4. Sequences and Prediction/Week 3.2 - LSTM.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "S+P Week 3 Lesson 4 - LSTM.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "D1J15Vh_1Jih",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 629
        },
        "outputId": "be39dc7e-4501-4665-87b1-76ac0a9f8c99"
      },
      "source": [
        "!pip install tf-nightly-2.0-preview"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting tf-nightly-2.0-preview\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/a4/28/7728145f8dc00d3f48164a36aea7f959e0161fc2046ea71952258910b046/tf_nightly_2.0_preview-2.0.0.dev20190807-cp36-cp36m-manylinux2010_x86_64.whl (86.4MB)\n",
            "\u001b[K     |████████████████████████████████| 86.4MB 1.2MB/s \n",
            "\u001b[?25hCollecting opt-einsum>=2.3.2 (from tf-nightly-2.0-preview)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/f6/d6/44792ec668bcda7d91913c75237314e688f70415ab2acd7172c845f0b24f/opt_einsum-2.3.2.tar.gz (59kB)\n",
            "\u001b[K     |████████████████████████████████| 61kB 23.4MB/s \n",
            "\u001b[?25hRequirement already satisfied: gast>=0.2.0 in /usr/local/lib/python3.6/dist-packages (from tf-nightly-2.0-preview) (0.2.2)\n",
            "Requirement already satisfied: google-pasta>=0.1.6 in /usr/local/lib/python3.6/dist-packages (from tf-nightly-2.0-preview) (0.1.7)\n",
            "Requirement already satisfied: numpy<2.0,>=1.16.0 in /usr/local/lib/python3.6/dist-packages (from tf-nightly-2.0-preview) (1.16.4)\n",
            "Requirement already satisfied: wrapt>=1.11.1 in /usr/local/lib/python3.6/dist-packages (from tf-nightly-2.0-preview) (1.11.2)\n",
            "Requirement already satisfied: wheel>=0.26 in /usr/local/lib/python3.6/dist-packages (from tf-nightly-2.0-preview) (0.33.4)\n",
            "Requirement already satisfied: absl-py>=0.7.0 in /usr/local/lib/python3.6/dist-packages (from tf-nightly-2.0-preview) (0.7.1)\n",
            "Requirement already satisfied: protobuf>=3.6.1 in /usr/local/lib/python3.6/dist-packages (from tf-nightly-2.0-preview) (3.7.1)\n",
            "Requirement already satisfied: grpcio>=1.8.6 in /usr/local/lib/python3.6/dist-packages (from tf-nightly-2.0-preview) (1.15.0)\n",
            "Requirement already satisfied: termcolor>=1.1.0 in /usr/local/lib/python3.6/dist-packages (from tf-nightly-2.0-preview) (1.1.0)\n",
            "Requirement already satisfied: keras-preprocessing>=1.0.5 in /usr/local/lib/python3.6/dist-packages (from tf-nightly-2.0-preview) (1.1.0)\n",
            "Requirement already satisfied: six>=1.10.0 in /usr/local/lib/python3.6/dist-packages (from tf-nightly-2.0-preview) (1.12.0)\n",
            "Collecting tb-nightly<1.16.0a0,>=1.15.0a0 (from tf-nightly-2.0-preview)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/bc/88/24b5fb7280e74c7cf65bde47c171547fd02afb3840cff41bcbe9270650f5/tb_nightly-1.15.0a20190806-py3-none-any.whl (4.3MB)\n",
            "\u001b[K     |████████████████████████████████| 4.3MB 37.1MB/s \n",
            "\u001b[?25hRequirement already satisfied: astor>=0.6.0 in /usr/local/lib/python3.6/dist-packages (from tf-nightly-2.0-preview) (0.8.0)\n",
            "Requirement already satisfied: keras-applications>=1.0.8 in /usr/local/lib/python3.6/dist-packages (from tf-nightly-2.0-preview) (1.0.8)\n",
            "Collecting tensorflow-estimator-2.0-preview (from tf-nightly-2.0-preview)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/c5/e4/96fb38795ceabcc05372f6f96524ef22c375c6545e6eef078f4c15e095cc/tensorflow_estimator_2.0_preview-1.14.0.dev2019080700-py2.py3-none-any.whl (449kB)\n",
            "\u001b[K     |████████████████████████████████| 450kB 45.7MB/s \n",
            "\u001b[?25hRequirement already satisfied: setuptools in /usr/local/lib/python3.6/dist-packages (from protobuf>=3.6.1->tf-nightly-2.0-preview) (41.0.1)\n",
            "Requirement already satisfied: markdown>=2.6.8 in /usr/local/lib/python3.6/dist-packages (from tb-nightly<1.16.0a0,>=1.15.0a0->tf-nightly-2.0-preview) (3.1.1)\n",
            "Requirement already satisfied: werkzeug>=0.11.15 in /usr/local/lib/python3.6/dist-packages (from tb-nightly<1.16.0a0,>=1.15.0a0->tf-nightly-2.0-preview) (0.15.5)\n",
            "Requirement already satisfied: h5py in /usr/local/lib/python3.6/dist-packages (from keras-applications>=1.0.8->tf-nightly-2.0-preview) (2.8.0)\n",
            "Building wheels for collected packages: opt-einsum\n",
            "  Building wheel for opt-einsum (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for opt-einsum: filename=opt_einsum-2.3.2-cp36-none-any.whl size=49882 sha256=f08cba468b473375548443da4b6b3562ffce0a3337519661cce347f192ada6f4\n",
            "  Stored in directory: /root/.cache/pip/wheels/51/3e/a3/b351fae0cbf15373c2136a54a70f43fea5fe91d8168a5faaa4\n",
            "Successfully built opt-einsum\n",
            "Installing collected packages: opt-einsum, tb-nightly, tensorflow-estimator-2.0-preview, tf-nightly-2.0-preview\n",
            "Successfully installed opt-einsum-2.3.2 tb-nightly-1.15.0a20190806 tensorflow-estimator-2.0-preview-1.14.0.dev2019080700 tf-nightly-2.0-preview-2.0.0.dev20190807\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BOjujz601HcS",
        "colab_type": "code",
        "outputId": "2bfe08b3-b6f5-42de-8a52-c24889a0c3ba",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "import tensorflow as tf\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "print(tf.__version__)"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2.0.0-dev20190807\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "Zswl7jRtGzkk",
        "colab": {}
      },
      "source": [
        "def plot_series(time, series, format=\"-\", start=0, end=None):\n",
        "    plt.plot(time[start:end], series[start:end], format)\n",
        "    plt.xlabel(\"Time\")\n",
        "    plt.ylabel(\"Value\")\n",
        "    plt.grid(True)\n",
        "\n",
        "def trend(time, slope=0):\n",
        "    return slope * time\n",
        "\n",
        "def seasonal_pattern(season_time):\n",
        "    \"\"\"Just an arbitrary pattern, you can change it if you wish\"\"\"\n",
        "    return np.where(season_time < 0.4,\n",
        "                    np.cos(season_time * 2 * np.pi),\n",
        "                    1 / np.exp(3 * season_time))\n",
        "\n",
        "def seasonality(time, period, amplitude=1, phase=0):\n",
        "    \"\"\"Repeats the same pattern at each period\"\"\"\n",
        "    season_time = ((time + phase) % period) / period\n",
        "    return amplitude * seasonal_pattern(season_time)\n",
        "\n",
        "def noise(time, noise_level=1, seed=None):\n",
        "    rnd = np.random.RandomState(seed)\n",
        "    return rnd.randn(len(time)) * noise_level\n",
        "\n",
        "time = np.arange(4 * 365 + 1, dtype=\"float32\")\n",
        "baseline = 10\n",
        "series = trend(time, 0.1)  \n",
        "baseline = 10\n",
        "amplitude = 40\n",
        "slope = 0.05\n",
        "noise_level = 5\n",
        "\n",
        "# Create the series\n",
        "series = baseline + trend(time, slope) + seasonality(time, period=365, amplitude=amplitude)\n",
        "# Update with noise\n",
        "series += noise(time, noise_level, seed=42)\n",
        "\n",
        "split_time = 1000\n",
        "time_train = time[:split_time]\n",
        "x_train = series[:split_time]\n",
        "time_valid = time[split_time:]\n",
        "x_valid = series[split_time:]\n",
        "\n",
        "window_size = 20\n",
        "batch_size = 32\n",
        "shuffle_buffer_size = 1000"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4sTTIOCbyShY",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def windowed_dataset(series, window_size, batch_size, shuffle_buffer):\n",
        "  dataset = tf.data.Dataset.from_tensor_slices(series)\n",
        "  dataset = dataset.window(window_size + 1, shift=1, drop_remainder=True)\n",
        "  dataset = dataset.flat_map(lambda window: window.batch(window_size + 1))\n",
        "  dataset = dataset.shuffle(shuffle_buffer).map(lambda window: (window[:-1], window[-1]))\n",
        "  dataset = dataset.batch(batch_size).prefetch(1)\n",
        "  return dataset"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "A1Hl39rklkLm",
        "colab_type": "code",
        "outputId": "0d178b66-f1d0-469f-abee-1a5439a6da0c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "tf.keras.backend.clear_session()\n",
        "tf.random.set_seed(51)\n",
        "np.random.seed(51)\n",
        "\n",
        "tf.keras.backend.clear_session()\n",
        "dataset = windowed_dataset(x_train, window_size, batch_size, shuffle_buffer_size)\n",
        "\n",
        "model = tf.keras.models.Sequential([\n",
        "  tf.keras.layers.Lambda(lambda x: tf.expand_dims(x, axis=-1),\n",
        "                      input_shape=[None]),\n",
        "    tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(32, return_sequences=True)),\n",
        "  tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(32)),\n",
        "  tf.keras.layers.Dense(1),\n",
        "  tf.keras.layers.Lambda(lambda x: x * 100.0)\n",
        "])\n",
        "\n",
        "lr_schedule = tf.keras.callbacks.LearningRateScheduler(\n",
        "    lambda epoch: 1e-8 * 10**(epoch / 20))\n",
        "optimizer = tf.keras.optimizers.SGD(lr=1e-8, momentum=0.9)\n",
        "model.compile(loss=tf.keras.losses.Huber(),\n",
        "              optimizer=optimizer,\n",
        "              metrics=[\"mae\"])\n",
        "history = model.fit(dataset, epochs=100, callbacks=[lr_schedule])"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING: Logging before flag parsing goes to stderr.\n",
            "W0807 11:03:24.441927 140073583458176 deprecation.py:323] From /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/data/util/random_seed.py:58: where (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.where in 2.0, which has the same broadcast rule as np.where\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/100\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "W0807 11:03:28.267247 140073583458176 deprecation.py:323] From /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/keras/optimizer_v2/optimizer_v2.py:468: BaseResourceVariable.constraint (from tensorflow.python.ops.resource_variable_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Apply a constraint manually following the optimizer update step.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "31/31 [==============================] - 8s 245ms/step - loss: 21.4949 - mae: 22.0115\n",
            "Epoch 2/100\n",
            "31/31 [==============================] - 1s 34ms/step - loss: 21.2155 - mae: 21.6440\n",
            "Epoch 3/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 20.7115 - mae: 21.2278\n",
            "Epoch 4/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 20.1586 - mae: 20.7739\n",
            "Epoch 5/100\n",
            "31/31 [==============================] - 1s 31ms/step - loss: 19.7903 - mae: 20.2843\n",
            "Epoch 6/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 19.2600 - mae: 19.6776\n",
            "Epoch 7/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 18.2675 - mae: 18.7308\n",
            "Epoch 8/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 17.4072 - mae: 17.9757\n",
            "Epoch 9/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 17.2251 - mae: 17.6581\n",
            "Epoch 10/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 16.8969 - mae: 17.3457\n",
            "Epoch 11/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 16.6153 - mae: 17.0459\n",
            "Epoch 12/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 16.2141 - mae: 16.7529\n",
            "Epoch 13/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 15.9663 - mae: 16.4573\n",
            "Epoch 14/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 15.6956 - mae: 16.1587\n",
            "Epoch 15/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 15.3736 - mae: 15.8571\n",
            "Epoch 16/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 15.1127 - mae: 15.5765\n",
            "Epoch 17/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 14.7840 - mae: 15.2950\n",
            "Epoch 18/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 14.4769 - mae: 15.0164\n",
            "Epoch 19/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 14.3353 - mae: 14.7612\n",
            "Epoch 20/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 14.0844 - mae: 14.5354\n",
            "Epoch 21/100\n",
            "31/31 [==============================] - 1s 34ms/step - loss: 13.8134 - mae: 14.3081\n",
            "Epoch 22/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 13.6388 - mae: 14.1062\n",
            "Epoch 23/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 13.3828 - mae: 13.8858\n",
            "Epoch 24/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 13.1957 - mae: 13.7037\n",
            "Epoch 25/100\n",
            "31/31 [==============================] - 1s 34ms/step - loss: 13.0152 - mae: 13.4526\n",
            "Epoch 26/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 12.7582 - mae: 13.2327\n",
            "Epoch 27/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 12.5275 - mae: 12.9716\n",
            "Epoch 28/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 12.2088 - mae: 12.7095\n",
            "Epoch 29/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 12.3934 - mae: 12.8460\n",
            "Epoch 30/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 11.8438 - mae: 12.3121\n",
            "Epoch 31/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 11.3302 - mae: 11.8148\n",
            "Epoch 32/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 10.9489 - mae: 11.4830\n",
            "Epoch 33/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 10.6270 - mae: 11.1361\n",
            "Epoch 34/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 10.7821 - mae: 11.3007\n",
            "Epoch 35/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 10.9719 - mae: 11.4606\n",
            "Epoch 36/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 10.4495 - mae: 10.9636\n",
