loss 0.03239771\n",
            "2019-02-11 09:06:58,649 epoch 3 - iter 8/27 - loss 0.03024590\n",
            "2019-02-11 09:06:59,409 epoch 3 - iter 10/27 - loss 0.02976412\n",
            "2019-02-11 09:07:00,001 epoch 3 - iter 12/27 - loss 0.02857572\n",
            "2019-02-11 09:07:00,924 epoch 3 - iter 14/27 - loss 0.02771817\n",
            "2019-02-11 09:07:01,605 epoch 3 - iter 16/27 - loss 0.02796333\n",
            "2019-02-11 09:07:02,408 epoch 3 - iter 18/27 - loss 0.02816957\n",
            "2019-02-11 09:07:02,971 epoch 3 - iter 20/27 - loss 0.02754134\n",
            "2019-02-11 09:07:04,104 epoch 3 - iter 22/27 - loss 0.02786350\n",
            "2019-02-11 09:07:04,753 epoch 3 - iter 24/27 - loss 0.02783432\n",
            "2019-02-11 09:07:05,218 epoch 3 - iter 26/27 - loss 0.02862690\n",
            "2019-02-11 09:07:05,230 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:07:05,231 EPOCH 3 done: loss 0.0286 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:07:05,875 DEV  : loss 0.01990434 - f-score 0.6105 - acc 0.6105\n",
            "2019-02-11 09:07:09,549 TEST : loss 0.02264152 - f-score 0.5812 - acc 0.5812\n",
            "2019-02-11 09:07:26,907 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:07:27,936 epoch 4 - iter 0/27 - loss 0.02435673\n",
            "2019-02-11 09:07:29,261 epoch 4 - iter 2/27 - loss 0.02469709\n",
            "2019-02-11 09:07:30,006 epoch 4 - iter 4/27 - loss 0.02786501\n",
            "2019-02-11 09:07:30,769 epoch 4 - iter 6/27 - loss 0.02726408\n",
            "2019-02-11 09:07:31,334 epoch 4 - iter 8/27 - loss 0.02690460\n",
            "2019-02-11 09:07:32,019 epoch 4 - iter 10/27 - loss 0.02750946\n",
            "2019-02-11 09:07:32,814 epoch 4 - iter 12/27 - loss 0.02686737\n",
            "2019-02-11 09:07:33,459 epoch 4 - iter 14/27 - loss 0.02613316\n",
            "2019-02-11 09:07:34,330 epoch 4 - iter 16/27 - loss 0.02532951\n",
            "2019-02-11 09:07:35,602 epoch 4 - iter 18/27 - loss 0.02497849\n",
            "2019-02-11 09:07:36,115 epoch 4 - iter 20/27 - loss 0.02532624\n",
            "2019-02-11 09:07:36,873 epoch 4 - iter 22/27 - loss 0.02496719\n",
            "2019-02-11 09:07:37,514 epoch 4 - iter 24/27 - loss 0.02497485\n",
            "2019-02-11 09:07:38,028 epoch 4 - iter 26/27 - loss 0.02625114\n",
            "2019-02-11 09:07:38,039 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:07:38,040 EPOCH 4 done: loss 0.0263 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:07:38,667 DEV  : loss 0.02775835 - f-score 0.5474 - acc 0.5474\n",
            "2019-02-11 09:07:42,463 TEST : loss 0.02623390 - f-score 0.5833 - acc 0.5833\n",
            "2019-02-11 09:07:59,828 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:08:00,350 epoch 5 - iter 0/27 - loss 0.02283942\n",
            "2019-02-11 09:08:01,683 epoch 5 - iter 2/27 - loss 0.02745639\n",
            "2019-02-11 09:08:02,431 epoch 5 - iter 4/27 - loss 0.02923805\n",
            "2019-02-11 09:08:03,333 epoch 5 - iter 6/27 - loss 0.02736958\n",
            "2019-02-11 09:08:04,086 epoch 5 - iter 8/27 - loss 0.02655009\n",
            "2019-02-11 09:08:04,752 epoch 5 - iter 10/27 - loss 0.02641313\n",
            "2019-02-11 09:08:05,716 epoch 5 - iter 12/27 - loss 0.02514618\n",
            "2019-02-11 09:08:06,376 epoch 5 - iter 14/27 - loss 0.02560251\n",
            "2019-02-11 09:08:07,093 epoch 5 - iter 16/27 - loss 0.02756771\n",
            "2019-02-11 09:08:07,673 epoch 5 - iter 18/27 - loss 0.02739925\n",
            "2019-02-11 09:08:08,295 epoch 5 - iter 20/27 - loss 0.02711753\n",
            "2019-02-11 09:08:09,432 epoch 5 - iter 22/27 - loss 0.02752313\n",
            "2019-02-11 09:08:10,332 epoch 5 - iter 24/27 - loss 0.02714558\n",
            "2019-02-11 09:08:10,879 epoch 5 - iter 26/27 - loss 0.02757683\n",
            "2019-02-11 09:08:10,889 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:08:10,890 EPOCH 5 done: loss 0.0276 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:08:11,527 DEV  : loss 0.02580723 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 09:08:15,213 TEST : loss 0.03385265 - f-score 0.5385 - acc 0.5385\n",
            "2019-02-11 09:08:15,216 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:08:15,605 epoch 6 - iter 0/27 - loss 0.03538768\n",
            "2019-02-11 09:08:16,412 epoch 6 - iter 2/27 - loss 0.03202225\n",
            "2019-02-11 09:08:17,003 epoch 6 - iter 4/27 - loss 0.02709737\n",
            "2019-02-11 09:08:17,626 epoch 6 - iter 6/27 - loss 0.02553298\n",
            "2019-02-11 09:08:18,522 epoch 6 - iter 8/27 - loss 0.02592263\n",
            "2019-02-11 09:08:19,326 epoch 6 - iter 10/27 - loss 0.02798195\n",
            "2019-02-11 09:08:19,985 epoch 6 - iter 12/27 - loss 0.02735014\n",
            "2019-02-11 09:08:21,163 epoch 6 - iter 14/27 - loss 0.02680315\n",
            "2019-02-11 09:08:22,260 epoch 6 - iter 16/27 - loss 0.02668239\n",
            "2019-02-11 09:08:22,977 epoch 6 - iter 18/27 - loss 0.02674789\n",
            "2019-02-11 09:08:23,918 epoch 6 - iter 20/27 - loss 0.02684002\n",
            "2019-02-11 09:08:24,754 epoch 6 - iter 22/27 - loss 0.02720761\n",
            "2019-02-11 09:08:25,406 epoch 6 - iter 24/27 - loss 0.02701312\n",
            "2019-02-11 09:08:25,832 epoch 6 - iter 26/27 - loss 0.02777198\n",
            "2019-02-11 09:08:25,844 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:08:25,845 EPOCH 6 done: loss 0.0278 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 09:08:26,483 DEV  : loss 0.02718912 - f-score 0.5053 - acc 0.5053\n",
            "2019-02-11 09:08:30,340 TEST : loss 0.02479745 - f-score 0.5470 - acc 0.5470\n",
            "2019-02-11 09:08:30,344 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:08:30,814 epoch 7 - iter 0/27 - loss 0.02360817\n",
            "2019-02-11 09:08:31,451 epoch 7 - iter 2/27 - loss 0.02341867\n",
            "2019-02-11 09:08:32,114 epoch 7 - iter 4/27 - loss 0.02344742\n",
            "2019-02-11 09:08:32,823 epoch 7 - iter 6/27 - loss 0.02288214\n",
            "2019-02-11 09:08:33,720 epoch 7 - iter 8/27 - loss 0.02208739\n",
            "2019-02-11 09:08:34,896 epoch 7 - iter 10/27 - loss 0.02138960\n",
            "2019-02-11 09:08:35,761 epoch 7 - iter 12/27 - loss 0.02295814\n",
            "2019-02-11 09:08:36,950 epoch 7 - iter 14/27 - loss 0.02369287\n",
            "2019-02-11 09:08:37,556 epoch 7 - iter 16/27 - loss 0.02472325\n",
            "2019-02-11 09:08:38,450 epoch 7 - iter 18/27 - loss 0.02504610\n",
            "2019-02-11 09:08:39,300 epoch 7 - iter 20/27 - loss 0.02495200\n",
            "2019-02-11 09:08:40,063 epoch 7 - iter 22/27 - loss 0.02455030\n",
            "2019-02-11 09:08:40,809 epoch 7 - iter 24/27 - loss 0.02445008\n",
            "2019-02-11 09:08:41,308 epoch 7 - iter 26/27 - loss 0.02571463\n",
            "2019-02-11 09:08:41,320 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:08:41,321 EPOCH 7 done: loss 0.0257 - lr 0.1000 - bad epochs 2\n",
            "2019-02-11 09:08:41,969 DEV  : loss 0.02779440 - f-score 0.5579 - acc 0.5579\n",
            "2019-02-11 09:08:45,631 TEST : loss 0.02557339 - f-score 0.5919 - acc 0.5919\n",
            "2019-02-11 09:09:02,832 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:09:03,315 epoch 8 - iter 0/27 - loss 0.02262785\n",
            "2019-02-11 09:09:04,115 epoch 8 - iter 2/27 - loss 0.02425937\n",
            "2019-02-11 09:09:04,733 epoch 8 - iter 4/27 - loss 0.02505806\n",
            "2019-02-11 09:09:05,573 epoch 8 - iter 6/27 - loss 0.02654816\n",
            "2019-02-11 09:09:06,275 epoch 8 - iter 8/27 - loss 0.02889123\n",
            "2019-02-11 09:09:06,980 epoch 8 - iter 10/27 - loss 0.02757066\n",
            "2019-02-11 09:09:07,735 epoch 8 - iter 12/27 - loss 0.02836785\n",
            "2019-02-11 09:09:08,442 epoch 8 - iter 14/27 - loss 0.02753526\n",
            "2019-02-11 09:09:09,599 epoch 8 - iter 16/27 - loss 0.02662532\n",
            "2019-02-11 09:09:10,430 epoch 8 - iter 18/27 - loss 0.02680213\n",
            "2019-02-11 09:09:11,118 epoch 8 - iter 20/27 - loss 0.02711672\n",
            "2019-02-11 09:09:11,884 epoch 8 - iter 22/27 - loss 0.02660383\n",
            "2019-02-11 09:09:13,308 epoch 8 - iter 24/27 - loss 0.02611284\n",
            "2019-02-11 09:09:13,579 epoch 8 - iter 26/27 - loss 0.02654680\n",
            "2019-02-11 09:09:13,590 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:09:13,591 EPOCH 8 done: loss 0.0265 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:09:14,226 DEV  : loss 0.02168892 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 09:09:17,973 TEST : loss 0.02553855 - f-score 0.6090 - acc 0.6090\n",
            "2019-02-11 09:09:17,976 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:09:18,359 epoch 9 - iter 0/27 - loss 0.02402565\n",
            "2019-02-11 09:09:19,202 epoch 9 - iter 2/27 - loss 0.03103104\n",
            "2019-02-11 09:09:20,059 epoch 9 - iter 4/27 - loss 0.02680983\n",
            "2019-02-11 09:09:20,631 epoch 9 - iter 6/27 - loss 0.02530889\n",
            "2019-02-11 09:09:21,524 epoch 9 - iter 8/27 - loss 0.02419553\n",
            "2019-02-11 09:09:22,294 epoch 9 - iter 10/27 - loss 0.02383218\n",
            "2019-02-11 09:09:22,977 epoch 9 - iter 12/27 - loss 0.02529032\n",
            "2019-02-11 09:09:24,161 epoch 9 - iter 14/27 - loss 0.02564087\n",
            "2019-02-11 09:09:24,835 epoch 9 - iter 16/27 - loss 0.02539272\n",
            "2019-02-11 09:09:25,573 epoch 9 - iter 18/27 - loss 0.02513257\n",
            "2019-02-11 09:09:26,758 epoch 9 - iter 20/27 - loss 0.02663252\n",
            "2019-02-11 09:09:27,427 epoch 9 - iter 22/27 - loss 0.02649825\n",
            "2019-02-11 09:09:28,342 epoch 9 - iter 24/27 - loss 0.02604511\n",
            "2019-02-11 09:09:28,820 epoch 9 - iter 26/27 - loss 0.02654430\n",
            "2019-02-11 09:09:28,831 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:09:28,833 EPOCH 9 done: loss 0.0265 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 09:09:29,479 DEV  : loss 0.02541904 - f-score 0.5579 - acc 0.5579\n",
            "2019-02-11 09:09:33,170 TEST : loss 0.02455773 - f-score 0.6026 - acc 0.6026\n",
            "2019-02-11 09:09:33,173 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:09:33,565 epoch 10 - iter 0/27 - loss 0.02651311\n",
            "2019-02-11 09:09:34,103 epoch 10 - iter 2/27 - loss 0.03462915\n",
            "2019-02-11 09:09:35,386 epoch 10 - iter 4/27 - loss 0.03219842\n",
            "2019-02-11 09:09:36,114 epoch 10 - iter 6/27 - loss 0.03008955\n",
            "2019-02-11 09:09:36,761 epoch 10 - iter 8/27 - loss 0.02938077\n",
            "2019-02-11 09:09:37,614 epoch 10 - iter 10/27 - loss 0.02805084\n",
            "2019-02-11 09:09:38,438 epoch 10 - iter 12/27 - loss 0.02739163\n",
            "2019-02-11 09:09:39,292 epoch 10 - iter 14/27 - loss 0.02706901\n",
            "2019-02-11 09:09:40,139 epoch 10 - iter 16/27 - loss 0.02824403\n",
            "2019-02-11 09:09:40,758 epoch 10 - iter 18/27 - loss 0.02792300\n",
            "2019-02-11 09:09:41,644 epoch 10 - iter 20/27 - loss 0.02726412\n",
            "2019-02-11 09:09:42,391 epoch 10 - iter 22/27 - loss 0.02671585\n",
            "2019-02-11 09:09:42,960 epoch 10 - iter 24/27 - loss 0.02703429\n",
            "2019-02-11 09:09:43,320 epoch 10 - iter 26/27 - loss 0.02720924\n",
            "2019-02-11 09:09:43,331 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:09:43,332 EPOCH 10 done: loss 0.0272 - lr 0.1000 - bad epochs 2\n",
            "2019-02-11 09:09:43,973 DEV  : loss 0.02539234 - f-score 0.5579 - acc 0.5579\n",
            "2019-02-11 09:09:47,634 TEST : loss 0.02730408 - f-score 0.6154 - acc 0.6154\n",
            "2019-02-11 09:09:47,637 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:09:47,988 epoch 11 - iter 0/27 - loss 0.02627794\n",
            "2019-02-11 09:09:48,893 epoch 11 - iter 2/27 - loss 0.02725339\n",
            "2019-02-11 09:09:50,016 epoch 11 - iter 4/27 - loss 0.02816862\n",
            "2019-02-11 09:09:50,809 epoch 11 - iter 6/27 - loss 0.02749806\n",
            "2019-02-11 09:09:51,505 epoch 11 - iter 8/27 - loss 0.02751088\n",
            "2019-02-11 09:09:52,100 epoch 11 - iter 10/27 - loss 0.02651603\n",
            "2019-02-11 09:09:52,803 epoch 11 - iter 12/27 - loss 0.02482370\n",
            "2019-02-11 09:09:53,555 epoch 11 - iter 14/27 - loss 0.02489268\n",
            "2019-02-11 09:09:54,218 epoch 11 - iter 16/27 - loss 0.02455535\n",
            "2019-02-11 09:09:55,041 epoch 11 - iter 18/27 - loss 0.02533367\n",
            "2019-02-11 09:09:55,701 epoch 11 - iter 20/27 - loss 0.02523997\n",
            "2019-02-11 09:09:56,544 epoch 11 - iter 22/27 - loss 0.02496715\n",
            "2019-02-11 09:09:57,205 epoch 11 - iter 24/27 - loss 0.02469331\n",
            "2019-02-11 09:09:58,085 epoch 11 - iter 26/27 - loss 0.02520428\n",
            "2019-02-11 09:09:58,097 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:09:58,098 EPOCH 11 done: loss 0.0252 - lr 0.1000 - bad epochs 3\n",
            "2019-02-11 09:09:58,741 DEV  : loss 0.02131963 - f-score 0.6526 - acc 0.6526\n",
            "2019-02-11 09:10:02,586 TEST : loss 0.02302863 - f-score 0.5812 - acc 0.5812\n",
            "2019-02-11 09:10:20,669 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:10:21,233 epoch 12 - iter 0/27 - loss 0.01724037\n",
            "2019-02-11 09:10:22,021 epoch 12 - iter 2/27 - loss 0.02836350\n",
            "2019-02-11 09:10:22,637 epoch 12 - iter 4/27 - loss 0.03140935\n",
            "2019-02-11 09:10:23,217 epoch 12 - iter 6/27 - loss 0.03043128\n",
            "2019-02-11 09:10:24,377 epoch 12 - iter 8/27 - loss 0.02825793\n",
            "2019-02-11 09:10:24,964 epoch 12 - iter 10/27 - loss 0.02681656\n",
            "2019-02-11 09:10:25,625 epoch 12 - iter 12/27 - loss 0.02780157\n",
            "2019-02-11 09:10:26,316 epoch 12 - iter 14/27 - loss 0.02800704\n",
            "2019-02-11 09:10:27,577 epoch 12 - iter 16/27 - loss 0.02781447\n",
            "2019-02-11 09:10:28,433 epoch 12 - iter 18/27 - loss 0.02706905\n",
            "2019-02-11 09:10:29,443 epoch 12 - iter 20/27 - loss 0.02706205\n",
            "2019-02-11 09:10:30,308 epoch 12 - iter 22/27 - loss 0.02761720\n",
            "2019-02-11 09:10:31,002 epoch 12 - iter 24/27 - loss 0.02741608\n",
            "2019-02-11 09:10:31,425 epoch 12 - iter 26/27 - loss 0.02777436\n",
            "2019-02-11 09:10:31,436 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:10:31,437 EPOCH 12 done: loss 0.0278 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:10:32,059 DEV  : loss 0.01966601 - f-score 0.6737 - acc 0.6737\n",
            "2019-02-11 09:10:35,678 TEST : loss 0.02073630 - f-score 0.6368 - acc 0.6368\n",
            "2019-02-11 09:10:35,681 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:10:36,532 epoch 13 - iter 0/27 - loss 0.01791424\n",
            "2019-02-11 09:10:37,693 epoch 13 - iter 2/27 - loss 0.02149406\n",
            "2019-02-11 09:10:38,410 epoch 13 - iter 4/27 - loss 0.02307835\n",
            "2019-02-11 09:10:39,242 epoch 13 - iter 6/27 - loss 0.02290625\n",
            "2019-02-11 09:10:39,918 epoch 13 - iter 8/27 - loss 0.02373755\n",
            "2019-02-11 09:10:40,616 epoch 13 - iter 10/27 - loss 0.02402838\n",
            "2019-02-11 09:10:41,317 epoch 13 - iter 12/27 - loss 0.02332024\n",
            "2019-02-11 09:10:42,037 epoch 13 - iter 14/27 - loss 0.02342736\n",
            "2019-02-11 09:10:42,753 epoch 13 - iter 16/27 - loss 0.02384242\n",
            "2019-02-11 09:10:43,420 epoch 13 - iter 18/27 - loss 0.02417434\n",
            "2019-02-11 09:10:44,102 epoch 13 - iter 20/27 - loss 0.02481260\n",
            "2019-02-11 09:10:44,746 epoch 13 - iter 22/27 - loss 0.02538249\n",
            "2019-02-11 09:10:45,645 epoch 13 - iter 24/27 - loss 0.02507284\n",
            "2019-02-11 09:10:46,380 epoch 13 - iter 26/27 - loss 0.02573418\n",
            "2019-02-11 09:10:46,393 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:10:46,394 EPOCH 13 done: loss 0.0257 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 09:10:47,013 DEV  : loss 0.03094561 - f-score 0.5579 - acc 0.5579\n",
            "2019-02-11 09:10:50,778 TEST : loss 0.02614973 - f-score 0.5983 - acc 0.5983\n",
            "2019-02-11 09:10:50,781 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:10:51,162 epoch 14 - iter 0/27 - loss 0.02157577\n",
            "2019-02-11 09:10:52,014 epoch 14 - iter 2/27 - loss 0.02331307\n",
            "2019-02-11 09:10:52,584 epoch 14 - iter 4/27 - loss 0.02460169\n",
            "2019-02-11 09:10:53,342 epoch 14 - iter 6/27 - loss 0.02422575\n",
            "2019-02-11 09:10:54,063 epoch 14 - iter 8/27 - loss 0.02314300\n",
            "2019-02-11 09:10:54,638 epoch 14 - iter 10/27 - loss 0.02309686\n",
            "2019-02-11 09:10:55,303 epoch 14 - iter 12/27 - loss 0.02353267\n",
            "2019-02-11 09:10:56,248 epoch 14 - iter 14/27 - loss 0.02380771\n",
            "2019-02-11 09:10:57,342 epoch 14 - iter 16/27 - loss 0.02341634\n",
            "2019-02-11 09:10:58,262 epoch 14 - iter 18/27 - loss 0.02341753\n",
            "2019-02-11 09:10:58,928 epoch 14 - iter 20/27 - loss 0.02387396\n",
            "2019-02-11 09:10:59,555 epoch 14 - iter 22/27 - loss 0.02394367\n",
            "2019-02-11 09:11:00,812 epoch 14 - iter 24/27 - loss 0.02446702\n",
            "2019-02-11 09:11:01,299 epoch 14 - iter 26/27 - loss 0.02501454\n",
            "2019-02-11 09:11:01,311 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:11:01,312 EPOCH 14 done: loss 0.0250 - lr 0.1000 - bad epochs 2\n",
            "2019-02-11 09:11:01,933 DEV  : loss 0.02492525 - f-score 0.5263 - acc 0.5263\n",
            "2019-02-11 09:11:05,517 TEST : loss 0.02628160 - f-score 0.5085 - acc 0.5085\n",
            "2019-02-11 09:11:22,611 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:11:23,057 epoch 15 - iter 0/27 - loss 0.02482169\n",
            "2019-02-11 09:11:23,953 epoch 15 - iter 2/27 - loss 0.02153028\n",
            "2019-02-11 09:11:24,702 epoch 15 - iter 4/27 - loss 0.02585299\n",
            "2019-02-11 09:11:25,394 epoch 15 - iter 6/27 - loss 0.02773795\n",
            "2019-02-11 09:11:26,128 epoch 15 - iter 8/27 - loss 0.02807577\n",
            "2019-02-11 09:11:26,676 epoch 15 - iter 10/27 - loss 0.02675268\n",
            "2019-02-11 09:11:27,284 epoch 15 - iter 12/27 - loss 0.02591005\n",
            "2019-02-11 09:11:28,415 epoch 15 - iter 14/27 - loss 0.02636276\n",
            "2019-02-11 09:11:29,220 epoch 15 - iter 16/27 - loss 0.02566216\n",
            "2019-02-11 09:11:30,047 epoch 15 - iter 18/27 - loss 0.02565498\n",
            "2019-02-11 09:11:31,221 epoch 15 - iter 20/27 - loss 0.02655557\n",
            "2019-02-11 09:11:32,132 epoch 15 - iter 22/27 - loss 0.02632350\n",
            "2019-02-11 09:11:32,820 epoch 15 - iter 24/27 - loss 0.02594905\n",
            "2019-02-11 09:11:33,203 epoch 15 - iter 26/27 - loss 0.02634541\n",
            "2019-02-11 09:11:33,215 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:11:33,216 EPOCH 15 done: loss 0.0263 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:11:33,857 DEV  : loss 0.02803041 - f-score 0.6737 - acc 0.6737\n",
            "2019-02-11 09:11:37,535 TEST : loss 0.03304282 - f-score 0.5855 - acc 0.5855\n",
            "2019-02-11 09:11:37,538 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:11:37,906 epoch 16 - iter 0/27 - loss 0.02985284\n",
            "2019-02-11 09:11:38,510 epoch 16 - iter 2/27 - loss 0.02473933\n",
            "2019-02-11 09:11:39,184 epoch 16 - iter 4/27 - loss 0.02414623\n",
            "2019-02-11 09:11:39,698 epoch 16 - iter 6/27 - loss 0.02443106\n",
            "2019-02-11 09:11:40,388 epoch 16 - iter 8/27 - loss 0.02320872\n",
            "2019-02-11 09:11:41,161 epoch 16 - iter 10/27 - loss 0.02333229\n",
            "2019-02-11 09:11:41,799 epoch 16 - iter 12/27 - loss 0.02456870\n",
            "2019-02-11 09:11:42,610 epoch 16 - iter 14/27 - loss 0.02425303\n",
            "2019-02-11 09:11:43,847 epoch 16 - iter 16/27 - loss 0.02360459\n",
            "2019-02-11 09:11:44,671 epoch 16 - iter 18/27 - loss 0.02354406\n",
            "2019-02-11 09:11:45,385 epoch 16 - iter 20/27 - loss 0.02327800\n",
            "2019-02-11 09:11:46,388 epoch 16 - iter 22/27 - loss 0.02379232\n",
            "2019-02-11 09:11:47,184 epoch 16 - iter 24/27 - loss 0.02388759\n",
            "2019-02-11 09:11:47,831 epoch 16 - iter 26/27 - loss 0.02475535\n",
            "2019-02-11 09:11:47,844 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:11:47,845 EPOCH 16 done: loss 0.0248 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 09:11:48,486 DEV  : loss 0.02274992 - f-score 0.6105 - acc 0.6105\n",
            "2019-02-11 09:11:52,136 TEST : loss 0.02134552 - f-score 0.6538 - acc 0.6538\n",
            "2019-02-11 09:12:09,228 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:12:09,698 epoch 17 - iter 0/27 - loss 0.02245127\n",
            "2019-02-11 09:12:10,337 epoch 17 - iter 2/27 - loss 0.02354535\n",
            "2019-02-11 09:12:11,121 epoch 17 - iter 4/27 - loss 0.02445200\n",
            "2019-02-11 09:12:11,730 epoch 17 - iter 6/27 - loss 0.02451940\n",
            "2019-02-11 09:12:12,431 epoch 17 - iter 8/27 - loss 0.02323875\n",
            "2019-02-11 09:12:13,663 epoch 17 - iter 10/27 - loss 0.02244838\n",
            "2019-02-11 09:12:14,336 epoch 17 - iter 12/27 - loss 0.02270135\n",
            "2019-02-11 09:12:15,089 epoch 17 - iter 14/27 - loss 0.02285584\n",
            "2019-02-11 09:12:15,699 epoch 17 - iter 16/27 - loss 0.02291369\n",
            "2019-02-11 09:12:16,693 epoch 17 - iter 18/27 - loss 0.02325168\n",
            "2019-02-11 09:12:17,611 epoch 17 - iter 20/27 - loss 0.02308776\n",
            "2019-02-11 09:12:18,224 epoch 17 - iter 22/27 - loss 0.02287868\n",
            "2019-02-11 09:12:19,003 epoch 17 - iter 24/27 - loss 0.02301745\n",
            "2019-02-11 09:12:19,562 epoch 17 - iter 26/27 - loss 0.02376589\n",
            "2019-02-11 09:12:19,573 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:12:19,575 EPOCH 17 done: loss 0.0238 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:12:20,197 DEV  : loss 0.01959715 - f-score 0.6526 - acc 0.6526\n",
            "2019-02-11 09:12:24,001 TEST : loss 0.02028417 - f-score 0.6432 - acc 0.6432\n",
            "2019-02-11 09:12:41,240 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:12:42,041 epoch 18 - iter 0/27 - loss 0.01804114\n",
            "2019-02-11 09:12:42,884 epoch 18 - iter 2/27 - loss 0.02196919\n",
            "2019-02-11 09:12:43,677 epoch 18 - iter 4/27 - loss 0.02336106\n",
            "2019-02-11 09:12:44,887 epoch 18 - iter 6/27 - loss 0.02202554\n",
            "2019-02-11 09:12:45,739 epoch 18 - iter 8/27 - loss 0.02140661\n",
            "2019-02-11 09:12:46,461 epoch 18 - iter 10/27 - loss 0.02161529\n",
            "2019-02-11 09:12:47,205 epoch 18 - iter 12/27 - loss 0.02141153\n",
            "2019-02-11 09:12:47,857 epoch 18 - iter 14/27 - loss 0.02158325\n",
            "2019-02-11 09:12:48,816 epoch 18 - iter 16/27 - loss 0.02312145\n",
            "2019-02-11 09:12:49,560 epoch 18 - iter 18/27 - loss 0.02338444\n",
            "2019-02-11 09:12:50,737 epoch 18 - iter 20/27 - loss 0.02296571\n",
            "2019-02-11 09:12:51,477 epoch 18 - iter 22/27 - loss 0.02222781\n",
            "2019-02-11 09:12:52,171 epoch 18 - iter 24/27 - loss 0.02183588\n",
            "2019-02-11 09:12:52,660 epoch 18 - iter 26/27 - loss 0.02287173\n",
            "2019-02-11 09:12:52,671 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:12:52,672 EPOCH 18 done: loss 0.0229 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:12:53,319 DEV  : loss 0.02671055 - f-score 0.6000 - acc 0.6000\n",
            "2019-02-11 09:12:56,977 TEST : loss 0.02289311 - f-score 0.6603 - acc 0.6603\n",
            "2019-02-11 09:13:14,203 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:13:14,642 epoch 19 - iter 0/27 - loss 0.01968366\n",
            "2019-02-11 09:13:15,545 epoch 19 - iter 2/27 - loss 0.02034834\n",
            "2019-02-11 09:13:16,358 epoch 19 - iter 4/27 - loss 0.02080614\n",
            "2019-02-11 09:13:17,003 epoch 19 - iter 6/27 - loss 0.02133378\n",
            "2019-02-11 09:13:18,102 epoch 19 - iter 8/27 - loss 0.02093741\n",
            "2019-02-11 09:13:19,268 epoch 19 - iter 10/27 - loss 0.02119749\n",
            "2019-02-11 09:13:20,010 epoch 19 - iter 12/27 - loss 0.02178906\n",
            "2019-02-11 09:13:20,841 epoch 19 - iter 14/27 - loss 0.02274874\n",
            "2019-02-11 09:13:21,490 epoch 19 - iter 16/27 - loss 0.02196292\n",
            "2019-02-11 09:13:22,092 epoch 19 - iter 18/27 - loss 0.02216809\n",
            "2019-02-11 09:13:22,879 epoch 19 - iter 20/27 - loss 0.02194569\n",
            "2019-02-11 09:13:23,782 epoch 19 - iter 22/27 - loss 0.02193283\n",
            "2019-02-11 09:13:24,676 epoch 19 - iter 24/27 - loss 0.02192197\n",
            "2019-02-11 09:13:24,987 epoch 19 - iter 26/27 - loss 0.02242875\n",
            "2019-02-11 09:13:25,001 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:13:25,005 EPOCH 19 done: loss 0.0224 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:13:25,632 DEV  : loss 0.02145507 - f-score 0.6842 - acc 0.6842\n",
            "2019-02-11 09:13:29,292 TEST : loss 0.02302735 - f-score 0.6111 - acc 0.6111\n",
            "2019-02-11 09:13:46,328 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:13:46,965 epoch 20 - iter 0/27 - loss 0.02565745\n",
            "2019-02-11 09:13:47,800 epoch 20 - iter 2/27 - loss 0.02267573\n",
            "2019-02-11 09:13:48,443 epoch 20 - iter 4/27 - loss 0.02350703\n",
            "2019-02-11 09:13:49,039 epoch 20 - iter 6/27 - loss 0.02387212\n",
            "2019-02-11 09:13:50,156 epoch 20 - iter 8/27 - loss 0.02224557\n",
            "2019-02-11 09:13:50,927 epoch 20 - iter 10/27 - loss 0.02080039\n",
            "2019-02-11 09:13:51,927 epoch 20 - iter 12/27 - loss 0.02097409\n",
            "2019-02-11 09:13:52,829 epoch 20 - iter 14/27 - loss 0.02081051\n",
            "2019-02-11 09:13:53,602 epoch 20 - iter 16/27 - loss 0.02061645\n",
            "2019-02-11 09:13:54,324 epoch 20 - iter 18/27 - loss 0.02029498\n",
            "2019-02-11 09:13:54,968 epoch 20 - iter 20/27 - loss 0.02102122\n",
            "2019-02-11 09:13:56,032 epoch 20 - iter 22/27 - loss 0.02110919\n",
            "2019-02-11 09:13:56,832 epoch 20 - iter 24/27 - loss 0.02064465\n",
            "2019-02-11 09:13:57,235 epoch 20 - iter 26/27 - loss 0.02116231\n",
            "2019-02-11 09:13:57,246 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:13:57,247 EPOCH 20 done: loss 0.0212 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:13:57,888 DEV  : loss 0.02336925 - f-score 0.6421 - acc 0.6421\n",
            "2019-02-11 09:14:01,634 TEST : loss 0.02017142 - f-score 0.6731 - acc 0.6731\n",
            "2019-02-11 09:14:18,727 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:14:19,587 epoch 21 - iter 0/27 - loss 0.02775727\n",
            "2019-02-11 09:14:20,285 epoch 21 - iter 2/27 - loss 0.02019998\n",
            "2019-02-11 09:14:21,473 epoch 21 - iter 4/27 - loss 0.02063517\n",
            "2019-02-11 09:14:22,258 epoch 21 - iter 6/27 - loss 0.02254906\n",
            "2019-02-11 09:14:22,987 epoch 21 - iter 8/27 - loss 0.02143728\n",
            "2019-02-11 09:14:23,656 epoch 21 - iter 10/27 - loss 0.02260510\n",
            "2019-02-11 09:14:24,567 epoch 21 - iter 12/27 - loss 0.02374522\n",
            "2019-02-11 09:14:25,173 epoch 21 - iter 14/27 - loss 0.02256388\n",
            "2019-02-11 09:14:25,941 epoch 21 - iter 16/27 - loss 0.02216261\n",
            "2019-02-11 09:14:26,660 epoch 21 - iter 18/27 - loss 0.02267815\n",
            "2019-02-11 09:14:27,186 epoch 21 - iter 20/27 - loss 0.02238684\n",
            "2019-02-11 09:14:27,987 epoch 21 - iter 22/27 - loss 0.02272319\n",
            "2019-02-11 09:14:29,234 epoch 21 - iter 24/27 - loss 0.02308478\n",
            "2019-02-11 09:14:29,798 epoch 21 - iter 26/27 - loss 0.02310395\n",
            "2019-02-11 09:14:29,810 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:14:29,811 EPOCH 21 done: loss 0.0231 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:14:30,441 DEV  : loss 0.03160428 - f-score 0.6842 - acc 0.6842\n",
            "2019-02-11 09:14:34,157 TEST : loss 0.03540790 - f-score 0.5962 - acc 0.5962\n",
            "2019-02-11 09:14:34,160 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:14:34,482 epoch 22 - iter 0/27 - loss 0.04187674\n",
            "2019-02-11 09:14:35,053 epoch 22 - iter 2/27 - loss 0.02951019\n",
            "2019-02-11 09:14:35,766 epoch 22 - iter 4/27 - loss 0.02474061\n",
            "2019-02-11 09:14:36,715 epoch 22 - iter 6/27 - loss 0.02368634\n",
            "2019-02-11 09:14:37,560 epoch 22 - iter 8/27 - loss 0.02205380\n",
            "2019-02-11 09:14:38,160 epoch 22 - iter 10/27 - loss 0.02257538\n",
            "2019-02-11 09:14:38,883 epoch 22 - iter 12/27 - loss 0.02313938\n",
            "2019-02-11 09:14:39,719 epoch 22 - iter 14/27 - loss 0.02329540\n",
            "2019-02-11 09:14:40,495 epoch 22 - iter 16/27 - loss 0.02322683\n",
            "2019-02-11 09:14:41,838 epoch 22 - iter 18/27 - loss 0.02268800\n",
            "2019-02-11 09:14:42,555 epoch 22 - iter 20/27 - loss 0.02255925\n",
            "2019-02-11 09:14:43,266 epoch 22 - iter 22/27 - loss 0.02205974\n",
            "2019-02-11 09:14:44,360 epoch 22 - iter 24/27 - loss 0.02229104\n",
            "2019-02-11 09:14:44,878 epoch 22 - iter 26/27 - loss 0.02272578\n",
            "2019-02-11 09:14:44,890 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:14:44,891 EPOCH 22 done: loss 0.0227 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 09:14:45,522 DEV  : loss 0.01776081 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 09:14:49,230 TEST : loss 0.01628502 - f-score 0.7350 - acc 0.7350\n",
            "2019-02-11 09:14:49,233 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:14:49,526 epoch 23 - iter 0/27 - loss 0.01540201\n",
            "2019-02-11 09:14:50,652 epoch 23 - iter 2/27 - loss 0.01915997\n",
            "2019-02-11 09:14:51,440 epoch 23 - iter 4/27 - loss 0.01875937\n",
            "2019-02-11 09:14:52,618 epoch 23 - iter 6/27 - loss 0.01948244\n",
            "2019-02-11 09:14:53,298 epoch 23 - iter 8/27 - loss 0.02033118\n",
            "2019-02-11 09:14:54,025 epoch 23 - iter 10/27 - loss 0.02016465\n",
            "2019-02-11 09:14:54,649 epoch 23 - iter 12/27 - loss 0.02036204\n",
            "2019-02-11 09:14:55,329 epoch 23 - iter 14/27 - loss 0.02003094\n",
            "2019-02-11 09:14:56,271 epoch 23 - iter 16/27 - loss 0.01959504\n",
            "2019-02-11 09:14:57,018 epoch 23 - iter 18/27 - loss 0.01893391\n",
            "2019-02-11 09:14:57,850 epoch 23 - iter 20/27 - loss 0.01914117\n",
            "2019-02-11 09:14:58,582 epoch 23 - iter 22/27 - loss 0.01944927\n",
            "2019-02-11 09:14:59,420 epoch 23 - iter 24/27 - loss 0.01952802\n",
            "2019-02-11 09:14:59,807 epoch 23 - iter 26/27 - loss 0.02044594\n",
            "2019-02-11 09:14:59,819 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:14:59,820 EPOCH 23 done: loss 0.0204 - lr 0.1000 - bad epochs 2\n",
            "2019-02-11 09:15:00,467 DEV  : loss 0.01861337 - f-score 0.6737 - acc 0.6737\n",
            "2019-02-11 09:15:04,101 TEST : loss 0.01747251 - f-score 0.7393 - acc 0.7393\n",
            "2019-02-11 09:15:21,219 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:15:21,677 epoch 24 - iter 0/27 - loss 0.02038885\n",
            "2019-02-11 09:15:22,519 epoch 24 - iter 2/27 - loss 0.02163857\n",
            "2019-02-11 09:15:23,359 epoch 24 - iter 4/27 - loss 0.02037006\n",
            "2019-02-11 09:15:24,502 epoch 24 - iter 6/27 - loss 0.02097987\n",
            "2019-02-11 09:15:25,380 epoch 24 - iter 8/27 - loss 0.02109095\n",
            "2019-02-11 09:15:25,933 epoch 24 - iter 10/27 - loss 0.02114702\n",
            "2019-02-11 09:15:26,635 epoch 24 - iter 12/27 - loss 0.02226244\n",
            "2019-02-11 09:15:27,366 epoch 24 - iter 14/27 - loss 0.02139169\n",
            "2019-02-11 09:15:27,961 epoch 24 - iter 16/27 - loss 0.02078687\n",
            "2019-02-11 09:15:29,386 epoch 24 - iter 18/27 - loss 0.02121243\n",
            "2019-02-11 09:15:30,104 epoch 24 - iter 20/27 - loss 0.02105761\n",
            "2019-02-11 09:15:30,838 epoch 24 - iter 22/27 - loss 0.02103621\n",
            "2019-02-11 09:15:31,461 epoch 24 - iter 24/27 - loss 0.02105296\n",
            "2019-02-11 09:15:31,988 epoch 24 - iter 26/27 - loss 0.02094322\n",
            "2019-02-11 09:15:31,999 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:15:32,000 EPOCH 24 done: loss 0.0209 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:15:32,625 DEV  : loss 0.01939945 - f-score 0.7263 - acc 0.7263\n",
            "2019-02-11 09:15:36,321 TEST : loss 0.01657525 - f-score 0.7607 - acc 0.7607\n",
            "2019-02-11 09:15:36,324 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:15:36,570 epoch 25 - iter 0/27 - loss 0.02180297\n",
            "2019-02-11 09:15:37,270 epoch 25 - iter 2/27 - loss 0.01798655\n",
            "2019-02-11 09:15:38,200 epoch 25 - iter 4/27 - loss 0.02026064\n",
            "2019-02-11 09:15:39,199 epoch 25 - iter 6/27 - loss 0.01912345\n",
            "2019-02-11 09:15:39,886 epoch 25 - iter 8/27 - loss 0.01857419\n",
            "2019-02-11 09:15:40,600 epoch 25 - iter 10/27 - loss 0.01900655\n",
            "2019-02-11 09:15:41,304 epoch 25 - iter 12/27 - loss 0.01859476\n",
            "2019-02-11 09:15:41,985 epoch 25 - iter 14/27 - loss 0.01865915\n",
            "2019-02-11 09:15:42,763 epoch 25 - iter 16/27 - loss 0.01903070\n",
            "2019-02-11 09:15:43,368 epoch 25 - iter 18/27 - loss 0.01892691\n",
            "2019-02-11 09:15:44,033 epoch 25 - iter 20/27 - loss 0.01908795\n",
            "2019-02-11 09:15:45,136 epoch 25 - iter 22/27 - loss 0.01902667\n",
            "2019-02-11 09:15:46,007 epoch 25 - iter 24/27 - loss 0.02064735\n",
            "2019-02-11 09:15:46,557 epoch 25 - iter 26/27 - loss 0.02057994\n",
            "2019-02-11 09:15:46,569 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:15:46,570 EPOCH 25 done: loss 0.0206 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 09:15:47,183 DEV  : loss 0.01881805 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 09:15:50,780 TEST : loss 0.01459801 - f-score 0.7927 - acc 0.7927\n",
            "2019-02-11 09:15:50,783 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:15:51,120 epoch 26 - iter 0/27 - loss 0.01407774\n",
            "2019-02-11 09:15:51,881 epoch 26 - iter 2/27 - loss 0.01696923\n",
            "2019-02-11 09:15:52,500 epoch 26 - iter 4/27 - loss 0.01970018\n",
            "2019-02-11 09:15:53,221 epoch 26 - iter 6/27 - loss 0.01906386\n",
            "2019-02-11 09:15:54,181 epoch 26 - iter 8/27 - loss 0.01855486\n",
            "2019-02-11 09:15:54,824 epoch 26 - iter 10/27 - loss 0.01935480\n",
            "2019-02-11 09:15:55,860 epoch 26 - iter 12/27 - loss 0.01952118\n",
            "2019-02-11 09:15:56,545 epoch 26 - iter 14/27 - loss 0.01962849\n",
            "2019-02-11 09:15:57,116 epoch 26 - iter 16/27 - loss 0.01977974\n",
            "2019-02-11 09:15:58,212 epoch 26 - iter 18/27 - loss 0.02010696\n",
            "2019-02-11 09:15:59,087 epoch 26 - iter 20/27 - loss 0.01983927\n",
            "2019-02-11 09:15:59,794 epoch 26 - iter 22/27 - loss 0.01957879\n",
            "2019-02-11 09:16:00,597 epoch 26 - iter 24/27 - loss 0.01961751\n",
            "2019-02-11 09:16:01,092 epoch 26 - iter 26/27 - loss 0.02067810\n",
            "2019-02-11 09:16:01,104 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:16:01,105 EPOCH 26 done: loss 0.0207 - lr 0.1000 - bad epochs 2\n",
            "2019-02-11 09:16:01,724 DEV  : loss 0.02801574 - f-score 0.6421 - acc 0.6421\n",
            "2019-02-11 09:16:05,333 TEST : loss 0.01850183 - f-score 0.7543 - acc 0.7543\n",
            "2019-02-11 09:16:05,335 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:16:05,677 epoch 27 - iter 0/27 - loss 0.01859924\n",
            "2019-02-11 09:16:06,453 epoch 27 - iter 2/27 - loss 0.02221231\n",
            "2019-02-11 09:16:07,293 epoch 27 - iter 4/27 - loss 0.01779397\n",
            "2019-02-11 09:16:08,084 epoch 27 - iter 6/27 - loss 0.01945968\n",
            "2019-02-11 09:16:08,702 epoch 27 - iter 8/27 - loss 0.01859227\n",
            "2019-02-11 09:16:09,534 epoch 27 - iter 10/27 - loss 0.01898384\n",
            "2019-02-11 09:16:10,687 epoch 27 - iter 12/27 - loss 0.01819535\n",
            "2019-02-11 09:16:11,267 epoch 27 - iter 14/27 - loss 0.01749056\n",
            "2019-02-11 09:16:11,806 epoch 27 - iter 16/27 - loss 0.01774914\n",
            "2019-02-11 09:16:12,794 epoch 27 - iter 18/27 - loss 0.01799482\n",
            "2019-02-11 09:16:13,469 epoch 27 - iter 20/27 - loss 0.01790022\n",
            "2019-02-11 09:16:14,199 epoch 27 - iter 22/27 - loss 0.01836447\n",
            "2019-02-11 09:16:14,850 epoch 27 - iter 24/27 - loss 0.01839803\n",
            "2019-02-11 09:16:15,755 epoch 27 - iter 26/27 - loss 0.01890583\n",
            "2019-02-11 09:16:15,767 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:16:15,768 EPOCH 27 done: loss 0.0189 - lr 0.1000 - bad epochs 3\n",
            "2019-02-11 09:16:16,393 DEV  : loss 0.02331731 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 09:16:19,992 TEST : loss 0.01685144 - f-score 0.7714 - acc 0.7714\n",
            "2019-02-11 09:16:36,987 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:16:37,589 epoch 28 - iter 0/27 - loss 0.01460131\n",
            "2019-02-11 09:16:38,333 epoch 28 - iter 2/27 - loss 0.01357943\n",
            "2019-02-11 09:16:39,156 epoch 28 - iter 4/27 - loss 0.01518082\n",
            "2019-02-11 09:16:40,045 epoch 28 - iter 6/27 - loss 0.01601600\n",