            "Epoch 37/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 9.9901 - mae: 10.4892\n",
            "Epoch 38/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 9.5775 - mae: 10.0283\n",
            "Epoch 39/100\n",
            "31/31 [==============================] - 1s 31ms/step - loss: 9.0663 - mae: 9.5461\n",
            "Epoch 40/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 8.6546 - mae: 9.1358\n",
            "Epoch 41/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 8.2880 - mae: 8.7356\n",
            "Epoch 42/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 7.9125 - mae: 8.4081\n",
            "Epoch 43/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 7.6783 - mae: 8.1137\n",
            "Epoch 44/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 7.3541 - mae: 7.8033\n",
            "Epoch 45/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 7.1044 - mae: 7.5770\n",
            "Epoch 46/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 6.9451 - mae: 7.3933\n",
            "Epoch 47/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 6.5533 - mae: 7.0771\n",
            "Epoch 48/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 6.4139 - mae: 6.9074\n",
            "Epoch 49/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 6.3932 - mae: 6.8757\n",
            "Epoch 50/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 6.0517 - mae: 6.5164\n",
            "Epoch 51/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 5.8718 - mae: 6.3634\n",
            "Epoch 52/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 5.7788 - mae: 6.2421\n",
            "Epoch 53/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 5.6698 - mae: 6.1419\n",
            "Epoch 54/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 5.5064 - mae: 5.9468\n",
            "Epoch 55/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 5.4382 - mae: 5.9006\n",
            "Epoch 56/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 5.2936 - mae: 5.7471\n",
            "Epoch 57/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 5.3198 - mae: 5.7881\n",
            "Epoch 58/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 5.1549 - mae: 5.6591\n",
            "Epoch 59/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 5.3084 - mae: 5.7545\n",
            "Epoch 60/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 5.0298 - mae: 5.5567\n",
            "Epoch 61/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 4.9550 - mae: 5.4250\n",
            "Epoch 62/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 5.1296 - mae: 5.5808\n",
            "Epoch 63/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 5.2112 - mae: 5.6630\n",
            "Epoch 64/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 5.0575 - mae: 5.5312\n",
            "Epoch 65/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 4.9258 - mae: 5.3900\n",
            "Epoch 66/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 5.3043 - mae: 5.7637\n",
            "Epoch 67/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 5.2753 - mae: 5.7446\n",
            "Epoch 68/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 5.2993 - mae: 5.7234\n",
            "Epoch 69/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 5.3742 - mae: 5.8346\n",
            "Epoch 70/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 6.1359 - mae: 6.6386\n",
            "Epoch 71/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 4.9126 - mae: 5.4135\n",
            "Epoch 72/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 5.6095 - mae: 6.0636\n",
            "Epoch 73/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 5.3411 - mae: 5.8208\n",
            "Epoch 74/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 6.6834 - mae: 7.1966\n",
            "Epoch 75/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 5.7113 - mae: 6.1957\n",
            "Epoch 76/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 5.1433 - mae: 5.6158\n",
            "Epoch 77/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 5.7382 - mae: 6.2385\n",
            "Epoch 78/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 7.2282 - mae: 7.6426\n",
            "Epoch 79/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 5.0664 - mae: 5.5955\n",
            "Epoch 80/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 6.4776 - mae: 7.0136\n",
            "Epoch 81/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 6.6766 - mae: 7.1079\n",
            "Epoch 82/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 6.7007 - mae: 7.2374\n",
            "Epoch 83/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 6.3450 - mae: 6.8347\n",
            "Epoch 84/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 5.8163 - mae: 6.3277\n",
            "Epoch 85/100\n",
            "31/31 [==============================] - 1s 34ms/step - loss: 7.0088 - mae: 7.6007\n",
            "Epoch 86/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 7.4649 - mae: 7.9559\n",
            "Epoch 87/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 8.0526 - mae: 8.5456\n",
            "Epoch 88/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 7.0569 - mae: 7.4773\n",
            "Epoch 89/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 6.4203 - mae: 6.9714\n",
            "Epoch 90/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 7.9175 - mae: 8.3901\n",
            "Epoch 91/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 5.9646 - mae: 6.4224\n",
            "Epoch 92/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 7.6886 - mae: 8.2557\n",
            "Epoch 93/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 10.2000 - mae: 10.6407\n",
            "Epoch 94/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 9.8386 - mae: 10.2520\n",
            "Epoch 95/100\n",
            "31/31 [==============================] - 1s 31ms/step - loss: 8.0082 - mae: 8.6051\n",
            "Epoch 96/100\n",
            "31/31 [==============================] - 1s 33ms/step - loss: 10.0811 - mae: 10.7728\n",
            "Epoch 97/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 12.3814 - mae: 12.8968\n",
            "Epoch 98/100\n",
            "31/31 [==============================] - 1s 32ms/step - loss: 11.2206 - mae: 11.6217\n",
            "Epoch 99/100\n",
            "31/31 [==============================] - 1s 31ms/step - loss: 14.2741 - mae: 14.6141\n",
            "Epoch 100/100\n",
            "31/31 [==============================] - 1s 31ms/step - loss: 10.7613 - mae: 11.1881\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AkBsrsXMzoWR",
        "colab_type": "code",
        "outputId": "9b29db35-45ae-4896-e01b-747a173131d2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 290
        }
      },
      "source": [
        "plt.semilogx(history.history[\"lr\"], history.history[\"loss\"])\n",
        "plt.axis([1e-8, 1e-4, 0, 30])"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1e-08, 0.0001, 0, 30]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEACAYAAABfxaZOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nX0TEQ/4PpUQ0+8caPg0AAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4uh-97bpLZCA",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "tf.keras.backend.clear_session()\n",
        "tf.random.set_seed(51)\n",
        "np.random.seed(51)\n",
        "\n",
        "tf.keras.backend.clear_session()\n",
        "dataset = windowed_dataset(x_train, window_size, batch_size, shuffle_buffer_size)\n",
        "\n",
        "model = tf.keras.models.Sequential([\n",
        "  tf.keras.layers.Lambda(lambda x: tf.expand_dims(x, axis=-1),\n",
        "                      input_shape=[None]),\n",
        "   tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(32, return_sequences=True)),\n",
        "  tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(32)),\n",
        "  tf.keras.layers.Dense(1),\n",
        "  tf.keras.layers.Lambda(lambda x: x * 100.0)\n",
        "])\n",
        "\n",
        "\n",
        "model.compile(loss=\"mse\", optimizer=tf.keras.optimizers.SGD(lr=1e-5, momentum=0.9),metrics=[\"mae\"])\n",
        "history = model.fit(dataset,epochs=500,verbose=0)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "icGDaND7z0ne",
        "colab_type": "code",
        "outputId": "275dc035-5ca0-413a-cf4f-7ced7e54b8c7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "forecast = []\n",
        "results = []\n",
        "for time in range(len(series) - window_size):\n",
        "  forecast.append(model.predict(series[time:time + window_size][np.newaxis]))\n",
        "\n",
        "forecast = forecast[split_time-window_size:]\n",
        "results = np.array(forecast)[:, 0, 0]\n",
        "\n",
        "\n",
        "plt.figure(figsize=(10, 6))\n",
        "\n",
        "plot_series(time_valid, x_valid)\n",
        "plot_series(time_valid, results)"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "W0807 11:12:33.224316 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.552937 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.572025 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.586395 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.599598 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.613220 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.626095 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.639695 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.652239 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.664435 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.675713 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.688113 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.700494 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.712087 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.723525 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.734841 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.747426 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.760809 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.774636 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.786829 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.802190 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.814870 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.827379 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.840476 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.853754 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.865581 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.877696 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.888548 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.901703 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.913182 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.926594 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.938226 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.952477 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.964540 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.976596 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:34.988209 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.002067 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.013063 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.024538 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.037256 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.051092 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.062191 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.074017 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.085711 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.097335 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.109683 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.123651 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.135181 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.146395 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.158809 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.172126 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.185188 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.196710 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.208060 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.220484 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.231959 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.243566 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.254265 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.266233 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.277580 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.288931 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.299833 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.311738 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.324032 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.336882 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.348622 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.361567 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.373736 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.386617 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.398854 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.412287 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.423272 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.436011 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.447668 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.459297 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.470443 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.482719 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.494261 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.505977 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.517024 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.528718 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.540054 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.551789 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.562790 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.574731 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.586810 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.599979 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.610877 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.621710 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.632989 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.648352 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.664475 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.677896 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.692121 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.706887 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.720029 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.732676 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.744216 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.755133 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.766927 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.779179 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.790663 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.802104 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.814601 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.826536 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.838754 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.851528 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.864589 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.878540 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.891568 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.904001 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.915868 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.926935 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.939712 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.952265 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.965807 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.978527 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:35.992119 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.005275 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.019024 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.032638 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.047797 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.061888 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.074447 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.087496 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.100180 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.112714 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.125258 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.136832 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.149525 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.161684 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.173718 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.185956 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.198350 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.211195 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.230761 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.244144 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.257281 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.269137 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.282576 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.296809 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.310516 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.322935 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.335408 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.347972 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.360550 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.372436 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.385589 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.397191 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.410935 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.423919 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.436211 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.447440 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.460646 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.473383 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.486406 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.497550 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.509841 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.522583 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.535477 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.548046 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.561596 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.574540 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.587415 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.598366 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.612816 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.625061 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.638893 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.650085 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.662919 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.675043 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.687114 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.700199 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.715347 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.728643 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.740637 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.752051 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.764210 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.776629 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.789045 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.800643 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.813428 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.826449 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.839145 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.854064 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.867521 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.879728 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.893401 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.906087 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.922387 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.934241 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.947228 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.959987 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.973192 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:36.986772 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.000800 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.013688 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.029256 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.040934 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.054058 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.065600 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.077579 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.091526 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.102759 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.114593 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.126461 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.138135 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.151169 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.163001 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.175808 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.189058 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.200478 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.211344 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.223364 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.234698 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.246518 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.258051 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.274457 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.290612 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.305184 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.316843 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.329654 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.341231 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.352958 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.364048 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.376094 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.389505 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.402996 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.414673 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.425950 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.436827 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.448720 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.460278 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.471872 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.482946 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.495420 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.507850 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.520303 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.531124 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.542691 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.554226 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.566274 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.578365 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.590598 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.603907 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.616011 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.627968 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.640532 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.651853 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.663260 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.675026 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.688718 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.701205 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.712944 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.723962 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.736042 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.748028 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.760793 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.772864 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.785991 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.798714 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.810490 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.822113 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.834127 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.845660 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.857571 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.869830 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.882207 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.894070 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.905663 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.917523 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.930213 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.941556 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.953426 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.965215 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.978389 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:37.990669 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.002812 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.013614 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.025400 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.038469 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.051068 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.063719 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.076197 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.088794 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.101481 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.113248 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.125710 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.137947 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.151214 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.165678 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.180938 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.196063 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.208516 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.220341 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.233987 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.246384 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.257967 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.269481 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.282712 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.294566 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.305793 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.316603 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.328688 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.340552 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.352197 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.363932 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.377524 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.391622 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.404930 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.416509 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.429510 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.441590 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.453602 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.465556 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.477675 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.489884 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.501127 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.513031 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.524451 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.535461 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.546372 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.558068 