            "2019-02-11 09:16:40,746 epoch 28 - iter 8/27 - loss 0.01595298\n",
            "2019-02-11 09:16:41,959 epoch 28 - iter 10/27 - loss 0.01631771\n",
            "2019-02-11 09:16:42,590 epoch 28 - iter 12/27 - loss 0.01674045\n",
            "2019-02-11 09:16:43,203 epoch 28 - iter 14/27 - loss 0.01656615\n",
            "2019-02-11 09:16:44,084 epoch 28 - iter 16/27 - loss 0.01732888\n",
            "2019-02-11 09:16:44,785 epoch 28 - iter 18/27 - loss 0.01785889\n",
            "2019-02-11 09:16:45,706 epoch 28 - iter 20/27 - loss 0.01792158\n",
            "2019-02-11 09:16:46,727 epoch 28 - iter 22/27 - loss 0.01783071\n",
            "2019-02-11 09:16:47,455 epoch 28 - iter 24/27 - loss 0.01804806\n",
            "2019-02-11 09:16:47,982 epoch 28 - iter 26/27 - loss 0.01846696\n",
            "2019-02-11 09:16:47,993 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:16:47,995 EPOCH 28 done: loss 0.0185 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:16:48,632 DEV  : loss 0.02336213 - f-score 0.6842 - acc 0.6842\n",
            "2019-02-11 09:16:52,301 TEST : loss 0.01299598 - f-score 0.8056 - acc 0.8056\n",
            "2019-02-11 09:17:09,554 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:17:10,218 epoch 29 - iter 0/27 - loss 0.02843907\n",
            "2019-02-11 09:17:11,011 epoch 29 - iter 2/27 - loss 0.01845918\n",
            "2019-02-11 09:17:11,587 epoch 29 - iter 4/27 - loss 0.01934906\n",
            "2019-02-11 09:17:12,207 epoch 29 - iter 6/27 - loss 0.01786653\n",
            "2019-02-11 09:17:13,016 epoch 29 - iter 8/27 - loss 0.01750434\n",
            "2019-02-11 09:17:13,911 epoch 29 - iter 10/27 - loss 0.01704490\n",
            "2019-02-11 09:17:14,616 epoch 29 - iter 12/27 - loss 0.01798324\n",
            "2019-02-11 09:17:15,897 epoch 29 - iter 14/27 - loss 0.01790741\n",
            "2019-02-11 09:17:16,470 epoch 29 - iter 16/27 - loss 0.01762363\n",
            "2019-02-11 09:17:17,163 epoch 29 - iter 18/27 - loss 0.01802112\n",
            "2019-02-11 09:17:18,026 epoch 29 - iter 20/27 - loss 0.01784124\n",
            "2019-02-11 09:17:18,713 epoch 29 - iter 22/27 - loss 0.01803428\n",
            "2019-02-11 09:17:19,568 epoch 29 - iter 24/27 - loss 0.01761296\n",
            "2019-02-11 09:17:19,913 epoch 29 - iter 26/27 - loss 0.01769952\n",
            "2019-02-11 09:17:19,924 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:17:19,924 EPOCH 29 done: loss 0.0177 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:17:20,543 DEV  : loss 0.02199452 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 09:17:24,134 TEST : loss 0.01227621 - f-score 0.8333 - acc 0.8333\n",
            "2019-02-11 09:17:41,201 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:17:41,909 epoch 30 - iter 0/27 - loss 0.01704455\n",
            "2019-02-11 09:17:42,596 epoch 30 - iter 2/27 - loss 0.01694974\n",
            "2019-02-11 09:17:43,214 epoch 30 - iter 4/27 - loss 0.01651840\n",
            "2019-02-11 09:17:43,995 epoch 30 - iter 6/27 - loss 0.01620300\n",
            "2019-02-11 09:17:45,598 epoch 30 - iter 8/27 - loss 0.01674239\n",
            "2019-02-11 09:17:46,297 epoch 30 - iter 10/27 - loss 0.01857608\n",
            "2019-02-11 09:17:46,879 epoch 30 - iter 12/27 - loss 0.01760342\n",
            "2019-02-11 09:17:47,624 epoch 30 - iter 14/27 - loss 0.01830122\n",
            "2019-02-11 09:17:48,330 epoch 30 - iter 16/27 - loss 0.01853877\n",
            "2019-02-11 09:17:49,123 epoch 30 - iter 18/27 - loss 0.01833597\n",
            "2019-02-11 09:17:49,998 epoch 30 - iter 20/27 - loss 0.01862028\n",
            "2019-02-11 09:17:50,642 epoch 30 - iter 22/27 - loss 0.01850411\n",
            "2019-02-11 09:17:51,392 epoch 30 - iter 24/27 - loss 0.01853437\n",
            "2019-02-11 09:17:51,952 epoch 30 - iter 26/27 - loss 0.01900200\n",
            "2019-02-11 09:17:51,963 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:17:51,965 EPOCH 30 done: loss 0.0190 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:17:52,595 DEV  : loss 0.01724760 - f-score 0.7474 - acc 0.7474\n",
            "2019-02-11 09:17:56,470 TEST : loss 0.01193300 - f-score 0.8269 - acc 0.8269\n",
            "2019-02-11 09:17:56,474 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:17:56,794 epoch 31 - iter 0/27 - loss 0.00928967\n",
            "2019-02-11 09:17:57,629 epoch 31 - iter 2/27 - loss 0.01320151\n",
            "2019-02-11 09:17:58,527 epoch 31 - iter 4/27 - loss 0.01418813\n",
            "2019-02-11 09:17:59,217 epoch 31 - iter 6/27 - loss 0.01715471\n",
            "2019-02-11 09:17:59,898 epoch 31 - iter 8/27 - loss 0.01666031\n",
            "2019-02-11 09:18:00,612 epoch 31 - iter 10/27 - loss 0.01829067\n",
            "2019-02-11 09:18:01,460 epoch 31 - iter 12/27 - loss 0.01833337\n",
            "2019-02-11 09:18:02,973 epoch 31 - iter 14/27 - loss 0.01908493\n",
            "2019-02-11 09:18:04,151 epoch 31 - iter 16/27 - loss 0.01894333\n",
            "2019-02-11 09:18:04,778 epoch 31 - iter 18/27 - loss 0.01872303\n",
            "2019-02-11 09:18:05,683 epoch 31 - iter 20/27 - loss 0.01860927\n",
            "2019-02-11 09:18:06,337 epoch 31 - iter 22/27 - loss 0.01884388\n",
            "2019-02-11 09:18:07,189 epoch 31 - iter 24/27 - loss 0.01841902\n",
            "2019-02-11 09:18:07,718 epoch 31 - iter 26/27 - loss 0.01908295\n",
            "2019-02-11 09:18:07,730 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:18:07,731 EPOCH 31 done: loss 0.0191 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 09:18:08,410 DEV  : loss 0.01853148 - f-score 0.7368 - acc 0.7368\n",
            "2019-02-11 09:18:12,409 TEST : loss 0.01029411 - f-score 0.8526 - acc 0.8526\n",
            "2019-02-11 09:18:12,414 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:18:12,759 epoch 32 - iter 0/27 - loss 0.01288839\n",
            "2019-02-11 09:18:13,507 epoch 32 - iter 2/27 - loss 0.01297720\n",
            "2019-02-11 09:18:14,311 epoch 32 - iter 4/27 - loss 0.01495164\n",
            "2019-02-11 09:18:15,041 epoch 32 - iter 6/27 - loss 0.01520074\n",
            "2019-02-11 09:18:15,803 epoch 32 - iter 8/27 - loss 0.01533300\n",
            "2019-02-11 09:18:16,742 epoch 32 - iter 10/27 - loss 0.01568601\n",
            "2019-02-11 09:18:17,426 epoch 32 - iter 12/27 - loss 0.01595413\n",
            "2019-02-11 09:18:18,153 epoch 32 - iter 14/27 - loss 0.01518379\n",
            "2019-02-11 09:18:19,289 epoch 32 - iter 16/27 - loss 0.01558633\n",
            "2019-02-11 09:18:20,175 epoch 32 - iter 18/27 - loss 0.01597825\n",
            "2019-02-11 09:18:21,416 epoch 32 - iter 20/27 - loss 0.01595549\n",
            "2019-02-11 09:18:22,041 epoch 32 - iter 22/27 - loss 0.01563911\n",
            "2019-02-11 09:18:23,319 epoch 32 - iter 24/27 - loss 0.01606211\n",
            "2019-02-11 09:18:23,779 epoch 32 - iter 26/27 - loss 0.01691992\n",
            "2019-02-11 09:18:23,792 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:18:23,793 EPOCH 32 done: loss 0.0169 - lr 0.1000 - bad epochs 2\n",
            "2019-02-11 09:18:24,481 DEV  : loss 0.02081171 - f-score 0.7263 - acc 0.7263\n",
            "2019-02-11 09:18:28,330 TEST : loss 0.01151794 - f-score 0.8483 - acc 0.8483\n",
            "2019-02-11 09:18:45,812 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:18:46,385 epoch 33 - iter 0/27 - loss 0.02129206\n",
            "2019-02-11 09:18:47,524 epoch 33 - iter 2/27 - loss 0.01770546\n",
            "2019-02-11 09:18:48,297 epoch 33 - iter 4/27 - loss 0.01623992\n",
            "2019-02-11 09:18:49,009 epoch 33 - iter 6/27 - loss 0.01467020\n",
            "2019-02-11 09:18:49,771 epoch 33 - iter 8/27 - loss 0.01523434\n",
            "2019-02-11 09:18:50,651 epoch 33 - iter 10/27 - loss 0.01510312\n",
            "2019-02-11 09:18:51,437 epoch 33 - iter 12/27 - loss 0.01527379\n",
            "2019-02-11 09:18:52,294 epoch 33 - iter 14/27 - loss 0.01559313\n",
            "2019-02-11 09:18:52,938 epoch 33 - iter 16/27 - loss 0.01626473\n",
            "2019-02-11 09:18:53,764 epoch 33 - iter 18/27 - loss 0.01647138\n",
            "2019-02-11 09:18:54,512 epoch 33 - iter 20/27 - loss 0.01682142\n",
            "2019-02-11 09:18:55,815 epoch 33 - iter 22/27 - loss 0.01685663\n",
            "2019-02-11 09:18:56,476 epoch 33 - iter 24/27 - loss 0.01715176\n",
            "2019-02-11 09:18:57,443 epoch 33 - iter 26/27 - loss 0.01747948\n",
            "2019-02-11 09:18:57,454 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:18:57,456 EPOCH 33 done: loss 0.0175 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:18:58,142 DEV  : loss 0.01540804 - f-score 0.7684 - acc 0.7684\n",
            "2019-02-11 09:19:02,147 TEST : loss 0.00979450 - f-score 0.8526 - acc 0.8526\n",
            "2019-02-11 09:19:02,151 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:19:02,669 epoch 34 - iter 0/27 - loss 0.01107875\n",
            "2019-02-11 09:19:03,892 epoch 34 - iter 2/27 - loss 0.01603505\n",
            "2019-02-11 09:19:04,571 epoch 34 - iter 4/27 - loss 0.01403631\n",
            "2019-02-11 09:19:05,370 epoch 34 - iter 6/27 - loss 0.01415046\n",
            "2019-02-11 09:19:06,079 epoch 34 - iter 8/27 - loss 0.01555042\n",
            "2019-02-11 09:19:06,872 epoch 34 - iter 10/27 - loss 0.01486664\n",
            "2019-02-11 09:19:07,727 epoch 34 - iter 12/27 - loss 0.01557785\n",
            "2019-02-11 09:19:08,417 epoch 34 - iter 14/27 - loss 0.01662285\n",
            "2019-02-11 09:19:09,253 epoch 34 - iter 16/27 - loss 0.01635601\n",
            "2019-02-11 09:19:10,020 epoch 34 - iter 18/27 - loss 0.01630656\n",
            "2019-02-11 09:19:11,111 epoch 34 - iter 20/27 - loss 0.01583696\n",
            "2019-02-11 09:19:11,894 epoch 34 - iter 22/27 - loss 0.01581853\n",
            "2019-02-11 09:19:12,673 epoch 34 - iter 24/27 - loss 0.01564882\n",
            "2019-02-11 09:19:13,184 epoch 34 - iter 26/27 - loss 0.01549392\n",
            "2019-02-11 09:19:13,196 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:19:13,198 EPOCH 34 done: loss 0.0155 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 09:19:13,874 DEV  : loss 0.01637536 - f-score 0.7895 - acc 0.7895\n",
            "2019-02-11 09:19:17,847 TEST : loss 0.01024047 - f-score 0.8632 - acc 0.8632\n",
            "2019-02-11 09:19:35,774 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:19:36,287 epoch 35 - iter 0/27 - loss 0.01083867\n",
            "2019-02-11 09:19:37,838 epoch 35 - iter 2/27 - loss 0.01268739\n",
            "2019-02-11 09:19:38,635 epoch 35 - iter 4/27 - loss 0.01543797\n",
            "2019-02-11 09:19:39,367 epoch 35 - iter 6/27 - loss 0.01573281\n",
            "2019-02-11 09:19:40,104 epoch 35 - iter 8/27 - loss 0.01525470\n",
            "2019-02-11 09:19:40,848 epoch 35 - iter 10/27 - loss 0.01509252\n",
            "2019-02-11 09:19:41,612 epoch 35 - iter 12/27 - loss 0.01467403\n",
            "2019-02-11 09:19:42,840 epoch 35 - iter 14/27 - loss 0.01477065\n",
            "2019-02-11 09:19:43,555 epoch 35 - iter 16/27 - loss 0.01483382\n",
            "2019-02-11 09:19:44,272 epoch 35 - iter 18/27 - loss 0.01435581\n",
            "2019-02-11 09:19:44,932 epoch 35 - iter 20/27 - loss 0.01501342\n",
            "2019-02-11 09:19:45,671 epoch 35 - iter 22/27 - loss 0.01478445\n",
            "2019-02-11 09:19:46,590 epoch 35 - iter 24/27 - loss 0.01462389\n",
            "2019-02-11 09:19:47,314 epoch 35 - iter 26/27 - loss 0.01508135\n",
            "2019-02-11 09:19:47,326 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:19:47,327 EPOCH 35 done: loss 0.0151 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:19:47,962 DEV  : loss 0.01621926 - f-score 0.7895 - acc 0.7895\n",
            "2019-02-11 09:19:51,661 TEST : loss 0.00922073 - f-score 0.8590 - acc 0.8590\n",
            "2019-02-11 09:20:09,028 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:20:09,936 epoch 36 - iter 0/27 - loss 0.00890816\n",
            "2019-02-11 09:20:11,199 epoch 36 - iter 2/27 - loss 0.01242831\n",
            "2019-02-11 09:20:11,741 epoch 36 - iter 4/27 - loss 0.01541715\n",
            "2019-02-11 09:20:12,455 epoch 36 - iter 6/27 - loss 0.01529284\n",
            "2019-02-11 09:20:13,362 epoch 36 - iter 8/27 - loss 0.01457358\n",
            "2019-02-11 09:20:14,014 epoch 36 - iter 10/27 - loss 0.01406785\n",
            "2019-02-11 09:20:14,802 epoch 36 - iter 12/27 - loss 0.01391006\n",
            "2019-02-11 09:20:15,569 epoch 36 - iter 14/27 - loss 0.01434390\n",
            "2019-02-11 09:20:16,193 epoch 36 - iter 16/27 - loss 0.01532543\n",
            "2019-02-11 09:20:16,931 epoch 36 - iter 18/27 - loss 0.01623809\n",
            "2019-02-11 09:20:17,764 epoch 36 - iter 20/27 - loss 0.01615588\n",
            "2019-02-11 09:20:18,525 epoch 36 - iter 22/27 - loss 0.01646081\n",
            "2019-02-11 09:20:19,777 epoch 36 - iter 24/27 - loss 0.01629747\n",
            "2019-02-11 09:20:20,202 epoch 36 - iter 26/27 - loss 0.01683339\n",
            "2019-02-11 09:20:20,213 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:20:20,214 EPOCH 36 done: loss 0.0168 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:20:20,863 DEV  : loss 0.01796615 - f-score 0.7474 - acc 0.7474\n",
            "2019-02-11 09:20:24,800 TEST : loss 0.00740903 - f-score 0.8953 - acc 0.8953\n",
            "2019-02-11 09:20:24,803 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:20:25,657 epoch 37 - iter 0/27 - loss 0.01698286\n",
            "2019-02-11 09:20:26,430 epoch 37 - iter 2/27 - loss 0.01440550\n",
            "2019-02-11 09:20:27,437 epoch 37 - iter 4/27 - loss 0.01665060\n",
            "2019-02-11 09:20:28,282 epoch 37 - iter 6/27 - loss 0.01563986\n",
            "2019-02-11 09:20:29,232 epoch 37 - iter 8/27 - loss 0.01388243\n",
            "2019-02-11 09:20:29,924 epoch 37 - iter 10/27 - loss 0.01449738\n",
            "2019-02-11 09:20:30,628 epoch 37 - iter 12/27 - loss 0.01370900\n",
            "2019-02-11 09:20:31,516 epoch 37 - iter 14/27 - loss 0.01387599\n",
            "2019-02-11 09:20:32,140 epoch 37 - iter 16/27 - loss 0.01397424\n",
            "2019-02-11 09:20:32,987 epoch 37 - iter 18/27 - loss 0.01353532\n",
            "2019-02-11 09:20:34,257 epoch 37 - iter 20/27 - loss 0.01351499\n",
            "2019-02-11 09:20:35,099 epoch 37 - iter 22/27 - loss 0.01380888\n",
            "2019-02-11 09:20:35,775 epoch 37 - iter 24/27 - loss 0.01384593\n",
            "2019-02-11 09:20:36,280 epoch 37 - iter 26/27 - loss 0.01474976\n",
            "2019-02-11 09:20:36,291 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:20:36,293 EPOCH 37 done: loss 0.0147 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 09:20:36,947 DEV  : loss 0.01991834 - f-score 0.7579 - acc 0.7579\n",
            "2019-02-11 09:20:40,735 TEST : loss 0.00639561 - f-score 0.9209 - acc 0.9209\n",
            "2019-02-11 09:20:57,800 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:20:58,716 epoch 38 - iter 0/27 - loss 0.00881709\n",
            "2019-02-11 09:20:59,253 epoch 38 - iter 2/27 - loss 0.01016542\n",
            "2019-02-11 09:21:00,103 epoch 38 - iter 4/27 - loss 0.01446397\n",
            "2019-02-11 09:21:00,847 epoch 38 - iter 6/27 - loss 0.01386842\n",
            "2019-02-11 09:21:01,614 epoch 38 - iter 8/27 - loss 0.01363063\n",
            "2019-02-11 09:21:02,278 epoch 38 - iter 10/27 - loss 0.01389005\n",
            "2019-02-11 09:21:02,850 epoch 38 - iter 12/27 - loss 0.01418111\n",
            "2019-02-11 09:21:03,492 epoch 38 - iter 14/27 - loss 0.01395542\n",
            "2019-02-11 09:21:04,234 epoch 38 - iter 16/27 - loss 0.01410072\n",
            "2019-02-11 09:21:05,141 epoch 38 - iter 18/27 - loss 0.01444032\n",
            "2019-02-11 09:21:05,618 epoch 38 - iter 20/27 - loss 0.01400777\n",
            "2019-02-11 09:21:06,395 epoch 38 - iter 22/27 - loss 0.01408212\n",
            "2019-02-11 09:21:07,621 epoch 38 - iter 24/27 - loss 0.01455201\n",
            "2019-02-11 09:21:08,052 epoch 38 - iter 26/27 - loss 0.01494773\n",
            "2019-02-11 09:21:08,063 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:21:08,064 EPOCH 38 done: loss 0.0149 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:21:08,701 DEV  : loss 0.01849354 - f-score 0.7684 - acc 0.7684\n",
            "2019-02-11 09:21:12,420 TEST : loss 0.00741557 - f-score 0.9103 - acc 0.9103\n",
            "2019-02-11 09:21:12,423 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:21:12,717 epoch 39 - iter 0/27 - loss 0.01388238\n",
            "2019-02-11 09:21:13,650 epoch 39 - iter 2/27 - loss 0.01164886\n",
            "2019-02-11 09:21:14,430 epoch 39 - iter 4/27 - loss 0.01344933\n",
            "2019-02-11 09:21:15,082 epoch 39 - iter 6/27 - loss 0.01303090\n",
            "2019-02-11 09:21:15,776 epoch 39 - iter 8/27 - loss 0.01231699\n",
            "2019-02-11 09:21:16,684 epoch 39 - iter 10/27 - loss 0.01276764\n",
            "2019-02-11 09:21:17,719 epoch 39 - iter 12/27 - loss 0.01424655\n",
            "2019-02-11 09:21:18,416 epoch 39 - iter 14/27 - loss 0.01492791\n",
            "2019-02-11 09:21:18,986 epoch 39 - iter 16/27 - loss 0.01473538\n",
            "2019-02-11 09:21:19,783 epoch 39 - iter 18/27 - loss 0.01456282\n",
            "2019-02-11 09:21:20,700 epoch 39 - iter 20/27 - loss 0.01428490\n",
            "2019-02-11 09:21:21,935 epoch 39 - iter 22/27 - loss 0.01421205\n",
            "2019-02-11 09:21:22,658 epoch 39 - iter 24/27 - loss 0.01408501\n",
            "2019-02-11 09:21:23,154 epoch 39 - iter 26/27 - loss 0.01405387\n",
            "2019-02-11 09:21:23,165 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:21:23,166 EPOCH 39 done: loss 0.0141 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 09:21:23,799 DEV  : loss 0.01970619 - f-score 0.7368 - acc 0.7368\n",
            "2019-02-11 09:21:27,424 TEST : loss 0.00758338 - f-score 0.8996 - acc 0.8996\n",
            "2019-02-11 09:21:44,769 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:21:45,317 epoch 40 - iter 0/27 - loss 0.00998211\n",
            "2019-02-11 09:21:46,079 epoch 40 - iter 2/27 - loss 0.01443239\n",
            "2019-02-11 09:21:46,702 epoch 40 - iter 4/27 - loss 0.01336845\n",
            "2019-02-11 09:21:47,458 epoch 40 - iter 6/27 - loss 0.01328638\n",
            "2019-02-11 09:21:48,651 epoch 40 - iter 8/27 - loss 0.01461491\n",
            "2019-02-11 09:21:49,620 epoch 40 - iter 10/27 - loss 0.01603004\n",
            "2019-02-11 09:21:50,367 epoch 40 - iter 12/27 - loss 0.01575717\n",
            "2019-02-11 09:21:51,005 epoch 40 - iter 14/27 - loss 0.01575670\n",
            "2019-02-11 09:21:52,164 epoch 40 - iter 16/27 - loss 0.01516080\n",
            "2019-02-11 09:21:52,944 epoch 40 - iter 18/27 - loss 0.01515810\n",
            "2019-02-11 09:21:53,566 epoch 40 - iter 20/27 - loss 0.01482927\n",
            "2019-02-11 09:21:54,352 epoch 40 - iter 22/27 - loss 0.01447057\n",
            "2019-02-11 09:21:55,070 epoch 40 - iter 24/27 - loss 0.01442632\n",
            "2019-02-11 09:21:55,506 epoch 40 - iter 26/27 - loss 0.01490726\n",
            "2019-02-11 09:21:55,518 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:21:55,519 EPOCH 40 done: loss 0.0149 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:21:56,152 DEV  : loss 0.01762581 - f-score 0.8000 - acc 0.8000\n",
            "2019-02-11 09:22:00,010 TEST : loss 0.00590300 - f-score 0.9338 - acc 0.9338\n",
            "2019-02-11 09:22:00,013 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:22:00,306 epoch 41 - iter 0/27 - loss 0.00979636\n",
            "2019-02-11 09:22:01,096 epoch 41 - iter 2/27 - loss 0.01902282\n",
            "2019-02-11 09:22:02,373 epoch 41 - iter 4/27 - loss 0.01584203\n",
            "2019-02-11 09:22:03,565 epoch 41 - iter 6/27 - loss 0.01416529\n",
            "2019-02-11 09:22:04,362 epoch 41 - iter 8/27 - loss 0.01347992\n",
            "2019-02-11 09:22:04,849 epoch 41 - iter 10/27 - loss 0.01309347\n",
            "2019-02-11 09:22:05,674 epoch 41 - iter 12/27 - loss 0.01254318\n",
            "2019-02-11 09:22:06,297 epoch 41 - iter 14/27 - loss 0.01271367\n",
            "2019-02-11 09:22:07,025 epoch 41 - iter 16/27 - loss 0.01231850\n",
            "2019-02-11 09:22:07,959 epoch 41 - iter 18/27 - loss 0.01230009\n",
            "2019-02-11 09:22:08,604 epoch 41 - iter 20/27 - loss 0.01239666\n",
            "2019-02-11 09:22:09,433 epoch 41 - iter 22/27 - loss 0.01274454\n",
            "2019-02-11 09:22:10,205 epoch 41 - iter 24/27 - loss 0.01250225\n",
            "2019-02-11 09:22:10,653 epoch 41 - iter 26/27 - loss 0.01317524\n",
            "2019-02-11 09:22:10,663 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:22:10,665 EPOCH 41 done: loss 0.0132 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 09:22:11,297 DEV  : loss 0.02292046 - f-score 0.7368 - acc 0.7368\n",
            "2019-02-11 09:22:14,953 TEST : loss 0.01128489 - f-score 0.8376 - acc 0.8376\n",
            "2019-02-11 09:22:32,085 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:22:32,683 epoch 42 - iter 0/27 - loss 0.03359424\n",
            "2019-02-11 09:22:33,380 epoch 42 - iter 2/27 - loss 0.01980117\n",
            "2019-02-11 09:22:34,070 epoch 42 - iter 4/27 - loss 0.01485917\n",
            "2019-02-11 09:22:34,949 epoch 42 - iter 6/27 - loss 0.01250189\n",
            "2019-02-11 09:22:35,677 epoch 42 - iter 8/27 - loss 0.01271642\n",
            "2019-02-11 09:22:36,457 epoch 42 - iter 10/27 - loss 0.01239657\n",
            "2019-02-11 09:22:37,151 epoch 42 - iter 12/27 - loss 0.01268360\n",
            "2019-02-11 09:22:38,052 epoch 42 - iter 14/27 - loss 0.01300949\n",
            "2019-02-11 09:22:39,383 epoch 42 - iter 16/27 - loss 0.01360158\n",
            "2019-02-11 09:22:40,070 epoch 42 - iter 18/27 - loss 0.01354230\n",
            "2019-02-11 09:22:41,281 epoch 42 - iter 20/27 - loss 0.01300938\n",
            "2019-02-11 09:22:42,050 epoch 42 - iter 22/27 - loss 0.01337191\n",
            "2019-02-11 09:22:42,721 epoch 42 - iter 24/27 - loss 0.01339650\n",
            "2019-02-11 09:22:43,103 epoch 42 - iter 26/27 - loss 0.01362439\n",
            "2019-02-11 09:22:43,115 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:22:43,116 EPOCH 42 done: loss 0.0136 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:22:43,753 DEV  : loss 0.01579545 - f-score 0.8105 - acc 0.8105\n",
            "2019-02-11 09:22:47,565 TEST : loss 0.00552813 - f-score 0.9316 - acc 0.9316\n",
            "2019-02-11 09:22:47,568 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:22:47,950 epoch 43 - iter 0/27 - loss 0.00740820\n",
            "2019-02-11 09:22:48,840 epoch 43 - iter 2/27 - loss 0.01400115\n",
            "2019-02-11 09:22:49,686 epoch 43 - iter 4/27 - loss 0.01387525\n",
            "2019-02-11 09:22:50,745 epoch 43 - iter 6/27 - loss 0.01295552\n",
            "2019-02-11 09:22:51,564 epoch 43 - iter 8/27 - loss 0.01313784\n",
            "2019-02-11 09:22:52,796 epoch 43 - iter 10/27 - loss 0.01254496\n",
            "2019-02-11 09:22:53,889 epoch 43 - iter 12/27 - loss 0.01250594\n",
            "2019-02-11 09:22:54,579 epoch 43 - iter 14/27 - loss 0.01268832\n",
            "2019-02-11 09:22:55,303 epoch 43 - iter 16/27 - loss 0.01292032\n",
            "2019-02-11 09:22:55,945 epoch 43 - iter 18/27 - loss 0.01306786\n",
            "2019-02-11 09:22:56,594 epoch 43 - iter 20/27 - loss 0.01341081\n",
            "2019-02-11 09:22:57,312 epoch 43 - iter 22/27 - loss 0.01340905\n",
            "2019-02-11 09:22:57,924 epoch 43 - iter 24/27 - loss 0.01312671\n",
            "2019-02-11 09:22:58,331 epoch 43 - iter 26/27 - loss 0.01387219\n",
            "2019-02-11 09:22:58,342 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:22:58,344 EPOCH 43 done: loss 0.0139 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 09:22:58,992 DEV  : loss 0.01788246 - f-score 0.7684 - acc 0.7684\n",
            "2019-02-11 09:23:02,662 TEST : loss 0.00655702 - f-score 0.9060 - acc 0.9060\n",
            "2019-02-11 09:23:02,665 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:23:03,159 epoch 44 - iter 0/27 - loss 0.01307180\n",
            "2019-02-11 09:23:03,980 epoch 44 - iter 2/27 - loss 0.01373130\n",
            "2019-02-11 09:23:04,509 epoch 44 - iter 4/27 - loss 0.01189105\n",
            "2019-02-11 09:23:05,246 epoch 44 - iter 6/27 - loss 0.01135476\n",
            "2019-02-11 09:23:06,135 epoch 44 - iter 8/27 - loss 0.01312344\n",
            "2019-02-11 09:23:06,776 epoch 44 - iter 10/27 - loss 0.01296149\n",
            "2019-02-11 09:23:07,531 epoch 44 - iter 12/27 - loss 0.01249477\n",
            "2019-02-11 09:23:08,179 epoch 44 - iter 14/27 - loss 0.01251502\n",
            "2019-02-11 09:23:09,118 epoch 44 - iter 16/27 - loss 0.01233015\n",
            "2019-02-11 09:23:09,794 epoch 44 - iter 18/27 - loss 0.01247797\n",
            "2019-02-11 09:23:10,979 epoch 44 - iter 20/27 - loss 0.01218725\n",
            "2019-02-11 09:23:12,211 epoch 44 - iter 22/27 - loss 0.01214766\n",
            "2019-02-11 09:23:12,902 epoch 44 - iter 24/27 - loss 0.01219689\n",
            "2019-02-11 09:23:13,284 epoch 44 - iter 26/27 - loss 0.01192474\n",
            "2019-02-11 09:23:13,297 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:23:13,299 EPOCH 44 done: loss 0.0119 - lr 0.1000 - bad epochs 2\n",
            "2019-02-11 09:23:13,935 DEV  : loss 0.01537817 - f-score 0.8000 - acc 0.8000\n",
            "2019-02-11 09:23:17,590 TEST : loss 0.00408632 - f-score 0.9530 - acc 0.9530\n",
            "2019-02-11 09:23:34,642 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:23:35,134 epoch 45 - iter 0/27 - loss 0.00498993\n",
            "2019-02-11 09:23:35,894 epoch 45 - iter 2/27 - loss 0.00766039\n",
            "2019-02-11 09:23:36,569 epoch 45 - iter 4/27 - loss 0.00812869\n",
            "2019-02-11 09:23:37,442 epoch 45 - iter 6/27 - loss 0.00758709\n",
            "2019-02-11 09:23:38,181 epoch 45 - iter 8/27 - loss 0.00948044\n",
            "2019-02-11 09:23:39,112 epoch 45 - iter 10/27 - loss 0.00975092\n",
            "2019-02-11 09:23:39,874 epoch 45 - iter 12/27 - loss 0.01023489\n",
            "2019-02-11 09:23:40,635 epoch 45 - iter 14/27 - loss 0.01118060\n",
            "2019-02-11 09:23:41,353 epoch 45 - iter 16/27 - loss 0.01076787\n",
            "2019-02-11 09:23:42,022 epoch 45 - iter 18/27 - loss 0.01081961\n",
            "2019-02-11 09:23:42,767 epoch 45 - iter 20/27 - loss 0.01055356\n",
            "2019-02-11 09:23:43,464 epoch 45 - iter 22/27 - loss 0.01062607\n",
            "2019-02-11 09:23:44,269 epoch 45 - iter 24/27 - loss 0.01083903\n",
            "2019-02-11 09:23:45,265 epoch 45 - iter 26/27 - loss 0.01259960\n",
            "2019-02-11 09:23:45,277 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:23:45,278 EPOCH 45 done: loss 0.0126 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:23:45,909 DEV  : loss 0.01490405 - f-score 0.7895 - acc 0.7895\n",
            "2019-02-11 09:23:49,596 TEST : loss 0.00397849 - f-score 0.9594 - acc 0.9594\n",
            "2019-02-11 09:23:49,598 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:23:49,868 epoch 46 - iter 0/27 - loss 0.00827464\n",
            "2019-02-11 09:23:50,535 epoch 46 - iter 2/27 - loss 0.00806441\n",
            "2019-02-11 09:23:51,231 epoch 46 - iter 4/27 - loss 0.00949467\n",
            "2019-02-11 09:23:52,180 epoch 46 - iter 6/27 - loss 0.01090834\n",
            "2019-02-11 09:23:53,023 epoch 46 - iter 8/27 - loss 0.01089751\n",
            "2019-02-11 09:23:54,247 epoch 46 - iter 10/27 - loss 0.01142391\n",
            "2019-02-11 09:23:54,979 epoch 46 - iter 12/27 - loss 0.01200310\n",
            "2019-02-11 09:23:55,733 epoch 46 - iter 14/27 - loss 0.01150384\n",
            "2019-02-11 09:23:56,436 epoch 46 - iter 16/27 - loss 0.01155449\n",
            "2019-02-11 09:23:57,319 epoch 46 - iter 18/27 - loss 0.01153030\n",
            "2019-02-11 09:23:58,529 epoch 46 - iter 20/27 - loss 0.01240307\n",
            "2019-02-11 09:23:59,237 epoch 46 - iter 22/27 - loss 0.01217659\n",
            "2019-02-11 09:24:00,077 epoch 46 - iter 24/27 - loss 0.01178104\n",
            "2019-02-11 09:24:00,451 epoch 46 - iter 26/27 - loss 0.01165447\n",
            "2019-02-11 09:24:00,462 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:24:00,464 EPOCH 46 done: loss 0.0117 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 09:24:01,084 DEV  : loss 0.01565753 - f-score 0.8105 - acc 0.8105\n",
            "2019-02-11 09:24:04,696 TEST : loss 0.00461775 - f-score 0.9509 - acc 0.9509\n",
            "2019-02-11 09:24:21,724 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:24:22,089 epoch 47 - iter 0/27 - loss 0.00493770\n",
            "2019-02-11 09:24:23,051 epoch 47 - iter 2/27 - loss 0.00888775\n",
            "2019-02-11 09:24:23,927 epoch 47 - iter 4/27 - loss 0.00982306\n",
            "2019-02-11 09:24:25,088 epoch 47 - iter 6/27 - loss 0.01030319\n",
            "2019-02-11 09:24:25,772 epoch 47 - iter 8/27 - loss 0.00920078\n",
            "2019-02-11 09:24:27,157 epoch 47 - iter 10/27 - loss 0.00968218\n",
            "2019-02-11 09:24:27,790 epoch 47 - iter 12/27 - loss 0.00991930\n",
            "2019-02-11 09:24:28,575 epoch 47 - iter 14/27 - loss 0.00975783\n",
            "2019-02-11 09:24:29,276 epoch 47 - iter 16/27 - loss 0.00937041\n",
            "2019-02-11 09:24:29,998 epoch 47 - iter 18/27 - loss 0.00957814\n",
            "2019-02-11 09:24:30,742 epoch 47 - iter 20/27 - loss 0.00958764\n",
            "2019-02-11 09:24:31,340 epoch 47 - iter 22/27 - loss 0.00953169\n",
            "2019-02-11 09:24:31,981 epoch 47 - iter 24/27 - loss 0.00988109\n",
            "2019-02-11 09:24:32,506 epoch 47 - iter 26/27 - loss 0.01036305\n",
            "2019-02-11 09:24:32,518 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:24:32,519 EPOCH 47 done: loss 0.0104 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:24:33,159 DEV  : loss 0.01492179 - f-score 0.8211 - acc 0.8211\n",
            "2019-02-11 09:24:36,825 TEST : loss 0.00299395 - f-score 0.9615 - acc 0.9615\n",
            "2019-02-11 09:24:53,885 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:24:54,656 epoch 48 - iter 0/27 - loss 0.01303393\n",
            "2019-02-11 09:24:55,320 epoch 48 - iter 2/27 - loss 0.01281843\n",
            "2019-02-11 09:24:55,973 epoch 48 - iter 4/27 - loss 0.01169141\n",
            "2019-02-11 09:24:56,718 epoch 48 - iter 6/27 - loss 0.01203886\n",
            "2019-02-11 09:24:57,428 epoch 48 - iter 8/27 - loss 0.01111901\n",
            "2019-02-11 09:24:58,114 epoch 48 - iter 10/27 - loss 0.01164570\n",
            "2019-02-11 09:24:58,836 epoch 48 - iter 12/27 - loss 0.01077580\n",
            "2019-02-11 09:25:00,023 epoch 48 - iter 14/27 - loss 0.01097153\n",
            "2019-02-11 09:25:00,803 epoch 48 - iter 16/27 - loss 0.01104787\n",
            "2019-02-11 09:25:01,577 epoch 48 - iter 18/27 - loss 0.01116345\n",
            "2019-02-11 09:25:02,387 epoch 48 - iter 20/27 - loss 0.01118805\n",
            "2019-02-11 09:25:03,662 epoch 48 - iter 22/27 - loss 0.01155201\n",
            "2019-02-11 09:25:04,293 epoch 48 - iter 24/27 - loss 0.01153636\n",
            "2019-02-11 09:25:04,722 epoch 48 - iter 26/27 - loss 0.01221736\n",
            "2019-02-11 09:25:04,733 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:25:04,734 EPOCH 48 done: loss 0.0122 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:25:05,362 DEV  : loss 0.01654482 - f-score 0.7789 - acc 0.7789\n",
            "2019-02-11 09:25:09,156 TEST : loss 0.00317202 - f-score 0.9701 - acc 0.9701\n",
            "2019-02-11 09:25:09,159 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:25:09,515 epoch 49 - iter 0/27 - loss 0.00804388\n",
            "2019-02-11 09:25:10,504 epoch 49 - iter 2/27 - loss 0.01035977\n",
            "2019-02-11 09:25:11,258 epoch 49 - iter 4/27 - loss 0.01208120\n",
            "2019-02-11 09:25:11,889 epoch 49 - iter 6/27 - loss 0.01050398\n",
            "2019-02-11 09:25:12,560 epoch 49 - iter 8/27 - loss 0.00942133\n",
            "2019-02-11 09:25:13,552 epoch 49 - iter 10/27 - loss 0.00916469\n",
            "2019-02-11 09:25:14,718 epoch 49 - iter 12/27 - loss 0.00910643\n",
            "2019-02-11 09:25:15,322 epoch 49 - iter 14/27 - loss 0.00934109\n",
            "2019-02-11 09:25:16,021 epoch 49 - iter 16/27 - loss 0.00961157\n",
            "2019-02-11 09:25:16,752 epoch 49 - iter 18/27 - loss 0.00926584\n",
            "2019-02-11 09:25:17,333 epoch 49 - iter 20/27 - loss 0.00991379\n",
            "2019-02-11 09:25:18,565 epoch 49 - iter 22/27 - loss 0.01006524\n",
            "2019-02-11 09:25:19,402 epoch 49 - iter 24/27 - loss 0.00977755\n",
            "2019-02-11 09:25:19,847 epoch 49 - iter 26/27 - loss 0.01101307\n",
            "2019-02-11 09:25:19,859 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:25:19,861 EPOCH 49 done: loss 0.0110 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 09:25:20,491 DEV  : loss 0.02107446 - f-score 0.7474 - acc 0.7474\n",
            "2019-02-11 09:25:24,156 TEST : loss 0.00376097 - f-score 0.9573 - acc 0.9573\n",
            "2019-02-11 09:25:24,159 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:25:24,485 epoch 50 - iter 0/27 - loss 0.00699106\n",
            "2019-02-11 09:25:25,006 epoch 50 - iter 2/27 - loss 0.00705506\n",
            "2019-02-11 09:25:26,109 epoch 50 - iter 4/27 - loss 0.00799880\n",
            "2019-02-11 09:25:27,151 epoch 50 - iter 6/27 - loss 0.00743727\n",
            "2019-02-11 09:25:27,885 epoch 50 - iter 8/27 - loss 0.00941087\n",
            "2019-02-11 09:25:28,984 epoch 50 - iter 10/27 - loss 0.00948080\n",
            "2019-02-11 09:25:29,727 epoch 50 - iter 12/27 - loss 0.01007677\n",
            "2019-02-11 09:25:30,457 epoch 50 - iter 14/27 - loss 0.00994098\n",
            "2019-02-11 09:25:31,017 epoch 50 - iter 16/27 - loss 0.00994680\n",
            "2019-02-11 09:25:31,822 epoch 50 - iter 18/27 - loss 0.00961085\n",
            "2019-02-11 09:25:32,705 epoch 50 - iter 20/27 - loss 0.00959188\n",
            "2019-02-11 09:25:33,289 epoch 50 - iter 22/27 - loss 0.00986664\n",
            "2019-02-11 09:25:33,953 epoch 50 - iter 24/27 - loss 0.01001904\n",
            "2019-02-11 09:25:34,400 epoch 50 - iter 26/27 - loss 0.01012256\n",
            "2019-02-11 09:25:34,412 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:25:34,413 EPOCH 50 done: loss 0.0101 - lr 0.1000 - bad epochs 2\n",
            "2019-02-11 09:25:35,093 DEV  : loss 0.02026895 - f-score 0.7789 - acc 0.7789\n",
            "2019-02-11 09:25:39,068 TEST : loss 0.00275730 - f-score 0.9744 - acc 0.9744\n",
            "2019-02-11 09:25:56,709 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:25:57,160 epoch 51 - iter 0/27 - loss 0.00846365\n",
            "2019-02-11 09:25:57,950 epoch 51 - iter 2/27 - loss 0.00643574\n",
            "2019-02-11 09:25:58,694 epoch 51 - iter 4/27 - loss 0.00832261\n",
            "2019-02-11 09:25:59,520 epoch 51 - iter 6/27 - loss 0.00910989\n",
            "2019-02-11 09:26:00,275 epoch 51 - iter 8/27 - loss 0.00969539\n",
            "2019-02-11 09:26:00,890 epoch 51 - iter 10/27 - loss 0.00890601\n",
            "2019-02-11 09:26:01,567 epoch 51 - iter 12/27 - loss 0.00840635\n",
            "2019-02-11 09:26:02,734 epoch 51 - iter 14/27 - loss 0.00789453\n",
            "2019-02-11 09:26:03,526 epoch 51 - iter 16/27 - loss 0.00875702\n",
            "2019-02-11 09:26:04,991 epoch 51 - iter 18/27 - loss 0.00844082\n",
            "2019-02-11 09:26:05,844 epoch 51 - iter 20/27 - loss 0.00853050\n",
            "2019-02-11 09:26:06,430 epoch 51 - iter 22/27 - loss 0.00869258\n",
            "2019-02-11 09:26:07,095 epoch 51 - iter 24/27 - loss 0.00852305\n",
            "2019-02-11 09:26:07,546 epoch 51 - iter 26/27 - loss 0.00851751\n",
            "2019-02-11 09:26:07,559 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:26:07,560 EPOCH 51 done: loss 0.0085 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:26:08,204 DEV  : loss 0.01990024 - f-score 0.8000 - acc 0.8000\n",
            "2019-02-11 09:26:11,879 TEST : loss 0.00321795 - f-score 0.9573 - acc 0.9573\n",
            "2019-02-11 09:26:28,991 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:26:29,685 epoch 52 - iter 0/27 - loss 0.00753980\n",
            "2019-02-11 09:26:30,402 epoch 52 - iter 2/27 - loss 0.00862962\n",
            "2019-02-11 09:26:31,302 epoch 52 - iter 4/27 - loss 0.00802962\n",
            "2019-02-11 09:26:32,525 epoch 52 - iter 6/27 - loss 0.00731717\n",
            "2019-02-11 09:26:33,407 epoch 52 - iter 8/27 - loss 0.00707223\n",
            "2019-02-11 09:26:34,064 epoch 52 - iter 10/27 - loss 0.00708911\n",
            "2019-02-11 09:26:34,848 epoch 52 - iter 12/27 - loss 0.00725987\n",
            "2019-02-11 09:26:35,522 epoch 52 - iter 14/27 - loss 0.00721202\n",
            "2019-02-11 09:26:36,299 epoch 52 - iter 16/27 - loss 0.00713449\n",
            "2019-02-11 09:26:37,020 epoch 52 - iter 18/27 - loss 0.00711999\n",
            "2019-02-11 09:26:37,752 epoch 52 - iter 20/27 - loss 0.00708493\n",
            "2019-02-11 09:26:38,662 epoch 52 - iter 22/27 - loss 0.00703172\n",
            "2019-02-11 09:26:39,296 epoch 52 - iter 24/27 - loss 0.00729684\n",
            "2019-02-11 09:26:40,142 epoch 52 - iter 26/27 - loss 0.00734652\n",
            "2019-02-11 09:26:40,153 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:26:40,154 EPOCH 52 done: loss 0.0073 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:26:40,779 DEV  : loss 0.01674315 - f-score 0.8211 - acc 0.8211\n",
            "2019-02-11 09:26:44,630 TEST : loss 0.00159845 - f-score 0.9808 - acc 0.9808\n",
            "2019-02-11 09:27:01,680 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:27:02,451 epoch 53 - iter 0/27 - loss 0.00556475\n",
            "2019-02-11 09:27:03,243 epoch 53 - iter 2/27 - loss 0.00721287\n",
            "2019-02-11 09:27:04,251 epoch 53 - iter 4/27 - loss 0.00583015\n",