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.570500 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.582192 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.594599 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.606528 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.619029 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.631077 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.644070 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.655173 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.666660 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.677354 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.690110 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.704465 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.716845 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.728845 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.740459 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.753722 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.765496 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.777862 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.788770 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.800544 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.811716 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.822786 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.833659 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.845274 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.857743 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.870594 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.881533 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.893146 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.904997 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.916357 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.928086 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.940529 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.953234 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.966756 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.978031 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:38.991758 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.003920 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.017001 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.028974 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.039668 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.052107 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.063736 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.074924 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.085903 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.097244 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.109530 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.122092 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.134100 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.147189 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.159864 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.171689 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.183492 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.195523 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.207072 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.218606 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.230534 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.242599 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.254413 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.265819 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.277186 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.289496 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.301242 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.315296 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.327239 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.339471 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.350895 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.363225 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.374226 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.386827 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.399683 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.411159 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.421856 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.433688 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.444744 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.456189 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.466734 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.478861 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.490788 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.502124 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.515101 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.527561 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.540661 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.552875 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.564563 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.576193 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.588645 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.600034 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.613719 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.625462 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.637575 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.649203 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.661493 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.673174 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.685277 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.696267 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.708477 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.720251 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.732297 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.743816 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.756595 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.767745 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.779006 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.789824 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.801874 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.813541 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.824455 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.835078 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.846934 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.858157 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.869765 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.881373 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.898556 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.910641 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.921902 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.932373 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.946770 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.957704 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.970386 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.982637 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:39.994198 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.005091 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.017437 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.029440 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.041609 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.053861 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.066448 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.078382 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.090717 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.101745 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.113669 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.125238 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.136691 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.147724 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.160195 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.171931 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.183462 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.195728 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.208928 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.220893 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.233885 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.247032 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.261255 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.274411 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.286954 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.298705 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.310285 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.321544 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.332568 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.343963 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.357434 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.369263 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.382005 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.393643 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.405125 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.417211 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.429389 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.441203 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.453544 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.466412 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.481375 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.499462 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.512927 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.524986 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.537804 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.550134 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.563776 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.575883 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.588729 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.601294 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.614220 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.626618 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.638908 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.651061 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.663956 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.675951 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.688287 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.700183 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.712260 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.723957 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.735140 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.746206 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.758695 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.770344 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.782101 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.794058 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.806691 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.818103 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.830346 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.841872 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.853957 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.865543 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.877117 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.888044 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.899943 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.911225 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.923203 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.934749 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.946911 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.961130 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.972959 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.985472 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:40.998339 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.011005 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.023403 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.035994 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.051681 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.065451 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.077096 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.090236 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.104132 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.115892 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.127238 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.138604 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.149611 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.161508 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.179435 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.193460 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.205597 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.217749 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.229951 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.242201 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.254736 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.266333 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.278950 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.291151 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.302609 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.314392 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.326470 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.338477 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.350088 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.361646 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.377836 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.391442 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.405290 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.416296 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.428192 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.439393 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.451019 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.461932 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.473997 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.485684 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.497080 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.508115 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.521255 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.533385 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.545116 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.556860 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.568935 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.580585 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.592736 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.604776 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.617842 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.629833 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.641682 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.652556 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.664489 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.675568 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.687121 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.698719 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.711594 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.723414 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.734592 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.745126 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.756465 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.767793 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.778657 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.789278 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.801133 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.813480 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.825720 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.837430 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.850997 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.863379 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.876772 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.887699 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.901562 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.912856 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.924279 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.936189 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.948229 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.961841 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.972804 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.984009 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:41.998580 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.010013 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.021681 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.032767 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.044777 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.056989 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.068700 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.080458 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.093986 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.106515 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.119481 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.131498 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.144299 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.156750 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.169402 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.181632 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.196151 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.208264 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.219957 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.231745 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.244758 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.256985 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.269425 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.280656 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.295150 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.307250 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.320604 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.333868 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.345633 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.356849 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.368760 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.380771 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.392439 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.404261 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.415176 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.427641 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.439626 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.451393 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.462715 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.474985 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.487970 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.499920 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.511567 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.525148 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.537016 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.549987 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.562047 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.574607 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.587131 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.599457 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.611700 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.623542 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.634791 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.646033 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.657025 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.668615 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.679619 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.690764 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.702456 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.715021 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.726717 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.738297 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.749290 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.764427 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.777665 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.791230 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.806154 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.819355 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.832326 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.844288 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.855846 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.869577 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.883454 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.896452 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.907759 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.921542 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.933642 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.946057 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.957441 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.969964 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.983354 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:42.996631 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.010591 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.026109 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.041878 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.056136 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.069675 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.086834 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.099444 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.112537 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.123776 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.136048 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.147847 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.160350 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.172388 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.187004 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.200480 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.212831 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.224214 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.236417 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.248965 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.261071 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.273221 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.288263 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.301756 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.314861 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.327961 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.342868 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.356189 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.368525 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.380727 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.394483 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.406588 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.419566 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.432400 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.446660 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.459787 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.472504 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.484839 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.497721 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.511400 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.524462 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.538331 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.552213 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.564645 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.582782 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.598045 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.611001 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.622514 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.635168 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.647202 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.661591 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.673211 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.685329 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.697332 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.710453 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.722816 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.734684 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.745507 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.757047 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.768283 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.779622 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.790386 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.804219 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.816474 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.827694 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.838444 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.851226 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.863746 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.875746 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.887167 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.899768 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.911900 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.924863 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.936567 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.948738 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.960151 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.971628 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.982764 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:43.994714 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.006218 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.018048 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.029841 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.049007 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.063030 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.076016 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.090140 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.103480 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.115768 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.127728 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.139734 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.151809 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.163959 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.176577 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.187714 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.199735 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.211261 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.223115 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.234200 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.246110 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.257547 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.268904 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.279691 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.292372 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.304574 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.317030 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.327922 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.342293 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.353911 