            "2019-02-11 09:27:05,102 epoch 53 - iter 6/27 - loss 0.00634881\n",
            "2019-02-11 09:27:05,664 epoch 53 - iter 8/27 - loss 0.00615257\n",
            "2019-02-11 09:27:06,350 epoch 53 - iter 10/27 - loss 0.00721306\n",
            "2019-02-11 09:27:07,197 epoch 53 - iter 12/27 - loss 0.00778388\n",
            "2019-02-11 09:27:08,323 epoch 53 - iter 14/27 - loss 0.00831427\n",
            "2019-02-11 09:27:09,556 epoch 53 - iter 16/27 - loss 0.00785833\n",
            "2019-02-11 09:27:10,326 epoch 53 - iter 18/27 - loss 0.00789363\n",
            "2019-02-11 09:27:11,228 epoch 53 - iter 20/27 - loss 0.00780444\n",
            "2019-02-11 09:27:11,728 epoch 53 - iter 22/27 - loss 0.00801233\n",
            "2019-02-11 09:27:12,320 epoch 53 - iter 24/27 - loss 0.00824656\n",
            "2019-02-11 09:27:12,711 epoch 53 - iter 26/27 - loss 0.00849452\n",
            "2019-02-11 09:27:12,722 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:27:12,723 EPOCH 53 done: loss 0.0085 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:27:13,348 DEV  : loss 0.01858117 - f-score 0.8000 - acc 0.8000\n",
            "2019-02-11 09:27:17,042 TEST : loss 0.00199941 - f-score 0.9808 - acc 0.9808\n",
            "2019-02-11 09:27:17,045 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:27:17,459 epoch 54 - iter 0/27 - loss 0.00263505\n",
            "2019-02-11 09:27:18,483 epoch 54 - iter 2/27 - loss 0.00816618\n",
            "2019-02-11 09:27:19,089 epoch 54 - iter 4/27 - loss 0.00814533\n",
            "2019-02-11 09:27:20,232 epoch 54 - iter 6/27 - loss 0.00912786\n",
            "2019-02-11 09:27:20,933 epoch 54 - iter 8/27 - loss 0.00774257\n",
            "2019-02-11 09:27:22,162 epoch 54 - iter 10/27 - loss 0.00818222\n",
            "2019-02-11 09:27:23,030 epoch 54 - iter 12/27 - loss 0.00808668\n",
            "2019-02-11 09:27:23,602 epoch 54 - iter 14/27 - loss 0.00721231\n",
            "2019-02-11 09:27:24,356 epoch 54 - iter 16/27 - loss 0.00764712\n",
            "2019-02-11 09:27:25,264 epoch 54 - iter 18/27 - loss 0.00755043\n",
            "2019-02-11 09:27:25,896 epoch 54 - iter 20/27 - loss 0.00783333\n",
            "2019-02-11 09:27:26,646 epoch 54 - iter 22/27 - loss 0.00840693\n",
            "2019-02-11 09:27:27,405 epoch 54 - iter 24/27 - loss 0.00815231\n",
            "2019-02-11 09:27:27,843 epoch 54 - iter 26/27 - loss 0.00845562\n",
            "2019-02-11 09:27:27,854 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:27:27,855 EPOCH 54 done: loss 0.0085 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 09:27:28,503 DEV  : loss 0.01686791 - f-score 0.8316 - acc 0.8316\n",
            "2019-02-11 09:27:32,360 TEST : loss 0.00160051 - f-score 0.9872 - acc 0.9872\n",
            "2019-02-11 09:27:32,363 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:27:32,751 epoch 55 - iter 0/27 - loss 0.00944628\n",
            "2019-02-11 09:27:33,521 epoch 55 - iter 2/27 - loss 0.00558567\n",
            "2019-02-11 09:27:34,141 epoch 55 - iter 4/27 - loss 0.00586137\n",
            "2019-02-11 09:27:34,846 epoch 55 - iter 6/27 - loss 0.00632118\n",
            "2019-02-11 09:27:35,513 epoch 55 - iter 8/27 - loss 0.00662297\n",
            "2019-02-11 09:27:36,412 epoch 55 - iter 10/27 - loss 0.00662258\n",
            "2019-02-11 09:27:37,167 epoch 55 - iter 12/27 - loss 0.00703539\n",
            "2019-02-11 09:27:38,157 epoch 55 - iter 14/27 - loss 0.00722424\n",
            "2019-02-11 09:27:38,892 epoch 55 - iter 16/27 - loss 0.00695698\n",
            "2019-02-11 09:27:39,790 epoch 55 - iter 18/27 - loss 0.00731446\n",
            "2019-02-11 09:27:40,901 epoch 55 - iter 20/27 - loss 0.00718554\n",
            "2019-02-11 09:27:42,167 epoch 55 - iter 22/27 - loss 0.00738758\n",
            "2019-02-11 09:27:42,874 epoch 55 - iter 24/27 - loss 0.00737922\n",
            "2019-02-11 09:27:43,337 epoch 55 - iter 26/27 - loss 0.00759668\n",
            "2019-02-11 09:27:43,349 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:27:43,351 EPOCH 55 done: loss 0.0076 - lr 0.1000 - bad epochs 2\n",
            "2019-02-11 09:27:43,987 DEV  : loss 0.01638570 - f-score 0.8421 - acc 0.8421\n",
            "2019-02-11 09:27:47,687 TEST : loss 0.00200693 - f-score 0.9744 - acc 0.9744\n",
            "2019-02-11 09:27:47,692 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:27:48,038 epoch 56 - iter 0/27 - loss 0.00459496\n",
            "2019-02-11 09:27:48,743 epoch 56 - iter 2/27 - loss 0.00506225\n",
            "2019-02-11 09:27:49,471 epoch 56 - iter 4/27 - loss 0.00466531\n",
            "2019-02-11 09:27:50,176 epoch 56 - iter 6/27 - loss 0.00465250\n",
            "2019-02-11 09:27:50,799 epoch 56 - iter 8/27 - loss 0.00490247\n",
            "2019-02-11 09:27:51,620 epoch 56 - iter 10/27 - loss 0.00511232\n",
            "2019-02-11 09:27:52,276 epoch 56 - iter 12/27 - loss 0.00561759\n",
            "2019-02-11 09:27:53,008 epoch 56 - iter 14/27 - loss 0.00571147\n",
            "2019-02-11 09:27:53,937 epoch 56 - iter 16/27 - loss 0.00578650\n",
            "2019-02-11 09:27:54,509 epoch 56 - iter 18/27 - loss 0.00684818\n",
            "2019-02-11 09:27:55,277 epoch 56 - iter 20/27 - loss 0.00697345\n",
            "2019-02-11 09:27:56,533 epoch 56 - iter 22/27 - loss 0.00719054\n",
            "2019-02-11 09:27:57,820 epoch 56 - iter 24/27 - loss 0.00771862\n",
            "2019-02-11 09:27:58,367 epoch 56 - iter 26/27 - loss 0.00786712\n",
            "2019-02-11 09:27:58,380 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:27:58,381 EPOCH 56 done: loss 0.0079 - lr 0.1000 - bad epochs 3\n",
            "2019-02-11 09:27:59,019 DEV  : loss 0.01504991 - f-score 0.8211 - acc 0.8211\n",
            "2019-02-11 09:28:02,703 TEST : loss 0.00138118 - f-score 0.9850 - acc 0.9850\n",
            "Epoch    55: reducing learning rate of group 0 to 5.0000e-02.\n",
            "2019-02-11 09:28:02,705 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:28:03,202 epoch 57 - iter 0/27 - loss 0.00151454\n",
            "2019-02-11 09:28:03,953 epoch 57 - iter 2/27 - loss 0.00278501\n",
            "2019-02-11 09:28:04,627 epoch 57 - iter 4/27 - loss 0.00568632\n",
            "2019-02-11 09:28:05,773 epoch 57 - iter 6/27 - loss 0.00526818\n",
            "2019-02-11 09:28:06,384 epoch 57 - iter 8/27 - loss 0.00536121\n",
            "2019-02-11 09:28:07,503 epoch 57 - iter 10/27 - loss 0.00508842\n",
            "2019-02-11 09:28:08,297 epoch 57 - iter 12/27 - loss 0.00515629\n",
            "2019-02-11 09:28:09,064 epoch 57 - iter 14/27 - loss 0.00568091\n",
            "2019-02-11 09:28:09,824 epoch 57 - iter 16/27 - loss 0.00546676\n",
            "2019-02-11 09:28:10,517 epoch 57 - iter 18/27 - loss 0.00583226\n",
            "2019-02-11 09:28:11,147 epoch 57 - iter 20/27 - loss 0.00548957\n",
            "2019-02-11 09:28:11,698 epoch 57 - iter 22/27 - loss 0.00599526\n",
            "2019-02-11 09:28:12,520 epoch 57 - iter 24/27 - loss 0.00623787\n",
            "2019-02-11 09:28:13,487 epoch 57 - iter 26/27 - loss 0.00659569\n",
            "2019-02-11 09:28:13,498 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:28:13,499 EPOCH 57 done: loss 0.0066 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 09:28:14,137 DEV  : loss 0.01586357 - f-score 0.8421 - acc 0.8421\n",
            "2019-02-11 09:28:17,941 TEST : loss 0.00143741 - f-score 0.9893 - acc 0.9893\n",
            "2019-02-11 09:28:34,962 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:28:35,386 epoch 58 - iter 0/27 - loss 0.00435689\n",
            "2019-02-11 09:28:36,385 epoch 58 - iter 2/27 - loss 0.00616395\n",
            "2019-02-11 09:28:37,306 epoch 58 - iter 4/27 - loss 0.00532646\n",
            "2019-02-11 09:28:38,150 epoch 58 - iter 6/27 - loss 0.00546760\n",
            "2019-02-11 09:28:38,831 epoch 58 - iter 8/27 - loss 0.00558954\n",
            "2019-02-11 09:28:39,539 epoch 58 - iter 10/27 - loss 0.00548375\n",
            "2019-02-11 09:28:40,147 epoch 58 - iter 12/27 - loss 0.00542297\n",
            "2019-02-11 09:28:41,382 epoch 58 - iter 14/27 - loss 0.00646396\n",
            "2019-02-11 09:28:42,124 epoch 58 - iter 16/27 - loss 0.00610314\n",
            "2019-02-11 09:28:42,834 epoch 58 - iter 18/27 - loss 0.00639905\n",
            "2019-02-11 09:28:43,443 epoch 58 - iter 20/27 - loss 0.00638804\n",
            "2019-02-11 09:28:44,528 epoch 58 - iter 22/27 - loss 0.00666044\n",
            "2019-02-11 09:28:45,449 epoch 58 - iter 24/27 - loss 0.00681873\n",
            "2019-02-11 09:28:45,980 epoch 58 - iter 26/27 - loss 0.00720847\n",
            "2019-02-11 09:28:45,992 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:28:45,993 EPOCH 58 done: loss 0.0072 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 09:28:46,629 DEV  : loss 0.01540492 - f-score 0.8737 - acc 0.8737\n",
            "2019-02-11 09:28:50,300 TEST : loss 0.00108924 - f-score 0.9936 - acc 0.9936\n",
            "2019-02-11 09:28:50,303 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:28:50,671 epoch 59 - iter 0/27 - loss 0.00374635\n",
            "2019-02-11 09:28:51,759 epoch 59 - iter 2/27 - loss 0.00909927\n",
            "2019-02-11 09:28:52,341 epoch 59 - iter 4/27 - loss 0.00771492\n",
            "2019-02-11 09:28:53,176 epoch 59 - iter 6/27 - loss 0.00808682\n",
            "2019-02-11 09:28:53,780 epoch 59 - iter 8/27 - loss 0.00727052\n",
            "2019-02-11 09:28:54,879 epoch 59 - iter 10/27 - loss 0.00690810\n",
            "2019-02-11 09:28:56,030 epoch 59 - iter 12/27 - loss 0.00668236\n",
            "2019-02-11 09:28:56,796 epoch 59 - iter 14/27 - loss 0.00639704\n",
            "2019-02-11 09:28:57,503 epoch 59 - iter 16/27 - loss 0.00633366\n",
            "2019-02-11 09:28:58,364 epoch 59 - iter 18/27 - loss 0.00644840\n",
            "2019-02-11 09:28:59,093 epoch 59 - iter 20/27 - loss 0.00635564\n",
            "2019-02-11 09:29:00,046 epoch 59 - iter 22/27 - loss 0.00615187\n",
            "2019-02-11 09:29:00,700 epoch 59 - iter 24/27 - loss 0.00591761\n",
            "2019-02-11 09:29:01,234 epoch 59 - iter 26/27 - loss 0.00588685\n",
            "2019-02-11 09:29:01,246 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:29:01,248 EPOCH 59 done: loss 0.0059 - lr 0.0500 - bad epochs 1\n",
            "2019-02-11 09:29:01,940 DEV  : loss 0.01560007 - f-score 0.8421 - acc 0.8421\n",
            "2019-02-11 09:29:05,852 TEST : loss 0.00085926 - f-score 0.9979 - acc 0.9979\n",
            "2019-02-11 09:29:22,882 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:29:23,572 epoch 60 - iter 0/27 - loss 0.00999487\n",
            "2019-02-11 09:29:24,398 epoch 60 - iter 2/27 - loss 0.00834002\n",
            "2019-02-11 09:29:25,278 epoch 60 - iter 4/27 - loss 0.00819738\n",
            "2019-02-11 09:29:25,785 epoch 60 - iter 6/27 - loss 0.00784360\n",
            "2019-02-11 09:29:26,470 epoch 60 - iter 8/27 - loss 0.00709606\n",
            "2019-02-11 09:29:27,639 epoch 60 - iter 10/27 - loss 0.00699111\n",
            "2019-02-11 09:29:28,374 epoch 60 - iter 12/27 - loss 0.00664644\n",
            "2019-02-11 09:29:29,086 epoch 60 - iter 14/27 - loss 0.00658717\n",
            "2019-02-11 09:29:29,794 epoch 60 - iter 16/27 - loss 0.00665462\n",
            "2019-02-11 09:29:30,856 epoch 60 - iter 18/27 - loss 0.00626690\n",
            "2019-02-11 09:29:31,717 epoch 60 - iter 20/27 - loss 0.00634120\n",
            "2019-02-11 09:29:32,464 epoch 60 - iter 22/27 - loss 0.00622188\n",
            "2019-02-11 09:29:33,252 epoch 60 - iter 24/27 - loss 0.00627851\n",
            "2019-02-11 09:29:33,851 epoch 60 - iter 26/27 - loss 0.00611535\n",
            "2019-02-11 09:29:33,862 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:29:33,863 EPOCH 60 done: loss 0.0061 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 09:29:34,497 DEV  : loss 0.01509268 - f-score 0.8526 - acc 0.8526\n",
            "2019-02-11 09:29:38,166 TEST : loss 0.00085430 - f-score 0.9915 - acc 0.9915\n",
            "2019-02-11 09:29:38,169 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:29:38,503 epoch 61 - iter 0/27 - loss 0.00471956\n",
            "2019-02-11 09:29:39,094 epoch 61 - iter 2/27 - loss 0.00558621\n",
            "2019-02-11 09:29:39,736 epoch 61 - iter 4/27 - loss 0.00590340\n",
            "2019-02-11 09:29:41,106 epoch 61 - iter 6/27 - loss 0.00538668\n",
            "2019-02-11 09:29:41,705 epoch 61 - iter 8/27 - loss 0.00480562\n",
            "2019-02-11 09:29:42,441 epoch 61 - iter 10/27 - loss 0.00514635\n",
            "2019-02-11 09:29:43,220 epoch 61 - iter 12/27 - loss 0.00450721\n",
            "2019-02-11 09:29:43,830 epoch 61 - iter 14/27 - loss 0.00436918\n",
            "2019-02-11 09:29:44,565 epoch 61 - iter 16/27 - loss 0.00507952\n",
            "2019-02-11 09:29:45,673 epoch 61 - iter 18/27 - loss 0.00515906\n",
            "2019-02-11 09:29:46,547 epoch 61 - iter 20/27 - loss 0.00518991\n",
            "2019-02-11 09:29:47,234 epoch 61 - iter 22/27 - loss 0.00526296\n",
            "2019-02-11 09:29:47,938 epoch 61 - iter 24/27 - loss 0.00527280\n",
            "2019-02-11 09:29:48,416 epoch 61 - iter 26/27 - loss 0.00514668\n",
            "2019-02-11 09:29:48,427 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:29:48,428 EPOCH 61 done: loss 0.0051 - lr 0.0500 - bad epochs 1\n",
            "2019-02-11 09:29:49,054 DEV  : loss 0.01553471 - f-score 0.8105 - acc 0.8105\n",
            "2019-02-11 09:29:52,894 TEST : loss 0.00071658 - f-score 0.9957 - acc 0.9957\n",
            "2019-02-11 09:30:10,197 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:30:10,602 epoch 62 - iter 0/27 - loss 0.00501481\n",
            "2019-02-11 09:30:11,481 epoch 62 - iter 2/27 - loss 0.00439808\n",
            "2019-02-11 09:30:12,348 epoch 62 - iter 4/27 - loss 0.00540474\n",
            "2019-02-11 09:30:13,236 epoch 62 - iter 6/27 - loss 0.00474521\n",
            "2019-02-11 09:30:14,464 epoch 62 - iter 8/27 - loss 0.00405989\n",
            "2019-02-11 09:30:15,219 epoch 62 - iter 10/27 - loss 0.00431473\n",
            "2019-02-11 09:30:15,798 epoch 62 - iter 12/27 - loss 0.00462240\n",
            "2019-02-11 09:30:16,547 epoch 62 - iter 14/27 - loss 0.00440083\n",
            "2019-02-11 09:30:17,346 epoch 62 - iter 16/27 - loss 0.00473474\n",
            "2019-02-11 09:30:17,940 epoch 62 - iter 18/27 - loss 0.00472008\n",
            "2019-02-11 09:30:18,662 epoch 62 - iter 20/27 - loss 0.00505273\n",
            "2019-02-11 09:30:19,503 epoch 62 - iter 22/27 - loss 0.00495274\n",
            "2019-02-11 09:30:20,786 epoch 62 - iter 24/27 - loss 0.00500491\n",
            "2019-02-11 09:30:21,251 epoch 62 - iter 26/27 - loss 0.00502571\n",
            "2019-02-11 09:30:21,263 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:30:21,264 EPOCH 62 done: loss 0.0050 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 09:30:21,897 DEV  : loss 0.01430917 - f-score 0.8421 - acc 0.8421\n",
            "2019-02-11 09:30:25,616 TEST : loss 0.00083775 - f-score 0.9915 - acc 0.9915\n",
            "2019-02-11 09:30:42,731 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:30:43,715 epoch 63 - iter 0/27 - loss 0.00343850\n",
            "2019-02-11 09:30:44,565 epoch 63 - iter 2/27 - loss 0.00339166\n",
            "2019-02-11 09:30:45,339 epoch 63 - iter 4/27 - loss 0.00336503\n",
            "2019-02-11 09:30:46,550 epoch 63 - iter 6/27 - loss 0.00385585\n",
            "2019-02-11 09:30:47,293 epoch 63 - iter 8/27 - loss 0.00421191\n",
            "2019-02-11 09:30:47,877 epoch 63 - iter 10/27 - loss 0.00420044\n",
            "2019-02-11 09:30:48,523 epoch 63 - iter 12/27 - loss 0.00462708\n",
            "2019-02-11 09:30:49,412 epoch 63 - iter 14/27 - loss 0.00489751\n",
            "2019-02-11 09:30:50,743 epoch 63 - iter 16/27 - loss 0.00492234\n",
            "2019-02-11 09:30:51,484 epoch 63 - iter 18/27 - loss 0.00496574\n",
            "2019-02-11 09:30:52,204 epoch 63 - iter 20/27 - loss 0.00491659\n",
            "2019-02-11 09:30:52,954 epoch 63 - iter 22/27 - loss 0.00485071\n",
            "2019-02-11 09:30:53,743 epoch 63 - iter 24/27 - loss 0.00474002\n",
            "2019-02-11 09:30:54,260 epoch 63 - iter 26/27 - loss 0.00481412\n",
            "2019-02-11 09:30:54,272 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:30:54,274 EPOCH 63 done: loss 0.0048 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 09:30:54,943 DEV  : loss 0.01494600 - f-score 0.8632 - acc 0.8632\n",
            "2019-02-11 09:30:58,729 TEST : loss 0.00071985 - f-score 0.9957 - acc 0.9957\n",
            "2019-02-11 09:31:16,186 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:31:17,111 epoch 64 - iter 0/27 - loss 0.01072983\n",
            "2019-02-11 09:31:18,018 epoch 64 - iter 2/27 - loss 0.00801347\n",
            "2019-02-11 09:31:19,236 epoch 64 - iter 4/27 - loss 0.00592834\n",
            "2019-02-11 09:31:20,024 epoch 64 - iter 6/27 - loss 0.00543377\n",
            "2019-02-11 09:31:20,785 epoch 64 - iter 8/27 - loss 0.00517616\n",
            "2019-02-11 09:31:21,481 epoch 64 - iter 10/27 - loss 0.00486538\n",
            "2019-02-11 09:31:22,091 epoch 64 - iter 12/27 - loss 0.00542951\n",
            "2019-02-11 09:31:22,848 epoch 64 - iter 14/27 - loss 0.00516409\n",
            "2019-02-11 09:31:23,482 epoch 64 - iter 16/27 - loss 0.00605212\n",
            "2019-02-11 09:31:24,255 epoch 64 - iter 18/27 - loss 0.00602955\n",
            "2019-02-11 09:31:25,067 epoch 64 - iter 20/27 - loss 0.00581680\n",
            "2019-02-11 09:31:25,683 epoch 64 - iter 22/27 - loss 0.00574340\n",
            "2019-02-11 09:31:26,337 epoch 64 - iter 24/27 - loss 0.00565053\n",
            "2019-02-11 09:31:26,762 epoch 64 - iter 26/27 - loss 0.00547930\n",
            "2019-02-11 09:31:26,774 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:31:26,776 EPOCH 64 done: loss 0.0055 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 09:31:27,410 DEV  : loss 0.01390643 - f-score 0.8421 - acc 0.8421\n",
            "2019-02-11 09:31:31,342 TEST : loss 0.00084370 - f-score 0.9936 - acc 0.9936\n",
            "2019-02-11 09:31:31,345 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:31:31,632 epoch 65 - iter 0/27 - loss 0.00134646\n",
            "2019-02-11 09:31:32,621 epoch 65 - iter 2/27 - loss 0.00608066\n",
            "2019-02-11 09:31:33,437 epoch 65 - iter 4/27 - loss 0.00452032\n",
            "2019-02-11 09:31:34,317 epoch 65 - iter 6/27 - loss 0.00380712\n",
            "2019-02-11 09:31:34,966 epoch 65 - iter 8/27 - loss 0.00365126\n",
            "2019-02-11 09:31:36,341 epoch 65 - iter 10/27 - loss 0.00388821\n",
            "2019-02-11 09:31:37,141 epoch 65 - iter 12/27 - loss 0.00365652\n",
            "2019-02-11 09:31:37,911 epoch 65 - iter 14/27 - loss 0.00409023\n",
            "2019-02-11 09:31:38,657 epoch 65 - iter 16/27 - loss 0.00448505\n",
            "2019-02-11 09:31:39,689 epoch 65 - iter 18/27 - loss 0.00457581\n",
            "2019-02-11 09:31:40,659 epoch 65 - iter 20/27 - loss 0.00435962\n",
            "2019-02-11 09:31:41,444 epoch 65 - iter 22/27 - loss 0.00450529\n",
            "2019-02-11 09:31:42,017 epoch 65 - iter 24/27 - loss 0.00448840\n",
            "2019-02-11 09:31:42,429 epoch 65 - iter 26/27 - loss 0.00450680\n",
            "2019-02-11 09:31:42,441 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:31:42,443 EPOCH 65 done: loss 0.0045 - lr 0.0500 - bad epochs 1\n",
            "2019-02-11 09:31:43,105 DEV  : loss 0.01377038 - f-score 0.8316 - acc 0.8316\n",
            "2019-02-11 09:31:46,894 TEST : loss 0.00065795 - f-score 0.9915 - acc 0.9915\n",
            "2019-02-11 09:32:04,441 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:32:05,406 epoch 66 - iter 0/27 - loss 0.00425503\n",
            "2019-02-11 09:32:06,075 epoch 66 - iter 2/27 - loss 0.00441433\n",
            "2019-02-11 09:32:06,814 epoch 66 - iter 4/27 - loss 0.00481405\n",
            "2019-02-11 09:32:07,617 epoch 66 - iter 6/27 - loss 0.00412585\n",
            "2019-02-11 09:32:08,297 epoch 66 - iter 8/27 - loss 0.00383583\n",
            "2019-02-11 09:32:09,158 epoch 66 - iter 10/27 - loss 0.00508660\n",
            "2019-02-11 09:32:09,870 epoch 66 - iter 12/27 - loss 0.00497916\n",
            "2019-02-11 09:32:10,518 epoch 66 - iter 14/27 - loss 0.00458939\n",
            "2019-02-11 09:32:11,226 epoch 66 - iter 16/27 - loss 0.00498539\n",
            "2019-02-11 09:32:11,811 epoch 66 - iter 18/27 - loss 0.00478326\n",
            "2019-02-11 09:32:12,988 epoch 66 - iter 20/27 - loss 0.00489965\n",
            "2019-02-11 09:32:14,027 epoch 66 - iter 22/27 - loss 0.00529279\n",
            "2019-02-11 09:32:15,218 epoch 66 - iter 24/27 - loss 0.00521410\n",
            "2019-02-11 09:32:15,640 epoch 66 - iter 26/27 - loss 0.00513431\n",
            "2019-02-11 09:32:15,652 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:32:15,654 EPOCH 66 done: loss 0.0051 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 09:32:16,305 DEV  : loss 0.01582218 - f-score 0.8316 - acc 0.8316\n",
            "2019-02-11 09:32:20,236 TEST : loss 0.00081121 - f-score 0.9979 - acc 0.9979\n",
            "2019-02-11 09:32:20,242 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:32:20,719 epoch 67 - iter 0/27 - loss 0.00197141\n",
            "2019-02-11 09:32:22,460 epoch 67 - iter 2/27 - loss 0.00356949\n",
            "2019-02-11 09:32:23,163 epoch 67 - iter 4/27 - loss 0.00370738\n",
            "2019-02-11 09:32:23,966 epoch 67 - iter 6/27 - loss 0.00430258\n",
            "2019-02-11 09:32:24,888 epoch 67 - iter 8/27 - loss 0.00447454\n",
            "2019-02-11 09:32:25,538 epoch 67 - iter 10/27 - loss 0.00447142\n",
            "2019-02-11 09:32:26,247 epoch 67 - iter 12/27 - loss 0.00419707\n",
            "2019-02-11 09:32:26,993 epoch 67 - iter 14/27 - loss 0.00414472\n",
            "2019-02-11 09:32:27,922 epoch 67 - iter 16/27 - loss 0.00394550\n",
            "2019-02-11 09:32:28,575 epoch 67 - iter 18/27 - loss 0.00388373\n",
            "2019-02-11 09:32:29,383 epoch 67 - iter 20/27 - loss 0.00382668\n",
            "2019-02-11 09:32:29,959 epoch 67 - iter 22/27 - loss 0.00403882\n",
            "2019-02-11 09:32:30,678 epoch 67 - iter 24/27 - loss 0.00426309\n",
            "2019-02-11 09:32:31,173 epoch 67 - iter 26/27 - loss 0.00422084\n",
            "2019-02-11 09:32:31,185 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:32:31,186 EPOCH 67 done: loss 0.0042 - lr 0.0500 - bad epochs 1\n",
            "2019-02-11 09:32:31,852 DEV  : loss 0.01465444 - f-score 0.8211 - acc 0.8211\n",
            "2019-02-11 09:32:35,678 TEST : loss 0.00080478 - f-score 0.9957 - acc 0.9957\n",
            "2019-02-11 09:32:53,317 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:32:53,693 epoch 68 - iter 0/27 - loss 0.00369037\n",
            "2019-02-11 09:32:54,673 epoch 68 - iter 2/27 - loss 0.00562552\n",
            "2019-02-11 09:32:55,400 epoch 68 - iter 4/27 - loss 0.00497801\n",
            "2019-02-11 09:32:56,139 epoch 68 - iter 6/27 - loss 0.00475333\n",
            "2019-02-11 09:32:56,692 epoch 68 - iter 8/27 - loss 0.00426969\n",
            "2019-02-11 09:32:57,677 epoch 68 - iter 10/27 - loss 0.00528870\n",
            "2019-02-11 09:32:58,555 epoch 68 - iter 12/27 - loss 0.00473565\n",
            "2019-02-11 09:32:59,258 epoch 68 - iter 14/27 - loss 0.00529103\n",
            "2019-02-11 09:33:00,543 epoch 68 - iter 16/27 - loss 0.00488139\n",
            "2019-02-11 09:33:01,201 epoch 68 - iter 18/27 - loss 0.00501223\n",
            "2019-02-11 09:33:01,883 epoch 68 - iter 20/27 - loss 0.00511597\n",
            "2019-02-11 09:33:02,654 epoch 68 - iter 22/27 - loss 0.00496556\n",
            "2019-02-11 09:33:03,857 epoch 68 - iter 24/27 - loss 0.00483308\n",
            "2019-02-11 09:33:04,313 epoch 68 - iter 26/27 - loss 0.00490772\n",
            "2019-02-11 09:33:04,325 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:33:04,326 EPOCH 68 done: loss 0.0049 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 09:33:04,987 DEV  : loss 0.01545472 - f-score 0.8105 - acc 0.8105\n",
            "2019-02-11 09:33:08,901 TEST : loss 0.00080106 - f-score 0.9936 - acc 0.9936\n",
            "2019-02-11 09:33:08,904 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:33:09,195 epoch 69 - iter 0/27 - loss 0.00724820\n",
            "2019-02-11 09:33:10,037 epoch 69 - iter 2/27 - loss 0.00529849\n",
            "2019-02-11 09:33:10,778 epoch 69 - iter 4/27 - loss 0.00402435\n",
            "2019-02-11 09:33:11,576 epoch 69 - iter 6/27 - loss 0.00510694\n",
            "2019-02-11 09:33:12,895 epoch 69 - iter 8/27 - loss 0.00630277\n",
            "2019-02-11 09:33:13,633 epoch 69 - iter 10/27 - loss 0.00591814\n",
            "2019-02-11 09:33:14,350 epoch 69 - iter 12/27 - loss 0.00547877\n",
            "2019-02-11 09:33:15,466 epoch 69 - iter 14/27 - loss 0.00520869\n",
            "2019-02-11 09:33:16,048 epoch 69 - iter 16/27 - loss 0.00497110\n",
            "2019-02-11 09:33:16,764 epoch 69 - iter 18/27 - loss 0.00457805\n",
            "2019-02-11 09:33:17,721 epoch 69 - iter 20/27 - loss 0.00450909\n",
            "2019-02-11 09:33:18,573 epoch 69 - iter 22/27 - loss 0.00484745\n",
            "2019-02-11 09:33:19,356 epoch 69 - iter 24/27 - loss 0.00469099\n",
            "2019-02-11 09:33:19,865 epoch 69 - iter 26/27 - loss 0.00476861\n",
            "2019-02-11 09:33:19,878 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:33:19,879 EPOCH 69 done: loss 0.0048 - lr 0.0500 - bad epochs 1\n",
            "2019-02-11 09:33:20,597 DEV  : loss 0.01524978 - f-score 0.8316 - acc 0.8316\n",
            "2019-02-11 09:33:24,420 TEST : loss 0.00108671 - f-score 0.9872 - acc 0.9872\n",
            "2019-02-11 09:33:24,423 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:33:24,850 epoch 70 - iter 0/27 - loss 0.00918929\n",
            "2019-02-11 09:33:25,677 epoch 70 - iter 2/27 - loss 0.00398677\n",
            "2019-02-11 09:33:26,883 epoch 70 - iter 4/27 - loss 0.00403217\n",
            "2019-02-11 09:33:27,568 epoch 70 - iter 6/27 - loss 0.00406961\n",
            "2019-02-11 09:33:28,809 epoch 70 - iter 8/27 - loss 0.00398962\n",
            "2019-02-11 09:33:29,581 epoch 70 - iter 10/27 - loss 0.00442136\n",
            "2019-02-11 09:33:30,252 epoch 70 - iter 12/27 - loss 0.00455765\n",
            "2019-02-11 09:33:31,115 epoch 70 - iter 14/27 - loss 0.00433853\n",
            "2019-02-11 09:33:31,932 epoch 70 - iter 16/27 - loss 0.00427165\n",
            "2019-02-11 09:33:32,654 epoch 70 - iter 18/27 - loss 0.00406458\n",
            "2019-02-11 09:33:33,256 epoch 70 - iter 20/27 - loss 0.00396742\n",
            "2019-02-11 09:33:33,924 epoch 70 - iter 22/27 - loss 0.00375994\n",
            "2019-02-11 09:33:34,895 epoch 70 - iter 24/27 - loss 0.00381777\n",
            "2019-02-11 09:33:35,303 epoch 70 - iter 26/27 - loss 0.00388940\n",
            "2019-02-11 09:33:35,316 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:33:35,317 EPOCH 70 done: loss 0.0039 - lr 0.0500 - bad epochs 2\n",
            "2019-02-11 09:33:36,015 DEV  : loss 0.01496134 - f-score 0.8000 - acc 0.8000\n",
            "2019-02-11 09:33:40,513 TEST : loss 0.00077691 - f-score 0.9915 - acc 0.9915\n",
            "2019-02-11 09:33:58,460 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:33:58,991 epoch 71 - iter 0/27 - loss 0.00127993\n",
            "2019-02-11 09:33:59,925 epoch 71 - iter 2/27 - loss 0.00797910\n",
            "2019-02-11 09:34:00,560 epoch 71 - iter 4/27 - loss 0.00642924\n",
            "2019-02-11 09:34:01,629 epoch 71 - iter 6/27 - loss 0.00523883\n",
            "2019-02-11 09:34:02,360 epoch 71 - iter 8/27 - loss 0.00647918\n",
            "2019-02-11 09:34:02,971 epoch 71 - iter 10/27 - loss 0.00607035\n",
            "2019-02-11 09:34:03,735 epoch 71 - iter 12/27 - loss 0.00537895\n",
            "2019-02-11 09:34:05,015 epoch 71 - iter 14/27 - loss 0.00492013\n",
            "2019-02-11 09:34:05,913 epoch 71 - iter 16/27 - loss 0.00490338\n",
            "2019-02-11 09:34:06,569 epoch 71 - iter 18/27 - loss 0.00474319\n",
            "2019-02-11 09:34:07,402 epoch 71 - iter 20/27 - loss 0.00449021\n",
            "2019-02-11 09:34:08,203 epoch 71 - iter 22/27 - loss 0.00438990\n",
            "2019-02-11 09:34:08,762 epoch 71 - iter 24/27 - loss 0.00480682\n",
            "2019-02-11 09:34:09,183 epoch 71 - iter 26/27 - loss 0.00471105\n",
            "2019-02-11 09:34:09,195 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:34:09,196 EPOCH 71 done: loss 0.0047 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 09:34:09,842 DEV  : loss 0.01533194 - f-score 0.8421 - acc 0.8421\n",
            "2019-02-11 09:34:13,614 TEST : loss 0.00058108 - f-score 0.9979 - acc 0.9979\n",
            "2019-02-11 09:34:13,617 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:34:13,951 epoch 72 - iter 0/27 - loss 0.00470975\n",
            "2019-02-11 09:34:14,674 epoch 72 - iter 2/27 - loss 0.00441554\n",
            "2019-02-11 09:34:15,370 epoch 72 - iter 4/27 - loss 0.00470677\n",
            "2019-02-11 09:34:16,289 epoch 72 - iter 6/27 - loss 0.00426329\n",
            "2019-02-11 09:34:17,086 epoch 72 - iter 8/27 - loss 0.00424635\n",
            "2019-02-11 09:34:17,912 epoch 72 - iter 10/27 - loss 0.00397181\n",
            "2019-02-11 09:34:18,754 epoch 72 - iter 12/27 - loss 0.00471950\n",
            "2019-02-11 09:34:19,543 epoch 72 - iter 14/27 - loss 0.00437218\n",
            "2019-02-11 09:34:20,460 epoch 72 - iter 16/27 - loss 0.00429756\n",
            "2019-02-11 09:34:21,114 epoch 72 - iter 18/27 - loss 0.00480707\n",
            "2019-02-11 09:34:22,151 epoch 72 - iter 20/27 - loss 0.00496164\n",
            "2019-02-11 09:34:23,332 epoch 72 - iter 22/27 - loss 0.00498216\n",
            "2019-02-11 09:34:23,964 epoch 72 - iter 24/27 - loss 0.00482858\n",
            "2019-02-11 09:34:24,502 epoch 72 - iter 26/27 - loss 0.00536238\n",
            "2019-02-11 09:34:24,514 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:34:24,515 EPOCH 72 done: loss 0.0054 - lr 0.0500 - bad epochs 1\n",
            "2019-02-11 09:34:25,172 DEV  : loss 0.01453209 - f-score 0.8632 - acc 0.8632\n",
            "2019-02-11 09:34:28,983 TEST : loss 0.00090990 - f-score 0.9936 - acc 0.9936\n",
            "2019-02-11 09:34:28,986 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:34:29,591 epoch 73 - iter 0/27 - loss 0.00493599\n",
            "2019-02-11 09:34:30,578 epoch 73 - iter 2/27 - loss 0.00243991\n",
            "2019-02-11 09:34:31,312 epoch 73 - iter 4/27 - loss 0.00261563\n",
            "2019-02-11 09:34:31,964 epoch 73 - iter 6/27 - loss 0.00296757\n",
            "2019-02-11 09:34:32,716 epoch 73 - iter 8/27 - loss 0.00306242\n",
            "2019-02-11 09:34:34,056 epoch 73 - iter 10/27 - loss 0.00284852\n",
            "2019-02-11 09:34:34,781 epoch 73 - iter 12/27 - loss 0.00304998\n",
            "2019-02-11 09:34:35,530 epoch 73 - iter 14/27 - loss 0.00344071\n",
            "2019-02-11 09:34:36,336 epoch 73 - iter 16/27 - loss 0.00316986\n",
            "2019-02-11 09:34:37,010 epoch 73 - iter 18/27 - loss 0.00330678\n",
            "2019-02-11 09:34:37,731 epoch 73 - iter 20/27 - loss 0.00327764\n",
            "2019-02-11 09:34:38,547 epoch 73 - iter 22/27 - loss 0.00317074\n",
            "2019-02-11 09:34:39,282 epoch 73 - iter 24/27 - loss 0.00351297\n",
            "2019-02-11 09:34:39,938 epoch 73 - iter 26/27 - loss 0.00353719\n",
            "2019-02-11 09:34:39,951 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:34:39,952 EPOCH 73 done: loss 0.0035 - lr 0.0500 - bad epochs 2\n",
            "2019-02-11 09:34:40,605 DEV  : loss 0.01453241 - f-score 0.8105 - acc 0.8105\n",
            "2019-02-11 09:34:44,541 TEST : loss 0.00077476 - f-score 0.9915 - acc 0.9915\n",
            "2019-02-11 09:35:02,502 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:35:04,117 epoch 74 - iter 0/27 - loss 0.01031112\n",
            "2019-02-11 09:35:05,511 epoch 74 - iter 2/27 - loss 0.00605782\n",
            "2019-02-11 09:35:06,239 epoch 74 - iter 4/27 - loss 0.00446026\n",
            "2019-02-11 09:35:06,943 epoch 74 - iter 6/27 - loss 0.00425010\n",
            "2019-02-11 09:35:07,644 epoch 74 - iter 8/27 - loss 0.00349995\n",
            "2019-02-11 09:35:08,450 epoch 74 - iter 10/27 - loss 0.00377025\n",
            "2019-02-11 09:35:09,702 epoch 74 - iter 12/27 - loss 0.00375948\n",
            "2019-02-11 09:35:10,384 epoch 74 - iter 14/27 - loss 0.00357022\n",
            "2019-02-11 09:35:11,176 epoch 74 - iter 16/27 - loss 0.00338676\n",
            "2019-02-11 09:35:11,860 epoch 74 - iter 18/27 - loss 0.00330288\n",
            "2019-02-11 09:35:12,796 epoch 74 - iter 20/27 - loss 0.00306476\n",
            "2019-02-11 09:35:13,697 epoch 74 - iter 22/27 - loss 0.00324575\n",
            "2019-02-11 09:35:14,780 epoch 74 - iter 24/27 - loss 0.00332361\n",
            "2019-02-11 09:35:15,511 epoch 74 - iter 26/27 - loss 0.00364999\n",
            "2019-02-11 09:35:15,525 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:35:15,526 EPOCH 74 done: loss 0.0036 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 09:35:16,210 DEV  : loss 0.01688338 - f-score 0.8105 - acc 0.8105\n",
            "2019-02-11 09:35:19,985 TEST : loss 0.00084142 - f-score 0.9936 - acc 0.9936\n",
            "2019-02-11 09:35:19,990 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:35:20,206 epoch 75 - iter 0/27 - loss 0.00560429\n",
            "2019-02-11 09:35:20,852 epoch 75 - iter 2/27 - loss 0.00419943\n",
            "2019-02-11 09:35:21,602 epoch 75 - iter 4/27 - loss 0.00450278\n",
            "2019-02-11 09:35:22,620 epoch 75 - iter 6/27 - loss 0.00375863\n",
            "2019-02-11 09:35:23,450 epoch 75 - iter 8/27 - loss 0.00347064\n",
            "2019-02-11 09:35:24,087 epoch 75 - iter 10/27 - loss 0.00353590\n",
            "2019-02-11 09:35:25,080 epoch 75 - iter 12/27 - loss 0.00344259\n",
            "2019-02-11 09:35:25,795 epoch 75 - iter 14/27 - loss 0.00308043\n",
            "2019-02-11 09:35:26,952 epoch 75 - iter 16/27 - loss 0.00307144\n",
            "2019-02-11 09:35:27,728 epoch 75 - iter 18/27 - loss 0.00308162\n",
            "2019-02-11 09:35:28,342 epoch 75 - iter 20/27 - loss 0.00309420\n",
            "2019-02-11 09:35:28,936 epoch 75 - iter 22/27 - loss 0.00327888\n",
            "2019-02-11 09:35:29,725 epoch 75 - iter 24/27 - loss 0.00346616\n",
            "2019-02-11 09:35:30,605 epoch 75 - iter 26/27 - loss 0.00347847\n",
            "2019-02-11 09:35:30,620 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:35:30,621 EPOCH 75 done: loss 0.0035 - lr 0.0500 - bad epochs 1\n",
            "2019-02-11 09:35:31,271 DEV  : loss 0.01731183 - f-score 0.8316 - acc 0.8316\n",
            "2019-02-11 09:35:35,110 TEST : loss 0.00060773 - f-score 0.9936 - acc 0.9936\n",
            "2019-02-11 09:35:52,834 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:35:53,794 epoch 76 - iter 0/27 - loss 0.00609017\n",
            "2019-02-11 09:35:54,643 epoch 76 - iter 2/27 - loss 0.00516316\n",
            "2019-02-11 09:35:55,461 epoch 76 - iter 4/27 - loss 0.00425641\n",
            "2019-02-11 09:35:56,643 epoch 76 - iter 6/27 - loss 0.00417028\n",
            "2019-02-11 09:35:57,345 epoch 76 - iter 8/27 - loss 0.00389098\n",
            "2019-02-11 09:35:58,121 epoch 76 - iter 10/27 - loss 0.00354283\n",
            "2019-02-11 09:35:59,325 epoch 76 - iter 12/27 - loss 0.00347125\n",
            "2019-02-11 09:35:59,962 epoch 76 - iter 14/27 - loss 0.00320206\n",
            "2019-02-11 09:36:00,745 epoch 76 - iter 16/27 - loss 0.00348574\n",
            "2019-02-11 09:36:01,570 epoch 76 - iter 18/27 - loss 0.00319053\n",
            "2019-02-11 09:36:02,391 epoch 76 - iter 20/27 - loss 0.00381089\n",
            "2019-02-11 09:36:04,656 epoch 76 - iter 22/27 - loss 0.00358355\n",
            "2019-02-11 09:36:05,413 epoch 76 - iter 24/27 - loss 0.00390446\n",
            "2019-02-11 09:36:05,830 epoch 76 - iter 26/27 - loss 0.00385339\n",
            "2019-02-11 09:36:05,844 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:36:05,846 EPOCH 76 done: loss 0.0039 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 09:36:06,511 DEV  : loss 0.01854851 - f-score 0.8211 - acc 0.8211\n",
            "2019-02-11 09:36:10,319 TEST : loss 0.00058288 - f-score 0.9936 - acc 0.9936\n",
            "2019-02-11 09:36:10,323 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:36:10,736 epoch 77 - iter 0/27 - loss 0.00444322\n",
            "2019-02-11 09:36:11,910 epoch 77 - iter 2/27 - loss 0.00684017\n",
            "2019-02-11 09:36:12,613 epoch 77 - iter 4/27 - loss 0.00548703\n",
            "2019-02-11 09:36:13,373 epoch 77 - iter 6/27 - loss 0.00535887\n",
            "2019-02-11 09:36:14,147 epoch 77 - iter 8/27 - loss 0.00538331\n",
            "2019-02-11 09:36:15,133 epoch 77 - iter 10/27 - loss 0.00519482\n",
            "2019-02-11 09:36:15,877 epoch 77 - iter 12/27 - loss 0.00455023\n",
            "2019-02-11 09:36:16,787 epoch 77 - iter 14/27 - loss 0.00409500\n",
            "2019-02-11 09:36:17,447 epoch 77 - iter 16/27 - loss 0.00375300\n",
            "2019-02-11 09:36:18,661 epoch 77 - iter 18/27 - loss 0.00354657\n",
            "2019-02-11 09:36:19,498 epoch 77 - iter 20/27 - loss 0.00342033\n",
            "2019-02-11 09:36:20,172 epoch 77 - iter 22/27 - loss 0.00368811\n",
            "2019-02-11 09:36:20,942 epoch 77 - iter 24/27 - loss 0.00355434\n",
            "2019-02-11 09:36:21,298 epoch 77 - iter 26/27 - loss 0.00361443\n",
            "2019-02-11 09:36:21,311 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:36:21,313 EPOCH 77 done: loss 0.0036 - lr 0.0500 - bad epochs 1\n",
            "2019-02-11 09:36:22,002 DEV  : loss 0.01715155 - f-score 0.8421 - acc 0.8421\n",
            "2019-02-11 09:36:25,996 TEST : loss 0.00042996 - f-score 0.9957 - acc 0.9957\n",
            "2019-02-11 09:36:25,999 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:36:26,251 epoch 78 - iter 0/27 - loss 0.00660181\n",
            "2019-02-11 09:36:27,500 epoch 78 - iter 2/27 - loss 0.00320340\n",
            "2019-02-11 09:36:28,286 epoch 78 - iter 4/27 - loss 0.00259594\n",
            "2019-02-11 09:36:29,118 epoch 78 - iter 6/27 - loss 0.00322864\n",
            "2019-02-11 09:36:29,804 epoch 78 - iter 8/27 - loss 0.00280546\n",
            "2019-02-11 09:36:30,840 epoch 78 - iter 10/27 - loss 0.00284159\n",
            "2019-02-11 09:36:31,638 epoch 78 - iter 12/27 - loss 0.00268270\n",
            "2019-02-11 09:36:32,528 epoch 78 - iter 14/27 - loss 0.00247725\n",
            "2019-02-11 09:36:33,546 epoch 78 - iter 16/27 - loss 0.00274368\n",
            "2019-02-11 09:36:34,319 epoch 78 - iter 18/27 - loss 0.00277083\n",
            "2019-02-11 09:36:35,327 epoch 78 - iter 20/27 - loss 0.00292480\n",
            "2019-02-11 09:36:36,120 epoch 78 - iter 22/27 - loss 0.00293153\n",
            "2019-02-11 09:36:37,533 epoch 78 - iter 24/27 - loss 0.00298662\n",
            "2019-02-11 09:36:38,146 epoch 78 - iter 26/27 - loss 0.00301363\n",
            "2019-02-11 09:36:38,162 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:36:38,164 EPOCH 78 done: loss 0.0030 - lr 0.0500 - bad epochs 2\n",
            "2019-02-11 09:36:38,939 DEV  : loss 0.01789583 - f-score 0.8526 - acc 0.8526\n",
            "2019-02-11 09:36:43,462 TEST : loss 0.00037235 - f-score 0.9979 - acc 0.9979\n",
            "2019-02-11 09:37:01,415 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:37:02,628 epoch 79 - iter 0/27 - loss 0.00614228\n",
            "2019-02-11 09:37:04,008 epoch 79 - iter 2/27 - loss 0.00549228\n",
            "2019-02-11 09:37:05,231 epoch 79 - iter 4/27 - loss 0.00464112\n",
            "2019-02-11 09:37:05,985 epoch 79 - iter 6/27 - loss 0.00442453\n",
            "2019-02-11 09:37:06,730 epoch 79 - iter 8/27 - loss 0.00461234\n",
            "2019-02-11 09:37:07,373 epoch 79 - iter 10/27 - loss 0.00572070\n",
            "2019-02-11 09:37:08,076 epoch 79 - iter 12/27 - loss 0.00512444\n",
            "2019-02-11 09:37:08,825 epoch 79 - iter 14/27 - loss 0.00484783\n",
            "2019-02-11 09:37:09,491 epoch 79 - iter 16/27 - loss 0.00455388\n",
            "2019-02-11 09:37:10,218 epoch 79 - iter 18/27 - loss 0.00419041\n",
            "2019-02-11 09:37:10,966 epoch 79 - iter 20/27 - loss 0.00440298\n",
            "2019-02-11 09:37:11,698 epoch 79 - iter 22/27 - loss 0.00452775\n",
            "2019-02-11 09:37:12,754 epoch 79 - iter 24/27 - loss 0.00434215\n",
            "2019-02-11 09:37:13,366 epoch 79 - iter 26/27 - loss 0.00435307\n",
            "2019-02-11 09:37:13,382 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:37:13,383 EPOCH 79 done: loss 0.0044 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 09:37:14,058 DEV  : loss 0.01574661 - f-score 0.8421 - acc 0.8421\n",
            "2019-02-11 09:37:18,082 TEST : loss 0.00062548 - f-score 0.9915 - acc 0.9915\n",
            "2019-02-11 09:37:18,085 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:37:18,460 epoch 80 - iter 0/27 - loss 0.00177486\n",
            "2019-02-11 09:37:19,293 epoch 80 - iter 2/27 - loss 0.00689184\n",
            "2019-02-11 09:37:20,185 epoch 80 - iter 4/27 - loss 0.00560511\n",
            "2019-02-11 09:37:21,032 epoch 80 - iter 6/27 - loss 0.00437897\n",
            "2019-02-11 09:37:22,110 epoch 80 - iter 8/27 - loss 0.00351124\n",
            "2019-02-11 09:37:22,819 epoch 80 - iter 10/27 - loss 0.00327492\n",
            "2019-02-11 09:37:23,698 epoch 80 - iter 12/27 - loss 0.00323220\n",
            "2019-02-11 09:37:25,091 epoch 80 - iter 14/27 - loss 0.00297813\n",
            "2019-02-11 09:37:25,797 epoch 80 - iter 16/27 - loss 0.00310365\n",
            "2019-02-11 09:37:27,007 epoch 80 - iter 18/27 - loss 0.00330070\n",
            "2019-02-11 09:37:27,722 epoch 80 - iter 20/27 - loss 0.00334595\n",
            "2019-02-11 09:37:28,339 epoch 80 - iter 22/27 - loss 0.00344587\n",
            "2019-02-11 09:37:29,102 epoch 80 - iter 24/27 - loss 0.00325304\n",
            "2019-02-11 09:37:29,581 epoch 80 - iter 26/27 - loss 0.00332090\n",
            "2019-02-11 09:37:29,595 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:37:29,596 EPOCH 80 done: loss 0.0033 - lr 0.0500 - bad epochs 1\n",
            "2019-02-11 09:37:30,268 DEV  : loss 0.01481516 - f-score 0.8632 - acc 0.8632\n",
            "2019-02-11 09:37:34,140 TEST : loss 0.00036990 - f-score 0.9979 - acc 0.9979\n",
            "2019-02-11 09:37:34,144 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:37:34,509 epoch 81 - iter 0/27 - loss 0.00058659\n",
            "2019-02-11 09:37:35,288 epoch 81 - iter 2/27 - loss 0.00454479\n",
            "2019-02-11 09:37:36,220 epoch 81 - iter 4/27 - loss 0.00383010\n",
            "2019-02-11 09:37:37,698 epoch 81 - iter 6/27 - loss 0.00347231\n",
            "2019-02-11 09:37:38,841 epoch 81 - iter 8/27 - loss 0.00353433\n",
            "2019-02-11 09:37:39,471 epoch 81 - iter 10/27 - loss 0.00326846\n",
            "2019-02-11 09:37:40,114 epoch 81 - iter 12/27 - loss 0.00337203\n",
            "2019-02-11 09:37:40,758 epoch 81 - iter 14/27 - loss 0.00366079\n",
            "2019-02-11 09:37:41,448 epoch 81 - iter 16/27 - loss 0.00356043\n",
            "2019-02-11 09:37:42,353 epoch 81 - iter 18/27 - loss 0.00350603\n",
            "2019-02-11 09:37:43,155 epoch 81 - iter 20/27 - loss 0.00365069\n",
            "2019-02-11 09:37:43,905 epoch 81 - iter 22/27 - loss 0.00385379\n",
            "2019-02-11 09:37:44,683 epoch 81 - iter 24/27 - loss 0.00405000\n",
            "2019-02-11 09:37:45,216 epoch 81 - iter 26/27 - loss 0.00396953\n",
            "2019-02-11 09:37:45,230 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:37:45,232 EPOCH 81 done: loss 0.0040 - lr 0.0500 - bad epochs 2\n",
            "2019-02-11 09:37:45,890 DEV  : loss 0.01354188 - f-score 0.8526 - acc 0.8526\n",
            "2019-02-11 09:37:50,306 TEST : loss 0.00035278 - f-score 0.9979 - acc 0.9979\n",
            "2019-02-11 09:37:50,310 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:37:50,696 epoch 82 - iter 0/27 - loss 0.00588801\n",
            "2019-02-11 09:37:51,333 epoch 82 - iter 2/27 - loss 0.00510663\n",
            "2019-02-11 09:37:52,189 epoch 82 - iter 4/27 - loss 0.00459191\n",
            "2019-02-11 09:37:52,854 epoch 82 - iter 6/27 - loss 0.00407627\n",
            "2019-02-11 09:37:53,611 epoch 82 - iter 8/27 - loss 0.00345011\n",
            "2019-02-11 09:37:54,208 epoch 82 - iter 10/27 - loss 0.00335240\n",
            "2019-02-11 09:37:55,051 epoch 82 - iter 12/27 - loss 0.00326054\n",
            "2019-02-11 09:37:56,304 epoch 82 - iter 14/27 - loss 0.00299570\n",
            "2019-02-11 09:37:57,016 epoch 82 - iter 16/27 - loss 0.00276979\n",
            "2019-02-11 09:37:58,284 epoch 82 - iter 18/27 - loss 0.00300369\n",
            "2019-02-11 09:37:59,102 epoch 82 - iter 20/27 - loss 0.00287368\n",
            "2019-02-11 09:37:59,932 epoch 82 - iter 22/27 - loss 0.00284309\n",
            "2019-02-11 09:38:00,703 epoch 82 - iter 24/27 - loss 0.00292740\n",
            "2019-02-11 09:38:01,130 epoch 82 - iter 26/27 - loss 0.00301959\n",
            "2019-02-11 09:38:01,142 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:38:01,144 EPOCH 82 done: loss 0.0030 - lr 0.0500 - bad epochs 3\n",
            "2019-02-11 09:38:01,821 DEV  : loss 0.01436732 - f-score 0.8316 - acc 0.8316\n",
            "2019-02-11 09:38:05,775 TEST : loss 0.00036950 - f-score 0.9979 - acc 0.9979\n",
            "Epoch    81: reducing learning rate of group 0 to 2.5000e-02.\n",
            "2019-02-11 09:38:05,778 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:38:06,675 epoch 83 - iter 0/27 - loss 0.00282649\n",
            "2019-02-11 09:38:07,596 epoch 83 - iter 2/27 - loss 0.00360236\n",
            "2019-02-11 09:38:08,350 epoch 83 - iter 4/27 - loss 0.00236204\n",
            "2019-02-11 09:38:09,557 epoch 83 - iter 6/27 - loss 0.00258546\n",
            "2019-02-11 09:38:10,504 epoch 83 - iter 8/27 - loss 0.00242303\n",
            "2019-02-11 09:38:11,397 epoch 83 - iter 10/27 - loss 0.00223605\n",
            "2019-02-11 09:38:12,160 epoch 83 - iter 12/27 - loss 0.00205240\n",
            "2019-02-11 09:38:12,948 epoch 83 - iter 14/27 - loss 0.00217436\n",
            "2019-02-11 09:38:13,804 epoch 83 - iter 16/27 - loss 0.00223424\n",
            "2019-02-11 09:38:14,406 epoch 83 - iter 18/27 - loss 0.00221570\n",
            "2019-02-11 09:38:15,077 epoch 83 - iter 20/27 - loss 0.00235192\n",
            "2019-02-11 09:38:15,682 epoch 83 - iter 22/27 - loss 0.00245976\n",
            "2019-02-11 09:38:16,311 epoch 83 - iter 24/27 - loss 0.00251681\n",
            "2019-02-11 09:38:16,725 epoch 83 - iter 26/27 - loss 0.00246061\n",
            "2019-02-11 09:38:16,738 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:38:16,739 EPOCH 83 done: loss 0.0025 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 09:38:17,415 DEV  : loss 0.01404338 - f-score 0.8421 - acc 0.8421\n",
            "2019-02-11 09:38:21,255 TEST : loss 0.00031579 - f-score 0.9957 - acc 0.9957\n",
            "2019-02-11 09:38:38,813 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:38:39,216 epoch 84 - iter 0/27 - loss 0.00056712\n",
            "2019-02-11 09:38:40,154 epoch 84 - iter 2/27 - loss 0.00255169\n",
            "2019-02-11 09:38:40,851 epoch 84 - iter 4/27 - loss 0.00219653\n",
            "2019-02-11 09:38:42,117 epoch 84 - iter 6/27 - loss 0.00209072\n",
            "2019-02-11 09:38:43,129 epoch 84 - iter 8/27 - loss 0.00218635\n",
            "2019-02-11 09:38:43,911 epoch 84 - iter 10/27 - loss 0.00193403\n",
            "2019-02-11 09:38:44,560 epoch 84 - iter 12/27 - loss 0.00189700\n",
            "2019-02-11 09:38:45,303 epoch 84 - iter 14/27 - loss 0.00187068\n",
            "2019-02-11 09:38:46,471 epoch 84 - iter 16/27 - loss 0.00184298\n",
            "2019-02-11 09:38:47,112 epoch 84 - iter 18/27 - loss 0.00210836\n",
            "2019-02-11 09:38:47,857 epoch 84 - iter 20/27 - loss 0.00202762\n",
            "2019-02-11 09:38:48,659 epoch 84 - iter 22/27 - loss 0.00207979\n",
            "2019-02-11 09:38:49,552 epoch 84 - iter 24/27 - loss 0.00196442\n",
            "2019-02-11 09:38:49,933 epoch 84 - iter 26/27 - loss 0.00210864\n",
            "2019-02-11 09:38:49,945 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:38:49,947 EPOCH 84 done: loss 0.0021 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 09:38:50,619 DEV  : loss 0.01469018 - f-score 0.8421 - acc 0.8421\n",
            "2019-02-11 09:38:54,643 TEST : loss 0.00028326 - f-score 0.9979 - acc 0.9979\n",
            "2019-02-11 09:39:12,582 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:39:13,117 epoch 85 - iter 0/27 - loss 0.00437086\n",
            "2019-02-11 09:39:14,027 epoch 85 - iter 2/27 - loss 0.00542499\n",
            "2019-02-11 09:39:14,934 epoch 85 - iter 4/27 - loss 0.00426632\n",
            "2019-02-11 09:39:15,604 epoch 85 - iter 6/27 - loss 0.00358912\n",
            "2019-02-11 09:39:16,351 epoch 85 - iter 8/27 - loss 0.00298252\n",
            "2019-02-11 09:39:17,057 epoch 85 - iter 10/27 - loss 0.00278893\n",
            "2019-02-11 09:39:17,710 epoch 85 - iter 12/27 - loss 0.00273290\n",
            "2019-02-11 09:39:18,519 epoch 85 - iter 14/27 - loss 0.00266470\n",
            "2019-02-11 09:39:19,230 epoch 85 - iter 16/27 - loss 0.00311589\n",
            "2019-02-11 09:39:20,409 epoch 85 - iter 18/27 - loss 0.00304711\n",
            "2019-02-11 09:39:21,110 epoch 85 - iter 20/27 - loss 0.00290348\n",
            "2019-02-11 09:39:22,381 epoch 85 - iter 22/27 - loss 0.00278651\n",
            "2019-02-11 09:39:23,213 epoch 85 - iter 24/27 - loss 0.00287198\n",
            "2019-02-11 09:39:23,818 epoch 85 - iter 26/27 - loss 0.00294222\n",
            "2019-02-11 09:39:23,831 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:39:23,832 EPOCH 85 done: loss 0.0029 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 09:39:24,517 DEV  : loss 0.01517433 - f-score 0.8421 - acc 0.8421\n",
            "2019-02-11 09:39:28,862 TEST : loss 0.00025290 - f-score 0.9979 - acc 0.9979\n",
            "2019-02-11 09:39:28,865 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:39:29,236 epoch 86 - iter 0/27 - loss 0.00445838\n",
            "2019-02-11 09:39:30,082 epoch 86 - iter 2/27 - loss 0.00250948\n",
            "2019-02-11 09:39:31,182 epoch 86 - iter 4/27 - loss 0.00191977\n",
            "2019-02-11 09:39:32,134 epoch 86 - iter 6/27 - loss 0.00156293\n",
            "2019-02-11 09:39:32,841 epoch 86 - iter 8/27 - loss 0.00178825\n",
            "2019-02-11 09:39:33,760 epoch 86 - iter 10/27 - loss 0.00219989\n",
            "2019-02-11 09:39:34,613 epoch 86 - iter 12/27 - loss 0.00198924\n",
            "2019-02-11 09:39:35,609 epoch 86 - iter 14/27 - loss 0.00187236\n",
            "2019-02-11 09:39:37,038 epoch 86 - iter 16/27 - loss 0.00192021\n",
            "2019-02-11 09:39:38,050 epoch 86 - iter 18/27 - loss 0.00197448\n",
            "2019-02-11 09:39:38,848 epoch 86 - iter 20/27 - loss 0.00196634\n",
            "2019-02-11 09:39:39,640 epoch 86 - iter 22/27 - loss 0.00207578\n",
            "2019-02-11 09:39:40,261 epoch 86 - iter 24/27 - loss 0.00224888\n",
            "2019-02-11 09:39:40,714 epoch 86 - iter 26/27 - loss 0.00222404\n",
            "2019-02-11 09:39:40,726 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:39:40,727 EPOCH 86 done: loss 0.0022 - lr 0.0250 - bad epochs 1\n",
            "2019-02-11 09:39:41,406 DEV  : loss 0.01502799 - f-score 0.8526 - acc 0.8526\n",
            "2019-02-11 09:39:45,370 TEST : loss 0.00029580 - f-score 0.9957 - acc 0.9957\n",
            "2019-02-11 09:39:45,373 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:39:45,672 epoch 87 - iter 0/27 - loss 0.00388385\n",
            "2019-02-11 09:39:46,368 epoch 87 - iter 2/27 - loss 0.00200047\n",
            "2019-02-11 09:39:47,218 epoch 87 - iter 4/27 - loss 0.00238693\n",
            "2019-02-11 09:39:48,459 epoch 87 - iter 6/27 - loss 0.00234253\n",
            "2019-02-11 09:39:49,230 epoch 87 - iter 8/27 - loss 0.00328397\n",
            "2019-02-11 09:39:50,023 epoch 87 - iter 10/27 - loss 0.00305056\n",
            "2019-02-11 09:39:50,750 epoch 87 - iter 12/27 - loss 0.00275058\n",
            "2019-02-11 09:39:51,456 epoch 87 - iter 14/27 - loss 0.00262030\n",
            "2019-02-11 09:39:52,501 epoch 87 - iter 16/27 - loss 0.00272267\n",
            "2019-02-11 09:39:53,223 epoch 87 - iter 18/27 - loss 0.00252864\n",
            "2019-02-11 09:39:53,977 epoch 87 - iter 20/27 - loss 0.00244224\n",
            "2019-02-11 09:39:54,597 epoch 87 - iter 22/27 - loss 0.00236291\n",
            "2019-02-11 09:39:55,807 epoch 87 - iter 24/27 - loss 0.00234702\n",
            "2019-02-11 09:39:56,469 epoch 87 - iter 26/27 - loss 0.00228795\n",
            "2019-02-11 09:39:56,481 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:39:56,482 EPOCH 87 done: loss 0.0023 - lr 0.0250 - bad epochs 2\n",
            "2019-02-11 09:39:57,142 DEV  : loss 0.01564783 - f-score 0.8526 - acc 0.8526\n",
            "2019-02-11 09:40:00,988 TEST : loss 0.00026785 - f-score 0.9979 - acc 0.9979\n",
            "2019-02-11 09:40:00,991 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:40:01,279 epoch 88 - iter 0/27 - loss 0.00227589\n",
            "2019-02-11 09:40:02,566 epoch 88 - iter 2/27 - loss 0.00136630\n",
            "2019-02-11 09:40:03,429 epoch 88 - iter 4/27 - loss 0.00125625\n",
            "2019-02-11 09:40:04,172 epoch 88 - iter 6/27 - loss 0.00108590\n",
            "2019-02-11 09:40:04,825 epoch 88 - iter 8/27 - loss 0.00131367\n",
            "2019-02-11 09:40:05,549 epoch 88 - iter 10/27 - loss 0.00121160\n",
            "2019-02-11 09:40:06,304 epoch 88 - iter 12/27 - loss 0.00117082\n",
            "2019-02-11 09:40:07,078 epoch 88 - iter 14/27 - loss 0.00120057\n",
            "2019-02-11 09:40:07,839 epoch 88 - iter 16/27 - loss 0.00126204\n",
            "2019-02-11 09:40:09,074 epoch 88 - iter 18/27 - loss 0.00150657\n",
            "2019-02-11 09:40:09,743 epoch 88 - iter 20/27 - loss 0.00163146\n",
            "2019-02-11 09:40:10,565 epoch 88 - iter 22/27 - loss 0.00161801\n",
            "2019-02-11 09:40:11,224 epoch 88 - iter 24/27 - loss 0.00151454\n",
            "2019-02-11 09:40:11,975 epoch 88 - iter 26/27 - loss 0.00165568\n",
            "2019-02-11 09:40:11,987 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:40:11,988 EPOCH 88 done: loss 0.0017 - lr 0.0250 - bad epochs 3\n",
            "2019-02-11 09:40:12,646 DEV  : loss 0.01514746 - f-score 0.8526 - acc 0.8526\n",
            "2019-02-11 09:40:16,474 TEST : loss 0.00028991 - f-score 0.9979 - acc 0.9979\n",
            "2019-02-11 09:40:34,180 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:40:34,724 epoch 89 - iter 0/27 - loss 0.00015999\n",
            "2019-02-11 09:40:35,440 epoch 89 - iter 2/27 - loss 0.00040900\n",
            "2019-02-11 09:40:36,505 epoch 89 - iter 4/27 - loss 0.00104820\n",
            "2019-02-11 09:40:37,733 epoch 89 - iter 6/27 - loss 0.00171073\n",
            "2019-02-11 09:40:38,522 epoch 89 - iter 8/27 - loss 0.00259779\n",
            "2019-02-11 09:40:39,790 epoch 89 - iter 10/27 - loss 0.00294137\n",
            "2019-02-11 09:40:40,503 epoch 89 - iter 12/27 - loss 0.00304840\n",
            "2019-02-11 09:40:41,166 epoch 89 - iter 14/27 - loss 0.00322043\n",
            "2019-02-11 09:40:42,001 epoch 89 - iter 16/27 - loss 0.00295402\n",
            "2019-02-11 09:40:42,915 epoch 89 - iter 18/27 - loss 0.00290127\n",
            "2019-02-11 09:40:43,521 epoch 89 - iter 20/27 - loss 0.00287448\n",
            "2019-02-11 09:40:44,194 epoch 89 - iter 22/27 - loss 0.00295649\n",
            "2019-02-11 09:40:45,007 epoch 89 - iter 24/27 - loss 0.00277628\n",
            "2019-02-11 09:40:45,432 epoch 89 - iter 26/27 - loss 0.00292658\n",
            "2019-02-11 09:40:45,444 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:40:45,445 EPOCH 89 done: loss 0.0029 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 09:40:46,111 DEV  : loss 0.01573514 - f-score 0.8526 - acc 0.8526\n",
            "2019-02-11 09:40:50,000 TEST : loss 0.00027585 - f-score 0.9957 - acc 0.9957\n",
            "2019-02-11 09:40:50,003 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:40:50,400 epoch 90 - iter 0/27 - loss 0.00137828\n",
            "2019-02-11 09:40:51,585 epoch 90 - iter 2/27 - loss 0.00226121\n",
            "2019-02-11 09:40:52,356 epoch 90 - iter 4/27 - loss 0.00180823\n",
            "2019-02-11 09:40:52,949 epoch 90 - iter 6/27 - loss 0.00205833\n",
            "2019-02-11 09:40:53,871 epoch 90 - iter 8/27 - loss 0.00216800\n",
            "2019-02-11 09:40:54,647 epoch 90 - iter 10/27 - loss 0.00203309\n",
            "2019-02-11 09:40:55,422 epoch 90 - iter 12/27 - loss 0.00187100\n",
            "2019-02-11 09:40:56,068 epoch 90 - iter 14/27 - loss 0.00179924\n",
            "2019-02-11 09:40:57,198 epoch 90 - iter 16/27 - loss 0.00194392\n",
            "2019-02-11 09:40:58,239 epoch 90 - iter 18/27 - loss 0.00179231\n",
            "2019-02-11 09:40:58,934 epoch 90 - iter 20/27 - loss 0.00174126\n",
            "2019-02-11 09:40:59,686 epoch 90 - iter 22/27 - loss 0.00177572\n",
            "2019-02-11 09:41:00,280 epoch 90 - iter 24/27 - loss 0.00217587\n",
            "2019-02-11 09:41:00,890 epoch 90 - iter 26/27 - loss 0.00255156\n",
            "2019-02-11 09:41:00,902 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:41:00,903 EPOCH 90 done: loss 0.0026 - lr 0.0250 - bad epochs 1\n",
            "2019-02-11 09:41:01,571 DEV  : loss 0.01471080 - f-score 0.8526 - acc 0.8526\n",
            "2019-02-11 09:41:05,378 TEST : loss 0.00028080 - f-score 0.9957 - acc 0.9957\n",
            "2019-02-11 09:41:05,381 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:41:05,707 epoch 91 - iter 0/27 - loss 0.00030583\n",
            "2019-02-11 09:41:06,870 epoch 91 - iter 2/27 - loss 0.00120423\n",
            "2019-02-11 09:41:07,736 epoch 91 - iter 4/27 - loss 0.00295110\n",
            "2019-02-11 09:41:08,491 epoch 91 - iter 6/27 - loss 0.00245890\n",
            "2019-02-11 09:41:09,055 epoch 91 - iter 8/27 - loss 0.00255025\n",
            "2019-02-11 09:41:09,801 epoch 91 - iter 10/27 - loss 0.00231229\n",
            "2019-02-11 09:41:10,585 epoch 91 - iter 12/27 - loss 0.00229943\n",
            "2019-02-11 09:41:11,644 epoch 91 - iter 14/27 - loss 0.00208023\n",
            "2019-02-11 09:41:12,358 epoch 91 - iter 16/27 - loss 0.00217367\n",
            "2019-02-11 09:41:13,069 epoch 91 - iter 18/27 - loss 0.00236208\n",
            "2019-02-11 09:41:13,655 epoch 91 - iter 20/27 - loss 0.00229150\n",
            "2019-02-11 09:41:14,700 epoch 91 - iter 22/27 - loss 0.00244463\n",
            "2019-02-11 09:41:15,521 epoch 91 - iter 24/27 - loss 0.00310558\n",
            "2019-02-11 09:41:15,930 epoch 91 - iter 26/27 - loss 0.00312599\n",
            "2019-02-11 09:41:15,941 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:41:15,942 EPOCH 91 done: loss 0.0031 - lr 0.0250 - bad epochs 2\n",
            "2019-02-11 09:41:16,600 DEV  : loss 0.01435954 - f-score 0.8526 - acc 0.8526\n",
            "2019-02-11 09:41:20,430 TEST : loss 0.00028437 - f-score 0.9957 - acc 0.9957\n",
            "2019-02-11 09:41:20,433 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:41:20,855 epoch 92 - iter 0/27 - loss 0.00086497\n",
            "2019-02-11 09:41:21,580 epoch 92 - iter 2/27 - loss 0.00196044\n",
            "2019-02-11 09:41:22,334 epoch 92 - iter 4/27 - loss 0.00145128\n",
            "2019-02-11 09:41:22,954 epoch 92 - iter 6/27 - loss 0.00228916\n",
            "2019-02-11 09:41:23,635 epoch 92 - iter 8/27 - loss 0.00242994\n",
            "2019-02-11 09:41:24,480 epoch 92 - iter 10/27 - loss 0.00226236\n",
            "2019-02-11 09:41:25,326 epoch 92 - iter 12/27 - loss 0.00241164\n",
            "2019-02-11 09:41:26,525 epoch 92 - iter 14/27 - loss 0.00238065\n",
            "2019-02-11 09:41:27,393 epoch 92 - iter 16/27 - loss 0.00232918\n",
            "2019-02-11 09:41:28,321 epoch 92 - iter 18/27 - loss 0.00214116\n",
            "2019-02-11 09:41:29,398 epoch 92 - iter 20/27 - loss 0.00197212\n",
            "2019-02-11 09:41:30,299 epoch 92 - iter 22/27 - loss 0.00195097\n",
            "2019-02-11 09:41:31,022 epoch 92 - iter 24/27 - loss 0.00213703\n",
            "2019-02-11 09:41:31,497 epoch 92 - iter 26/27 - loss 0.00223909\n",
            "2019-02-11 09:41:31,509 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:41:31,511 EPOCH 92 done: loss 0.0022 - lr 0.0250 - bad epochs 3\n",
            "2019-02-11 09:41:32,181 DEV  : loss 0.01458220 - f-score 0.8526 - acc 0.8526\n",
            "2019-02-11 09:41:36,019 TEST : loss 0.00033178 - f-score 0.9957 - acc 0.9957\n",
            "Epoch    91: reducing learning rate of group 0 to 1.2500e-02.\n",
            "2019-02-11 09:41:36,023 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:41:36,463 epoch 93 - iter 0/27 - loss 0.00021445\n",
            "2019-02-11 09:41:37,129 epoch 93 - iter 2/27 - loss 0.00062802\n",
            "2019-02-11 09:41:37,900 epoch 93 - iter 4/27 - loss 0.00246773\n",
            "2019-02-11 09:41:39,051 epoch 93 - iter 6/27 - loss 0.00299199\n",
            "2019-02-11 09:41:39,919 epoch 93 - iter 8/27 - loss 0.00259307\n",
            "2019-02-11 09:41:40,567 epoch 93 - iter 10/27 - loss 0.00243949\n",
            "2019-02-11 09:41:41,320 epoch 93 - iter 12/27 - loss 0.00242641\n",
            "2019-02-11 09:41:42,071 epoch 93 - iter 14/27 - loss 0.00255541\n",
            "2019-02-11 09:41:42,795 epoch 93 - iter 16/27 - loss 0.00289266\n",
            "2019-02-11 09:41:43,556 epoch 93 - iter 18/27 - loss 0.00271977\n",
            "2019-02-11 09:41:44,761 epoch 93 - iter 20/27 - loss 0.00254531\n",
            "2019-02-11 09:41:45,509 epoch 93 - iter 22/27 - loss 0.00242760\n",
            "2019-02-11 09:41:46,432 epoch 93 - iter 24/27 - loss 0.00252490\n",
            "2019-02-11 09:41:47,137 epoch 93 - iter 26/27 - loss 0.00241810\n",
            "2019-02-11 09:41:47,149 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:41:47,150 EPOCH 93 done: loss 0.0024 - lr 0.0125 - bad epochs 0\n",
            "2019-02-11 09:41:47,802 DEV  : loss 0.01434751 - f-score 0.8632 - acc 0.8632\n",
            "2019-02-11 09:41:51,632 TEST : loss 0.00031827 - f-score 0.9957 - acc 0.9957\n",
            "2019-02-11 09:41:51,635 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:41:51,939 epoch 94 - iter 0/27 - loss 0.00276535\n",
            "2019-02-11 09:41:53,186 epoch 94 - iter 2/27 - loss 0.00359418\n",
            "2019-02-11 09:41:54,032 epoch 94 - iter 4/27 - loss 0.00237010\n",
            "2019-02-11 09:41:54,782 epoch 94 - iter 6/27 - loss 0.00192886\n",
            "2019-02-11 09:41:55,434 epoch 94 - iter 8/27 - loss 0.00166017\n",
            "2019-02-11 09:41:56,153 epoch 94 - iter 10/27 - loss 0.00187085\n",
            "2019-02-11 09:41:56,791 epoch 94 - iter 12/27 - loss 0.00198473\n",
            "2019-02-11 09:41:57,557 epoch 94 - iter 14/27 - loss 0.00188799\n",
            "2019-02-11 09:41:58,226 epoch 94 - iter 16/27 - loss 0.00205858\n",
            "2019-02-11 09:41:59,102 epoch 94 - iter 18/27 - loss 0.00208770\n",
            "2019-02-11 09:42:00,059 epoch 94 - iter 20/27 - loss 0.00195358\n",
            "2019-02-11 09:42:00,735 epoch 94 - iter 22/27 - loss 0.00192742\n",
            "2019-02-11 09:42:02,168 epoch 94 - iter 24/27 - loss 0.00201743\n",
            "2019-02-11 09:42:02,646 epoch 94 - iter 26/27 - loss 0.00217671\n",
            "2019-02-11 09:42:02,658 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:42:02,659 EPOCH 94 done: loss 0.0022 - lr 0.0125 - bad epochs 1\n",
            "2019-02-11 09:42:03,342 DEV  : loss 0.01425955 - f-score 0.8632 - acc 0.8632\n",
            "2019-02-11 09:42:08,026 TEST : loss 0.00036732 - f-score 0.9957 - acc 0.9957\n",
            "2019-02-11 09:42:08,030 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:42:08,507 epoch 95 - iter 0/27 - loss 0.00033888\n",
            "2019-02-11 09:42:09,590 epoch 95 - iter 2/27 - loss 0.00060079\n",
            "2019-02-11 09:42:10,471 epoch 95 - iter 4/27 - loss 0.00292466\n",
            "2019-02-11 09:42:11,285 epoch 95 - iter 6/27 - loss 0.00227573\n",
            "2019-02-11 09:42:12,055 epoch 95 - iter 8/27 - loss 0.00209190\n",
            "2019-02-11 09:42:12,864 epoch 95 - iter 10/27 - loss 0.00218357\n",
            "2019-02-11 09:42:14,019 epoch 95 - iter 12/27 - loss 0.00196141\n",
            "2019-02-11 09:42:14,692 epoch 95 - iter 14/27 - loss 0.00185082\n",
            "2019-02-11 09:42:15,458 epoch 95 - iter 16/27 - loss 0.00180809\n",
            "2019-02-11 09:42:16,784 epoch 95 - iter 18/27 - loss 0.00185861\n",
            "2019-02-11 09:42:17,566 epoch 95 - iter 20/27 - loss 0.00203231\n",
            "2019-02-11 09:42:18,521 epoch 95 - iter 22/27 - loss 0.00196486\n",
            "2019-02-11 09:42:19,191 epoch 95 - iter 24/27 - loss 0.00186144\n",
            "2019-02-11 09:42:19,594 epoch 95 - iter 26/27 - loss 0.00187277\n",
            "2019-02-11 09:42:19,606 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:42:19,608 EPOCH 95 done: loss 0.0019 - lr 0.0125 - bad epochs 2\n",
            "2019-02-11 09:42:20,303 DEV  : loss 0.01460176 - f-score 0.8526 - acc 0.8526\n",
            "2019-02-11 09:42:24,315 TEST : loss 0.00033101 - f-score 0.9957 - acc 0.9957\n",
            "2019-02-11 09:42:24,318 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:42:24,689 epoch 96 - iter 0/27 - loss 0.00110297\n",
            "2019-02-11 09:42:25,459 epoch 96 - iter 2/27 - loss 0.00104180\n",
            "2019-02-11 09:42:26,257 epoch 96 - iter 4/27 - loss 0.00191235\n",
            "2019-02-11 09:42:27,125 epoch 96 - iter 6/27 - loss 0.00146119\n",
            "2019-02-11 09:42:27,972 epoch 96 - iter 8/27 - loss 0.00148431\n",
            "2019-02-11 09:42:28,656 epoch 96 - iter 10/27 - loss 0.00141120\n",
            "2019-02-11 09:42:29,377 epoch 96 - iter 12/27 - loss 0.00130123\n",
            "2019-02-11 09:42:30,407 epoch 96 - iter 14/27 - loss 0.00136553\n",
            "2019-02-11 09:42:31,179 epoch 96 - iter 16/27 - loss 0.00148147\n",
            "2019-02-11 09:42:32,468 epoch 96 - iter 18/27 - loss 0.00149857\n",
            "2019-02-11 09:42:33,566 epoch 96 - iter 20/27 - loss 0.00151146\n",
            "2019-02-11 09:42:34,123 epoch 96 - iter 22/27 - loss 0.00152458\n",
            "2019-02-11 09:42:34,805 epoch 96 - iter 24/27 - loss 0.00159038\n",
            "2019-02-11 09:42:35,358 epoch 96 - iter 26/27 - loss 0.00175006\n",
            "2019-02-11 09:42:35,370 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:42:35,371 EPOCH 96 done: loss 0.0018 - lr 0.0125 - bad epochs 3\n",
            "2019-02-11 09:42:36,032 DEV  : loss 0.01482404 - f-score 0.8526 - acc 0.8526\n",
            "2019-02-11 09:42:39,870 TEST : loss 0.00035710 - f-score 0.9957 - acc 0.9957\n",
            "Epoch    95: reducing learning rate of group 0 to 6.2500e-03.\n",
            "2019-02-11 09:42:39,873 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:42:40,250 epoch 97 - iter 0/27 - loss 0.00124806\n",
            "2019-02-11 09:42:40,994 epoch 97 - iter 2/27 - loss 0.00238392\n",
            "2019-02-11 09:42:41,681 epoch 97 - iter 4/27 - loss 0.00186391\n",
            "2019-02-11 09:42:42,797 epoch 97 - iter 6/27 - loss 0.00179380\n",
            "2019-02-11 09:42:43,501 epoch 97 - iter 8/27 - loss 0.00170548\n",
            "2019-02-11 09:42:44,342 epoch 97 - iter 10/27 - loss 0.00227974\n",
            "2019-02-11 09:42:44,992 epoch 97 - iter 12/27 - loss 0.00220341\n",
            "2019-02-11 09:42:45,853 epoch 97 - iter 14/27 - loss 0.00259365\n",
            "2019-02-11 09:42:46,675 epoch 97 - iter 16/27 - loss 0.00253228\n",
            "2019-02-11 09:42:47,409 epoch 97 - iter 18/27 - loss 0.00234560\n",
            "2019-02-11 09:42:48,695 epoch 97 - iter 20/27 - loss 0.00245537\n",
            "2019-02-11 09:42:49,635 epoch 97 - iter 22/27 - loss 0.00228538\n",
            "2019-02-11 09:42:50,412 epoch 97 - iter 24/27 - loss 0.00230954\n",
            "2019-02-11 09:42:50,897 epoch 97 - iter 26/27 - loss 0.00246353\n",
            "2019-02-11 09:42:50,909 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:42:50,911 EPOCH 97 done: loss 0.0025 - lr 0.0063 - bad epochs 0\n",
            "2019-02-11 09:42:51,568 DEV  : loss 0.01493072 - f-score 0.8421 - acc 0.8421\n",
            "2019-02-11 09:42:55,522 TEST : loss 0.00033902 - f-score 0.9957 - acc 0.9957\n",
            "2019-02-11 09:42:55,525 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:42:55,837 epoch 98 - iter 0/27 - loss 0.00015428\n",
            "2019-02-11 09:42:56,661 epoch 98 - iter 2/27 - loss 0.00047047\n",
            "2019-02-11 09:42:57,402 epoch 98 - iter 4/27 - loss 0.00179261\n",
            "2019-02-11 09:42:58,553 epoch 98 - iter 6/27 - loss 0.00193020\n",
            "2019-02-11 09:42:59,230 epoch 98 - iter 8/27 - loss 0.00200499\n",
            "2019-02-11 09:43:00,541 epoch 98 - iter 10/27 - loss 0.00243724\n",
            "2019-02-11 09:43:01,453 epoch 98 - iter 12/27 - loss 0.00226022\n",
            "2019-02-11 09:43:02,229 epoch 98 - iter 14/27 - loss 0.00226214\n",
            "2019-02-11 09:43:03,145 epoch 98 - iter 16/27 - loss 0.00205614\n",
            "2019-02-11 09:43:03,751 epoch 98 - iter 18/27 - loss 0.00197931\n",
            "2019-02-11 09:43:04,317 epoch 98 - iter 20/27 - loss 0.00200474\n",
            "2019-02-11 09:43:05,004 epoch 98 - iter 22/27 - loss 0.00210095\n",
            "2019-02-11 09:43:05,961 epoch 98 - iter 24/27 - loss 0.00199993\n",
            "2019-02-11 09:43:06,441 epoch 98 - iter 26/27 - loss 0.00192688\n",
            "2019-02-11 09:43:06,453 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:43:06,455 EPOCH 98 done: loss 0.0019 - lr 0.0063 - bad epochs 1\n",
            "2019-02-11 09:43:07,117 DEV  : loss 0.01493442 - f-score 0.8526 - acc 0.8526\n",
            "2019-02-11 09:43:10,962 TEST : loss 0.00033162 - f-score 0.9957 - acc 0.9957\n",
            "2019-02-11 09:43:10,965 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:43:11,302 epoch 99 - iter 0/27 - loss 0.00106037\n",
            "2019-02-11 09:43:12,018 epoch 99 - iter 2/27 - loss 0.00096117\n",
            "2019-02-11 09:43:12,984 epoch 99 - iter 4/27 - loss 0.00106065\n",
            "2019-02-11 09:43:13,753 epoch 99 - iter 6/27 - loss 0.00126713\n",
            "2019-02-11 09:43:14,375 epoch 99 - iter 8/27 - loss 0.00207127\n",
            "2019-02-11 09:43:15,128 epoch 99 - iter 10/27 - loss 0.00186745\n",
            "2019-02-11 09:43:16,166 epoch 99 - iter 12/27 - loss 0.00220354\n",
            "2019-02-11 09:43:16,893 epoch 99 - iter 14/27 - loss 0.00215708\n",
            "2019-02-11 09:43:17,660 epoch 99 - iter 16/27 - loss 0.00212847\n",
            "2019-02-11 09:43:18,382 epoch 99 - iter 18/27 - loss 0.00196369\n",
            "2019-02-11 09:43:19,520 epoch 99 - iter 20/27 - loss 0.00188763\n",
            "2019-02-11 09:43:20,693 epoch 99 - iter 22/27 - loss 0.00182957\n",
            "2019-02-11 09:43:21,634 epoch 99 - iter 24/27 - loss 0.00179459\n",
            "2019-02-11 09:43:22,037 epoch 99 - iter 26/27 - loss 0.00196185\n",
            "2019-02-11 09:43:22,049 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:43:22,051 EPOCH 99 done: loss 0.0020 - lr 0.0063 - bad epochs 2\n",
            "2019-02-11 09:43:22,714 DEV  : loss 0.01499727 - f-score 0.8526 - acc 0.8526\n",
            "2019-02-11 09:43:26,551 TEST : loss 0.00027063 - f-score 0.9957 - acc 0.9957\n",
            "2019-02-11 09:43:26,554 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:43:26,867 epoch 100 - iter 0/27 - loss 0.00292115\n",
            "2019-02-11 09:43:27,616 epoch 100 - iter 2/27 - loss 0.00240631\n",
            "2019-02-11 09:43:28,356 epoch 100 - iter 4/27 - loss 0.00197155\n",
            "2019-02-11 09:43:29,673 epoch 100 - iter 6/27 - loss 0.00200016\n",
            "2019-02-11 09:43:30,793 epoch 100 - iter 8/27 - loss 0.00225565\n",
            "2019-02-11 09:43:31,693 epoch 100 - iter 10/27 - loss 0.00243274\n",
            "2019-02-11 09:43:32,408 epoch 100 - iter 12/27 - loss 0.00270138\n",
            "2019-02-11 09:43:33,165 epoch 100 - iter 14/27 - loss 0.00246267\n",
            "2019-02-11 09:43:33,916 epoch 100 - iter 16/27 - loss 0.00272752\n",
            "2019-02-11 09:43:34,634 epoch 100 - iter 18/27 - loss 0.00277109\n",
            "2019-02-11 09:43:35,373 epoch 100 - iter 20/27 - loss 0.00257893\n",
            "2019-02-11 09:43:36,165 epoch 100 - iter 22/27 - loss 0.00244884\n",
            "2019-02-11 09:43:37,138 epoch 100 - iter 24/27 - loss 0.00266892\n",
            "2019-02-11 09:43:37,581 epoch 100 - iter 26/27 - loss 0.00280347\n",
            "2019-02-11 09:43:37,592 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:43:37,594 EPOCH 100 done: loss 0.0028 - lr 0.0063 - bad epochs 3\n",
            "2019-02-11 09:43:38,255 DEV  : loss 0.01478276 - f-score 0.8526 - acc 0.8526\n",
            "2019-02-11 09:43:42,230 TEST : loss 0.00027685 - f-score 0.9957 - acc 0.9957\n",
            "Epoch    99: reducing learning rate of group 0 to 3.1250e-03.\n",
            "2019-02-11 09:43:59,952 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:43:59,954 Testing using best model ...\n",
            "2019-02-11 09:44:15,774 MICRO_AVG: acc 0.9979 - f1-score 0.9979\n",
            "2019-02-11 09:44:15,776 MARCO_AVG: acc 0.9979 - f1-score 0.9977\n",
            "2019-02-11 09:44:15,779 0          tp: 216 - fp: 1 - fn: 0 - tn: 216 - precision: 0.9954 - recall: 1.0000 - accuracy: 0.9977 - f1-score: 0.9977\n",
            "2019-02-11 09:44:15,784 1          tp: 251 - fp: 0 - fn: 1 - tn: 251 - precision: 1.0000 - recall: 0.9960 - accuracy: 0.9980 - f1-score: 0.9980\n",
            "2019-02-11 09:44:15,786 ----------------------------------------------------------------------------------------------------\n",
            "0.996 0.996 0.996 0.996 0.996 0.996 0.996\n",
            "841 95 466\n",
            "2019-02-11 09:44:21,032 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:44:21,034 Evaluation method: MICRO_F1_SCORE\n",
            "2019-02-11 09:44:21,039 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:44:21,452 epoch 1 - iter 0/27 - loss 0.02829367\n",
            "2019-02-11 09:44:22,256 epoch 1 - iter 2/27 - loss 0.07939318\n",
            "2019-02-11 09:44:22,923 epoch 1 - iter 4/27 - loss 0.07758305\n",
            "2019-02-11 09:44:23,636 epoch 1 - iter 6/27 - loss 0.08193175\n",
            "2019-02-11 09:44:24,317 epoch 1 - iter 8/27 - loss 0.07947498\n",
            "2019-02-11 09:44:25,586 epoch 1 - iter 10/27 - loss 0.07637887\n",
            "2019-02-11 09:44:26,300 epoch 1 - iter 12/27 - loss 0.07295545\n",
            "2019-02-11 09:44:27,099 epoch 1 - iter 14/27 - loss 0.06750558\n",
            "2019-02-11 09:44:27,773 epoch 1 - iter 16/27 - loss 0.06232175\n",
            "2019-02-11 09:44:28,728 epoch 1 - iter 18/27 - loss 0.05914163\n",
            "2019-02-11 09:44:29,624 epoch 1 - iter 20/27 - loss 0.05699963\n",
            "2019-02-11 09:44:30,342 epoch 1 - iter 22/27 - loss 0.05440823\n",
            "2019-02-11 09:44:31,732 epoch 1 - iter 24/27 - loss 0.05240695\n",
            "2019-02-11 09:44:32,323 epoch 1 - iter 26/27 - loss 0.05319146\n",
            "2019-02-11 09:44:32,335 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:44:32,336 EPOCH 1 done: loss 0.0532 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:44:32,992 DEV  : loss 0.02355347 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 09:44:39,582 TEST : loss 0.03022930 - f-score 0.5386 - acc 0.5386\n",
            "2019-02-11 09:44:57,566 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:44:57,967 epoch 2 - iter 0/27 - loss 0.02551029\n",
            "2019-02-11 09:44:59,063 epoch 2 - iter 2/27 - loss 0.02999046\n",