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.366377 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.381588 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.397512 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.410866 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.423838 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.437973 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.451225 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.462708 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.477485 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.489875 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.502357 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.514398 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.526956 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.537977 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.550786 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.562797 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.574574 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.585695 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.597727 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.611947 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.626857 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.638693 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.653007 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.665946 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.678082 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.689750 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.702831 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.715597 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.729273 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.740951 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.754254 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.766127 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.777533 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.788584 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.801065 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.812597 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.825427 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.836722 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.848422 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.861170 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.874840 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.886666 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.900613 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.912846 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.924758 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.936509 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.948034 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.960631 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.972884 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.984371 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:44.995591 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.006513 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.018699 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.030213 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.041350 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.053936 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.067016 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.080891 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.093850 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.105788 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.118747 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.130744 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.142986 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.154408 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.166711 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.178206 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.191540 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.203046 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.215551 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.227360 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.239843 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.250955 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.263339 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.275549 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.287993 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.299961 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.312429 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.324580 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.337874 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.350221 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.364289 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.376706 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.388754 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.400295 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.412840 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.425901 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.438665 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.449772 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.462711 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.475791 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.490080 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.501680 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.513510 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.524401 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.537543 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.550333 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.565439 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.577739 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.589775 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.602666 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.615433 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.627352 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.639436 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.653515 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.667222 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.679814 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.694922 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.708337 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.723226 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.736300 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.748513 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.761476 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.776395 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.792102 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.807507 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.822077 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.837283 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.851700 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.864290 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.876239 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.889851 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.903374 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.917907 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.930777 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.945099 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.959578 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.972243 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.984291 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:45.997860 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.011406 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.024771 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.036597 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.050202 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.063487 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.076642 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.091652 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.106287 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.119154 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.130909 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.143220 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.155829 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.167838 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.179553 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.191277 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.203666 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.216738 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.228462 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.239634 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.252778 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.265612 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.277045 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.289530 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.301461 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.312658 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.323917 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.336131 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.351601 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.366798 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.378535 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.391095 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.406266 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.418522 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.430275 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.441155 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.453179 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.465412 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.477430 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.489768 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.504799 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.516520 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.529001 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.540909 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.554778 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.567579 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.580784 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.592469 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.605742 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.617204 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.628808 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.639778 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.653257 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.669212 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.682109 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.697084 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.710858 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.725148 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.737755 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.750182 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.763576 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.775987 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.789046 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.801026 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.813892 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.826649 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.838757 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.851589 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.866140 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.880607 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.894675 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.908580 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.922752 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.935167 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.948287 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.959701 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.976266 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:46.991284 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.003415 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.016788 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.028333 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.041910 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.055997 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.068748 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.081249 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.095990 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.109303 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.125600 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.139426 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.152478 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.165860 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.180160 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.197562 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.211416 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.223526 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.236263 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.247895 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.259613 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.270798 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.282718 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.296153 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.307853 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.319487 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.331557 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.343616 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.354944 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.373062 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.384881 