            "2019-02-11 09:45:00,032 epoch 2 - iter 4/27 - loss 0.03648085\n",
            "2019-02-11 09:45:00,830 epoch 2 - iter 6/27 - loss 0.03993749\n",
            "2019-02-11 09:45:01,842 epoch 2 - iter 8/27 - loss 0.03902192\n",
            "2019-02-11 09:45:02,961 epoch 2 - iter 10/27 - loss 0.03744782\n",
            "2019-02-11 09:45:03,888 epoch 2 - iter 12/27 - loss 0.03596844\n",
            "2019-02-11 09:45:04,929 epoch 2 - iter 14/27 - loss 0.03397012\n",
            "2019-02-11 09:45:06,384 epoch 2 - iter 16/27 - loss 0.03372244\n",
            "2019-02-11 09:45:07,240 epoch 2 - iter 18/27 - loss 0.03308734\n",
            "2019-02-11 09:45:08,040 epoch 2 - iter 20/27 - loss 0.03238069\n",
            "2019-02-11 09:45:09,421 epoch 2 - iter 22/27 - loss 0.03129221\n",
            "2019-02-11 09:45:10,183 epoch 2 - iter 24/27 - loss 0.03057714\n",
            "2019-02-11 09:45:10,707 epoch 2 - iter 26/27 - loss 0.03116284\n",
            "2019-02-11 09:45:10,719 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:45:10,720 EPOCH 2 done: loss 0.0312 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:45:11,513 DEV  : loss 0.01983558 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 09:45:16,413 TEST : loss 0.02312489 - f-score 0.5386 - acc 0.5386\n",
            "2019-02-11 09:45:34,638 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:45:35,516 epoch 3 - iter 0/27 - loss 0.02531972\n",
            "2019-02-11 09:45:36,780 epoch 3 - iter 2/27 - loss 0.02178314\n",
            "2019-02-11 09:45:37,932 epoch 3 - iter 4/27 - loss 0.02298508\n",
            "2019-02-11 09:45:38,892 epoch 3 - iter 6/27 - loss 0.02361737\n",
            "2019-02-11 09:45:39,724 epoch 3 - iter 8/27 - loss 0.02532925\n",
            "2019-02-11 09:45:40,555 epoch 3 - iter 10/27 - loss 0.02576703\n",
            "2019-02-11 09:45:41,326 epoch 3 - iter 12/27 - loss 0.02580680\n",
            "2019-02-11 09:45:42,196 epoch 3 - iter 14/27 - loss 0.02603701\n",
            "2019-02-11 09:45:43,074 epoch 3 - iter 16/27 - loss 0.02557416\n",
            "2019-02-11 09:45:43,957 epoch 3 - iter 18/27 - loss 0.02544342\n",
            "2019-02-11 09:45:44,797 epoch 3 - iter 20/27 - loss 0.02606070\n",
            "2019-02-11 09:45:45,831 epoch 3 - iter 22/27 - loss 0.02622206\n",
            "2019-02-11 09:45:46,623 epoch 3 - iter 24/27 - loss 0.02580848\n",
            "2019-02-11 09:45:47,213 epoch 3 - iter 26/27 - loss 0.02599243\n",
            "2019-02-11 09:45:47,225 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:45:47,227 EPOCH 3 done: loss 0.0260 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:45:47,992 DEV  : loss 0.03759937 - f-score 0.5579 - acc 0.5579\n",
            "2019-02-11 09:45:52,818 TEST : loss 0.03073651 - f-score 0.6330 - acc 0.6330\n",
            "2019-02-11 09:46:10,807 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:46:11,359 epoch 4 - iter 0/27 - loss 0.01921150\n",
            "2019-02-11 09:46:12,579 epoch 4 - iter 2/27 - loss 0.02607061\n",
            "2019-02-11 09:46:13,467 epoch 4 - iter 4/27 - loss 0.02441552\n",
            "2019-02-11 09:46:14,194 epoch 4 - iter 6/27 - loss 0.02333362\n",
            "2019-02-11 09:46:14,907 epoch 4 - iter 8/27 - loss 0.02222041\n",
            "2019-02-11 09:46:16,161 epoch 4 - iter 10/27 - loss 0.02226645\n",
            "2019-02-11 09:46:16,799 epoch 4 - iter 12/27 - loss 0.02311046\n",
            "2019-02-11 09:46:17,527 epoch 4 - iter 14/27 - loss 0.02417325\n",
            "2019-02-11 09:46:18,765 epoch 4 - iter 16/27 - loss 0.02366992\n",
            "2019-02-11 09:46:19,524 epoch 4 - iter 18/27 - loss 0.02348143\n",
            "2019-02-11 09:46:20,423 epoch 4 - iter 20/27 - loss 0.02373427\n",
            "2019-02-11 09:46:21,236 epoch 4 - iter 22/27 - loss 0.02439080\n",
            "2019-02-11 09:46:22,164 epoch 4 - iter 24/27 - loss 0.02484776\n",
            "2019-02-11 09:46:22,638 epoch 4 - iter 26/27 - loss 0.02558263\n",
            "2019-02-11 09:46:22,651 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:46:22,652 EPOCH 4 done: loss 0.0256 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:46:23,422 DEV  : loss 0.02409314 - f-score 0.5368 - acc 0.5368\n",
            "2019-02-11 09:46:28,529 TEST : loss 0.02176826 - f-score 0.6459 - acc 0.6459\n",
            "2019-02-11 09:46:46,510 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:46:47,057 epoch 5 - iter 0/27 - loss 0.02512185\n",
            "2019-02-11 09:46:48,512 epoch 5 - iter 2/27 - loss 0.02677488\n",
            "2019-02-11 09:46:49,432 epoch 5 - iter 4/27 - loss 0.02929952\n",
            "2019-02-11 09:46:50,089 epoch 5 - iter 6/27 - loss 0.02833216\n",
            "2019-02-11 09:46:50,852 epoch 5 - iter 8/27 - loss 0.02622261\n",
            "2019-02-11 09:46:51,695 epoch 5 - iter 10/27 - loss 0.02624908\n",
            "2019-02-11 09:46:52,313 epoch 5 - iter 12/27 - loss 0.02741200\n",
            "2019-02-11 09:46:53,215 epoch 5 - iter 14/27 - loss 0.02672665\n",
            "2019-02-11 09:46:54,059 epoch 5 - iter 16/27 - loss 0.02629693\n",
            "2019-02-11 09:46:54,904 epoch 5 - iter 18/27 - loss 0.02674624\n",
            "2019-02-11 09:46:56,081 epoch 5 - iter 20/27 - loss 0.02727110\n",
            "2019-02-11 09:46:56,933 epoch 5 - iter 22/27 - loss 0.02682860\n",
            "2019-02-11 09:46:57,556 epoch 5 - iter 24/27 - loss 0.02774947\n",
            "2019-02-11 09:46:58,074 epoch 5 - iter 26/27 - loss 0.02873739\n",
            "2019-02-11 09:46:58,086 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:46:58,088 EPOCH 5 done: loss 0.0287 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:46:58,826 DEV  : loss 0.01991179 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 09:47:03,582 TEST : loss 0.02310546 - f-score 0.5451 - acc 0.5451\n",
            "2019-02-11 09:47:03,585 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:47:04,040 epoch 6 - iter 0/27 - loss 0.02036749\n",
            "2019-02-11 09:47:04,842 epoch 6 - iter 2/27 - loss 0.02593153\n",
            "2019-02-11 09:47:06,080 epoch 6 - iter 4/27 - loss 0.02424065\n",
            "2019-02-11 09:47:07,152 epoch 6 - iter 6/27 - loss 0.02394546\n",
            "2019-02-11 09:47:07,885 epoch 6 - iter 8/27 - loss 0.02441845\n",
            "2019-02-11 09:47:09,101 epoch 6 - iter 10/27 - loss 0.02516384\n",
            "2019-02-11 09:47:09,954 epoch 6 - iter 12/27 - loss 0.02458528\n",
            "2019-02-11 09:47:10,827 epoch 6 - iter 14/27 - loss 0.02488825\n",
            "2019-02-11 09:47:11,721 epoch 6 - iter 16/27 - loss 0.02532651\n",
            "2019-02-11 09:47:12,563 epoch 6 - iter 18/27 - loss 0.02543578\n",
            "2019-02-11 09:47:13,209 epoch 6 - iter 20/27 - loss 0.02514860\n",
            "2019-02-11 09:47:14,009 epoch 6 - iter 22/27 - loss 0.02578836\n",
            "2019-02-11 09:47:14,767 epoch 6 - iter 24/27 - loss 0.02551695\n",
            "2019-02-11 09:47:15,160 epoch 6 - iter 26/27 - loss 0.02538360\n",
            "2019-02-11 09:47:15,173 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:47:15,174 EPOCH 6 done: loss 0.0254 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 09:47:15,920 DEV  : loss 0.02515960 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 09:47:20,922 TEST : loss 0.03252517 - f-score 0.5429 - acc 0.5429\n",
            "2019-02-11 09:47:39,305 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:47:40,107 epoch 7 - iter 0/27 - loss 0.03079190\n",
            "2019-02-11 09:47:41,049 epoch 7 - iter 2/27 - loss 0.02406313\n",
            "2019-02-11 09:47:42,122 epoch 7 - iter 4/27 - loss 0.02463342\n",
            "2019-02-11 09:47:42,948 epoch 7 - iter 6/27 - loss 0.02304524\n",
            "2019-02-11 09:47:43,736 epoch 7 - iter 8/27 - loss 0.02414049\n",
            "2019-02-11 09:47:44,460 epoch 7 - iter 10/27 - loss 0.02387447\n",
            "2019-02-11 09:47:45,325 epoch 7 - iter 12/27 - loss 0.02328145\n",
            "2019-02-11 09:47:46,175 epoch 7 - iter 14/27 - loss 0.02328220\n",
            "2019-02-11 09:47:46,993 epoch 7 - iter 16/27 - loss 0.02406140\n",
            "2019-02-11 09:47:47,764 epoch 7 - iter 18/27 - loss 0.02456916\n",
            "2019-02-11 09:47:49,002 epoch 7 - iter 20/27 - loss 0.02411600\n",
            "2019-02-11 09:47:49,804 epoch 7 - iter 22/27 - loss 0.02390819\n",
            "2019-02-11 09:47:51,154 epoch 7 - iter 24/27 - loss 0.02405945\n",
            "2019-02-11 09:47:51,627 epoch 7 - iter 26/27 - loss 0.02571416\n",
            "2019-02-11 09:47:51,639 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:47:51,641 EPOCH 7 done: loss 0.0257 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:47:52,399 DEV  : loss 0.02050157 - f-score 0.6526 - acc 0.6526\n",
            "2019-02-11 09:47:57,301 TEST : loss 0.02292879 - f-score 0.6459 - acc 0.6459\n",
            "2019-02-11 09:47:57,304 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:47:57,624 epoch 8 - iter 0/27 - loss 0.02762554\n",
            "2019-02-11 09:47:58,689 epoch 8 - iter 2/27 - loss 0.03308471\n",
            "2019-02-11 09:47:59,576 epoch 8 - iter 4/27 - loss 0.02984669\n",
            "2019-02-11 09:48:00,425 epoch 8 - iter 6/27 - loss 0.02749662\n",
            "2019-02-11 09:48:01,261 epoch 8 - iter 8/27 - loss 0.02615116\n",
            "2019-02-11 09:48:02,551 epoch 8 - iter 10/27 - loss 0.02647334\n",
            "2019-02-11 09:48:03,250 epoch 8 - iter 12/27 - loss 0.02669572\n",
            "2019-02-11 09:48:04,151 epoch 8 - iter 14/27 - loss 0.02590560\n",
            "2019-02-11 09:48:05,059 epoch 8 - iter 16/27 - loss 0.02582794\n",
            "2019-02-11 09:48:05,804 epoch 8 - iter 18/27 - loss 0.02606447\n",
            "2019-02-11 09:48:07,079 epoch 8 - iter 20/27 - loss 0.02546394\n",
            "2019-02-11 09:48:07,887 epoch 8 - iter 22/27 - loss 0.02522737\n",
            "2019-02-11 09:48:08,842 epoch 8 - iter 24/27 - loss 0.02499015\n",
            "2019-02-11 09:48:09,408 epoch 8 - iter 26/27 - loss 0.02620339\n",
            "2019-02-11 09:48:09,420 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:48:09,421 EPOCH 8 done: loss 0.0262 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 09:48:10,188 DEV  : loss 0.02716739 - f-score 0.6842 - acc 0.6842\n",
            "2019-02-11 09:48:15,152 TEST : loss 0.03202183 - f-score 0.5258 - acc 0.5258\n",
            "2019-02-11 09:48:15,155 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:48:15,584 epoch 9 - iter 0/27 - loss 0.02540569\n",
            "2019-02-11 09:48:16,439 epoch 9 - iter 2/27 - loss 0.02598258\n",
            "2019-02-11 09:48:17,442 epoch 9 - iter 4/27 - loss 0.02883714\n",
            "2019-02-11 09:48:18,271 epoch 9 - iter 6/27 - loss 0.02584676\n",
            "2019-02-11 09:48:18,974 epoch 9 - iter 8/27 - loss 0.02464850\n",
            "2019-02-11 09:48:19,821 epoch 9 - iter 10/27 - loss 0.02535138\n",
            "2019-02-11 09:48:21,265 epoch 9 - iter 12/27 - loss 0.02555195\n",
            "2019-02-11 09:48:22,019 epoch 9 - iter 14/27 - loss 0.02557882\n",
            "2019-02-11 09:48:23,211 epoch 9 - iter 16/27 - loss 0.02491898\n",
            "2019-02-11 09:48:24,104 epoch 9 - iter 18/27 - loss 0.02458144\n",
            "2019-02-11 09:48:25,157 epoch 9 - iter 20/27 - loss 0.02418441\n",
            "2019-02-11 09:48:25,965 epoch 9 - iter 22/27 - loss 0.02395486\n",
            "2019-02-11 09:48:26,712 epoch 9 - iter 24/27 - loss 0.02425166\n",
            "2019-02-11 09:48:27,351 epoch 9 - iter 26/27 - loss 0.02520037\n",
            "2019-02-11 09:48:27,363 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:48:27,364 EPOCH 9 done: loss 0.0252 - lr 0.1000 - bad epochs 2\n",
            "2019-02-11 09:48:28,135 DEV  : loss 0.02021463 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 09:48:33,011 TEST : loss 0.02318612 - f-score 0.6030 - acc 0.6030\n",
            "2019-02-11 09:48:51,022 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:48:51,379 epoch 10 - iter 0/27 - loss 0.01960653\n",
            "2019-02-11 09:48:52,246 epoch 10 - iter 2/27 - loss 0.02409938\n",
            "2019-02-11 09:48:52,998 epoch 10 - iter 4/27 - loss 0.02756160\n",
            "2019-02-11 09:48:53,762 epoch 10 - iter 6/27 - loss 0.02764590\n",
            "2019-02-11 09:48:54,672 epoch 10 - iter 8/27 - loss 0.02668244\n",
            "2019-02-11 09:48:56,291 epoch 10 - iter 10/27 - loss 0.02701456\n",
            "2019-02-11 09:48:57,202 epoch 10 - iter 12/27 - loss 0.02645764\n",
            "2019-02-11 09:48:57,868 epoch 10 - iter 14/27 - loss 0.02526852\n",
            "2019-02-11 09:48:58,720 epoch 10 - iter 16/27 - loss 0.02517591\n",
            "2019-02-11 09:48:59,469 epoch 10 - iter 18/27 - loss 0.02579002\n",
            "2019-02-11 09:49:00,202 epoch 10 - iter 20/27 - loss 0.02560303\n",
            "2019-02-11 09:49:00,972 epoch 10 - iter 22/27 - loss 0.02612988\n",
            "2019-02-11 09:49:01,962 epoch 10 - iter 24/27 - loss 0.02571036\n",
            "2019-02-11 09:49:02,472 epoch 10 - iter 26/27 - loss 0.02700245\n",
            "2019-02-11 09:49:02,484 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:49:02,485 EPOCH 10 done: loss 0.0270 - lr 0.1000 - bad epochs 0\n",
            "2019-02-11 09:49:03,245 DEV  : loss 0.02107807 - f-score 0.6737 - acc 0.6737\n",
            "2019-02-11 09:49:08,254 TEST : loss 0.02247533 - f-score 0.6094 - acc 0.6094\n",
            "2019-02-11 09:49:08,257 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:49:08,680 epoch 11 - iter 0/27 - loss 0.01980860\n",
            "2019-02-11 09:49:09,648 epoch 11 - iter 2/27 - loss 0.02008031\n",
            "2019-02-11 09:49:10,500 epoch 11 - iter 4/27 - loss 0.02678327\n",
            "2019-02-11 09:49:11,263 epoch 11 - iter 6/27 - loss 0.02727984\n",
            "2019-02-11 09:49:11,927 epoch 11 - iter 8/27 - loss 0.02578445\n",
            "2019-02-11 09:49:13,643 epoch 11 - iter 10/27 - loss 0.02581531\n",
            "2019-02-11 09:49:14,423 epoch 11 - iter 12/27 - loss 0.02579880\n",
            "2019-02-11 09:49:15,261 epoch 11 - iter 14/27 - loss 0.02652832\n",
            "2019-02-11 09:49:16,261 epoch 11 - iter 16/27 - loss 0.02578749\n",
            "2019-02-11 09:49:17,117 epoch 11 - iter 18/27 - loss 0.02516432\n",
            "2019-02-11 09:49:18,193 epoch 11 - iter 20/27 - loss 0.02475374\n",
            "2019-02-11 09:49:19,005 epoch 11 - iter 22/27 - loss 0.02459626\n",
            "2019-02-11 09:49:19,914 epoch 11 - iter 24/27 - loss 0.02517531\n",
            "2019-02-11 09:49:20,470 epoch 11 - iter 26/27 - loss 0.02587905\n",
            "2019-02-11 09:49:20,482 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:49:20,483 EPOCH 11 done: loss 0.0259 - lr 0.1000 - bad epochs 1\n",
            "2019-02-11 09:49:21,241 DEV  : loss 0.02353608 - f-score 0.5368 - acc 0.5368\n",
            "2019-02-11 09:49:26,047 TEST : loss 0.02150565 - f-score 0.6330 - acc 0.6330\n",
            "2019-02-11 09:49:26,050 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:49:26,433 epoch 12 - iter 0/27 - loss 0.02263664\n",
            "2019-02-11 09:49:27,152 epoch 12 - iter 2/27 - loss 0.02491762\n",
            "2019-02-11 09:49:28,218 epoch 12 - iter 4/27 - loss 0.02570560\n",
            "2019-02-11 09:49:29,578 epoch 12 - iter 6/27 - loss 0.02502053\n",
            "2019-02-11 09:49:30,420 epoch 12 - iter 8/27 - loss 0.02600087\n",
            "2019-02-11 09:49:31,170 epoch 12 - iter 10/27 - loss 0.02605400\n",
            "2019-02-11 09:49:32,026 epoch 12 - iter 12/27 - loss 0.02478262\n",
            "2019-02-11 09:49:33,229 epoch 12 - iter 14/27 - loss 0.02494622\n",
            "2019-02-11 09:49:34,083 epoch 12 - iter 16/27 - loss 0.02535273\n",
            "2019-02-11 09:49:34,874 epoch 12 - iter 18/27 - loss 0.02495629\n",
            "2019-02-11 09:49:35,562 epoch 12 - iter 20/27 - loss 0.02478326\n",
            "2019-02-11 09:49:36,202 epoch 12 - iter 22/27 - loss 0.02432256\n",
            "2019-02-11 09:49:37,031 epoch 12 - iter 24/27 - loss 0.02408451\n",
            "2019-02-11 09:49:37,482 epoch 12 - iter 26/27 - loss 0.02531125\n",
            "2019-02-11 09:49:37,493 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:49:37,494 EPOCH 12 done: loss 0.0253 - lr 0.1000 - bad epochs 2\n",
            "2019-02-11 09:49:38,226 DEV  : loss 0.02486476 - f-score 0.5474 - acc 0.5474\n",
            "2019-02-11 09:49:42,835 TEST : loss 0.02338766 - f-score 0.6545 - acc 0.6545\n",
            "2019-02-11 09:49:42,838 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:49:43,127 epoch 13 - iter 0/27 - loss 0.01970611\n",
            "2019-02-11 09:49:43,904 epoch 13 - iter 2/27 - loss 0.02254052\n",
            "2019-02-11 09:49:44,688 epoch 13 - iter 4/27 - loss 0.02209285\n",
            "2019-02-11 09:49:45,873 epoch 13 - iter 6/27 - loss 0.02412952\n",
            "2019-02-11 09:49:46,674 epoch 13 - iter 8/27 - loss 0.02543917\n",
            "2019-02-11 09:49:47,550 epoch 13 - iter 10/27 - loss 0.02495964\n",
            "2019-02-11 09:49:48,259 epoch 13 - iter 12/27 - loss 0.02718171\n",
            "2019-02-11 09:49:49,332 epoch 13 - iter 14/27 - loss 0.02735535\n",
            "2019-02-11 09:49:50,255 epoch 13 - iter 16/27 - loss 0.02686021\n",
            "2019-02-11 09:49:51,093 epoch 13 - iter 18/27 - loss 0.02807468\n",
            "2019-02-11 09:49:52,306 epoch 13 - iter 20/27 - loss 0.02777609\n",
            "2019-02-11 09:49:53,159 epoch 13 - iter 22/27 - loss 0.02728890\n",
            "2019-02-11 09:49:53,923 epoch 13 - iter 24/27 - loss 0.02709590\n",
            "2019-02-11 09:49:54,512 epoch 13 - iter 26/27 - loss 0.02751350\n",
            "2019-02-11 09:49:54,524 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:49:54,525 EPOCH 13 done: loss 0.0275 - lr 0.1000 - bad epochs 3\n",
            "2019-02-11 09:49:55,269 DEV  : loss 0.02222247 - f-score 0.6737 - acc 0.6737\n",
            "2019-02-11 09:50:00,255 TEST : loss 0.02455069 - f-score 0.6288 - acc 0.6288\n",
            "Epoch    12: reducing learning rate of group 0 to 5.0000e-02.\n",
            "2019-02-11 09:50:00,260 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:50:01,143 epoch 14 - iter 0/27 - loss 0.02072010\n",
            "2019-02-11 09:50:02,108 epoch 14 - iter 2/27 - loss 0.02243531\n",
            "2019-02-11 09:50:03,010 epoch 14 - iter 4/27 - loss 0.02163795\n",
            "2019-02-11 09:50:03,826 epoch 14 - iter 6/27 - loss 0.02069180\n",
            "2019-02-11 09:50:04,548 epoch 14 - iter 8/27 - loss 0.02000927\n",
            "2019-02-11 09:50:05,390 epoch 14 - iter 10/27 - loss 0.01999994\n",
            "2019-02-11 09:50:06,143 epoch 14 - iter 12/27 - loss 0.02023883\n",
            "2019-02-11 09:50:07,119 epoch 14 - iter 14/27 - loss 0.02053724\n",
            "2019-02-11 09:50:08,523 epoch 14 - iter 16/27 - loss 0.02092459\n",
            "2019-02-11 09:50:09,581 epoch 14 - iter 18/27 - loss 0.02074346\n",
            "2019-02-11 09:50:10,640 epoch 14 - iter 20/27 - loss 0.02080296\n",
            "2019-02-11 09:50:11,579 epoch 14 - iter 22/27 - loss 0.02098404\n",
            "2019-02-11 09:50:12,416 epoch 14 - iter 24/27 - loss 0.02101045\n",
            "2019-02-11 09:50:13,025 epoch 14 - iter 26/27 - loss 0.02145029\n",
            "2019-02-11 09:50:13,037 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:50:13,039 EPOCH 14 done: loss 0.0215 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 09:50:13,802 DEV  : loss 0.01875665 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 09:50:18,632 TEST : loss 0.02101262 - f-score 0.6352 - acc 0.6352\n",
            "2019-02-11 09:50:38,072 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:50:38,384 epoch 15 - iter 0/27 - loss 0.02288427\n",
            "2019-02-11 09:50:39,221 epoch 15 - iter 2/27 - loss 0.02336027\n",
            "2019-02-11 09:50:40,065 epoch 15 - iter 4/27 - loss 0.02161285\n",
            "2019-02-11 09:50:41,088 epoch 15 - iter 6/27 - loss 0.02168114\n",
            "2019-02-11 09:50:41,756 epoch 15 - iter 8/27 - loss 0.02135985\n",
            "2019-02-11 09:50:42,622 epoch 15 - iter 10/27 - loss 0.02108085\n",
            "2019-02-11 09:50:43,478 epoch 15 - iter 12/27 - loss 0.02102752\n",
            "2019-02-11 09:50:44,174 epoch 15 - iter 14/27 - loss 0.02088306\n",
            "2019-02-11 09:50:45,070 epoch 15 - iter 16/27 - loss 0.02126079\n",
            "2019-02-11 09:50:45,970 epoch 15 - iter 18/27 - loss 0.02118162\n",
            "2019-02-11 09:50:47,131 epoch 15 - iter 20/27 - loss 0.02091955\n",
            "2019-02-11 09:50:48,062 epoch 15 - iter 22/27 - loss 0.02072505\n",
            "2019-02-11 09:50:49,412 epoch 15 - iter 24/27 - loss 0.02094133\n",
            "2019-02-11 09:50:49,958 epoch 15 - iter 26/27 - loss 0.02148432\n",
            "2019-02-11 09:50:49,971 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:50:49,973 EPOCH 15 done: loss 0.0215 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 09:50:50,741 DEV  : loss 0.01969429 - f-score 0.6737 - acc 0.6737\n",
            "2019-02-11 09:50:55,827 TEST : loss 0.02101159 - f-score 0.6524 - acc 0.6524\n",
            "2019-02-11 09:50:55,833 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:50:56,460 epoch 16 - iter 0/27 - loss 0.01949861\n",
            "2019-02-11 09:50:57,205 epoch 16 - iter 2/27 - loss 0.01874307\n",
            "2019-02-11 09:50:58,414 epoch 16 - iter 4/27 - loss 0.02113071\n",
            "2019-02-11 09:50:59,246 epoch 16 - iter 6/27 - loss 0.02136881\n",
            "2019-02-11 09:51:00,469 epoch 16 - iter 8/27 - loss 0.02112092\n",
            "2019-02-11 09:51:01,379 epoch 16 - iter 10/27 - loss 0.02141616\n",
            "2019-02-11 09:51:02,214 epoch 16 - iter 12/27 - loss 0.02082055\n",
            "2019-02-11 09:51:02,962 epoch 16 - iter 14/27 - loss 0.02018776\n",
            "2019-02-11 09:51:03,784 epoch 16 - iter 16/27 - loss 0.02045287\n",
            "2019-02-11 09:51:04,609 epoch 16 - iter 18/27 - loss 0.02030998\n",
            "2019-02-11 09:51:05,365 epoch 16 - iter 20/27 - loss 0.02043868\n",
            "2019-02-11 09:51:06,163 epoch 16 - iter 22/27 - loss 0.02014623\n",
            "2019-02-11 09:51:07,478 epoch 16 - iter 24/27 - loss 0.02005141\n",
            "2019-02-11 09:51:08,013 epoch 16 - iter 26/27 - loss 0.02064713\n",
            "2019-02-11 09:51:08,026 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:51:08,028 EPOCH 16 done: loss 0.0206 - lr 0.0500 - bad epochs 1\n",
            "2019-02-11 09:51:08,795 DEV  : loss 0.02656222 - f-score 0.5579 - acc 0.5579\n",
            "2019-02-11 09:51:13,624 TEST : loss 0.02475790 - f-score 0.6373 - acc 0.6373\n",
            "2019-02-11 09:51:31,715 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:51:32,384 epoch 17 - iter 0/27 - loss 0.02926105\n",
            "2019-02-11 09:51:33,375 epoch 17 - iter 2/27 - loss 0.02377609\n",
            "2019-02-11 09:51:34,043 epoch 17 - iter 4/27 - loss 0.02048594\n",
            "2019-02-11 09:51:34,925 epoch 17 - iter 6/27 - loss 0.02026414\n",
            "2019-02-11 09:51:35,894 epoch 17 - iter 8/27 - loss 0.01981911\n",
            "2019-02-11 09:51:37,074 epoch 17 - iter 10/27 - loss 0.02024855\n",
            "2019-02-11 09:51:37,975 epoch 17 - iter 12/27 - loss 0.02071158\n",
            "2019-02-11 09:51:38,766 epoch 17 - iter 14/27 - loss 0.02014014\n",
            "2019-02-11 09:51:40,102 epoch 17 - iter 16/27 - loss 0.02020887\n",
            "2019-02-11 09:51:40,695 epoch 17 - iter 18/27 - loss 0.02028653\n",
            "2019-02-11 09:51:41,404 epoch 17 - iter 20/27 - loss 0.02035793\n",
            "2019-02-11 09:51:42,180 epoch 17 - iter 22/27 - loss 0.02018866\n",
            "2019-02-11 09:51:42,997 epoch 17 - iter 24/27 - loss 0.02040067\n",
            "2019-02-11 09:51:43,514 epoch 17 - iter 26/27 - loss 0.02101720\n",
            "2019-02-11 09:51:43,527 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:51:43,528 EPOCH 17 done: loss 0.0210 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 09:51:44,284 DEV  : loss 0.02049340 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 09:51:49,230 TEST : loss 0.02530762 - f-score 0.6052 - acc 0.6052\n",
            "2019-02-11 09:51:49,233 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:51:49,650 epoch 18 - iter 0/27 - loss 0.02759937\n",
            "2019-02-11 09:51:50,763 epoch 18 - iter 2/27 - loss 0.02738031\n",
            "2019-02-11 09:51:51,443 epoch 18 - iter 4/27 - loss 0.02496928\n",
            "2019-02-11 09:51:52,308 epoch 18 - iter 6/27 - loss 0.02382523\n",
            "2019-02-11 09:51:53,140 epoch 18 - iter 8/27 - loss 0.02389844\n",
            "2019-02-11 09:51:54,197 epoch 18 - iter 10/27 - loss 0.02301495\n",
            "2019-02-11 09:51:54,879 epoch 18 - iter 12/27 - loss 0.02264932\n",
            "2019-02-11 09:51:55,608 epoch 18 - iter 14/27 - loss 0.02264256\n",
            "2019-02-11 09:51:56,478 epoch 18 - iter 16/27 - loss 0.02239838\n",
            "2019-02-11 09:51:57,249 epoch 18 - iter 18/27 - loss 0.02198321\n",
            "2019-02-11 09:51:58,577 epoch 18 - iter 20/27 - loss 0.02196632\n",
            "2019-02-11 09:51:59,819 epoch 18 - iter 22/27 - loss 0.02177875\n",
            "2019-02-11 09:52:00,579 epoch 18 - iter 24/27 - loss 0.02152414\n",
            "2019-02-11 09:52:01,059 epoch 18 - iter 26/27 - loss 0.02197053\n",
            "2019-02-11 09:52:01,072 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:52:01,074 EPOCH 18 done: loss 0.0220 - lr 0.0500 - bad epochs 1\n",
            "2019-02-11 09:52:01,817 DEV  : loss 0.01962078 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 09:52:06,544 TEST : loss 0.02068865 - f-score 0.6609 - acc 0.6609\n",
            "2019-02-11 09:52:06,547 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:52:06,931 epoch 19 - iter 0/27 - loss 0.01751054\n",
            "2019-02-11 09:52:07,756 epoch 19 - iter 2/27 - loss 0.01900393\n",
            "2019-02-11 09:52:08,652 epoch 19 - iter 4/27 - loss 0.01931244\n",
            "2019-02-11 09:52:09,434 epoch 19 - iter 6/27 - loss 0.01997476\n",
            "2019-02-11 09:52:10,155 epoch 19 - iter 8/27 - loss 0.02029828\n",
            "2019-02-11 09:52:10,891 epoch 19 - iter 10/27 - loss 0.02012741\n",
            "2019-02-11 09:52:11,779 epoch 19 - iter 12/27 - loss 0.02010256\n",
            "2019-02-11 09:52:12,507 epoch 19 - iter 14/27 - loss 0.01960773\n",
            "2019-02-11 09:52:13,343 epoch 19 - iter 16/27 - loss 0.01970718\n",
            "2019-02-11 09:52:14,154 epoch 19 - iter 18/27 - loss 0.01958893\n",
            "2019-02-11 09:52:14,896 epoch 19 - iter 20/27 - loss 0.01895649\n",
            "2019-02-11 09:52:16,405 epoch 19 - iter 22/27 - loss 0.01957753\n",
            "2019-02-11 09:52:17,520 epoch 19 - iter 24/27 - loss 0.01949733\n",
            "2019-02-11 09:52:18,054 epoch 19 - iter 26/27 - loss 0.01990386\n",
            "2019-02-11 09:52:18,067 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:52:18,068 EPOCH 19 done: loss 0.0199 - lr 0.0500 - bad epochs 2\n",
            "2019-02-11 09:52:18,829 DEV  : loss 0.02008854 - f-score 0.6526 - acc 0.6526\n",
            "2019-02-11 09:52:23,600 TEST : loss 0.02118174 - f-score 0.6524 - acc 0.6524\n",
            "2019-02-11 09:52:41,439 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:52:42,182 epoch 20 - iter 0/27 - loss 0.02132495\n",
            "2019-02-11 09:52:42,834 epoch 20 - iter 2/27 - loss 0.01832624\n",
            "2019-02-11 09:52:43,826 epoch 20 - iter 4/27 - loss 0.01821933\n",
            "2019-02-11 09:52:44,454 epoch 20 - iter 6/27 - loss 0.01764760\n",
            "2019-02-11 09:52:45,283 epoch 20 - iter 8/27 - loss 0.01931460\n",
            "2019-02-11 09:52:46,112 epoch 20 - iter 10/27 - loss 0.01917840\n",
            "2019-02-11 09:52:47,382 epoch 20 - iter 12/27 - loss 0.01893119\n",
            "2019-02-11 09:52:48,140 epoch 20 - iter 14/27 - loss 0.01912019\n",
            "2019-02-11 09:52:49,384 epoch 20 - iter 16/27 - loss 0.01966055\n",
            "2019-02-11 09:52:50,166 epoch 20 - iter 18/27 - loss 0.01988865\n",
            "2019-02-11 09:52:50,941 epoch 20 - iter 20/27 - loss 0.01985775\n",
            "2019-02-11 09:52:51,697 epoch 20 - iter 22/27 - loss 0.01988091\n",
            "2019-02-11 09:52:52,654 epoch 20 - iter 24/27 - loss 0.01990474\n",
            "2019-02-11 09:52:53,144 epoch 20 - iter 26/27 - loss 0.02051418\n",
            "2019-02-11 09:52:53,157 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:52:53,158 EPOCH 20 done: loss 0.0205 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 09:52:53,909 DEV  : loss 0.01825978 - f-score 0.6737 - acc 0.6737\n",
            "2019-02-11 09:52:58,955 TEST : loss 0.02182743 - f-score 0.6330 - acc 0.6330\n",
            "2019-02-11 09:52:58,958 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:52:59,826 epoch 21 - iter 0/27 - loss 0.02006727\n",
            "2019-02-11 09:53:00,584 epoch 21 - iter 2/27 - loss 0.01990023\n",
            "2019-02-11 09:53:01,776 epoch 21 - iter 4/27 - loss 0.01903804\n",
            "2019-02-11 09:53:02,648 epoch 21 - iter 6/27 - loss 0.01928717\n",
            "2019-02-11 09:53:03,403 epoch 21 - iter 8/27 - loss 0.01916771\n",
            "2019-02-11 09:53:04,318 epoch 21 - iter 10/27 - loss 0.01944130\n",
            "2019-02-11 09:53:05,336 epoch 21 - iter 12/27 - loss 0.01930166\n",
            "2019-02-11 09:53:06,272 epoch 21 - iter 14/27 - loss 0.01889063\n",
            "2019-02-11 09:53:07,016 epoch 21 - iter 16/27 - loss 0.01848814\n",
            "2019-02-11 09:53:07,737 epoch 21 - iter 18/27 - loss 0.01808847\n",
            "2019-02-11 09:53:08,530 epoch 21 - iter 20/27 - loss 0.01804042\n",
            "2019-02-11 09:53:09,262 epoch 21 - iter 22/27 - loss 0.01810014\n",
            "2019-02-11 09:53:10,006 epoch 21 - iter 24/27 - loss 0.01832336\n",
            "2019-02-11 09:53:10,712 epoch 21 - iter 26/27 - loss 0.01891088\n",
            "2019-02-11 09:53:10,726 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:53:10,727 EPOCH 21 done: loss 0.0189 - lr 0.0500 - bad epochs 1\n",
            "2019-02-11 09:53:11,568 DEV  : loss 0.01883742 - f-score 0.6842 - acc 0.6842\n",
            "2019-02-11 09:53:17,186 TEST : loss 0.02288541 - f-score 0.6245 - acc 0.6245\n",
            "2019-02-11 09:53:35,584 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:53:36,732 epoch 22 - iter 0/27 - loss 0.02671734\n",
            "2019-02-11 09:53:37,702 epoch 22 - iter 2/27 - loss 0.02200223\n",
            "2019-02-11 09:53:38,558 epoch 22 - iter 4/27 - loss 0.01915222\n",
            "2019-02-11 09:53:39,375 epoch 22 - iter 6/27 - loss 0.01896451\n",
            "2019-02-11 09:53:40,501 epoch 22 - iter 8/27 - loss 0.01934859\n",
            "2019-02-11 09:53:41,292 epoch 22 - iter 10/27 - loss 0.01929231\n",
            "2019-02-11 09:53:42,793 epoch 22 - iter 12/27 - loss 0.01969214\n",
            "2019-02-11 09:53:43,479 epoch 22 - iter 14/27 - loss 0.02013827\n",
            "2019-02-11 09:53:44,249 epoch 22 - iter 16/27 - loss 0.01963023\n",
            "2019-02-11 09:53:45,109 epoch 22 - iter 18/27 - loss 0.01920518\n",
            "2019-02-11 09:53:45,725 epoch 22 - iter 20/27 - loss 0.01927777\n",
            "2019-02-11 09:53:46,467 epoch 22 - iter 22/27 - loss 0.01925013\n",
            "2019-02-11 09:53:47,273 epoch 22 - iter 24/27 - loss 0.01936028\n",
            "2019-02-11 09:53:47,876 epoch 22 - iter 26/27 - loss 0.01967728\n",
            "2019-02-11 09:53:47,892 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:53:47,893 EPOCH 22 done: loss 0.0197 - lr 0.0500 - bad epochs 0\n",
            "2019-02-11 09:53:48,695 DEV  : loss 0.01839861 - f-score 0.7158 - acc 0.7158\n",
            "2019-02-11 09:53:53,633 TEST : loss 0.02263857 - f-score 0.6524 - acc 0.6524\n",
            "2019-02-11 09:53:53,639 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:53:53,981 epoch 23 - iter 0/27 - loss 0.01792913\n",
            "2019-02-11 09:53:54,748 epoch 23 - iter 2/27 - loss 0.01400354\n",
            "2019-02-11 09:53:55,554 epoch 23 - iter 4/27 - loss 0.01841109\n",
            "2019-02-11 09:53:56,265 epoch 23 - iter 6/27 - loss 0.01820549\n",
            "2019-02-11 09:53:57,608 epoch 23 - iter 8/27 - loss 0.01850528\n",
            "2019-02-11 09:53:58,471 epoch 23 - iter 10/27 - loss 0.01866888\n",
            "2019-02-11 09:53:59,355 epoch 23 - iter 12/27 - loss 0.01950123\n",
            "2019-02-11 09:54:00,259 epoch 23 - iter 14/27 - loss 0.01924089\n",
            "2019-02-11 09:54:01,362 epoch 23 - iter 16/27 - loss 0.01908633\n",
            "2019-02-11 09:54:02,095 epoch 23 - iter 18/27 - loss 0.01891081\n",
            "2019-02-11 09:54:02,962 epoch 23 - iter 20/27 - loss 0.01889661\n",
            "2019-02-11 09:54:03,940 epoch 23 - iter 22/27 - loss 0.01898652\n",
            "2019-02-11 09:54:05,110 epoch 23 - iter 24/27 - loss 0.01899746\n",
            "2019-02-11 09:54:05,575 epoch 23 - iter 26/27 - loss 0.01955592\n",
            "2019-02-11 09:54:05,588 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:54:05,589 EPOCH 23 done: loss 0.0196 - lr 0.0500 - bad epochs 1\n",
            "2019-02-11 09:54:06,348 DEV  : loss 0.01790283 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 09:54:11,287 TEST : loss 0.02170985 - f-score 0.6330 - acc 0.6330\n",
            "2019-02-11 09:54:11,290 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:54:11,679 epoch 24 - iter 0/27 - loss 0.01630982\n",
            "2019-02-11 09:54:12,704 epoch 24 - iter 2/27 - loss 0.02023973\n",
            "2019-02-11 09:54:13,560 epoch 24 - iter 4/27 - loss 0.01905415\n",
            "2019-02-11 09:54:14,436 epoch 24 - iter 6/27 - loss 0.01807357\n",
            "2019-02-11 09:54:15,356 epoch 24 - iter 8/27 - loss 0.01772653\n",
            "2019-02-11 09:54:16,607 epoch 24 - iter 10/27 - loss 0.01781425\n",
            "2019-02-11 09:54:17,283 epoch 24 - iter 12/27 - loss 0.01798098\n",
            "2019-02-11 09:54:18,134 epoch 24 - iter 14/27 - loss 0.01775569\n",
            "2019-02-11 09:54:19,405 epoch 24 - iter 16/27 - loss 0.01825222\n",
            "2019-02-11 09:54:20,319 epoch 24 - iter 18/27 - loss 0.01818262\n",
            "2019-02-11 09:54:21,096 epoch 24 - iter 20/27 - loss 0.01822229\n",
            "2019-02-11 09:54:22,092 epoch 24 - iter 22/27 - loss 0.01805142\n",
            "2019-02-11 09:54:22,847 epoch 24 - iter 24/27 - loss 0.01832772\n",
            "2019-02-11 09:54:23,348 epoch 24 - iter 26/27 - loss 0.01950872\n",
            "2019-02-11 09:54:23,362 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:54:23,363 EPOCH 24 done: loss 0.0195 - lr 0.0500 - bad epochs 2\n",
            "2019-02-11 09:54:24,142 DEV  : loss 0.01869557 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 09:54:29,037 TEST : loss 0.02138491 - f-score 0.6502 - acc 0.6502\n",
            "2019-02-11 09:54:29,040 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:54:29,400 epoch 25 - iter 0/27 - loss 0.02135670\n",
            "2019-02-11 09:54:30,349 epoch 25 - iter 2/27 - loss 0.02130526\n",
            "2019-02-11 09:54:31,585 epoch 25 - iter 4/27 - loss 0.02083593\n",
            "2019-02-11 09:54:32,481 epoch 25 - iter 6/27 - loss 0.02216258\n",
            "2019-02-11 09:54:33,315 epoch 25 - iter 8/27 - loss 0.02100216\n",
            "2019-02-11 09:54:34,152 epoch 25 - iter 10/27 - loss 0.02111144\n",
            "2019-02-11 09:54:34,843 epoch 25 - iter 12/27 - loss 0.02112109\n",
            "2019-02-11 09:54:35,516 epoch 25 - iter 14/27 - loss 0.02075178\n",
            "2019-02-11 09:54:36,507 epoch 25 - iter 16/27 - loss 0.02081892\n",
            "2019-02-11 09:54:37,524 epoch 25 - iter 18/27 - loss 0.02048770\n",
            "2019-02-11 09:54:38,436 epoch 25 - iter 20/27 - loss 0.02021156\n",
            "2019-02-11 09:54:39,239 epoch 25 - iter 22/27 - loss 0.01995304\n",
            "2019-02-11 09:54:39,954 epoch 25 - iter 24/27 - loss 0.01966312\n",
            "2019-02-11 09:54:40,899 epoch 25 - iter 26/27 - loss 0.01970837\n",
            "2019-02-11 09:54:40,912 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:54:40,913 EPOCH 25 done: loss 0.0197 - lr 0.0500 - bad epochs 3\n",
            "2019-02-11 09:54:41,670 DEV  : loss 0.01888865 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 09:54:46,551 TEST : loss 0.02584060 - f-score 0.6309 - acc 0.6309\n",
            "Epoch    24: reducing learning rate of group 0 to 2.5000e-02.\n",
            "2019-02-11 09:54:46,554 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:54:46,988 epoch 26 - iter 0/27 - loss 0.01091679\n",
            "2019-02-11 09:54:47,732 epoch 26 - iter 2/27 - loss 0.01545452\n",
            "2019-02-11 09:54:48,623 epoch 26 - iter 4/27 - loss 0.01796254\n",
            "2019-02-11 09:54:49,686 epoch 26 - iter 6/27 - loss 0.01844692\n",
            "2019-02-11 09:54:50,472 epoch 26 - iter 8/27 - loss 0.01740154\n",
            "2019-02-11 09:54:51,835 epoch 26 - iter 10/27 - loss 0.01808423\n",
            "2019-02-11 09:54:52,816 epoch 26 - iter 12/27 - loss 0.01807213\n",
            "2019-02-11 09:54:53,643 epoch 26 - iter 14/27 - loss 0.01804106\n",
            "2019-02-11 09:54:54,380 epoch 26 - iter 16/27 - loss 0.01768686\n",
            "2019-02-11 09:54:55,527 epoch 26 - iter 18/27 - loss 0.01758705\n",
            "2019-02-11 09:54:56,274 epoch 26 - iter 20/27 - loss 0.01764223\n",
            "2019-02-11 09:54:57,056 epoch 26 - iter 22/27 - loss 0.01789674\n",
            "2019-02-11 09:54:57,844 epoch 26 - iter 24/27 - loss 0.01797324\n",
            "2019-02-11 09:54:58,325 epoch 26 - iter 26/27 - loss 0.01828050\n",
            "2019-02-11 09:54:58,339 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:54:58,340 EPOCH 26 done: loss 0.0183 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 09:54:59,107 DEV  : loss 0.01876432 - f-score 0.6737 - acc 0.6737\n",
            "2019-02-11 09:55:04,163 TEST : loss 0.02135186 - f-score 0.6674 - acc 0.6674\n",
            "2019-02-11 09:55:22,263 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:55:23,321 epoch 27 - iter 0/27 - loss 0.01841355\n",