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.397501 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.410005 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.421376 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.432720 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.445500 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.458423 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.471276 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.483943 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.498808 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.511658 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.523533 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.534989 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.549233 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.561853 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.574805 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.587651 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.601012 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.614426 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.626656 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.639057 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.654112 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.668203 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.681055 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.694733 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.708791 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.722457 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.736345 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.749142 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.762910 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.777230 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.789327 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.801791 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.815050 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.829899 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.843022 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.855552 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.867182 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.879824 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.892083 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.904343 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.917300 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.930906 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.943158 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.955552 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.966986 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.980067 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:47.993149 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.007039 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.020342 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.035214 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.049335 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.065817 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.079561 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.096072 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.109985 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.123665 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.136422 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.150738 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.164371 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.176733 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.188274 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.200877 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.213278 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.227423 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.238951 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.252573 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.265001 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.278007 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.290007 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.306032 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.319907 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.334396 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.346269 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.359453 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.371797 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.383976 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.396183 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.409808 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.422814 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.435604 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.447479 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.459734 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.471474 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.484755 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.497529 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.509952 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.521662 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.534728 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.545713 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.559496 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.572400 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.586113 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.597398 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.610797 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.621713 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.633590 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.644708 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.657176 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.667963 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.679557 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.691050 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.702641 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.713425 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.725365 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.737722 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.749448 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.761699 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.774392 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.786096 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.797339 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.809432 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.820898 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.833178 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.844512 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.858657 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.874901 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.889018 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.903634 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.916509 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.929036 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.941167 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.952739 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.964531 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.977554 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:48.989410 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.001767 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.012755 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.026249 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.039249 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.052409 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.064435 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.076535 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.088264 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.101715 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.113510 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.126426 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.141258 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.156815 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.171091 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.185245 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.199506 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.212985 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.227854 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.240056 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.254396 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.266999 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.280221 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.293620 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.306825 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.319589 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.332784 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.345235 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.358362 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.371456 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.385693 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.400163 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.412038 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.423472 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.434694 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.446912 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.460272 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.473753 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.486449 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.499630 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.511886 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.525551 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.537849 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.551913 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.564669 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.576209 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.586992 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.599011 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.610750 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.624428 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.636064 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.648186 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.660574 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.674450 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.691717 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.704434 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.716433 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.727639 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.744718 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.757357 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.769996 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.781405 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.796415 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.808716 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.820921 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.833092 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.845709 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.859129 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.872784 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.885550 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.898480 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.909863 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.922591 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.935889 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.949015 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.961645 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.972639 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.983359 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:49.994674 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.006831 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.019046 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.030730 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.043660 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.056099 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.067768 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.079276 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.092725 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.105434 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.118495 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.130840 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.143677 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.156998 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.169026 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.180390 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.193814 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.206176 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.219576 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.232222 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.246557 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.261299 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.276193 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.289139 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.303401 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.315539 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.327724 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.339978 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.353075 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.365229 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.378051 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.390521 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.405622 