            "2019-02-11 09:55:24,269 epoch 27 - iter 2/27 - loss 0.01543337\n",
            "2019-02-11 09:55:25,022 epoch 27 - iter 4/27 - loss 0.01655308\n",
            "2019-02-11 09:55:25,817 epoch 27 - iter 6/27 - loss 0.01658607\n",
            "2019-02-11 09:55:26,629 epoch 27 - iter 8/27 - loss 0.01713757\n",
            "2019-02-11 09:55:27,512 epoch 27 - iter 10/27 - loss 0.01710368\n",
            "2019-02-11 09:55:28,819 epoch 27 - iter 12/27 - loss 0.01733429\n",
            "2019-02-11 09:55:29,631 epoch 27 - iter 14/27 - loss 0.01705048\n",
            "2019-02-11 09:55:30,520 epoch 27 - iter 16/27 - loss 0.01696521\n",
            "2019-02-11 09:55:31,544 epoch 27 - iter 18/27 - loss 0.01674132\n",
            "2019-02-11 09:55:32,733 epoch 27 - iter 20/27 - loss 0.01683753\n",
            "2019-02-11 09:55:33,550 epoch 27 - iter 22/27 - loss 0.01667191\n",
            "2019-02-11 09:55:34,396 epoch 27 - iter 24/27 - loss 0.01683254\n",
            "2019-02-11 09:55:34,974 epoch 27 - iter 26/27 - loss 0.01738058\n",
            "2019-02-11 09:55:34,987 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:55:34,988 EPOCH 27 done: loss 0.0174 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 09:55:35,760 DEV  : loss 0.01888749 - f-score 0.6842 - acc 0.6842\n",
            "2019-02-11 09:55:40,608 TEST : loss 0.02132142 - f-score 0.6652 - acc 0.6652\n",
            "2019-02-11 09:55:58,501 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:55:59,307 epoch 28 - iter 0/27 - loss 0.01602476\n",
            "2019-02-11 09:56:00,212 epoch 28 - iter 2/27 - loss 0.01734416\n",
            "2019-02-11 09:56:01,091 epoch 28 - iter 4/27 - loss 0.01581956\n",
            "2019-02-11 09:56:02,065 epoch 28 - iter 6/27 - loss 0.01684143\n",
            "2019-02-11 09:56:02,860 epoch 28 - iter 8/27 - loss 0.01729766\n",
            "2019-02-11 09:56:03,965 epoch 28 - iter 10/27 - loss 0.01696310\n",
            "2019-02-11 09:56:04,849 epoch 28 - iter 12/27 - loss 0.01681971\n",
            "2019-02-11 09:56:05,752 epoch 28 - iter 14/27 - loss 0.01704826\n",
            "2019-02-11 09:56:06,522 epoch 28 - iter 16/27 - loss 0.01718188\n",
            "2019-02-11 09:56:07,750 epoch 28 - iter 18/27 - loss 0.01722896\n",
            "2019-02-11 09:56:09,068 epoch 28 - iter 20/27 - loss 0.01729614\n",
            "2019-02-11 09:56:09,868 epoch 28 - iter 22/27 - loss 0.01710752\n",
            "2019-02-11 09:56:10,736 epoch 28 - iter 24/27 - loss 0.01692466\n",
            "2019-02-11 09:56:11,339 epoch 28 - iter 26/27 - loss 0.01743397\n",
            "2019-02-11 09:56:11,352 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:56:11,353 EPOCH 28 done: loss 0.0174 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 09:56:12,148 DEV  : loss 0.01747039 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 09:56:17,756 TEST : loss 0.02151505 - f-score 0.6760 - acc 0.6760\n",
            "2019-02-11 09:56:17,762 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:56:18,229 epoch 29 - iter 0/27 - loss 0.01336246\n",
            "2019-02-11 09:56:19,070 epoch 29 - iter 2/27 - loss 0.01405837\n",
            "2019-02-11 09:56:20,047 epoch 29 - iter 4/27 - loss 0.01647196\n",
            "2019-02-11 09:56:21,424 epoch 29 - iter 6/27 - loss 0.01631370\n",
            "2019-02-11 09:56:22,549 epoch 29 - iter 8/27 - loss 0.01630218\n",
            "2019-02-11 09:56:23,368 epoch 29 - iter 10/27 - loss 0.01655362\n",
            "2019-02-11 09:56:24,598 epoch 29 - iter 12/27 - loss 0.01676919\n",
            "2019-02-11 09:56:26,119 epoch 29 - iter 14/27 - loss 0.01737136\n",
            "2019-02-11 09:56:26,816 epoch 29 - iter 16/27 - loss 0.01739693\n",
            "2019-02-11 09:56:27,813 epoch 29 - iter 18/27 - loss 0.01739326\n",
            "2019-02-11 09:56:28,653 epoch 29 - iter 20/27 - loss 0.01716589\n",
            "2019-02-11 09:56:29,629 epoch 29 - iter 22/27 - loss 0.01709897\n",
            "2019-02-11 09:56:30,398 epoch 29 - iter 24/27 - loss 0.01689856\n",
            "2019-02-11 09:56:30,961 epoch 29 - iter 26/27 - loss 0.01727499\n",
            "2019-02-11 09:56:30,973 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:56:30,975 EPOCH 29 done: loss 0.0173 - lr 0.0250 - bad epochs 1\n",
            "2019-02-11 09:56:31,745 DEV  : loss 0.01815686 - f-score 0.6632 - acc 0.6632\n",
            "2019-02-11 09:56:36,753 TEST : loss 0.02154045 - f-score 0.6545 - acc 0.6545\n",
            "2019-02-11 09:56:54,834 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:56:55,764 epoch 30 - iter 0/27 - loss 0.01807879\n",
            "2019-02-11 09:56:56,647 epoch 30 - iter 2/27 - loss 0.01889143\n",
            "2019-02-11 09:56:57,485 epoch 30 - iter 4/27 - loss 0.01806172\n",
            "2019-02-11 09:56:58,247 epoch 30 - iter 6/27 - loss 0.01745363\n",
            "2019-02-11 09:56:59,663 epoch 30 - iter 8/27 - loss 0.01735484\n",
            "2019-02-11 09:57:00,471 epoch 30 - iter 10/27 - loss 0.01683800\n",
            "2019-02-11 09:57:01,227 epoch 30 - iter 12/27 - loss 0.01668457\n",
            "2019-02-11 09:57:02,564 epoch 30 - iter 14/27 - loss 0.01650117\n",
            "2019-02-11 09:57:03,331 epoch 30 - iter 16/27 - loss 0.01620022\n",
            "2019-02-11 09:57:04,130 epoch 30 - iter 18/27 - loss 0.01638615\n",
            "2019-02-11 09:57:04,984 epoch 30 - iter 20/27 - loss 0.01628579\n",
            "2019-02-11 09:57:05,714 epoch 30 - iter 22/27 - loss 0.01613200\n",
            "2019-02-11 09:57:06,498 epoch 30 - iter 24/27 - loss 0.01625039\n",
            "2019-02-11 09:57:06,989 epoch 30 - iter 26/27 - loss 0.01643436\n",
            "2019-02-11 09:57:07,001 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:57:07,003 EPOCH 30 done: loss 0.0164 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 09:57:07,754 DEV  : loss 0.01641442 - f-score 0.7263 - acc 0.7263\n",
            "2019-02-11 09:57:12,536 TEST : loss 0.02233866 - f-score 0.6781 - acc 0.6781\n",
            "2019-02-11 09:57:30,465 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:57:31,421 epoch 31 - iter 0/27 - loss 0.01601576\n",
            "2019-02-11 09:57:32,194 epoch 31 - iter 2/27 - loss 0.01621106\n",
            "2019-02-11 09:57:32,948 epoch 31 - iter 4/27 - loss 0.01650844\n",
            "2019-02-11 09:57:33,878 epoch 31 - iter 6/27 - loss 0.01665923\n",
            "2019-02-11 09:57:34,898 epoch 31 - iter 8/27 - loss 0.01671016\n",
            "2019-02-11 09:57:35,695 epoch 31 - iter 10/27 - loss 0.01626624\n",
            "2019-02-11 09:57:36,517 epoch 31 - iter 12/27 - loss 0.01663117\n",
            "2019-02-11 09:57:37,366 epoch 31 - iter 14/27 - loss 0.01659396\n",
            "2019-02-11 09:57:38,278 epoch 31 - iter 16/27 - loss 0.01629929\n",
            "2019-02-11 09:57:39,005 epoch 31 - iter 18/27 - loss 0.01615639\n",
            "2019-02-11 09:57:40,250 epoch 31 - iter 20/27 - loss 0.01627973\n",
            "2019-02-11 09:57:41,042 epoch 31 - iter 22/27 - loss 0.01603336\n",
            "2019-02-11 09:57:41,768 epoch 31 - iter 24/27 - loss 0.01596597\n",
            "2019-02-11 09:57:42,773 epoch 31 - iter 26/27 - loss 0.01674929\n",
            "2019-02-11 09:57:42,785 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:57:42,787 EPOCH 31 done: loss 0.0167 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 09:57:43,543 DEV  : loss 0.01661437 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 09:57:48,311 TEST : loss 0.02140111 - f-score 0.6674 - acc 0.6674\n",
            "2019-02-11 09:57:48,316 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:57:48,624 epoch 32 - iter 0/27 - loss 0.01586039\n",
            "2019-02-11 09:57:49,459 epoch 32 - iter 2/27 - loss 0.01816718\n",
            "2019-02-11 09:57:50,230 epoch 32 - iter 4/27 - loss 0.01853557\n",
            "2019-02-11 09:57:51,009 epoch 32 - iter 6/27 - loss 0.01748327\n",
            "2019-02-11 09:57:51,700 epoch 32 - iter 8/27 - loss 0.01717025\n",
            "2019-02-11 09:57:52,621 epoch 32 - iter 10/27 - loss 0.01684995\n",
            "2019-02-11 09:57:53,595 epoch 32 - iter 12/27 - loss 0.01735687\n",
            "2019-02-11 09:57:54,302 epoch 32 - iter 14/27 - loss 0.01707819\n",
            "2019-02-11 09:57:55,689 epoch 32 - iter 16/27 - loss 0.01798812\n",
            "2019-02-11 09:57:56,557 epoch 32 - iter 18/27 - loss 0.01788829\n",
            "2019-02-11 09:57:57,738 epoch 32 - iter 20/27 - loss 0.01774736\n",
            "2019-02-11 09:57:58,596 epoch 32 - iter 22/27 - loss 0.01745478\n",
            "2019-02-11 09:57:59,298 epoch 32 - iter 24/27 - loss 0.01742024\n",
            "2019-02-11 09:57:59,844 epoch 32 - iter 26/27 - loss 0.01744820\n",
            "2019-02-11 09:57:59,855 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:57:59,857 EPOCH 32 done: loss 0.0174 - lr 0.0250 - bad epochs 1\n",
            "2019-02-11 09:58:00,616 DEV  : loss 0.01859396 - f-score 0.6842 - acc 0.6842\n",
            "2019-02-11 09:58:05,414 TEST : loss 0.02192953 - f-score 0.6459 - acc 0.6459\n",
            "2019-02-11 09:58:05,417 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:58:06,060 epoch 33 - iter 0/27 - loss 0.01676391\n",
            "2019-02-11 09:58:06,802 epoch 33 - iter 2/27 - loss 0.01549409\n",
            "2019-02-11 09:58:07,964 epoch 33 - iter 4/27 - loss 0.01473410\n",
            "2019-02-11 09:58:08,782 epoch 33 - iter 6/27 - loss 0.01477212\n",
            "2019-02-11 09:58:09,527 epoch 33 - iter 8/27 - loss 0.01491926\n",
            "2019-02-11 09:58:10,333 epoch 33 - iter 10/27 - loss 0.01585594\n",
            "2019-02-11 09:58:11,157 epoch 33 - iter 12/27 - loss 0.01614320\n",
            "2019-02-11 09:58:11,965 epoch 33 - iter 14/27 - loss 0.01626898\n",
            "2019-02-11 09:58:13,373 epoch 33 - iter 16/27 - loss 0.01650591\n",
            "2019-02-11 09:58:14,050 epoch 33 - iter 18/27 - loss 0.01662103\n",
            "2019-02-11 09:58:14,997 epoch 33 - iter 20/27 - loss 0.01633276\n",
            "2019-02-11 09:58:15,828 epoch 33 - iter 22/27 - loss 0.01625490\n",
            "2019-02-11 09:58:16,720 epoch 33 - iter 24/27 - loss 0.01644265\n",
            "2019-02-11 09:58:17,148 epoch 33 - iter 26/27 - loss 0.01652545\n",
            "2019-02-11 09:58:17,160 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:58:17,161 EPOCH 33 done: loss 0.0165 - lr 0.0250 - bad epochs 2\n",
            "2019-02-11 09:58:17,907 DEV  : loss 0.01634101 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 09:58:22,870 TEST : loss 0.02180263 - f-score 0.6674 - acc 0.6674\n",
            "2019-02-11 09:58:22,873 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:58:23,375 epoch 34 - iter 0/27 - loss 0.01714170\n",
            "2019-02-11 09:58:24,369 epoch 34 - iter 2/27 - loss 0.01554805\n",
            "2019-02-11 09:58:25,445 epoch 34 - iter 4/27 - loss 0.01674321\n",
            "2019-02-11 09:58:26,179 epoch 34 - iter 6/27 - loss 0.01677258\n",
            "2019-02-11 09:58:27,478 epoch 34 - iter 8/27 - loss 0.01618132\n",
            "2019-02-11 09:58:28,231 epoch 34 - iter 10/27 - loss 0.01577257\n",
            "2019-02-11 09:58:28,891 epoch 34 - iter 12/27 - loss 0.01543826\n",
            "2019-02-11 09:58:29,843 epoch 34 - iter 14/27 - loss 0.01577278\n",
            "2019-02-11 09:58:31,094 epoch 34 - iter 16/27 - loss 0.01581080\n",
            "2019-02-11 09:58:31,794 epoch 34 - iter 18/27 - loss 0.01578861\n",
            "2019-02-11 09:58:32,735 epoch 34 - iter 20/27 - loss 0.01559147\n",
            "2019-02-11 09:58:33,537 epoch 34 - iter 22/27 - loss 0.01545429\n",
            "2019-02-11 09:58:34,283 epoch 34 - iter 24/27 - loss 0.01519786\n",
            "2019-02-11 09:58:34,855 epoch 34 - iter 26/27 - loss 0.01534064\n",
            "2019-02-11 09:58:34,868 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:58:34,869 EPOCH 34 done: loss 0.0153 - lr 0.0250 - bad epochs 3\n",
            "2019-02-11 09:58:35,625 DEV  : loss 0.01631505 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 09:58:40,398 TEST : loss 0.02252679 - f-score 0.6674 - acc 0.6674\n",
            "2019-02-11 09:58:58,599 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:58:59,043 epoch 35 - iter 0/27 - loss 0.01324552\n",
            "2019-02-11 09:59:00,082 epoch 35 - iter 2/27 - loss 0.01351468\n",
            "2019-02-11 09:59:01,489 epoch 35 - iter 4/27 - loss 0.01647566\n",
            "2019-02-11 09:59:02,286 epoch 35 - iter 6/27 - loss 0.01649971\n",
            "2019-02-11 09:59:03,082 epoch 35 - iter 8/27 - loss 0.01599974\n",
            "2019-02-11 09:59:04,335 epoch 35 - iter 10/27 - loss 0.01573351\n",
            "2019-02-11 09:59:05,049 epoch 35 - iter 12/27 - loss 0.01536655\n",
            "2019-02-11 09:59:05,900 epoch 35 - iter 14/27 - loss 0.01586950\n",
            "2019-02-11 09:59:06,828 epoch 35 - iter 16/27 - loss 0.01574069\n",
            "2019-02-11 09:59:07,632 epoch 35 - iter 18/27 - loss 0.01544345\n",
            "2019-02-11 09:59:08,432 epoch 35 - iter 20/27 - loss 0.01521441\n",
            "2019-02-11 09:59:09,419 epoch 35 - iter 22/27 - loss 0.01538093\n",
            "2019-02-11 09:59:10,249 epoch 35 - iter 24/27 - loss 0.01530590\n",
            "2019-02-11 09:59:10,754 epoch 35 - iter 26/27 - loss 0.01565785\n",
            "2019-02-11 09:59:10,765 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:59:10,766 EPOCH 35 done: loss 0.0157 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 09:59:11,532 DEV  : loss 0.01648341 - f-score 0.7263 - acc 0.7263\n",
            "2019-02-11 09:59:16,491 TEST : loss 0.02309309 - f-score 0.6567 - acc 0.6567\n",
            "2019-02-11 09:59:16,494 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:59:16,947 epoch 36 - iter 0/27 - loss 0.00907724\n",
            "2019-02-11 09:59:17,787 epoch 36 - iter 2/27 - loss 0.01250435\n",
            "2019-02-11 09:59:18,592 epoch 36 - iter 4/27 - loss 0.01394670\n",
            "2019-02-11 09:59:19,542 epoch 36 - iter 6/27 - loss 0.01551857\n",
            "2019-02-11 09:59:20,765 epoch 36 - iter 8/27 - loss 0.01534740\n",
            "2019-02-11 09:59:21,461 epoch 36 - iter 10/27 - loss 0.01511365\n",
            "2019-02-11 09:59:22,684 epoch 36 - iter 12/27 - loss 0.01546686\n",
            "2019-02-11 09:59:23,457 epoch 36 - iter 14/27 - loss 0.01514634\n",
            "2019-02-11 09:59:24,170 epoch 36 - iter 16/27 - loss 0.01525523\n",
            "2019-02-11 09:59:24,978 epoch 36 - iter 18/27 - loss 0.01516499\n",
            "2019-02-11 09:59:25,804 epoch 36 - iter 20/27 - loss 0.01516260\n",
            "2019-02-11 09:59:27,002 epoch 36 - iter 22/27 - loss 0.01516030\n",
            "2019-02-11 09:59:27,862 epoch 36 - iter 24/27 - loss 0.01548260\n",
            "2019-02-11 09:59:28,393 epoch 36 - iter 26/27 - loss 0.01656402\n",
            "2019-02-11 09:59:28,405 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:59:28,406 EPOCH 36 done: loss 0.0166 - lr 0.0250 - bad epochs 1\n",
            "2019-02-11 09:59:29,181 DEV  : loss 0.01612257 - f-score 0.7158 - acc 0.7158\n",
            "2019-02-11 09:59:34,074 TEST : loss 0.02299827 - f-score 0.6609 - acc 0.6609\n",
            "2019-02-11 09:59:34,077 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:59:34,560 epoch 37 - iter 0/27 - loss 0.01322237\n",
            "2019-02-11 09:59:35,680 epoch 37 - iter 2/27 - loss 0.01440376\n",
            "2019-02-11 09:59:36,451 epoch 37 - iter 4/27 - loss 0.01582957\n",
            "2019-02-11 09:59:37,381 epoch 37 - iter 6/27 - loss 0.01575353\n",
            "2019-02-11 09:59:38,167 epoch 37 - iter 8/27 - loss 0.01560850\n",
            "2019-02-11 09:59:38,924 epoch 37 - iter 10/27 - loss 0.01521225\n",
            "2019-02-11 09:59:39,600 epoch 37 - iter 12/27 - loss 0.01534991\n",
            "2019-02-11 09:59:40,443 epoch 37 - iter 14/27 - loss 0.01504337\n",
            "2019-02-11 09:59:41,270 epoch 37 - iter 16/27 - loss 0.01544943\n",
            "2019-02-11 09:59:42,079 epoch 37 - iter 18/27 - loss 0.01590055\n",
            "2019-02-11 09:59:42,942 epoch 37 - iter 20/27 - loss 0.01613495\n",
            "2019-02-11 09:59:44,302 epoch 37 - iter 22/27 - loss 0.01592769\n",
            "2019-02-11 09:59:45,139 epoch 37 - iter 24/27 - loss 0.01577695\n",
            "2019-02-11 09:59:46,321 epoch 37 - iter 26/27 - loss 0.01599770\n",
            "2019-02-11 09:59:46,334 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:59:46,335 EPOCH 37 done: loss 0.0160 - lr 0.0250 - bad epochs 2\n",
            "2019-02-11 09:59:47,195 DEV  : loss 0.01637846 - f-score 0.7263 - acc 0.7263\n",
            "2019-02-11 09:59:52,838 TEST : loss 0.02262130 - f-score 0.6695 - acc 0.6695\n",
            "2019-02-11 09:59:52,842 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 09:59:53,173 epoch 38 - iter 0/27 - loss 0.02014475\n",
            "2019-02-11 09:59:54,082 epoch 38 - iter 2/27 - loss 0.01863201\n",
            "2019-02-11 09:59:55,637 epoch 38 - iter 4/27 - loss 0.01692801\n",
            "2019-02-11 09:59:56,481 epoch 38 - iter 6/27 - loss 0.01647074\n",
            "2019-02-11 09:59:57,491 epoch 38 - iter 8/27 - loss 0.01656843\n",
            "2019-02-11 09:59:58,757 epoch 38 - iter 10/27 - loss 0.01662282\n",
            "2019-02-11 09:59:59,567 epoch 38 - iter 12/27 - loss 0.01590426\n",
            "2019-02-11 10:00:00,401 epoch 38 - iter 14/27 - loss 0.01573515\n",
            "2019-02-11 10:00:01,241 epoch 38 - iter 16/27 - loss 0.01553364\n",
            "2019-02-11 10:00:02,423 epoch 38 - iter 18/27 - loss 0.01557801\n",
            "2019-02-11 10:00:03,238 epoch 38 - iter 20/27 - loss 0.01536797\n",
            "2019-02-11 10:00:04,060 epoch 38 - iter 22/27 - loss 0.01534673\n",
            "2019-02-11 10:00:04,968 epoch 38 - iter 24/27 - loss 0.01512564\n",
            "2019-02-11 10:00:05,468 epoch 38 - iter 26/27 - loss 0.01524017\n",
            "2019-02-11 10:00:05,481 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:00:05,483 EPOCH 38 done: loss 0.0152 - lr 0.0250 - bad epochs 3\n",
            "2019-02-11 10:00:06,246 DEV  : loss 0.01932263 - f-score 0.6842 - acc 0.6842\n",
            "2019-02-11 10:00:10,997 TEST : loss 0.02289432 - f-score 0.6545 - acc 0.6545\n",
            "2019-02-11 10:00:28,712 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:00:29,160 epoch 39 - iter 0/27 - loss 0.02294486\n",
            "2019-02-11 10:00:30,050 epoch 39 - iter 2/27 - loss 0.01647847\n",
            "2019-02-11 10:00:31,034 epoch 39 - iter 4/27 - loss 0.01655568\n",
            "2019-02-11 10:00:31,799 epoch 39 - iter 6/27 - loss 0.01633934\n",
            "2019-02-11 10:00:32,580 epoch 39 - iter 8/27 - loss 0.01600783\n",
            "2019-02-11 10:00:33,394 epoch 39 - iter 10/27 - loss 0.01573068\n",
            "2019-02-11 10:00:34,417 epoch 39 - iter 12/27 - loss 0.01551501\n",
            "2019-02-11 10:00:35,270 epoch 39 - iter 14/27 - loss 0.01580348\n",
            "2019-02-11 10:00:36,393 epoch 39 - iter 16/27 - loss 0.01585608\n",
            "2019-02-11 10:00:37,143 epoch 39 - iter 18/27 - loss 0.01568851\n",
            "2019-02-11 10:00:38,344 epoch 39 - iter 20/27 - loss 0.01588277\n",
            "2019-02-11 10:00:39,140 epoch 39 - iter 22/27 - loss 0.01580927\n",
            "2019-02-11 10:00:39,949 epoch 39 - iter 24/27 - loss 0.01577868\n",
            "2019-02-11 10:00:40,880 epoch 39 - iter 26/27 - loss 0.01590457\n",
            "2019-02-11 10:00:40,892 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:00:40,894 EPOCH 39 done: loss 0.0159 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 10:00:41,633 DEV  : loss 0.01776355 - f-score 0.6737 - acc 0.6737\n",
            "2019-02-11 10:00:46,405 TEST : loss 0.02218152 - f-score 0.6588 - acc 0.6588\n",
            "2019-02-11 10:00:46,410 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:00:46,842 epoch 40 - iter 0/27 - loss 0.01685577\n",
            "2019-02-11 10:00:48,030 epoch 40 - iter 2/27 - loss 0.01438897\n",
            "2019-02-11 10:00:48,880 epoch 40 - iter 4/27 - loss 0.01486443\n",
            "2019-02-11 10:00:49,589 epoch 40 - iter 6/27 - loss 0.01508335\n",
            "2019-02-11 10:00:50,279 epoch 40 - iter 8/27 - loss 0.01467706\n",
            "2019-02-11 10:00:51,534 epoch 40 - iter 10/27 - loss 0.01482068\n",
            "2019-02-11 10:00:52,330 epoch 40 - iter 12/27 - loss 0.01457651\n",
            "2019-02-11 10:00:53,071 epoch 40 - iter 14/27 - loss 0.01451837\n",
            "2019-02-11 10:00:53,862 epoch 40 - iter 16/27 - loss 0.01441738\n",
            "2019-02-11 10:00:54,736 epoch 40 - iter 18/27 - loss 0.01469456\n",
            "2019-02-11 10:00:55,579 epoch 40 - iter 20/27 - loss 0.01468197\n",
            "2019-02-11 10:00:56,697 epoch 40 - iter 22/27 - loss 0.01477103\n",
            "2019-02-11 10:00:57,379 epoch 40 - iter 24/27 - loss 0.01483010\n",
            "2019-02-11 10:00:57,952 epoch 40 - iter 26/27 - loss 0.01483113\n",
            "2019-02-11 10:00:57,963 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:00:57,965 EPOCH 40 done: loss 0.0148 - lr 0.0250 - bad epochs 1\n",
            "2019-02-11 10:00:58,717 DEV  : loss 0.01679452 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 10:01:03,748 TEST : loss 0.02260106 - f-score 0.6652 - acc 0.6652\n",
            "2019-02-11 10:01:21,608 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:01:22,974 epoch 41 - iter 0/27 - loss 0.01141125\n",
            "2019-02-11 10:01:23,746 epoch 41 - iter 2/27 - loss 0.01177069\n",
            "2019-02-11 10:01:24,580 epoch 41 - iter 4/27 - loss 0.01303822\n",
            "2019-02-11 10:01:25,344 epoch 41 - iter 6/27 - loss 0.01287285\n",
            "2019-02-11 10:01:26,714 epoch 41 - iter 8/27 - loss 0.01333090\n",
            "2019-02-11 10:01:27,396 epoch 41 - iter 10/27 - loss 0.01403301\n",
            "2019-02-11 10:01:28,373 epoch 41 - iter 12/27 - loss 0.01426889\n",
            "2019-02-11 10:01:29,297 epoch 41 - iter 14/27 - loss 0.01434951\n",
            "2019-02-11 10:01:30,170 epoch 41 - iter 16/27 - loss 0.01435081\n",
            "2019-02-11 10:01:31,064 epoch 41 - iter 18/27 - loss 0.01430243\n",
            "2019-02-11 10:01:32,178 epoch 41 - iter 20/27 - loss 0.01427223\n",
            "2019-02-11 10:01:32,990 epoch 41 - iter 22/27 - loss 0.01405864\n",
            "2019-02-11 10:01:33,671 epoch 41 - iter 24/27 - loss 0.01428769\n",
            "2019-02-11 10:01:35,150 epoch 41 - iter 26/27 - loss 0.01462980\n",
            "2019-02-11 10:01:35,161 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:01:35,162 EPOCH 41 done: loss 0.0146 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 10:01:35,929 DEV  : loss 0.01651603 - f-score 0.7263 - acc 0.7263\n",
            "2019-02-11 10:01:40,789 TEST : loss 0.02309692 - f-score 0.6738 - acc 0.6738\n",
            "2019-02-11 10:01:58,789 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:01:59,351 epoch 42 - iter 0/27 - loss 0.01366298\n",
            "2019-02-11 10:02:00,383 epoch 42 - iter 2/27 - loss 0.01436572\n",
            "2019-02-11 10:02:01,026 epoch 42 - iter 4/27 - loss 0.01397461\n",
            "2019-02-11 10:02:01,945 epoch 42 - iter 6/27 - loss 0.01425910\n",
            "2019-02-11 10:02:02,713 epoch 42 - iter 8/27 - loss 0.01516209\n",
            "2019-02-11 10:02:03,720 epoch 42 - iter 10/27 - loss 0.01418180\n",
            "2019-02-11 10:02:04,548 epoch 42 - iter 12/27 - loss 0.01396017\n",
            "2019-02-11 10:02:05,349 epoch 42 - iter 14/27 - loss 0.01438352\n",
            "2019-02-11 10:02:06,199 epoch 42 - iter 16/27 - loss 0.01429573\n",
            "2019-02-11 10:02:07,543 epoch 42 - iter 18/27 - loss 0.01421200\n",
            "2019-02-11 10:02:08,635 epoch 42 - iter 20/27 - loss 0.01420778\n",
            "2019-02-11 10:02:09,314 epoch 42 - iter 22/27 - loss 0.01462162\n",
            "2019-02-11 10:02:09,919 epoch 42 - iter 24/27 - loss 0.01455844\n",
            "2019-02-11 10:02:10,714 epoch 42 - iter 26/27 - loss 0.01466856\n",
            "2019-02-11 10:02:10,726 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:02:10,728 EPOCH 42 done: loss 0.0147 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 10:02:11,497 DEV  : loss 0.01646642 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 10:02:16,436 TEST : loss 0.02283703 - f-score 0.6674 - acc 0.6674\n",
            "2019-02-11 10:02:16,439 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:02:16,755 epoch 43 - iter 0/27 - loss 0.00964262\n",
            "2019-02-11 10:02:18,349 epoch 43 - iter 2/27 - loss 0.01275165\n",
            "2019-02-11 10:02:19,463 epoch 43 - iter 4/27 - loss 0.01280438\n",
            "2019-02-11 10:02:20,704 epoch 43 - iter 6/27 - loss 0.01390530\n",
            "2019-02-11 10:02:21,412 epoch 43 - iter 8/27 - loss 0.01415872\n",
            "2019-02-11 10:02:22,397 epoch 43 - iter 10/27 - loss 0.01425684\n",
            "2019-02-11 10:02:23,602 epoch 43 - iter 12/27 - loss 0.01397800\n",
            "2019-02-11 10:02:24,691 epoch 43 - iter 14/27 - loss 0.01339151\n",
            "2019-02-11 10:02:25,499 epoch 43 - iter 16/27 - loss 0.01379789\n",
            "2019-02-11 10:02:26,225 epoch 43 - iter 18/27 - loss 0.01386361\n",
            "2019-02-11 10:02:27,182 epoch 43 - iter 20/27 - loss 0.01377888\n",
            "2019-02-11 10:02:28,062 epoch 43 - iter 22/27 - loss 0.01380058\n",
            "2019-02-11 10:02:28,832 epoch 43 - iter 24/27 - loss 0.01402397\n",
            "2019-02-11 10:02:29,380 epoch 43 - iter 26/27 - loss 0.01432328\n",
            "2019-02-11 10:02:29,393 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:02:29,394 EPOCH 43 done: loss 0.0143 - lr 0.0250 - bad epochs 1\n",
            "2019-02-11 10:02:30,171 DEV  : loss 0.01611337 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 10:02:35,173 TEST : loss 0.02413367 - f-score 0.6631 - acc 0.6631\n",
            "2019-02-11 10:02:53,406 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:02:53,955 epoch 44 - iter 0/27 - loss 0.01359891\n",
            "2019-02-11 10:02:55,585 epoch 44 - iter 2/27 - loss 0.01122708\n",
            "2019-02-11 10:02:56,504 epoch 44 - iter 4/27 - loss 0.01199842\n",
            "2019-02-11 10:02:57,356 epoch 44 - iter 6/27 - loss 0.01231237\n",
            "2019-02-11 10:02:58,042 epoch 44 - iter 8/27 - loss 0.01328595\n",
            "2019-02-11 10:02:58,817 epoch 44 - iter 10/27 - loss 0.01307393\n",
            "2019-02-11 10:02:59,639 epoch 44 - iter 12/27 - loss 0.01406912\n",
            "2019-02-11 10:03:00,892 epoch 44 - iter 14/27 - loss 0.01464689\n",
            "2019-02-11 10:03:01,819 epoch 44 - iter 16/27 - loss 0.01507816\n",
            "2019-02-11 10:03:02,591 epoch 44 - iter 18/27 - loss 0.01476749\n",
            "2019-02-11 10:03:03,376 epoch 44 - iter 20/27 - loss 0.01463331\n",
            "2019-02-11 10:03:04,204 epoch 44 - iter 22/27 - loss 0.01474560\n",
            "2019-02-11 10:03:04,943 epoch 44 - iter 24/27 - loss 0.01497038\n",
            "2019-02-11 10:03:05,460 epoch 44 - iter 26/27 - loss 0.01540988\n",
            "2019-02-11 10:03:05,474 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:03:05,475 EPOCH 44 done: loss 0.0154 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 10:03:06,233 DEV  : loss 0.01662895 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 10:03:11,051 TEST : loss 0.02314767 - f-score 0.6738 - acc 0.6738\n",
            "2019-02-11 10:03:11,055 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:03:11,489 epoch 45 - iter 0/27 - loss 0.00927885\n",
            "2019-02-11 10:03:12,367 epoch 45 - iter 2/27 - loss 0.01205046\n",
            "2019-02-11 10:03:13,189 epoch 45 - iter 4/27 - loss 0.01222138\n",
            "2019-02-11 10:03:14,064 epoch 45 - iter 6/27 - loss 0.01281173\n",
            "2019-02-11 10:03:14,767 epoch 45 - iter 8/27 - loss 0.01387235\n",
            "2019-02-11 10:03:15,452 epoch 45 - iter 10/27 - loss 0.01353107\n",
            "2019-02-11 10:03:16,415 epoch 45 - iter 12/27 - loss 0.01364454\n",
            "2019-02-11 10:03:17,396 epoch 45 - iter 14/27 - loss 0.01308405\n",
            "2019-02-11 10:03:18,187 epoch 45 - iter 16/27 - loss 0.01316008\n",
            "2019-02-11 10:03:19,193 epoch 45 - iter 18/27 - loss 0.01308497\n",
            "2019-02-11 10:03:20,453 epoch 45 - iter 20/27 - loss 0.01398888\n",
            "2019-02-11 10:03:21,742 epoch 45 - iter 22/27 - loss 0.01384391\n",
            "2019-02-11 10:03:22,419 epoch 45 - iter 24/27 - loss 0.01382477\n",
            "2019-02-11 10:03:22,950 epoch 45 - iter 26/27 - loss 0.01419297\n",
            "2019-02-11 10:03:22,964 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:03:22,966 EPOCH 45 done: loss 0.0142 - lr 0.0250 - bad epochs 1\n",
            "2019-02-11 10:03:23,792 DEV  : loss 0.01643978 - f-score 0.7579 - acc 0.7579\n",
            "2019-02-11 10:03:29,522 TEST : loss 0.02391488 - f-score 0.6781 - acc 0.6781\n",
            "2019-02-11 10:03:48,003 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:03:48,536 epoch 46 - iter 0/27 - loss 0.01588681\n",
            "2019-02-11 10:03:49,550 epoch 46 - iter 2/27 - loss 0.01316824\n",
            "2019-02-11 10:03:50,485 epoch 46 - iter 4/27 - loss 0.01388938\n",
            "2019-02-11 10:03:51,631 epoch 46 - iter 6/27 - loss 0.01385262\n",
            "2019-02-11 10:03:52,414 epoch 46 - iter 8/27 - loss 0.01355302\n",
            "2019-02-11 10:03:53,077 epoch 46 - iter 10/27 - loss 0.01448232\n",
            "2019-02-11 10:03:54,321 epoch 46 - iter 12/27 - loss 0.01417906\n",
            "2019-02-11 10:03:55,186 epoch 46 - iter 14/27 - loss 0.01370183\n",
            "2019-02-11 10:03:56,029 epoch 46 - iter 16/27 - loss 0.01415068\n",
            "2019-02-11 10:03:56,964 epoch 46 - iter 18/27 - loss 0.01417269\n",
            "2019-02-11 10:03:57,795 epoch 46 - iter 20/27 - loss 0.01418346\n",
            "2019-02-11 10:03:58,569 epoch 46 - iter 22/27 - loss 0.01438881\n",
            "2019-02-11 10:03:59,320 epoch 46 - iter 24/27 - loss 0.01433691\n",
            "2019-02-11 10:04:00,086 epoch 46 - iter 26/27 - loss 0.01475078\n",
            "2019-02-11 10:04:00,099 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:04:00,100 EPOCH 46 done: loss 0.0148 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 10:04:00,838 DEV  : loss 0.01612357 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 10:04:05,495 TEST : loss 0.02390808 - f-score 0.6631 - acc 0.6631\n",
            "2019-02-11 10:04:05,499 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:04:05,928 epoch 47 - iter 0/27 - loss 0.01248497\n",
            "2019-02-11 10:04:07,111 epoch 47 - iter 2/27 - loss 0.01256046\n",
            "2019-02-11 10:04:07,829 epoch 47 - iter 4/27 - loss 0.01292733\n",
            "2019-02-11 10:04:08,825 epoch 47 - iter 6/27 - loss 0.01317449\n",
            "2019-02-11 10:04:09,597 epoch 47 - iter 8/27 - loss 0.01418220\n",
            "2019-02-11 10:04:10,504 epoch 47 - iter 10/27 - loss 0.01407212\n",
            "2019-02-11 10:04:11,362 epoch 47 - iter 12/27 - loss 0.01407740\n",
            "2019-02-11 10:04:12,222 epoch 47 - iter 14/27 - loss 0.01345676\n",
            "2019-02-11 10:04:13,430 epoch 47 - iter 16/27 - loss 0.01324265\n",
            "2019-02-11 10:04:14,145 epoch 47 - iter 18/27 - loss 0.01386316\n",
            "2019-02-11 10:04:14,914 epoch 47 - iter 20/27 - loss 0.01371654\n",
            "2019-02-11 10:04:15,733 epoch 47 - iter 22/27 - loss 0.01396915\n",
            "2019-02-11 10:04:16,537 epoch 47 - iter 24/27 - loss 0.01379515\n",
            "2019-02-11 10:04:17,083 epoch 47 - iter 26/27 - loss 0.01511104\n",
            "2019-02-11 10:04:17,096 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:04:17,098 EPOCH 47 done: loss 0.0151 - lr 0.0250 - bad epochs 1\n",
            "2019-02-11 10:04:17,818 DEV  : loss 0.01552224 - f-score 0.7158 - acc 0.7158\n",
            "2019-02-11 10:04:22,552 TEST : loss 0.02369394 - f-score 0.6760 - acc 0.6760\n",
            "2019-02-11 10:04:22,555 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:04:23,026 epoch 48 - iter 0/27 - loss 0.01707504\n",
            "2019-02-11 10:04:23,883 epoch 48 - iter 2/27 - loss 0.01711594\n",
            "2019-02-11 10:04:25,130 epoch 48 - iter 4/27 - loss 0.01668960\n",
            "2019-02-11 10:04:26,053 epoch 48 - iter 6/27 - loss 0.01579826\n",
            "2019-02-11 10:04:27,186 epoch 48 - iter 8/27 - loss 0.01457012\n",
            "2019-02-11 10:04:28,260 epoch 48 - iter 10/27 - loss 0.01451305\n",
            "2019-02-11 10:04:29,115 epoch 48 - iter 12/27 - loss 0.01462220\n",
            "2019-02-11 10:04:30,069 epoch 48 - iter 14/27 - loss 0.01400045\n",
            "2019-02-11 10:04:30,826 epoch 48 - iter 16/27 - loss 0.01420139\n",
            "2019-02-11 10:04:31,529 epoch 48 - iter 18/27 - loss 0.01413286\n",
            "2019-02-11 10:04:32,239 epoch 48 - iter 20/27 - loss 0.01404010\n",
            "2019-02-11 10:04:33,008 epoch 48 - iter 22/27 - loss 0.01372033\n",
            "2019-02-11 10:04:33,869 epoch 48 - iter 24/27 - loss 0.01366478\n",
            "2019-02-11 10:04:34,352 epoch 48 - iter 26/27 - loss 0.01400651\n",
            "2019-02-11 10:04:34,364 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:04:34,365 EPOCH 48 done: loss 0.0140 - lr 0.0250 - bad epochs 2\n",
            "2019-02-11 10:04:35,109 DEV  : loss 0.01597350 - f-score 0.7158 - acc 0.7158\n",
            "2019-02-11 10:04:40,080 TEST : loss 0.02379516 - f-score 0.6695 - acc 0.6695\n",
            "2019-02-11 10:04:57,966 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:04:58,509 epoch 49 - iter 0/27 - loss 0.01305535\n",
            "2019-02-11 10:04:59,570 epoch 49 - iter 2/27 - loss 0.01262670\n",
            "2019-02-11 10:05:00,431 epoch 49 - iter 4/27 - loss 0.01327324\n",
            "2019-02-11 10:05:01,278 epoch 49 - iter 6/27 - loss 0.01403280\n",
            "2019-02-11 10:05:02,415 epoch 49 - iter 8/27 - loss 0.01358340\n",
            "2019-02-11 10:05:03,282 epoch 49 - iter 10/27 - loss 0.01441311\n",
            "2019-02-11 10:05:03,993 epoch 49 - iter 12/27 - loss 0.01443633\n",
            "2019-02-11 10:05:04,745 epoch 49 - iter 14/27 - loss 0.01489689\n",
            "2019-02-11 10:05:05,405 epoch 49 - iter 16/27 - loss 0.01500872\n",
            "2019-02-11 10:05:06,152 epoch 49 - iter 18/27 - loss 0.01489272\n",
            "2019-02-11 10:05:06,833 epoch 49 - iter 20/27 - loss 0.01436484\n",
            "2019-02-11 10:05:07,655 epoch 49 - iter 22/27 - loss 0.01413252\n",
            "2019-02-11 10:05:09,086 epoch 49 - iter 24/27 - loss 0.01420271\n",
            "2019-02-11 10:05:09,616 epoch 49 - iter 26/27 - loss 0.01465395\n",
            "2019-02-11 10:05:09,630 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:05:09,632 EPOCH 49 done: loss 0.0147 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 10:05:10,363 DEV  : loss 0.01606321 - f-score 0.7368 - acc 0.7368\n",
            "2019-02-11 10:05:15,297 TEST : loss 0.02412006 - f-score 0.6738 - acc 0.6738\n",
            "2019-02-11 10:05:15,300 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:05:15,672 epoch 50 - iter 0/27 - loss 0.00849226\n",
            "2019-02-11 10:05:16,718 epoch 50 - iter 2/27 - loss 0.01174338\n",
            "2019-02-11 10:05:18,084 epoch 50 - iter 4/27 - loss 0.01258022\n",
            "2019-02-11 10:05:19,152 epoch 50 - iter 6/27 - loss 0.01303160\n",
            "2019-02-11 10:05:19,878 epoch 50 - iter 8/27 - loss 0.01248569\n",
            "2019-02-11 10:05:20,688 epoch 50 - iter 10/27 - loss 0.01262155\n",
            "2019-02-11 10:05:21,475 epoch 50 - iter 12/27 - loss 0.01261558\n",
            "2019-02-11 10:05:22,187 epoch 50 - iter 14/27 - loss 0.01251940\n",
            "2019-02-11 10:05:23,010 epoch 50 - iter 16/27 - loss 0.01260466\n",
            "2019-02-11 10:05:23,679 epoch 50 - iter 18/27 - loss 0.01281814\n",
            "2019-02-11 10:05:24,497 epoch 50 - iter 20/27 - loss 0.01312363\n",
            "2019-02-11 10:05:25,345 epoch 50 - iter 22/27 - loss 0.01313724\n",
            "2019-02-11 10:05:26,079 epoch 50 - iter 24/27 - loss 0.01307287\n",
            "2019-02-11 10:05:26,631 epoch 50 - iter 26/27 - loss 0.01349238\n",
            "2019-02-11 10:05:26,643 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:05:26,645 EPOCH 50 done: loss 0.0135 - lr 0.0250 - bad epochs 1\n",
            "2019-02-11 10:05:27,365 DEV  : loss 0.01636772 - f-score 0.7368 - acc 0.7368\n",
            "2019-02-11 10:05:32,016 TEST : loss 0.02419789 - f-score 0.6738 - acc 0.6738\n",
            "2019-02-11 10:05:49,469 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:05:50,109 epoch 51 - iter 0/27 - loss 0.01479109\n",
            "2019-02-11 10:05:51,163 epoch 51 - iter 2/27 - loss 0.01396758\n",
            "2019-02-11 10:05:52,027 epoch 51 - iter 4/27 - loss 0.01298283\n",
            "2019-02-11 10:05:53,264 epoch 51 - iter 6/27 - loss 0.01350617\n",
            "2019-02-11 10:05:54,081 epoch 51 - iter 8/27 - loss 0.01397984\n",
            "2019-02-11 10:05:54,813 epoch 51 - iter 10/27 - loss 0.01368736\n",
            "2019-02-11 10:05:55,526 epoch 51 - iter 12/27 - loss 0.01304897\n",
            "2019-02-11 10:05:56,301 epoch 51 - iter 14/27 - loss 0.01332663\n",
            "2019-02-11 10:05:57,815 epoch 51 - iter 16/27 - loss 0.01340853\n",
            "2019-02-11 10:05:58,645 epoch 51 - iter 18/27 - loss 0.01315606\n",
            "2019-02-11 10:05:59,345 epoch 51 - iter 20/27 - loss 0.01362677\n",
            "2019-02-11 10:05:59,903 epoch 51 - iter 22/27 - loss 0.01350146\n",
            "2019-02-11 10:06:00,716 epoch 51 - iter 24/27 - loss 0.01337495\n",
            "2019-02-11 10:06:01,311 epoch 51 - iter 26/27 - loss 0.01392793\n",
            "2019-02-11 10:06:01,323 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:06:01,325 EPOCH 51 done: loss 0.0139 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 10:06:02,052 DEV  : loss 0.01632983 - f-score 0.7158 - acc 0.7158\n",
            "2019-02-11 10:06:07,035 TEST : loss 0.02453473 - f-score 0.6760 - acc 0.6760\n",
            "2019-02-11 10:06:07,038 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:06:07,443 epoch 52 - iter 0/27 - loss 0.01442986\n",
            "2019-02-11 10:06:08,284 epoch 52 - iter 2/27 - loss 0.01274073\n",
            "2019-02-11 10:06:09,140 epoch 52 - iter 4/27 - loss 0.01307380\n",
            "2019-02-11 10:06:10,010 epoch 52 - iter 6/27 - loss 0.01222833\n",
            "2019-02-11 10:06:11,032 epoch 52 - iter 8/27 - loss 0.01246999\n",
            "2019-02-11 10:06:11,764 epoch 52 - iter 10/27 - loss 0.01218392\n",
            "2019-02-11 10:06:13,451 epoch 52 - iter 12/27 - loss 0.01217867\n",
            "2019-02-11 10:06:14,362 epoch 52 - iter 14/27 - loss 0.01212264\n",
            "2019-02-11 10:06:15,171 epoch 52 - iter 16/27 - loss 0.01209539\n",
            "2019-02-11 10:06:16,112 epoch 52 - iter 18/27 - loss 0.01237646\n",
            "2019-02-11 10:06:16,996 epoch 52 - iter 20/27 - loss 0.01246936\n",
            "2019-02-11 10:06:17,948 epoch 52 - iter 22/27 - loss 0.01279385\n",
            "2019-02-11 10:06:18,588 epoch 52 - iter 24/27 - loss 0.01276969\n",
            "2019-02-11 10:06:19,037 epoch 52 - iter 26/27 - loss 0.01277062\n",
            "2019-02-11 10:06:19,049 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:06:19,051 EPOCH 52 done: loss 0.0128 - lr 0.0250 - bad epochs 1\n",
            "2019-02-11 10:06:19,842 DEV  : loss 0.01602848 - f-score 0.7789 - acc 0.7789\n",
            "2019-02-11 10:06:24,818 TEST : loss 0.02694236 - f-score 0.6502 - acc 0.6502\n",
            "2019-02-11 10:06:43,138 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:06:44,142 epoch 53 - iter 0/27 - loss 0.01542300\n",
            "2019-02-11 10:06:44,878 epoch 53 - iter 2/27 - loss 0.01338188\n",
            "2019-02-11 10:06:45,553 epoch 53 - iter 4/27 - loss 0.01292114\n",
            "2019-02-11 10:06:46,409 epoch 53 - iter 6/27 - loss 0.01275589\n",
            "2019-02-11 10:06:47,337 epoch 53 - iter 8/27 - loss 0.01280296\n",
            "2019-02-11 10:06:48,380 epoch 53 - iter 10/27 - loss 0.01370562\n",
            "2019-02-11 10:06:49,291 epoch 53 - iter 12/27 - loss 0.01339769\n",
            "2019-02-11 10:06:50,615 epoch 53 - iter 14/27 - loss 0.01369647\n",
            "2019-02-11 10:06:51,394 epoch 53 - iter 16/27 - loss 0.01336764\n",
            "2019-02-11 10:06:52,137 epoch 53 - iter 18/27 - loss 0.01358651\n",
            "2019-02-11 10:06:52,929 epoch 53 - iter 20/27 - loss 0.01317079\n",
            "2019-02-11 10:06:53,804 epoch 53 - iter 22/27 - loss 0.01341182\n",
            "2019-02-11 10:06:54,819 epoch 53 - iter 24/27 - loss 0.01343890\n",
            "2019-02-11 10:06:55,394 epoch 53 - iter 26/27 - loss 0.01375197\n",
            "2019-02-11 10:06:55,408 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:06:55,410 EPOCH 53 done: loss 0.0138 - lr 0.0250 - bad epochs 0\n",
            "2019-02-11 10:06:56,186 DEV  : loss 0.01590475 - f-score 0.7579 - acc 0.7579\n",
            "2019-02-11 10:07:01,278 TEST : loss 0.02475328 - f-score 0.6695 - acc 0.6695\n",
            "2019-02-11 10:07:01,283 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:07:01,685 epoch 54 - iter 0/27 - loss 0.01328906\n",
            "2019-02-11 10:07:02,769 epoch 54 - iter 2/27 - loss 0.01283562\n",
            "2019-02-11 10:07:03,734 epoch 54 - iter 4/27 - loss 0.01233517\n",
            "2019-02-11 10:07:04,571 epoch 54 - iter 6/27 - loss 0.01128673\n",
            "2019-02-11 10:07:05,469 epoch 54 - iter 8/27 - loss 0.01093970\n",
            "2019-02-11 10:07:06,684 epoch 54 - iter 10/27 - loss 0.01124821\n",
            "2019-02-11 10:07:07,576 epoch 54 - iter 12/27 - loss 0.01189013\n",
            "2019-02-11 10:07:08,466 epoch 54 - iter 14/27 - loss 0.01180493\n",
            "2019-02-11 10:07:09,433 epoch 54 - iter 16/27 - loss 0.01205756\n",
            "2019-02-11 10:07:10,065 epoch 54 - iter 18/27 - loss 0.01242671\n",
            "2019-02-11 10:07:10,841 epoch 54 - iter 20/27 - loss 0.01275062\n",
            "2019-02-11 10:07:11,635 epoch 54 - iter 22/27 - loss 0.01264826\n",
            "2019-02-11 10:07:12,494 epoch 54 - iter 24/27 - loss 0.01233947\n",
            "2019-02-11 10:07:13,503 epoch 54 - iter 26/27 - loss 0.01311633\n",
            "2019-02-11 10:07:13,515 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:07:13,517 EPOCH 54 done: loss 0.0131 - lr 0.0250 - bad epochs 1\n",
            "2019-02-11 10:07:14,300 DEV  : loss 0.01647725 - f-score 0.7263 - acc 0.7263\n",
            "2019-02-11 10:07:19,272 TEST : loss 0.02513763 - f-score 0.6867 - acc 0.6867\n",
            "2019-02-11 10:07:19,276 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:07:19,725 epoch 55 - iter 0/27 - loss 0.01002084\n",
            "2019-02-11 10:07:20,687 epoch 55 - iter 2/27 - loss 0.01090684\n",
            "2019-02-11 10:07:21,537 epoch 55 - iter 4/27 - loss 0.01306498\n",
            "2019-02-11 10:07:22,340 epoch 55 - iter 6/27 - loss 0.01228622\n",
            "2019-02-11 10:07:23,261 epoch 55 - iter 8/27 - loss 0.01223983\n",
            "2019-02-11 10:07:23,937 epoch 55 - iter 10/27 - loss 0.01244694\n",
            "2019-02-11 10:07:24,851 epoch 55 - iter 12/27 - loss 0.01235676\n",
            "2019-02-11 10:07:25,852 epoch 55 - iter 14/27 - loss 0.01223124\n",
            "2019-02-11 10:07:26,442 epoch 55 - iter 16/27 - loss 0.01197202\n",
            "2019-02-11 10:07:27,079 epoch 55 - iter 18/27 - loss 0.01210323\n",
            "2019-02-11 10:07:27,725 epoch 55 - iter 20/27 - loss 0.01221145\n",
            "2019-02-11 10:07:28,769 epoch 55 - iter 22/27 - loss 0.01235336\n",
            "2019-02-11 10:07:30,415 epoch 55 - iter 24/27 - loss 0.01258604\n",
            "2019-02-11 10:07:30,956 epoch 55 - iter 26/27 - loss 0.01283990\n",
            "2019-02-11 10:07:30,968 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:07:30,970 EPOCH 55 done: loss 0.0128 - lr 0.0250 - bad epochs 2\n",
            "2019-02-11 10:07:31,715 DEV  : loss 0.01721407 - f-score 0.7158 - acc 0.7158\n",
            "2019-02-11 10:07:36,347 TEST : loss 0.02524756 - f-score 0.6738 - acc 0.6738\n",
            "2019-02-11 10:07:36,350 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:07:36,760 epoch 56 - iter 0/27 - loss 0.01117645\n",
            "2019-02-11 10:07:37,482 epoch 56 - iter 2/27 - loss 0.01368231\n",
            "2019-02-11 10:07:38,790 epoch 56 - iter 4/27 - loss 0.01460988\n",
            "2019-02-11 10:07:39,612 epoch 56 - iter 6/27 - loss 0.01425536\n",
            "2019-02-11 10:07:40,393 epoch 56 - iter 8/27 - loss 0.01389326\n",
            "2019-02-11 10:07:41,238 epoch 56 - iter 10/27 - loss 0.01332281\n",
            "2019-02-11 10:07:42,376 epoch 56 - iter 12/27 - loss 0.01313776\n",
            "2019-02-11 10:07:43,071 epoch 56 - iter 14/27 - loss 0.01314125\n",
            "2019-02-11 10:07:43,826 epoch 56 - iter 16/27 - loss 0.01320319\n",
            "2019-02-11 10:07:44,443 epoch 56 - iter 18/27 - loss 0.01268846\n",
            "2019-02-11 10:07:45,647 epoch 56 - iter 20/27 - loss 0.01263711\n",
            "2019-02-11 10:07:46,559 epoch 56 - iter 22/27 - loss 0.01259193\n",
            "2019-02-11 10:07:47,381 epoch 56 - iter 24/27 - loss 0.01292045\n",
            "2019-02-11 10:07:47,883 epoch 56 - iter 26/27 - loss 0.01322960\n",
            "2019-02-11 10:07:47,895 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:07:47,896 EPOCH 56 done: loss 0.0132 - lr 0.0250 - bad epochs 3\n",
            "2019-02-11 10:07:48,642 DEV  : loss 0.01501276 - f-score 0.8000 - acc 0.8000\n",
            "2019-02-11 10:07:53,415 TEST : loss 0.02612285 - f-score 0.6631 - acc 0.6631\n",
            "Epoch    55: reducing learning rate of group 0 to 1.2500e-02.\n",
            "2019-02-11 10:07:53,418 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:07:53,814 epoch 57 - iter 0/27 - loss 0.01551327\n",
            "2019-02-11 10:07:54,703 epoch 57 - iter 2/27 - loss 0.01344371\n",
            "2019-02-11 10:07:55,926 epoch 57 - iter 4/27 - loss 0.01449491\n",
            "2019-02-11 10:07:56,759 epoch 57 - iter 6/27 - loss 0.01404881\n",
            "2019-02-11 10:07:57,474 epoch 57 - iter 8/27 - loss 0.01372707\n",
            "2019-02-11 10:07:58,169 epoch 57 - iter 10/27 - loss 0.01275754\n",
            "2019-02-11 10:07:59,182 epoch 57 - iter 12/27 - loss 0.01292091\n",
            "2019-02-11 10:08:00,002 epoch 57 - iter 14/27 - loss 0.01278153\n",
            "2019-02-11 10:08:01,133 epoch 57 - iter 16/27 - loss 0.01246661\n",
            "2019-02-11 10:08:01,835 epoch 57 - iter 18/27 - loss 0.01234681\n",
            "2019-02-11 10:08:02,793 epoch 57 - iter 20/27 - loss 0.01250821\n",
            "2019-02-11 10:08:03,515 epoch 57 - iter 22/27 - loss 0.01256295\n",
            "2019-02-11 10:08:04,463 epoch 57 - iter 24/27 - loss 0.01257887\n",
            "2019-02-11 10:08:04,902 epoch 57 - iter 26/27 - loss 0.01280241\n",
            "2019-02-11 10:08:04,914 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:08:04,915 EPOCH 57 done: loss 0.0128 - lr 0.0125 - bad epochs 0\n",
            "2019-02-11 10:08:05,642 DEV  : loss 0.01510284 - f-score 0.7684 - acc 0.7684\n",
            "2019-02-11 10:08:10,204 TEST : loss 0.02597170 - f-score 0.6609 - acc 0.6609\n",
            "2019-02-11 10:08:10,207 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:08:10,477 epoch 58 - iter 0/27 - loss 0.00803200\n",
            "2019-02-11 10:08:11,681 epoch 58 - iter 2/27 - loss 0.00927287\n",
            "2019-02-11 10:08:12,334 epoch 58 - iter 4/27 - loss 0.00931058\n",
            "2019-02-11 10:08:13,149 epoch 58 - iter 6/27 - loss 0.00985627\n",
            "2019-02-11 10:08:13,848 epoch 58 - iter 8/27 - loss 0.00988888\n",
            "2019-02-11 10:08:15,106 epoch 58 - iter 10/27 - loss 0.01114296\n",
            "2019-02-11 10:08:15,873 epoch 58 - iter 12/27 - loss 0.01151320\n",
            "2019-02-11 10:08:16,629 epoch 58 - iter 14/27 - loss 0.01146380\n",
            "2019-02-11 10:08:17,421 epoch 58 - iter 16/27 - loss 0.01171286\n",
            "2019-02-11 10:08:18,321 epoch 58 - iter 18/27 - loss 0.01160975\n",
            "2019-02-11 10:08:19,040 epoch 58 - iter 20/27 - loss 0.01157504\n",
            "2019-02-11 10:08:19,946 epoch 58 - iter 22/27 - loss 0.01166867\n",
            "2019-02-11 10:08:20,831 epoch 58 - iter 24/27 - loss 0.01177856\n",
            "2019-02-11 10:08:21,368 epoch 58 - iter 26/27 - loss 0.01222706\n",
            "2019-02-11 10:08:21,380 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:08:21,382 EPOCH 58 done: loss 0.0122 - lr 0.0125 - bad epochs 1\n",
            "2019-02-11 10:08:22,098 DEV  : loss 0.01510924 - f-score 0.7579 - acc 0.7579\n",
            "2019-02-11 10:08:26,603 TEST : loss 0.02566404 - f-score 0.6803 - acc 0.6803\n",
            "2019-02-11 10:08:43,999 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:08:44,423 epoch 59 - iter 0/27 - loss 0.00845190\n",
            "2019-02-11 10:08:45,602 epoch 59 - iter 2/27 - loss 0.01421762\n",
            "2019-02-11 10:08:46,384 epoch 59 - iter 4/27 - loss 0.01365648\n",
            "2019-02-11 10:08:47,184 epoch 59 - iter 6/27 - loss 0.01352172\n",
            "2019-02-11 10:08:47,995 epoch 59 - iter 8/27 - loss 0.01245704\n",
            "2019-02-11 10:08:48,851 epoch 59 - iter 10/27 - loss 0.01264966\n",
            "2019-02-11 10:08:49,594 epoch 59 - iter 12/27 - loss 0.01255738\n",
            "2019-02-11 10:08:50,838 epoch 59 - iter 14/27 - loss 0.01263621\n",
            "2019-02-11 10:08:51,748 epoch 59 - iter 16/27 - loss 0.01276754\n",
            "2019-02-11 10:08:52,632 epoch 59 - iter 18/27 - loss 0.01263940\n",
            "2019-02-11 10:08:53,295 epoch 59 - iter 20/27 - loss 0.01250444\n",
            "2019-02-11 10:08:54,494 epoch 59 - iter 22/27 - loss 0.01258635\n",
            "2019-02-11 10:08:55,194 epoch 59 - iter 24/27 - loss 0.01257444\n",
            "2019-02-11 10:08:55,835 epoch 59 - iter 26/27 - loss 0.01281388\n",
            "2019-02-11 10:08:55,848 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:08:55,849 EPOCH 59 done: loss 0.0128 - lr 0.0125 - bad epochs 0\n",
            "2019-02-11 10:08:56,586 DEV  : loss 0.01551234 - f-score 0.7474 - acc 0.7474\n",
            "2019-02-11 10:09:01,262 TEST : loss 0.02551949 - f-score 0.6931 - acc 0.6931\n",
            "2019-02-11 10:09:01,269 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:09:01,689 epoch 60 - iter 0/27 - loss 0.00977299\n",
            "2019-02-11 10:09:02,518 epoch 60 - iter 2/27 - loss 0.01013432\n",
            "2019-02-11 10:09:03,490 epoch 60 - iter 4/27 - loss 0.00924638\n",
            "2019-02-11 10:09:04,203 epoch 60 - iter 6/27 - loss 0.01085320\n",
            "2019-02-11 10:09:05,020 epoch 60 - iter 8/27 - loss 0.01172705\n",
            "2019-02-11 10:09:05,717 epoch 60 - iter 10/27 - loss 0.01149086\n",
            "2019-02-11 10:09:06,546 epoch 60 - iter 12/27 - loss 0.01136867\n",
            "2019-02-11 10:09:07,318 epoch 60 - iter 14/27 - loss 0.01147790\n",
            "2019-02-11 10:09:08,028 epoch 60 - iter 16/27 - loss 0.01114640\n",
            "2019-02-11 10:09:09,212 epoch 60 - iter 18/27 - loss 0.01132087\n",
            "2019-02-11 10:09:10,018 epoch 60 - iter 20/27 - loss 0.01128837\n",
            "2019-02-11 10:09:11,182 epoch 60 - iter 22/27 - loss 0.01147649\n",
            "2019-02-11 10:09:11,980 epoch 60 - iter 24/27 - loss 0.01129705\n",
            "2019-02-11 10:09:12,785 epoch 60 - iter 26/27 - loss 0.01185542\n",
            "2019-02-11 10:09:12,799 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:09:12,800 EPOCH 60 done: loss 0.0119 - lr 0.0125 - bad epochs 1\n",
            "2019-02-11 10:09:13,535 DEV  : loss 0.01559701 - f-score 0.7474 - acc 0.7474\n",
            "2019-02-11 10:09:18,158 TEST : loss 0.02607115 - f-score 0.6845 - acc 0.6845\n",
            "2019-02-11 10:09:35,463 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:09:36,150 epoch 61 - iter 0/27 - loss 0.01155001\n",
            "2019-02-11 10:09:36,929 epoch 61 - iter 2/27 - loss 0.01337656\n",
            "2019-02-11 10:09:37,782 epoch 61 - iter 4/27 - loss 0.01262638\n",
            "2019-02-11 10:09:38,623 epoch 61 - iter 6/27 - loss 0.01271781\n",
            "2019-02-11 10:09:39,327 epoch 61 - iter 8/27 - loss 0.01238887\n",
            "2019-02-11 10:09:40,035 epoch 61 - iter 10/27 - loss 0.01223533\n",
            "2019-02-11 10:09:41,005 epoch 61 - iter 12/27 - loss 0.01232411\n",
            "2019-02-11 10:09:42,353 epoch 61 - iter 14/27 - loss 0.01255540\n",
            "2019-02-11 10:09:43,015 epoch 61 - iter 16/27 - loss 0.01210513\n",
            "2019-02-11 10:09:44,484 epoch 61 - iter 18/27 - loss 0.01234583\n",
            "2019-02-11 10:09:45,125 epoch 61 - iter 20/27 - loss 0.01197817\n",
            "2019-02-11 10:09:45,825 epoch 61 - iter 22/27 - loss 0.01223923\n",
            "2019-02-11 10:09:46,780 epoch 61 - iter 24/27 - loss 0.01230810\n",
            "2019-02-11 10:09:47,366 epoch 61 - iter 26/27 - loss 0.01230785\n",
            "2019-02-11 10:09:47,378 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:09:47,379 EPOCH 61 done: loss 0.0123 - lr 0.0125 - bad epochs 0\n",
            "2019-02-11 10:09:48,106 DEV  : loss 0.01603663 - f-score 0.7579 - acc 0.7579\n",
            "2019-02-11 10:09:52,742 TEST : loss 0.02602866 - f-score 0.6845 - acc 0.6845\n",
            "2019-02-11 10:09:52,746 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:09:53,082 epoch 62 - iter 0/27 - loss 0.01166291\n",
            "2019-02-11 10:09:53,985 epoch 62 - iter 2/27 - loss 0.01142768\n",
            "2019-02-11 10:09:54,770 epoch 62 - iter 4/27 - loss 0.01265405\n",
            "2019-02-11 10:09:55,811 epoch 62 - iter 6/27 - loss 0.01198701\n",
            "2019-02-11 10:09:56,954 epoch 62 - iter 8/27 - loss 0.01180983\n",
            "2019-02-11 10:09:57,755 epoch 62 - iter 10/27 - loss 0.01116630\n",
            "2019-02-11 10:09:58,721 epoch 62 - iter 12/27 - loss 0.01082316\n",
            "2019-02-11 10:09:59,381 epoch 62 - iter 14/27 - loss 0.01068231\n",
            "2019-02-11 10:10:00,181 epoch 62 - iter 16/27 - loss 0.01056590\n",
            "2019-02-11 10:10:00,876 epoch 62 - iter 18/27 - loss 0.01055603\n",
            "2019-02-11 10:10:01,755 epoch 62 - iter 20/27 - loss 0.01051578\n",
            "2019-02-11 10:10:02,526 epoch 62 - iter 22/27 - loss 0.01101254\n",
            "2019-02-11 10:10:03,403 epoch 62 - iter 24/27 - loss 0.01148763\n",
            "2019-02-11 10:10:04,076 epoch 62 - iter 26/27 - loss 0.01197142\n",
            "2019-02-11 10:10:04,088 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:10:04,089 EPOCH 62 done: loss 0.0120 - lr 0.0125 - bad epochs 1\n",
            "2019-02-11 10:10:04,820 DEV  : loss 0.01550706 - f-score 0.7579 - acc 0.7579\n",
            "2019-02-11 10:10:09,439 TEST : loss 0.02635141 - f-score 0.6867 - acc 0.6867\n",
            "2019-02-11 10:10:09,442 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:10:09,821 epoch 63 - iter 0/27 - loss 0.00783733\n",
            "2019-02-11 10:10:10,719 epoch 63 - iter 2/27 - loss 0.01056485\n",
            "2019-02-11 10:10:11,749 epoch 63 - iter 4/27 - loss 0.01126154\n",
            "2019-02-11 10:10:12,605 epoch 63 - iter 6/27 - loss 0.01179677\n",
            "2019-02-11 10:10:13,801 epoch 63 - iter 8/27 - loss 0.01148696\n",
            "2019-02-11 10:10:14,983 epoch 63 - iter 10/27 - loss 0.01137269\n",
            "2019-02-11 10:10:15,796 epoch 63 - iter 12/27 - loss 0.01214408\n",
            "2019-02-11 10:10:16,422 epoch 63 - iter 14/27 - loss 0.01230966\n",
            "2019-02-11 10:10:17,168 epoch 63 - iter 16/27 - loss 0.01220357\n",
            "2019-02-11 10:10:17,985 epoch 63 - iter 18/27 - loss 0.01224792\n",
            "2019-02-11 10:10:18,766 epoch 63 - iter 20/27 - loss 0.01230391\n",
            "2019-02-11 10:10:19,466 epoch 63 - iter 22/27 - loss 0.01221904\n",
            "2019-02-11 10:10:20,194 epoch 63 - iter 24/27 - loss 0.01225900\n",
            "2019-02-11 10:10:20,910 epoch 63 - iter 26/27 - loss 0.01288285\n",
            "2019-02-11 10:10:20,922 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:10:20,923 EPOCH 63 done: loss 0.0129 - lr 0.0125 - bad epochs 2\n",
            "2019-02-11 10:10:21,634 DEV  : loss 0.01486128 - f-score 0.7789 - acc 0.7789\n",
            "2019-02-11 10:10:26,275 TEST : loss 0.02686661 - f-score 0.6717 - acc 0.6717\n",
            "2019-02-11 10:10:26,278 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:10:26,706 epoch 64 - iter 0/27 - loss 0.01463551\n",
            "2019-02-11 10:10:27,865 epoch 64 - iter 2/27 - loss 0.01171298\n",
            "2019-02-11 10:10:28,439 epoch 64 - iter 4/27 - loss 0.01228607\n",
            "2019-02-11 10:10:29,324 epoch 64 - iter 6/27 - loss 0.01240987\n",
            "2019-02-11 10:10:30,436 epoch 64 - iter 8/27 - loss 0.01165628\n",
            "2019-02-11 10:10:31,253 epoch 64 - iter 10/27 - loss 0.01128129\n",
            "2019-02-11 10:10:32,097 epoch 64 - iter 12/27 - loss 0.01086978\n",
            "2019-02-11 10:10:32,889 epoch 64 - iter 14/27 - loss 0.01138351\n",
            "2019-02-11 10:10:34,158 epoch 64 - iter 16/27 - loss 0.01180138\n",
            "2019-02-11 10:10:34,796 epoch 64 - iter 18/27 - loss 0.01149347\n",
            "2019-02-11 10:10:35,523 epoch 64 - iter 20/27 - loss 0.01129739\n",
            "2019-02-11 10:10:36,387 epoch 64 - iter 22/27 - loss 0.01130839\n",
            "2019-02-11 10:10:37,194 epoch 64 - iter 24/27 - loss 0.01138746\n",
            "2019-02-11 10:10:37,720 epoch 64 - iter 26/27 - loss 0.01167638\n",
            "2019-02-11 10:10:37,733 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:10:37,734 EPOCH 64 done: loss 0.0117 - lr 0.0125 - bad epochs 3\n",
            "2019-02-11 10:10:38,498 DEV  : loss 0.01464818 - f-score 0.8000 - acc 0.8000\n",
            "2019-02-11 10:10:43,209 TEST : loss 0.02694628 - f-score 0.6738 - acc 0.6738\n",
            "2019-02-11 10:11:00,734 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:11:01,129 epoch 65 - iter 0/27 - loss 0.01267433\n",
            "2019-02-11 10:11:02,016 epoch 65 - iter 2/27 - loss 0.01188544\n",
            "2019-02-11 10:11:02,938 epoch 65 - iter 4/27 - loss 0.01088585\n",
            "2019-02-11 10:11:03,719 epoch 65 - iter 6/27 - loss 0.01030328\n",
            "2019-02-11 10:11:04,407 epoch 65 - iter 8/27 - loss 0.01167047\n",
            "2019-02-11 10:11:05,435 epoch 65 - iter 10/27 - loss 0.01195710\n",
            "2019-02-11 10:11:06,328 epoch 65 - iter 12/27 - loss 0.01205301\n",
            "2019-02-11 10:11:07,042 epoch 65 - iter 14/27 - loss 0.01190488\n",
            "2019-02-11 10:11:08,241 epoch 65 - iter 16/27 - loss 0.01172319\n",
            "2019-02-11 10:11:10,780 epoch 65 - iter 18/27 - loss 0.01203836\n",
            "2019-02-11 10:11:11,590 epoch 65 - iter 20/27 - loss 0.01196849\n",
            "2019-02-11 10:11:12,819 epoch 65 - iter 22/27 - loss 0.01199217\n",
            "2019-02-11 10:11:13,485 epoch 65 - iter 24/27 - loss 0.01211844\n",
            "2019-02-11 10:11:14,176 epoch 65 - iter 26/27 - loss 0.01256497\n",
            "2019-02-11 10:11:14,188 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:11:14,189 EPOCH 65 done: loss 0.0126 - lr 0.0125 - bad epochs 0\n",
            "2019-02-11 10:11:14,945 DEV  : loss 0.01476769 - f-score 0.7684 - acc 0.7684\n",
            "2019-02-11 10:11:19,720 TEST : loss 0.02658813 - f-score 0.6953 - acc 0.6953\n",
            "2019-02-11 10:11:19,723 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:11:20,188 epoch 66 - iter 0/27 - loss 0.01039831\n",
            "2019-02-11 10:11:21,846 epoch 66 - iter 2/27 - loss 0.01121905\n",
            "2019-02-11 10:11:22,634 epoch 66 - iter 4/27 - loss 0.01168444\n",
            "2019-02-11 10:11:23,671 epoch 66 - iter 6/27 - loss 0.01171224\n",
            "2019-02-11 10:11:24,561 epoch 66 - iter 8/27 - loss 0.01131388\n",
            "2019-02-11 10:11:25,237 epoch 66 - iter 10/27 - loss 0.01091288\n",
            "2019-02-11 10:11:26,071 epoch 66 - iter 12/27 - loss 0.01114315\n",
            "2019-02-11 10:11:27,032 epoch 66 - iter 14/27 - loss 0.01152154\n",
            "2019-02-11 10:11:27,824 epoch 66 - iter 16/27 - loss 0.01101111\n",
            "2019-02-11 10:11:28,576 epoch 66 - iter 18/27 - loss 0.01085269\n",
            "2019-02-11 10:11:29,593 epoch 66 - iter 20/27 - loss 0.01097861\n",
            "2019-02-11 10:11:30,245 epoch 66 - iter 22/27 - loss 0.01122000\n",
            "2019-02-11 10:11:31,007 epoch 66 - iter 24/27 - loss 0.01135494\n",
            "2019-02-11 10:11:31,595 epoch 66 - iter 26/27 - loss 0.01147430\n",
            "2019-02-11 10:11:31,607 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:11:31,608 EPOCH 66 done: loss 0.0115 - lr 0.0125 - bad epochs 1\n",
            "2019-02-11 10:11:32,358 DEV  : loss 0.01506003 - f-score 0.7474 - acc 0.7474\n",
            "2019-02-11 10:11:37,057 TEST : loss 0.02643765 - f-score 0.6867 - acc 0.6867\n",
            "2019-02-11 10:11:54,541 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:11:55,278 epoch 67 - iter 0/27 - loss 0.01002736\n",
            "2019-02-11 10:11:56,055 epoch 67 - iter 2/27 - loss 0.01038341\n",
            "2019-02-11 10:11:56,785 epoch 67 - iter 4/27 - loss 0.00941478\n",
            "2019-02-11 10:11:57,730 epoch 67 - iter 6/27 - loss 0.00931766\n",
            "2019-02-11 10:11:58,486 epoch 67 - iter 8/27 - loss 0.00960919\n",
            "2019-02-11 10:11:59,279 epoch 67 - iter 10/27 - loss 0.00954309\n",
            "2019-02-11 10:12:00,333 epoch 67 - iter 12/27 - loss 0.01002786\n",
            "2019-02-11 10:12:01,694 epoch 67 - iter 14/27 - loss 0.01038779\n",
            "2019-02-11 10:12:02,536 epoch 67 - iter 16/27 - loss 0.01070978\n",
            "2019-02-11 10:12:03,274 epoch 67 - iter 18/27 - loss 0.01085824\n",
            "2019-02-11 10:12:04,141 epoch 67 - iter 20/27 - loss 0.01091991\n",
            "2019-02-11 10:12:04,826 epoch 67 - iter 22/27 - loss 0.01105723\n",
            "2019-02-11 10:12:05,680 epoch 67 - iter 24/27 - loss 0.01134308\n",
            "2019-02-11 10:12:06,690 epoch 67 - iter 26/27 - loss 0.01175838\n",
            "2019-02-11 10:12:06,703 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:12:06,705 EPOCH 67 done: loss 0.0118 - lr 0.0125 - bad epochs 0\n",
            "2019-02-11 10:12:07,488 DEV  : loss 0.01510954 - f-score 0.7895 - acc 0.7895\n",
            "2019-02-11 10:12:12,397 TEST : loss 0.02719260 - f-score 0.6781 - acc 0.6781\n",
            "2019-02-11 10:12:12,402 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:12:13,294 epoch 68 - iter 0/27 - loss 0.01239527\n",
            "2019-02-11 10:12:13,999 epoch 68 - iter 2/27 - loss 0.01077155\n",
            "2019-02-11 10:12:14,868 epoch 68 - iter 4/27 - loss 0.01120702\n",
            "2019-02-11 10:12:15,972 epoch 68 - iter 6/27 - loss 0.01149405\n",
            "2019-02-11 10:12:16,665 epoch 68 - iter 8/27 - loss 0.01119849\n",
            "2019-02-11 10:12:17,304 epoch 68 - iter 10/27 - loss 0.01150134\n",
            "2019-02-11 10:12:17,954 epoch 68 - iter 12/27 - loss 0.01151623\n",
            "2019-02-11 10:12:18,794 epoch 68 - iter 14/27 - loss 0.01152293\n",
            "2019-02-11 10:12:19,429 epoch 68 - iter 16/27 - loss 0.01156216\n",
            "2019-02-11 10:12:20,281 epoch 68 - iter 18/27 - loss 0.01157998\n",
            "2019-02-11 10:12:21,156 epoch 68 - iter 20/27 - loss 0.01157064\n",
            "2019-02-11 10:12:22,511 epoch 68 - iter 22/27 - loss 0.01150387\n",
            "2019-02-11 10:12:23,437 epoch 68 - iter 24/27 - loss 0.01116850\n",
            "2019-02-11 10:12:24,003 epoch 68 - iter 26/27 - loss 0.01152432\n",
            "2019-02-11 10:12:24,016 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:12:24,017 EPOCH 68 done: loss 0.0115 - lr 0.0125 - bad epochs 1\n",
            "2019-02-11 10:12:24,784 DEV  : loss 0.01531997 - f-score 0.7684 - acc 0.7684\n",
            "2019-02-11 10:12:29,681 TEST : loss 0.02678370 - f-score 0.6652 - acc 0.6652\n",
            "2019-02-11 10:12:29,683 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:12:30,175 epoch 69 - iter 0/27 - loss 0.01161675\n",
            "2019-02-11 10:12:31,072 epoch 69 - iter 2/27 - loss 0.01256987\n",
            "2019-02-11 10:12:32,244 epoch 69 - iter 4/27 - loss 0.01225019\n",
            "2019-02-11 10:12:32,911 epoch 69 - iter 6/27 - loss 0.01147645\n",
            "2019-02-11 10:12:34,212 epoch 69 - iter 8/27 - loss 0.01176231\n",
            "2019-02-11 10:12:34,982 epoch 69 - iter 10/27 - loss 0.01109476\n",
            "2019-02-11 10:12:35,752 epoch 69 - iter 12/27 - loss 0.01115383\n",
            "2019-02-11 10:12:36,563 epoch 69 - iter 14/27 - loss 0.01114893\n",
            "2019-02-11 10:12:37,416 epoch 69 - iter 16/27 - loss 0.01111199\n",
            "2019-02-11 10:12:38,093 epoch 69 - iter 18/27 - loss 0.01096088\n",
            "2019-02-11 10:12:39,289 epoch 69 - iter 20/27 - loss 0.01054072\n",
            "2019-02-11 10:12:40,063 epoch 69 - iter 22/27 - loss 0.01063288\n",
            "2019-02-11 10:12:40,803 epoch 69 - iter 24/27 - loss 0.01059100\n",
            "2019-02-11 10:12:41,459 epoch 69 - iter 26/27 - loss 0.01089516\n",
            "2019-02-11 10:12:41,472 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:12:41,473 EPOCH 69 done: loss 0.0109 - lr 0.0125 - bad epochs 2\n",
            "2019-02-11 10:12:42,196 DEV  : loss 0.01631265 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 10:12:46,825 TEST : loss 0.02666731 - f-score 0.6824 - acc 0.6824\n",
            "2019-02-11 10:13:04,316 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:13:04,805 epoch 70 - iter 0/27 - loss 0.01399495\n",
            "2019-02-11 10:13:05,661 epoch 70 - iter 2/27 - loss 0.01290791\n",
            "2019-02-11 10:13:06,385 epoch 70 - iter 4/27 - loss 0.01218406\n",
            "2019-02-11 10:13:07,166 epoch 70 - iter 6/27 - loss 0.01139200\n",
            "2019-02-11 10:13:08,385 epoch 70 - iter 8/27 - loss 0.01224012\n",
            "2019-02-11 10:13:09,201 epoch 70 - iter 10/27 - loss 0.01237277\n",
            "2019-02-11 10:13:10,117 epoch 70 - iter 12/27 - loss 0.01210822\n",
            "2019-02-11 10:13:10,849 epoch 70 - iter 14/27 - loss 0.01202713\n",
            "2019-02-11 10:13:11,486 epoch 70 - iter 16/27 - loss 0.01215075\n",
            "2019-02-11 10:13:12,811 epoch 70 - iter 18/27 - loss 0.01192719\n",
            "2019-02-11 10:13:13,780 epoch 70 - iter 20/27 - loss 0.01170714\n",
            "2019-02-11 10:13:14,556 epoch 70 - iter 22/27 - loss 0.01147965\n",
            "2019-02-11 10:13:15,275 epoch 70 - iter 24/27 - loss 0.01141901\n",
            "2019-02-11 10:13:15,863 epoch 70 - iter 26/27 - loss 0.01188835\n",
            "2019-02-11 10:13:15,876 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:13:15,877 EPOCH 70 done: loss 0.0119 - lr 0.0125 - bad epochs 0\n",
            "2019-02-11 10:13:16,616 DEV  : loss 0.01577650 - f-score 0.7474 - acc 0.7474\n",
            "2019-02-11 10:13:21,423 TEST : loss 0.02683226 - f-score 0.6738 - acc 0.6738\n",
            "2019-02-11 10:13:21,426 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:13:21,839 epoch 71 - iter 0/27 - loss 0.01304343\n",
            "2019-02-11 10:13:22,798 epoch 71 - iter 2/27 - loss 0.01329223\n",
            "2019-02-11 10:13:23,560 epoch 71 - iter 4/27 - loss 0.01367064\n",
            "2019-02-11 10:13:24,189 epoch 71 - iter 6/27 - loss 0.01217352\n",
            "2019-02-11 10:13:25,045 epoch 71 - iter 8/27 - loss 0.01172281\n",
            "2019-02-11 10:13:25,781 epoch 71 - iter 10/27 - loss 0.01213812\n",
            "2019-02-11 10:13:26,517 epoch 71 - iter 12/27 - loss 0.01219366\n",
            "2019-02-11 10:13:27,410 epoch 71 - iter 14/27 - loss 0.01155893\n",
            "2019-02-11 10:13:28,658 epoch 71 - iter 16/27 - loss 0.01167428\n",
            "2019-02-11 10:13:29,741 epoch 71 - iter 18/27 - loss 0.01192531\n",
            "2019-02-11 10:13:30,947 epoch 71 - iter 20/27 - loss 0.01181277\n",
            "2019-02-11 10:13:31,659 epoch 71 - iter 22/27 - loss 0.01200151\n",
            "2019-02-11 10:13:32,428 epoch 71 - iter 24/27 - loss 0.01183090\n",
            "2019-02-11 10:13:32,917 epoch 71 - iter 26/27 - loss 0.01199329\n",
            "2019-02-11 10:13:32,930 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:13:32,931 EPOCH 71 done: loss 0.0120 - lr 0.0125 - bad epochs 1\n",
            "2019-02-11 10:13:33,663 DEV  : loss 0.01587136 - f-score 0.7368 - acc 0.7368\n",
            "2019-02-11 10:13:38,315 TEST : loss 0.02746323 - f-score 0.6845 - acc 0.6845\n",
            "2019-02-11 10:13:38,319 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:13:38,742 epoch 72 - iter 0/27 - loss 0.00833892\n",
            "2019-02-11 10:13:39,536 epoch 72 - iter 2/27 - loss 0.00863508\n",
            "2019-02-11 10:13:40,265 epoch 72 - iter 4/27 - loss 0.00962176\n",
            "2019-02-11 10:13:40,880 epoch 72 - iter 6/27 - loss 0.00983011\n",
            "2019-02-11 10:13:41,846 epoch 72 - iter 8/27 - loss 0.00980615\n",
            "2019-02-11 10:13:42,898 epoch 72 - iter 10/27 - loss 0.01030467\n",
            "2019-02-11 10:13:43,753 epoch 72 - iter 12/27 - loss 0.01014624\n",
            "2019-02-11 10:13:44,415 epoch 72 - iter 14/27 - loss 0.01064341\n",
            "2019-02-11 10:13:45,195 epoch 72 - iter 16/27 - loss 0.01035764\n",
            "2019-02-11 10:13:46,096 epoch 72 - iter 18/27 - loss 0.01019202\n",
            "2019-02-11 10:13:46,885 epoch 72 - iter 20/27 - loss 0.01008494\n",
            "2019-02-11 10:13:48,101 epoch 72 - iter 22/27 - loss 0.01034784\n",
            "2019-02-11 10:13:49,195 epoch 72 - iter 24/27 - loss 0.01027105\n",
            "2019-02-11 10:13:49,700 epoch 72 - iter 26/27 - loss 0.01066416\n",
            "2019-02-11 10:13:49,714 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:13:49,715 EPOCH 72 done: loss 0.0107 - lr 0.0125 - bad epochs 2\n",
            "2019-02-11 10:13:50,449 DEV  : loss 0.01722598 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 10:13:55,053 TEST : loss 0.02670673 - f-score 0.6760 - acc 0.6760\n",
            "2019-02-11 10:14:12,394 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:14:12,891 epoch 73 - iter 0/27 - loss 0.01224857\n",
            "2019-02-11 10:14:13,846 epoch 73 - iter 2/27 - loss 0.00907563\n",
            "2019-02-11 10:14:14,604 epoch 73 - iter 4/27 - loss 0.01103344\n",
            "2019-02-11 10:14:15,774 epoch 73 - iter 6/27 - loss 0.01064384\n",
            "2019-02-11 10:14:16,742 epoch 73 - iter 8/27 - loss 0.01000593\n",
            "2019-02-11 10:14:17,872 epoch 73 - iter 10/27 - loss 0.01028466\n",
            "2019-02-11 10:14:18,806 epoch 73 - iter 12/27 - loss 0.01077792\n",
            "2019-02-11 10:14:19,570 epoch 73 - iter 14/27 - loss 0.01015268\n",
            "2019-02-11 10:14:20,549 epoch 73 - iter 16/27 - loss 0.01001046\n",
            "2019-02-11 10:14:21,321 epoch 73 - iter 18/27 - loss 0.00998619\n",
            "2019-02-11 10:14:22,052 epoch 73 - iter 20/27 - loss 0.01054434\n",
            "2019-02-11 10:14:22,836 epoch 73 - iter 22/27 - loss 0.01040344\n",
            "2019-02-11 10:14:23,649 epoch 73 - iter 24/27 - loss 0.01042608\n",
            "2019-02-11 10:14:24,152 epoch 73 - iter 26/27 - loss 0.01111949\n",
            "2019-02-11 10:14:24,164 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:14:24,166 EPOCH 73 done: loss 0.0111 - lr 0.0125 - bad epochs 0\n",
            "2019-02-11 10:14:24,885 DEV  : loss 0.01614165 - f-score 0.7158 - acc 0.7158\n",
            "2019-02-11 10:14:29,522 TEST : loss 0.02748873 - f-score 0.6781 - acc 0.6781\n",
            "2019-02-11 10:14:29,526 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:14:29,960 epoch 74 - iter 0/27 - loss 0.01086825\n",
            "2019-02-11 10:14:30,647 epoch 74 - iter 2/27 - loss 0.00844239\n",
            "2019-02-11 10:14:31,378 epoch 74 - iter 4/27 - loss 0.00942357\n",
            "2019-02-11 10:14:32,249 epoch 74 - iter 6/27 - loss 0.00959202\n",
            "2019-02-11 10:14:33,412 epoch 74 - iter 8/27 - loss 0.00985710\n",
            "2019-02-11 10:14:34,169 epoch 74 - iter 10/27 - loss 0.00983625\n",
            "2019-02-11 10:14:35,007 epoch 74 - iter 12/27 - loss 0.01042350\n",
            "2019-02-11 10:14:36,033 epoch 74 - iter 14/27 - loss 0.01022002\n",
            "2019-02-11 10:14:36,910 epoch 74 - iter 16/27 - loss 0.01011055\n",
            "2019-02-11 10:14:37,783 epoch 74 - iter 18/27 - loss 0.00988186\n",
            "2019-02-11 10:14:38,605 epoch 74 - iter 20/27 - loss 0.01008248\n",
            "2019-02-11 10:14:39,833 epoch 74 - iter 22/27 - loss 0.01028939\n",
            "2019-02-11 10:14:40,578 epoch 74 - iter 24/27 - loss 0.01025817\n",
            "2019-02-11 10:14:40,992 epoch 74 - iter 26/27 - loss 0.01054748\n",
            "2019-02-11 10:14:41,003 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:14:41,005 EPOCH 74 done: loss 0.0105 - lr 0.0125 - bad epochs 1\n",
            "2019-02-11 10:14:41,743 DEV  : loss 0.01659570 - f-score 0.7158 - acc 0.7158\n",
            "2019-02-11 10:14:46,409 TEST : loss 0.02729525 - f-score 0.6781 - acc 0.6781\n",
            "2019-02-11 10:15:04,344 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:15:04,886 epoch 75 - iter 0/27 - loss 0.01191024\n",
            "2019-02-11 10:15:05,830 epoch 75 - iter 2/27 - loss 0.00982545\n",
            "2019-02-11 10:15:06,462 epoch 75 - iter 4/27 - loss 0.00962918\n",
            "2019-02-11 10:15:07,254 epoch 75 - iter 6/27 - loss 0.00984834\n",
            "2019-02-11 10:15:07,964 epoch 75 - iter 8/27 - loss 0.00963202\n",
            "2019-02-11 10:15:08,691 epoch 75 - iter 10/27 - loss 0.00930823\n",
            "2019-02-11 10:15:09,930 epoch 75 - iter 12/27 - loss 0.00959100\n",
            "2019-02-11 10:15:10,638 epoch 75 - iter 14/27 - loss 0.00968032\n",
            "2019-02-11 10:15:11,625 epoch 75 - iter 16/27 - loss 0.00987789\n",
            "2019-02-11 10:15:12,496 epoch 75 - iter 18/27 - loss 0.01026850\n",
            "2019-02-11 10:15:13,433 epoch 75 - iter 20/27 - loss 0.01036856\n",
            "2019-02-11 10:15:14,191 epoch 75 - iter 22/27 - loss 0.01063969\n",
            "2019-02-11 10:15:15,419 epoch 75 - iter 24/27 - loss 0.01062061\n",
            "2019-02-11 10:15:16,013 epoch 75 - iter 26/27 - loss 0.01069098\n",
            "2019-02-11 10:15:16,026 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:15:16,027 EPOCH 75 done: loss 0.0107 - lr 0.0125 - bad epochs 0\n",
            "2019-02-11 10:15:16,732 DEV  : loss 0.01666993 - f-score 0.7158 - acc 0.7158\n",
            "2019-02-11 10:15:21,346 TEST : loss 0.02793451 - f-score 0.6824 - acc 0.6824\n",
            "2019-02-11 10:15:21,349 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:15:21,610 epoch 76 - iter 0/27 - loss 0.00842099\n",
            "2019-02-11 10:15:22,309 epoch 76 - iter 2/27 - loss 0.01048530\n",
            "2019-02-11 10:15:23,099 epoch 76 - iter 4/27 - loss 0.01071400\n",
            "2019-02-11 10:15:24,315 epoch 76 - iter 6/27 - loss 0.01099399\n",
            "2019-02-11 10:15:25,147 epoch 76 - iter 8/27 - loss 0.01076850\n",
            "2019-02-11 10:15:25,779 epoch 76 - iter 10/27 - loss 0.01009134\n",
            "2019-02-11 10:15:27,193 epoch 76 - iter 12/27 - loss 0.00988985\n",
            "2019-02-11 10:15:27,864 epoch 76 - iter 14/27 - loss 0.00982492\n",
            "2019-02-11 10:15:28,594 epoch 76 - iter 16/27 - loss 0.00993674\n",
            "2019-02-11 10:15:29,328 epoch 76 - iter 18/27 - loss 0.00961174\n",
            "2019-02-11 10:15:30,255 epoch 76 - iter 20/27 - loss 0.00963356\n",
            "2019-02-11 10:15:30,951 epoch 76 - iter 22/27 - loss 0.00981173\n",
            "2019-02-11 10:15:31,697 epoch 76 - iter 24/27 - loss 0.00997065\n",
            "2019-02-11 10:15:32,376 epoch 76 - iter 26/27 - loss 0.01058801\n",
            "2019-02-11 10:15:32,390 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:15:32,392 EPOCH 76 done: loss 0.0106 - lr 0.0125 - bad epochs 1\n",
            "2019-02-11 10:15:33,118 DEV  : loss 0.01690721 - f-score 0.7053 - acc 0.7053\n",
            "2019-02-11 10:15:37,711 TEST : loss 0.02777359 - f-score 0.6824 - acc 0.6824\n",
            "2019-02-11 10:15:37,714 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:15:38,131 epoch 77 - iter 0/27 - loss 0.00836998\n",
            "2019-02-11 10:15:38,854 epoch 77 - iter 2/27 - loss 0.00885110\n",
            "2019-02-11 10:15:39,609 epoch 77 - iter 4/27 - loss 0.00874657\n",
            "2019-02-11 10:15:40,415 epoch 77 - iter 6/27 - loss 0.00933659\n",
            "2019-02-11 10:15:41,197 epoch 77 - iter 8/27 - loss 0.00921646\n",
            "2019-02-11 10:15:41,913 epoch 77 - iter 10/27 - loss 0.00942956\n",
            "2019-02-11 10:15:42,549 epoch 77 - iter 12/27 - loss 0.00957320\n",
            "2019-02-11 10:15:43,302 epoch 77 - iter 14/27 - loss 0.00965938\n",
            "2019-02-11 10:15:44,268 epoch 77 - iter 16/27 - loss 0.00989325\n",
            "2019-02-11 10:15:45,502 epoch 77 - iter 18/27 - loss 0.01013165\n",
            "2019-02-11 10:15:46,195 epoch 77 - iter 20/27 - loss 0.00985682\n",
            "2019-02-11 10:15:47,239 epoch 77 - iter 22/27 - loss 0.00997494\n",
            "2019-02-11 10:15:48,091 epoch 77 - iter 24/27 - loss 0.01005887\n",
            "2019-02-11 10:15:49,070 epoch 77 - iter 26/27 - loss 0.01036163\n",
            "2019-02-11 10:15:49,081 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:15:49,082 EPOCH 77 done: loss 0.0104 - lr 0.0125 - bad epochs 2\n",
            "2019-02-11 10:15:49,802 DEV  : loss 0.01839883 - f-score 0.7158 - acc 0.7158\n",
            "2019-02-11 10:15:54,444 TEST : loss 0.02783993 - f-score 0.6717 - acc 0.6717\n",
            "2019-02-11 10:16:11,651 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:16:12,195 epoch 78 - iter 0/27 - loss 0.01357456\n",
            "2019-02-11 10:16:13,056 epoch 78 - iter 2/27 - loss 0.01032820\n",
            "2019-02-11 10:16:13,826 epoch 78 - iter 4/27 - loss 0.00938197\n",
            "2019-02-11 10:16:14,505 epoch 78 - iter 6/27 - loss 0.00981911\n",
            "2019-02-11 10:16:15,323 epoch 78 - iter 8/27 - loss 0.00982877\n",
            "2019-02-11 10:16:16,136 epoch 78 - iter 10/27 - loss 0.01009216\n",
            "2019-02-11 10:16:16,955 epoch 78 - iter 12/27 - loss 0.00999379\n",
            "2019-02-11 10:16:18,620 epoch 78 - iter 14/27 - loss 0.01023146\n",
            "2019-02-11 10:16:19,472 epoch 78 - iter 16/27 - loss 0.00966027\n",
            "2019-02-11 10:16:20,283 epoch 78 - iter 18/27 - loss 0.00992705\n",
            "2019-02-11 10:16:20,975 epoch 78 - iter 20/27 - loss 0.01029641\n",
            "2019-02-11 10:16:21,588 epoch 78 - iter 22/27 - loss 0.01069948\n",
            "2019-02-11 10:16:22,316 epoch 78 - iter 24/27 - loss 0.01042868\n",
            "2019-02-11 10:16:22,979 epoch 78 - iter 26/27 - loss 0.01108764\n",
            "2019-02-11 10:16:22,990 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:16:22,992 EPOCH 78 done: loss 0.0111 - lr 0.0125 - bad epochs 0\n",
            "2019-02-11 10:16:23,712 DEV  : loss 0.01712353 - f-score 0.7263 - acc 0.7263\n",
            "2019-02-11 10:16:28,247 TEST : loss 0.02846095 - f-score 0.6845 - acc 0.6845\n",
            "2019-02-11 10:16:28,250 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:16:28,674 epoch 79 - iter 0/27 - loss 0.00983803\n",
            "2019-02-11 10:16:29,574 epoch 79 - iter 2/27 - loss 0.01299120\n",
            "2019-02-11 10:16:30,421 epoch 79 - iter 4/27 - loss 0.01240876\n",
            "2019-02-11 10:16:31,044 epoch 79 - iter 6/27 - loss 0.01196417\n",
            "2019-02-11 10:16:32,008 epoch 79 - iter 8/27 - loss 0.01127235\n",
            "2019-02-11 10:16:33,209 epoch 79 - iter 10/27 - loss 0.01065767\n",
            "2019-02-11 10:16:33,874 epoch 79 - iter 12/27 - loss 0.01018626\n",
            "2019-02-11 10:16:34,568 epoch 79 - iter 14/27 - loss 0.01061639\n",
            "2019-02-11 10:16:35,739 epoch 79 - iter 16/27 - loss 0.01074704\n",
            "2019-02-11 10:16:36,514 epoch 79 - iter 18/27 - loss 0.01049783\n",
            "2019-02-11 10:16:37,219 epoch 79 - iter 20/27 - loss 0.01084398\n",
            "2019-02-11 10:16:37,932 epoch 79 - iter 22/27 - loss 0.01093880\n",
            "2019-02-11 10:16:38,784 epoch 79 - iter 24/27 - loss 0.01074748\n",
            "2019-02-11 10:16:39,525 epoch 79 - iter 26/27 - loss 0.01077690\n",
            "2019-02-11 10:16:39,538 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:16:39,539 EPOCH 79 done: loss 0.0108 - lr 0.0125 - bad epochs 1\n",
            "2019-02-11 10:16:40,252 DEV  : loss 0.01737165 - f-score 0.7263 - acc 0.7263\n",
            "2019-02-11 10:16:45,105 TEST : loss 0.02807749 - f-score 0.6760 - acc 0.6760\n",
            "2019-02-11 10:16:45,108 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:16:45,526 epoch 80 - iter 0/27 - loss 0.01184240\n",
            "2019-02-11 10:16:46,806 epoch 80 - iter 2/27 - loss 0.01191557\n",
            "2019-02-11 10:16:47,464 epoch 80 - iter 4/27 - loss 0.01260927\n",
            "2019-02-11 10:16:48,246 epoch 80 - iter 6/27 - loss 0.01184725\n",
            "2019-02-11 10:16:49,512 epoch 80 - iter 8/27 - loss 0.01083738\n",
            "2019-02-11 10:16:50,380 epoch 80 - iter 10/27 - loss 0.01052510\n",
            "2019-02-11 10:16:51,162 epoch 80 - iter 12/27 - loss 0.01015165\n",
            "2019-02-11 10:16:51,931 epoch 80 - iter 14/27 - loss 0.00991880\n",
            "2019-02-11 10:16:52,792 epoch 80 - iter 16/27 - loss 0.00953045\n",
            "2019-02-11 10:16:53,641 epoch 80 - iter 18/27 - loss 0.00981924\n",
            "2019-02-11 10:16:54,637 epoch 80 - iter 20/27 - loss 0.00975368\n",
            "2019-02-11 10:16:55,379 epoch 80 - iter 22/27 - loss 0.00981943\n",
            "2019-02-11 10:16:56,181 epoch 80 - iter 24/27 - loss 0.01017432\n",
            "2019-02-11 10:16:56,653 epoch 80 - iter 26/27 - loss 0.01036227\n",
            "2019-02-11 10:16:56,664 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:16:56,666 EPOCH 80 done: loss 0.0104 - lr 0.0125 - bad epochs 2\n",
            "2019-02-11 10:16:57,393 DEV  : loss 0.01844784 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 10:17:02,019 TEST : loss 0.02784070 - f-score 0.6717 - acc 0.6717\n",
            "2019-02-11 10:17:02,022 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:17:02,281 epoch 81 - iter 0/27 - loss 0.00713311\n",
            "2019-02-11 10:17:03,102 epoch 81 - iter 2/27 - loss 0.00844357\n",
            "2019-02-11 10:17:03,717 epoch 81 - iter 4/27 - loss 0.00886061\n",
            "2019-02-11 10:17:04,492 epoch 81 - iter 6/27 - loss 0.00885842\n",
            "2019-02-11 10:17:05,752 epoch 81 - iter 8/27 - loss 0.01024248\n",
            "2019-02-11 10:17:06,526 epoch 81 - iter 10/27 - loss 0.00984768\n",
            "2019-02-11 10:17:07,683 epoch 81 - iter 12/27 - loss 0.00981682\n",
            "2019-02-11 10:17:08,236 epoch 81 - iter 14/27 - loss 0.00956324\n",
            "2019-02-11 10:17:08,913 epoch 81 - iter 16/27 - loss 0.00983340\n",
            "2019-02-11 10:17:09,720 epoch 81 - iter 18/27 - loss 0.01007793\n",
            "2019-02-11 10:17:10,519 epoch 81 - iter 20/27 - loss 0.01007007\n",
            "2019-02-11 10:17:11,599 epoch 81 - iter 22/27 - loss 0.01003285\n",
            "2019-02-11 10:17:12,359 epoch 81 - iter 24/27 - loss 0.01005105\n",
            "2019-02-11 10:17:12,963 epoch 81 - iter 26/27 - loss 0.01014368\n",
            "2019-02-11 10:17:12,974 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:17:12,975 EPOCH 81 done: loss 0.0101 - lr 0.0125 - bad epochs 3\n",
            "2019-02-11 10:17:13,697 DEV  : loss 0.01651240 - f-score 0.7684 - acc 0.7684\n",
            "2019-02-11 10:17:18,373 TEST : loss 0.02952607 - f-score 0.6888 - acc 0.6888\n",
            "2019-02-11 10:17:36,439 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:17:36,813 epoch 82 - iter 0/27 - loss 0.00735506\n",
            "2019-02-11 10:17:37,997 epoch 82 - iter 2/27 - loss 0.00958541\n",
            "2019-02-11 10:17:38,749 epoch 82 - iter 4/27 - loss 0.00956923\n",
            "2019-02-11 10:17:39,566 epoch 82 - iter 6/27 - loss 0.01083835\n",
            "2019-02-11 10:17:40,329 epoch 82 - iter 8/27 - loss 0.01011502\n",
            "2019-02-11 10:17:41,018 epoch 82 - iter 10/27 - loss 0.01047725\n",
            "2019-02-11 10:17:41,763 epoch 82 - iter 12/27 - loss 0.01050124\n",
            "2019-02-11 10:17:42,360 epoch 82 - iter 14/27 - loss 0.01023585\n",
            "2019-02-11 10:17:43,624 epoch 82 - iter 16/27 - loss 0.01005457\n",
            "2019-02-11 10:17:44,842 epoch 82 - iter 18/27 - loss 0.00990555\n",
            "2019-02-11 10:17:45,571 epoch 82 - iter 20/27 - loss 0.00993841\n",
            "2019-02-11 10:17:46,427 epoch 82 - iter 22/27 - loss 0.00985795\n",
            "2019-02-11 10:17:47,332 epoch 82 - iter 24/27 - loss 0.00998922\n",
            "2019-02-11 10:17:47,810 epoch 82 - iter 26/27 - loss 0.01016888\n",
            "2019-02-11 10:17:47,822 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:17:47,823 EPOCH 82 done: loss 0.0102 - lr 0.0125 - bad epochs 0\n",
            "2019-02-11 10:17:48,544 DEV  : loss 0.01738102 - f-score 0.7158 - acc 0.7158\n",
            "2019-02-11 10:17:53,142 TEST : loss 0.02849377 - f-score 0.6760 - acc 0.6760\n",
            "2019-02-11 10:17:53,145 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:17:53,561 epoch 83 - iter 0/27 - loss 0.00822914\n",
            "2019-02-11 10:17:54,360 epoch 83 - iter 2/27 - loss 0.00736465\n",
            "2019-02-11 10:17:55,108 epoch 83 - iter 4/27 - loss 0.00924487\n",
            "2019-02-11 10:17:55,897 epoch 83 - iter 6/27 - loss 0.00982798\n",
            "2019-02-11 10:17:56,677 epoch 83 - iter 8/27 - loss 0.00971818\n",
            "2019-02-11 10:17:57,329 epoch 83 - iter 10/27 - loss 0.01028627\n",
            "2019-02-11 10:17:58,665 epoch 83 - iter 12/27 - loss 0.01055896\n",
            "2019-02-11 10:17:59,409 epoch 83 - iter 14/27 - loss 0.01035494\n",
            "2019-02-11 10:18:00,079 epoch 83 - iter 16/27 - loss 0.00996015\n",
            "2019-02-11 10:18:00,792 epoch 83 - iter 18/27 - loss 0.00998968\n",
            "2019-02-11 10:18:01,845 epoch 83 - iter 20/27 - loss 0.00979686\n",
            "2019-02-11 10:18:03,123 epoch 83 - iter 22/27 - loss 0.00980282\n",
            "2019-02-11 10:18:04,094 epoch 83 - iter 24/27 - loss 0.00972308\n",
            "2019-02-11 10:18:04,747 epoch 83 - iter 26/27 - loss 0.01004279\n",
            "2019-02-11 10:18:04,759 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:18:04,760 EPOCH 83 done: loss 0.0100 - lr 0.0125 - bad epochs 1\n",
            "2019-02-11 10:18:05,489 DEV  : loss 0.01808615 - f-score 0.6947 - acc 0.6947\n",
            "2019-02-11 10:18:10,093 TEST : loss 0.02846401 - f-score 0.6760 - acc 0.6760\n",
            "2019-02-11 10:18:27,397 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:18:28,016 epoch 84 - iter 0/27 - loss 0.00861080\n",
            "2019-02-11 10:18:28,893 epoch 84 - iter 2/27 - loss 0.01089154\n",
            "2019-02-11 10:18:29,773 epoch 84 - iter 4/27 - loss 0.01002222\n",
            "2019-02-11 10:18:31,028 epoch 84 - iter 6/27 - loss 0.01001703\n",
            "2019-02-11 10:18:31,954 epoch 84 - iter 8/27 - loss 0.01008017\n",
            "2019-02-11 10:18:32,688 epoch 84 - iter 10/27 - loss 0.00999018\n",
            "2019-02-11 10:18:33,569 epoch 84 - iter 12/27 - loss 0.00990380\n",
            "2019-02-11 10:18:34,322 epoch 84 - iter 14/27 - loss 0.00965592\n",
            "2019-02-11 10:18:35,072 epoch 84 - iter 16/27 - loss 0.00981883\n",
            "2019-02-11 10:18:36,341 epoch 84 - iter 18/27 - loss 0.00979303\n",
            "2019-02-11 10:18:37,113 epoch 84 - iter 20/27 - loss 0.00964642\n",
            "2019-02-11 10:18:37,854 epoch 84 - iter 22/27 - loss 0.00969463\n",
            "2019-02-11 10:18:38,764 epoch 84 - iter 24/27 - loss 0.00965977\n",
            "2019-02-11 10:18:39,425 epoch 84 - iter 26/27 - loss 0.01035405\n",
            "2019-02-11 10:18:39,436 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:18:39,437 EPOCH 84 done: loss 0.0104 - lr 0.0125 - bad epochs 0\n",
            "2019-02-11 10:18:40,165 DEV  : loss 0.01739520 - f-score 0.7263 - acc 0.7263\n",
            "2019-02-11 10:18:44,784 TEST : loss 0.02901635 - f-score 0.6760 - acc 0.6760\n",
            "2019-02-11 10:18:44,787 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:18:45,102 epoch 85 - iter 0/27 - loss 0.00897877\n",
            "2019-02-11 10:18:45,856 epoch 85 - iter 2/27 - loss 0.01000264\n",
            "2019-02-11 10:18:46,631 epoch 85 - iter 4/27 - loss 0.00956518\n",
            "2019-02-11 10:18:47,728 epoch 85 - iter 6/27 - loss 0.00950984\n",
            "2019-02-11 10:18:48,590 epoch 85 - iter 8/27 - loss 0.00964786\n",
            "2019-02-11 10:18:49,261 epoch 85 - iter 10/27 - loss 0.00978801\n",
            "2019-02-11 10:18:50,212 epoch 85 - iter 12/27 - loss 0.00985980\n",
            "2019-02-11 10:18:51,336 epoch 85 - iter 14/27 - loss 0.00997523\n",
            "2019-02-11 10:18:52,417 epoch 85 - iter 16/27 - loss 0.00990341\n",
            "2019-02-11 10:18:53,300 epoch 85 - iter 18/27 - loss 0.01002836\n",
            "2019-02-11 10:18:54,119 epoch 85 - iter 20/27 - loss 0.00996292\n",
            "2019-02-11 10:18:54,864 epoch 85 - iter 22/27 - loss 0.00985908\n",
            "2019-02-11 10:18:55,467 epoch 85 - iter 24/27 - loss 0.00981577\n",
            "2019-02-11 10:18:55,937 epoch 85 - iter 26/27 - loss 0.00999464\n",
            "2019-02-11 10:18:55,948 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:18:55,954 EPOCH 85 done: loss 0.0100 - lr 0.0125 - bad epochs 1\n",
            "2019-02-11 10:18:56,666 DEV  : loss 0.01639700 - f-score 0.7474 - acc 0.7474\n",
            "2019-02-11 10:19:01,282 TEST : loss 0.02975036 - f-score 0.6781 - acc 0.6781\n",
            "2019-02-11 10:19:18,525 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:19:19,224 epoch 86 - iter 0/27 - loss 0.00741743\n",
            "2019-02-11 10:19:20,158 epoch 86 - iter 2/27 - loss 0.01087503\n",
            "2019-02-11 10:19:20,988 epoch 86 - iter 4/27 - loss 0.01085353\n",
            "2019-02-11 10:19:22,219 epoch 86 - iter 6/27 - loss 0.01062419\n",
            "2019-02-11 10:19:23,206 epoch 86 - iter 8/27 - loss 0.01055356\n",
            "2019-02-11 10:19:24,300 epoch 86 - iter 10/27 - loss 0.00993029\n",
            "2019-02-11 10:19:25,075 epoch 86 - iter 12/27 - loss 0.00965781\n",
            "2019-02-11 10:19:25,704 epoch 86 - iter 14/27 - loss 0.00971740\n",
            "2019-02-11 10:19:26,536 epoch 86 - iter 16/27 - loss 0.00963998\n",
            "2019-02-11 10:19:27,319 epoch 86 - iter 18/27 - loss 0.00974715\n",
            "2019-02-11 10:19:28,125 epoch 86 - iter 20/27 - loss 0.00972939\n",
            "2019-02-11 10:19:28,721 epoch 86 - iter 22/27 - loss 0.00960016\n",
            "2019-02-11 10:19:29,538 epoch 86 - iter 24/27 - loss 0.00967516\n",
            "2019-02-11 10:19:29,993 epoch 86 - iter 26/27 - loss 0.00984101\n",
            "2019-02-11 10:19:30,004 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:19:30,005 EPOCH 86 done: loss 0.0098 - lr 0.0125 - bad epochs 0\n",
            "2019-02-11 10:19:30,724 DEV  : loss 0.01630718 - f-score 0.7579 - acc 0.7579\n",
            "2019-02-11 10:19:35,320 TEST : loss 0.03032060 - f-score 0.6738 - acc 0.6738\n",
            "2019-02-11 10:19:52,611 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:19:53,525 epoch 87 - iter 0/27 - loss 0.00717371\n",
            "2019-02-11 10:19:54,446 epoch 87 - iter 2/27 - loss 0.00719156\n",
            "2019-02-11 10:19:55,353 epoch 87 - iter 4/27 - loss 0.00840326\n",
            "2019-02-11 10:19:57,051 epoch 87 - iter 6/27 - loss 0.00834373\n",
            "2019-02-11 10:19:57,836 epoch 87 - iter 8/27 - loss 0.00871435\n",
            "2019-02-11 10:19:58,685 epoch 87 - iter 10/27 - loss 0.00902695\n",
            "2019-02-11 10:19:59,426 epoch 87 - iter 12/27 - loss 0.00885294\n",
            "2019-02-11 10:20:00,212 epoch 87 - iter 14/27 - loss 0.00888097\n",
            "2019-02-11 10:20:01,179 epoch 87 - iter 16/27 - loss 0.00935201\n",
            "2019-02-11 10:20:01,908 epoch 87 - iter 18/27 - loss 0.00911110\n",
            "2019-02-11 10:20:02,595 epoch 87 - iter 20/27 - loss 0.00911082\n",
            "2019-02-11 10:20:03,385 epoch 87 - iter 22/27 - loss 0.00930899\n",
            "2019-02-11 10:20:04,117 epoch 87 - iter 24/27 - loss 0.00987627\n",
            "2019-02-11 10:20:04,696 epoch 87 - iter 26/27 - loss 0.01032995\n",
            "2019-02-11 10:20:04,708 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:20:04,710 EPOCH 87 done: loss 0.0103 - lr 0.0125 - bad epochs 0\n",
            "2019-02-11 10:20:05,426 DEV  : loss 0.01644115 - f-score 0.7158 - acc 0.7158\n",
            "2019-02-11 10:20:10,034 TEST : loss 0.03031643 - f-score 0.6760 - acc 0.6760\n",
            "2019-02-11 10:20:10,037 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:20:10,465 epoch 88 - iter 0/27 - loss 0.00595906\n",
            "2019-02-11 10:20:11,586 epoch 88 - iter 2/27 - loss 0.00866232\n",
            "2019-02-11 10:20:12,497 epoch 88 - iter 4/27 - loss 0.00952472\n",
            "2019-02-11 10:20:13,126 epoch 88 - iter 6/27 - loss 0.00967565\n",
            "2019-02-11 10:20:14,305 epoch 88 - iter 8/27 - loss 0.00932313\n",
            "2019-02-11 10:20:15,169 epoch 88 - iter 10/27 - loss 0.00977755\n",
            "2019-02-11 10:20:16,089 epoch 88 - iter 12/27 - loss 0.00922823\n",
            "2019-02-11 10:20:16,749 epoch 88 - iter 14/27 - loss 0.00878295\n",
            "2019-02-11 10:20:17,411 epoch 88 - iter 16/27 - loss 0.00955341\n",
            "2019-02-11 10:20:18,353 epoch 88 - iter 18/27 - loss 0.00953141\n",
            "2019-02-11 10:20:19,130 epoch 88 - iter 20/27 - loss 0.00965655\n",
            "2019-02-11 10:20:20,084 epoch 88 - iter 22/27 - loss 0.00961423\n",
            "2019-02-11 10:20:20,698 epoch 88 - iter 24/27 - loss 0.00974712\n",
            "2019-02-11 10:20:21,279 epoch 88 - iter 26/27 - loss 0.01096669\n",
            "2019-02-11 10:20:21,291 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:20:21,292 EPOCH 88 done: loss 0.0110 - lr 0.0125 - bad epochs 1\n",
            "2019-02-11 10:20:22,012 DEV  : loss 0.01633541 - f-score 0.7368 - acc 0.7368\n",
            "2019-02-11 10:20:26,602 TEST : loss 0.03041471 - f-score 0.6760 - acc 0.6760\n",
            "2019-02-11 10:20:26,608 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:20:26,939 epoch 89 - iter 0/27 - loss 0.00558114\n",
            "2019-02-11 10:20:27,667 epoch 89 - iter 2/27 - loss 0.00894967\n",
            "2019-02-11 10:20:28,790 epoch 89 - iter 4/27 - loss 0.00881461\n",
            "2019-02-11 10:20:29,618 epoch 89 - iter 6/27 - loss 0.00911451\n",
            "2019-02-11 10:20:30,393 epoch 89 - iter 8/27 - loss 0.00878037\n",
            "2019-02-11 10:20:31,146 epoch 89 - iter 10/27 - loss 0.00856487\n",
            "2019-02-11 10:20:31,920 epoch 89 - iter 12/27 - loss 0.00855247\n",
            "2019-02-11 10:20:32,690 epoch 89 - iter 14/27 - loss 0.00851559\n",
            "2019-02-11 10:20:33,489 epoch 89 - iter 16/27 - loss 0.00850083\n",
            "2019-02-11 10:20:34,348 epoch 89 - iter 18/27 - loss 0.00868809\n",
            "2019-02-11 10:20:35,537 epoch 89 - iter 20/27 - loss 0.00900611\n",
            "2019-02-11 10:20:36,805 epoch 89 - iter 22/27 - loss 0.00909214\n",
            "2019-02-11 10:20:37,406 epoch 89 - iter 24/27 - loss 0.00882625\n",
            "2019-02-11 10:20:37,932 epoch 89 - iter 26/27 - loss 0.00923000\n",
            "2019-02-11 10:20:37,943 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:20:37,944 EPOCH 89 done: loss 0.0092 - lr 0.0125 - bad epochs 2\n",
            "2019-02-11 10:20:38,711 DEV  : loss 0.01777793 - f-score 0.7263 - acc 0.7263\n",
            "2019-02-11 10:20:43,513 TEST : loss 0.02947478 - f-score 0.6674 - acc 0.6674\n",
            "2019-02-11 10:21:00,988 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:21:01,707 epoch 90 - iter 0/27 - loss 0.01128838\n",
            "2019-02-11 10:21:02,551 epoch 90 - iter 2/27 - loss 0.01086703\n",
            "2019-02-11 10:21:03,226 epoch 90 - iter 4/27 - loss 0.01024562\n",
            "2019-02-11 10:21:04,025 epoch 90 - iter 6/27 - loss 0.01005729\n",
            "2019-02-11 10:21:05,016 epoch 90 - iter 8/27 - loss 0.00978677\n",
            "2019-02-11 10:21:05,730 epoch 90 - iter 10/27 - loss 0.00983113\n",
            "2019-02-11 10:21:06,380 epoch 90 - iter 12/27 - loss 0.00932410\n",
            "2019-02-11 10:21:07,106 epoch 90 - iter 14/27 - loss 0.00946025\n",
            "2019-02-11 10:21:08,208 epoch 90 - iter 16/27 - loss 0.00923390\n",
            "2019-02-11 10:21:08,951 epoch 90 - iter 18/27 - loss 0.00886701\n",
            "2019-02-11 10:21:09,862 epoch 90 - iter 20/27 - loss 0.00914243\n",
            "2019-02-11 10:21:10,716 epoch 90 - iter 22/27 - loss 0.00884123\n",
            "2019-02-11 10:21:12,048 epoch 90 - iter 24/27 - loss 0.00879044\n",
            "2019-02-11 10:21:12,626 epoch 90 - iter 26/27 - loss 0.00885934\n",
            "2019-02-11 10:21:12,637 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:21:12,638 EPOCH 90 done: loss 0.0089 - lr 0.0125 - bad epochs 0\n",
            "2019-02-11 10:21:13,360 DEV  : loss 0.01807166 - f-score 0.7158 - acc 0.7158\n",
            "2019-02-11 10:21:17,930 TEST : loss 0.02968613 - f-score 0.6738 - acc 0.6738\n",
            "2019-02-11 10:21:35,071 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:21:35,821 epoch 91 - iter 0/27 - loss 0.00640349\n",
            "2019-02-11 10:21:36,811 epoch 91 - iter 2/27 - loss 0.00807111\n",
            "2019-02-11 10:21:37,503 epoch 91 - iter 4/27 - loss 0.00770070\n",
            "2019-02-11 10:21:38,844 epoch 91 - iter 6/27 - loss 0.00855983\n",
            "2019-02-11 10:21:39,650 epoch 91 - iter 8/27 - loss 0.00830131\n",
            "2019-02-11 10:21:40,408 epoch 91 - iter 10/27 - loss 0.00879483\n",
            "2019-02-11 10:21:41,316 epoch 91 - iter 12/27 - loss 0.00891142\n",
            "2019-02-11 10:21:42,118 epoch 91 - iter 14/27 - loss 0.00910462\n",
            "2019-02-11 10:21:42,901 epoch 91 - iter 16/27 - loss 0.00954626\n",
            "2019-02-11 10:21:43,929 epoch 91 - iter 18/27 - loss 0.00937229\n",
            "2019-02-11 10:21:44,656 epoch 91 - iter 20/27 - loss 0.00893584\n",
            "2019-02-11 10:21:45,520 epoch 91 - iter 22/27 - loss 0.00880944\n",
            "2019-02-11 10:21:46,620 epoch 91 - iter 24/27 - loss 0.00891174\n",
            "2019-02-11 10:21:47,205 epoch 91 - iter 26/27 - loss 0.00907934\n",
            "2019-02-11 10:21:47,216 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:21:47,218 EPOCH 91 done: loss 0.0091 - lr 0.0125 - bad epochs 0\n",
            "2019-02-11 10:21:47,941 DEV  : loss 0.01661372 - f-score 0.7474 - acc 0.7474\n",
            "2019-02-11 10:21:52,530 TEST : loss 0.03146503 - f-score 0.6803 - acc 0.6803\n",
            "2019-02-11 10:21:52,533 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:21:53,321 epoch 92 - iter 0/27 - loss 0.00405145\n",
            "2019-02-11 10:21:54,188 epoch 92 - iter 2/27 - loss 0.00688955\n",
            "2019-02-11 10:21:54,944 epoch 92 - iter 4/27 - loss 0.00792545\n",
            "2019-02-11 10:21:56,173 epoch 92 - iter 6/27 - loss 0.00863426\n",
            "2019-02-11 10:21:56,751 epoch 92 - iter 8/27 - loss 0.00823959\n",
            "2019-02-11 10:21:57,678 epoch 92 - iter 10/27 - loss 0.00848313\n",
            "2019-02-11 10:21:58,581 epoch 92 - iter 12/27 - loss 0.00858747\n",
            "2019-02-11 10:21:59,339 epoch 92 - iter 14/27 - loss 0.00849282\n",
            "2019-02-11 10:22:00,223 epoch 92 - iter 16/27 - loss 0.00893422\n",
            "2019-02-11 10:22:00,959 epoch 92 - iter 18/27 - loss 0.00884278\n",
            "2019-02-11 10:22:01,571 epoch 92 - iter 20/27 - loss 0.00970624\n",
            "2019-02-11 10:22:02,341 epoch 92 - iter 22/27 - loss 0.00969036\n",
            "2019-02-11 10:22:03,062 epoch 92 - iter 24/27 - loss 0.00969726\n",
            "2019-02-11 10:22:03,679 epoch 92 - iter 26/27 - loss 0.00971414\n",
            "2019-02-11 10:22:03,692 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:22:03,693 EPOCH 92 done: loss 0.0097 - lr 0.0125 - bad epochs 1\n",
            "2019-02-11 10:22:04,421 DEV  : loss 0.01752393 - f-score 0.7368 - acc 0.7368\n",
            "2019-02-11 10:22:09,052 TEST : loss 0.03039699 - f-score 0.6652 - acc 0.6652\n",
            "2019-02-11 10:22:09,055 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:22:09,521 epoch 93 - iter 0/27 - loss 0.01394181\n",
            "2019-02-11 10:22:10,554 epoch 93 - iter 2/27 - loss 0.01280543\n",
            "2019-02-11 10:22:11,750 epoch 93 - iter 4/27 - loss 0.01144045\n",
            "2019-02-11 10:22:12,599 epoch 93 - iter 6/27 - loss 0.00933933\n",
            "2019-02-11 10:22:13,477 epoch 93 - iter 8/27 - loss 0.00970441\n",
            "2019-02-11 10:22:14,235 epoch 93 - iter 10/27 - loss 0.00958293\n",
            "2019-02-11 10:22:15,075 epoch 93 - iter 12/27 - loss 0.01012264\n",
            "2019-02-11 10:22:15,846 epoch 93 - iter 14/27 - loss 0.00994937\n",
            "2019-02-11 10:22:16,717 epoch 93 - iter 16/27 - loss 0.00967604\n",
            "2019-02-11 10:22:17,339 epoch 93 - iter 18/27 - loss 0.00992783\n",
            "2019-02-11 10:22:18,108 epoch 93 - iter 20/27 - loss 0.00973376\n",
            "2019-02-11 10:22:18,875 epoch 93 - iter 22/27 - loss 0.00986765\n",
            "2019-02-11 10:22:19,560 epoch 93 - iter 24/27 - loss 0.00962927\n",
            "2019-02-11 10:22:20,477 epoch 93 - iter 26/27 - loss 0.00977724\n",
            "2019-02-11 10:22:20,489 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:22:20,490 EPOCH 93 done: loss 0.0098 - lr 0.0125 - bad epochs 2\n",
            "2019-02-11 10:22:21,208 DEV  : loss 0.01661381 - f-score 0.7368 - acc 0.7368\n",
            "2019-02-11 10:22:25,935 TEST : loss 0.03168552 - f-score 0.6760 - acc 0.6760\n",
            "2019-02-11 10:22:25,937 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:22:26,307 epoch 94 - iter 0/27 - loss 0.01552465\n",
            "2019-02-11 10:22:27,577 epoch 94 - iter 2/27 - loss 0.00897532\n",
            "2019-02-11 10:22:28,355 epoch 94 - iter 4/27 - loss 0.00931331\n",
            "2019-02-11 10:22:29,128 epoch 94 - iter 6/27 - loss 0.00887735\n",
            "2019-02-11 10:22:30,059 epoch 94 - iter 8/27 - loss 0.00937053\n",
            "2019-02-11 10:22:31,318 epoch 94 - iter 10/27 - loss 0.00947485\n",
            "2019-02-11 10:22:32,250 epoch 94 - iter 12/27 - loss 0.00970775\n",
            "2019-02-11 10:22:33,105 epoch 94 - iter 14/27 - loss 0.00962099\n",
            "2019-02-11 10:22:33,982 epoch 94 - iter 16/27 - loss 0.01001996\n",
            "2019-02-11 10:22:34,844 epoch 94 - iter 18/27 - loss 0.00982650\n",
            "2019-02-11 10:22:36,023 epoch 94 - iter 20/27 - loss 0.00960015\n",
            "2019-02-11 10:22:36,683 epoch 94 - iter 22/27 - loss 0.00958541\n",
            "2019-02-11 10:22:37,302 epoch 94 - iter 24/27 - loss 0.00971108\n",
            "2019-02-11 10:22:37,936 epoch 94 - iter 26/27 - loss 0.00980583\n",
            "2019-02-11 10:22:37,948 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:22:37,949 EPOCH 94 done: loss 0.0098 - lr 0.0125 - bad epochs 3\n",
            "2019-02-11 10:22:38,667 DEV  : loss 0.01695760 - f-score 0.7474 - acc 0.7474\n",
            "2019-02-11 10:22:43,259 TEST : loss 0.03130500 - f-score 0.6781 - acc 0.6781\n",
            "Epoch    93: reducing learning rate of group 0 to 6.2500e-03.\n",
            "2019-02-11 10:22:43,265 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:22:43,865 epoch 95 - iter 0/27 - loss 0.00632108\n",
            "2019-02-11 10:22:44,729 epoch 95 - iter 2/27 - loss 0.00757346\n",
            "2019-02-11 10:22:45,930 epoch 95 - iter 4/27 - loss 0.00875071\n",
            "2019-02-11 10:22:46,624 epoch 95 - iter 6/27 - loss 0.00762003\n",
            "2019-02-11 10:22:47,388 epoch 95 - iter 8/27 - loss 0.00781748\n",
            "2019-02-11 10:22:48,347 epoch 95 - iter 10/27 - loss 0.00903269\n",
            "2019-02-11 10:22:49,045 epoch 95 - iter 12/27 - loss 0.00846342\n",
            "2019-02-11 10:22:49,905 epoch 95 - iter 14/27 - loss 0.00851726\n",
            "2019-02-11 10:22:51,252 epoch 95 - iter 16/27 - loss 0.00830546\n",
            "2019-02-11 10:22:51,993 epoch 95 - iter 18/27 - loss 0.00845004\n",
            "2019-02-11 10:22:52,672 epoch 95 - iter 20/27 - loss 0.00857501\n",
            "2019-02-11 10:22:53,359 epoch 95 - iter 22/27 - loss 0.00868634\n",
            "2019-02-11 10:22:54,147 epoch 95 - iter 24/27 - loss 0.00869348\n",
            "2019-02-11 10:22:54,637 epoch 95 - iter 26/27 - loss 0.00940183\n",
            "2019-02-11 10:22:54,648 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:22:54,650 EPOCH 95 done: loss 0.0094 - lr 0.0063 - bad epochs 0\n",
            "2019-02-11 10:22:55,359 DEV  : loss 0.01698597 - f-score 0.7474 - acc 0.7474\n",
            "2019-02-11 10:22:59,916 TEST : loss 0.03130771 - f-score 0.6631 - acc 0.6631\n",
            "2019-02-11 10:22:59,919 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:23:00,708 epoch 96 - iter 0/27 - loss 0.00890305\n",
            "2019-02-11 10:23:01,569 epoch 96 - iter 2/27 - loss 0.01060473\n",
            "2019-02-11 10:23:02,226 epoch 96 - iter 4/27 - loss 0.00922384\n",
            "2019-02-11 10:23:03,089 epoch 96 - iter 6/27 - loss 0.00935894\n",
            "2019-02-11 10:23:03,860 epoch 96 - iter 8/27 - loss 0.00946340\n",
            "2019-02-11 10:23:04,529 epoch 96 - iter 10/27 - loss 0.00964456\n",
            "2019-02-11 10:23:05,261 epoch 96 - iter 12/27 - loss 0.00982179\n",
            "2019-02-11 10:23:06,021 epoch 96 - iter 14/27 - loss 0.00962996\n",
            "2019-02-11 10:23:06,958 epoch 96 - iter 16/27 - loss 0.00966049\n",
            "2019-02-11 10:23:07,714 epoch 96 - iter 18/27 - loss 0.00954337\n",
            "2019-02-11 10:23:08,728 epoch 96 - iter 20/27 - loss 0.00930729\n",
            "2019-02-11 10:23:09,451 epoch 96 - iter 22/27 - loss 0.00902541\n",
            "2019-02-11 10:23:10,642 epoch 96 - iter 24/27 - loss 0.00930229\n",
            "2019-02-11 10:23:11,143 epoch 96 - iter 26/27 - loss 0.00928822\n",
            "2019-02-11 10:23:11,155 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:23:11,156 EPOCH 96 done: loss 0.0093 - lr 0.0063 - bad epochs 1\n",
            "2019-02-11 10:23:11,878 DEV  : loss 0.01737863 - f-score 0.7263 - acc 0.7263\n",
            "2019-02-11 10:23:16,614 TEST : loss 0.03096506 - f-score 0.6652 - acc 0.6652\n",
            "2019-02-11 10:23:16,617 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:23:17,131 epoch 97 - iter 0/27 - loss 0.01208847\n",
            "2019-02-11 10:23:17,947 epoch 97 - iter 2/27 - loss 0.00974929\n",
            "2019-02-11 10:23:19,013 epoch 97 - iter 4/27 - loss 0.01063357\n",
            "2019-02-11 10:23:19,823 epoch 97 - iter 6/27 - loss 0.00960884\n",
            "2019-02-11 10:23:21,046 epoch 97 - iter 8/27 - loss 0.00923109\n",
            "2019-02-11 10:23:21,784 epoch 97 - iter 10/27 - loss 0.00905039\n",
            "2019-02-11 10:23:22,500 epoch 97 - iter 12/27 - loss 0.00925936\n",
            "2019-02-11 10:23:23,634 epoch 97 - iter 14/27 - loss 0.00914894\n",
            "2019-02-11 10:23:24,340 epoch 97 - iter 16/27 - loss 0.00892917\n",
            "2019-02-11 10:23:25,084 epoch 97 - iter 18/27 - loss 0.00911613\n",
            "2019-02-11 10:23:25,952 epoch 97 - iter 20/27 - loss 0.00896831\n",
            "2019-02-11 10:23:26,713 epoch 97 - iter 22/27 - loss 0.00890153\n",
            "2019-02-11 10:23:27,540 epoch 97 - iter 24/27 - loss 0.00888434\n",
            "2019-02-11 10:23:28,068 epoch 97 - iter 26/27 - loss 0.00884043\n",
            "2019-02-11 10:23:28,079 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:23:28,080 EPOCH 97 done: loss 0.0088 - lr 0.0063 - bad epochs 2\n",
            "2019-02-11 10:23:28,796 DEV  : loss 0.01677432 - f-score 0.7368 - acc 0.7368\n",
            "2019-02-11 10:23:33,362 TEST : loss 0.03166955 - f-score 0.6717 - acc 0.6717\n",
            "2019-02-11 10:23:50,504 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:23:50,858 epoch 98 - iter 0/27 - loss 0.00697470\n",
            "2019-02-11 10:23:51,585 epoch 98 - iter 2/27 - loss 0.00842850\n",
            "2019-02-11 10:23:52,379 epoch 98 - iter 4/27 - loss 0.00898587\n",
            "2019-02-11 10:23:53,242 epoch 98 - iter 6/27 - loss 0.00881501\n",
            "2019-02-11 10:23:54,454 epoch 98 - iter 8/27 - loss 0.00926064\n",
            "2019-02-11 10:23:55,404 epoch 98 - iter 10/27 - loss 0.00926747\n",
            "2019-02-11 10:23:56,429 epoch 98 - iter 12/27 - loss 0.00873939\n",
            "2019-02-11 10:23:57,281 epoch 98 - iter 14/27 - loss 0.00859756\n",
            "2019-02-11 10:23:58,047 epoch 98 - iter 16/27 - loss 0.00865207\n",
            "2019-02-11 10:23:58,718 epoch 98 - iter 18/27 - loss 0.00860037\n",
            "2019-02-11 10:23:59,548 epoch 98 - iter 20/27 - loss 0.00880312\n",
            "2019-02-11 10:24:00,435 epoch 98 - iter 22/27 - loss 0.00889610\n",
            "2019-02-11 10:24:01,629 epoch 98 - iter 24/27 - loss 0.00894495\n",
            "2019-02-11 10:24:02,189 epoch 98 - iter 26/27 - loss 0.00899461\n",
            "2019-02-11 10:24:02,200 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:24:02,202 EPOCH 98 done: loss 0.0090 - lr 0.0063 - bad epochs 0\n",
            "2019-02-11 10:24:02,919 DEV  : loss 0.01718080 - f-score 0.7368 - acc 0.7368\n",
            "2019-02-11 10:24:07,684 TEST : loss 0.03123854 - f-score 0.6652 - acc 0.6652\n",
            "2019-02-11 10:24:07,687 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:24:08,094 epoch 99 - iter 0/27 - loss 0.00976670\n",
            "2019-02-11 10:24:09,395 epoch 99 - iter 2/27 - loss 0.00928266\n",
            "2019-02-11 10:24:10,624 epoch 99 - iter 4/27 - loss 0.00940488\n",
            "2019-02-11 10:24:11,242 epoch 99 - iter 6/27 - loss 0.00974315\n",
            "2019-02-11 10:24:12,042 epoch 99 - iter 8/27 - loss 0.00969682\n",
            "2019-02-11 10:24:12,884 epoch 99 - iter 10/27 - loss 0.00921487\n",
            "2019-02-11 10:24:13,901 epoch 99 - iter 12/27 - loss 0.00924182\n",
            "2019-02-11 10:24:14,677 epoch 99 - iter 14/27 - loss 0.00932769\n",
            "2019-02-11 10:24:15,615 epoch 99 - iter 16/27 - loss 0.00927314\n",
            "2019-02-11 10:24:16,500 epoch 99 - iter 18/27 - loss 0.00915023\n",
            "2019-02-11 10:24:17,187 epoch 99 - iter 20/27 - loss 0.00908889\n",
            "2019-02-11 10:24:17,936 epoch 99 - iter 22/27 - loss 0.00909176\n",
            "2019-02-11 10:24:18,736 epoch 99 - iter 24/27 - loss 0.00915019\n",
            "2019-02-11 10:24:19,403 epoch 99 - iter 26/27 - loss 0.00982219\n",
            "2019-02-11 10:24:19,414 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:24:19,416 EPOCH 99 done: loss 0.0098 - lr 0.0063 - bad epochs 1\n",
            "2019-02-11 10:24:20,144 DEV  : loss 0.01665823 - f-score 0.7368 - acc 0.7368\n",
            "2019-02-11 10:24:24,775 TEST : loss 0.03180768 - f-score 0.6695 - acc 0.6695\n",
            "2019-02-11 10:24:24,778 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:24:25,206 epoch 100 - iter 0/27 - loss 0.01412515\n",
            "2019-02-11 10:24:26,003 epoch 100 - iter 2/27 - loss 0.00887598\n",
            "2019-02-11 10:24:26,825 epoch 100 - iter 4/27 - loss 0.00932146\n",
            "2019-02-11 10:24:27,623 epoch 100 - iter 6/27 - loss 0.00948557\n",
            "2019-02-11 10:24:28,903 epoch 100 - iter 8/27 - loss 0.00855172\n",
            "2019-02-11 10:24:29,943 epoch 100 - iter 10/27 - loss 0.00886428\n",
            "2019-02-11 10:24:31,060 epoch 100 - iter 12/27 - loss 0.00884209\n",
            "2019-02-11 10:24:31,725 epoch 100 - iter 14/27 - loss 0.00868070\n",
            "2019-02-11 10:24:32,705 epoch 100 - iter 16/27 - loss 0.00881997\n",
            "2019-02-11 10:24:33,490 epoch 100 - iter 18/27 - loss 0.00864762\n",
            "2019-02-11 10:24:34,366 epoch 100 - iter 20/27 - loss 0.00890726\n",
            "2019-02-11 10:24:35,151 epoch 100 - iter 22/27 - loss 0.00869480\n",
            "2019-02-11 10:24:36,010 epoch 100 - iter 24/27 - loss 0.00880721\n",
            "2019-02-11 10:24:36,541 epoch 100 - iter 26/27 - loss 0.00908710\n",
            "2019-02-11 10:24:36,553 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:24:36,554 EPOCH 100 done: loss 0.0091 - lr 0.0063 - bad epochs 2\n",
            "2019-02-11 10:24:37,305 DEV  : loss 0.01686654 - f-score 0.7263 - acc 0.7263\n",
            "2019-02-11 10:24:41,973 TEST : loss 0.03149960 - f-score 0.6760 - acc 0.6760\n",
            "2019-02-11 10:24:59,231 ----------------------------------------------------------------------------------------------------\n",
            "2019-02-11 10:24:59,614 Testing using best model ...\n",
            "2019-02-11 10:25:15,222 MICRO_AVG: acc 0.6717 - f1-score 0.6717\n",
            "2019-02-11 10:25:15,238 MARCO_AVG: acc 0.6629 - f1-score 0.6725\n",
            "2019-02-11 10:25:15,239 0          tp: 119 - fp: 57 - fn: 96 - tn: 119 - precision: 0.6761 - recall: 0.5535 - accuracy: 0.6087 - f1-score: 0.6087\n",
            "2019-02-11 10:25:15,241 1          tp: 194 - fp: 96 - fn: 57 - tn: 194 - precision: 0.6690 - recall: 0.7729 - accuracy: 0.7172 - f1-score: 0.7172\n",
            "2019-02-11 10:25:15,243 ----------------------------------------------------------------------------------------------------\n",
            "0.676 0.679 0.757 0.676 0.67 0.716 0.669\n",
            "[0.9416342412451363, 0.996031746031746, 0.7156308851224106]\n",
            "Mean:  0.8844322907997643 Std:  0.12140897289942568\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}