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.420244 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.435741 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.449206 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.462430 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.475442 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.488841 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.502412 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.515011 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.528189 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.543607 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.557409 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.570110 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.582094 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.594983 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.606956 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.618814 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.629808 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.643652 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.657959 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.671262 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.683537 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.697810 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.709614 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.721406 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.732522 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.746821 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.759216 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.771203 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.782871 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.796487 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.808936 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.821684 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.833354 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.847517 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.861893 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.875711 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.887746 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.900166 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.912807 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.924736 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.936195 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.948350 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.960459 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.972930 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.984676 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:50.997962 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.010534 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.024679 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.037228 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.051536 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.066507 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.079647 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.091839 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.105125 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.117616 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.129502 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.141912 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.155372 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.167681 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.179397 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.192809 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.206398 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.219236 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.232856 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.245868 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.262687 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.278060 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.294763 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.308069 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.321448 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.334550 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.348168 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.359434 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.372780 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.385615 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.397531 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.408856 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.422388 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.434216 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.446152 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.459967 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.472712 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.484642 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.499460 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.511931 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.524290 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.535862 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.549260 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.561682 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.575683 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.588414 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.600980 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.612294 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.624800 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.637269 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.649968 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.661876 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.674823 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.687538 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.701178 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.714543 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.728709 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.740672 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.753644 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.764606 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.776265 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.790900 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.808283 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.821218 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.833673 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.846991 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.859874 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.872885 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.884401 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.897414 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.909889 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.922192 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.934873 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.947402 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.960194 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.972194 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.983691 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:51.996759 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.008635 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.022197 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.034182 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.049007 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.061613 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.073424 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.090034 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.108508 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.121934 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.136519 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.153477 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.166494 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.179129 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.192602 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.204118 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.216533 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.228672 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.241713 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.254272 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.268325 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.282433 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.297564 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.311684 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.324875 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.338473 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.350944 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.362601 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.375767 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.387765 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.400598 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.413987 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.426688 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.438553 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.450743 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.462142 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.474802 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.486609 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.499551 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.512194 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.525508 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.540992 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.552694 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.564905 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.577087 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.590899 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.604188 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.616711 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.628836 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.641102 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.653092 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.666178 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n",
            "W0807 11:12:52.677817 140073583458176 training_utils.py:1211] When passing input data as arrays, do not specify `steps_per_epoch`/`steps` argument. Please use `batch_size` instead.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmcAAAF3CAYAAADgjOwXAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRx5wzIEAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KfPeqI7rz4LD",
        "colab_type": "code",
        "outputId": "af81f61c-4a36-44cd-e02d-b9cbe563df8e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "tf.keras.metrics.mean_absolute_error(x_valid, results).numpy()"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "6.6017175"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JUsdZB_tzDLe",
        "colab_type": "code",
        "outputId": "461032e2-c6d4-42db-8f2b-85b5f77d354d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 607
        }
      },
      "source": [
        "import matplotlib.image  as mpimg\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "#-----------------------------------------------------------\n",
        "# Retrieve a list of list results on training and test data\n",
        "# sets for each training epoch\n",
        "#-----------------------------------------------------------\n",
        "mae=history.history['mae']\n",
        "loss=history.history['loss']\n",
        "\n",
        "epochs=range(len(loss)) # Get number of epochs\n",
        "\n",
        "#------------------------------------------------\n",
        "# Plot MAE and Loss\n",
        "#------------------------------------------------\n",
        "plt.plot(epochs, mae, 'r')\n",
        "plt.plot(epochs, loss, 'b')\n",
        "plt.title('MAE and Loss')\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.ylabel(\"Accuracy\")\n",
        "plt.legend([\"MAE\", \"Loss\"])\n",
        "\n",
        "plt.figure()\n",
        "\n",
        "epochs_zoom = epochs[200:]\n",
        "mae_zoom = mae[200:]\n",
        "loss_zoom = loss[200:]\n",
        "\n",
        "#------------------------------------------------\n",
        "# Plot Zoomed MAE and Loss\n",
        "#------------------------------------------------\n",
        "plt.plot(epochs_zoom, mae_zoom, 'r')\n",
        "plt.plot(epochs_zoom, loss_zoom, 'b')\n",
        "plt.title('MAE and Loss')\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.ylabel(\"Accuracy\")\n",
        "plt.legend([\"MAE\", \"Loss\"])\n",
        "\n",
        "plt.figure()"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEWCAYAAAB8LwAVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3CGaYFxXNEAK",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "af1ff2a0-6c45-47a7-9866-3c217989437f"
      },
      "source": [
        "tf.keras.backend.clear_session()\n",
        "dataset = windowed_dataset(x_train, window_size, batch_size, shuffle_buffer_size)\n",
        "\n",
        "model = tf.keras.models.Sequential([\n",
        "  tf.keras.layers.Lambda(lambda x: tf.expand_dims(x, axis=-1),\n",
        "                      input_shape=[None]),\n",
        "  tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(32, return_sequences=True)),\n",
        "  tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(32)),\n",
        "  tf.keras.layers.Dense(1),\n",
        "  tf.keras.layers.Lambda(lambda x: x * 100.0)\n",
        "])\n",
        "\n",
        "\n",
        "model.compile(loss=\"mse\", optimizer=tf.keras.optimizers.SGD(lr=1e-6, momentum=0.9))\n",
        "model.fit(dataset,epochs=100, verbose=0)"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7f6516d36f28>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FJ3R8ysauz9e",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "b3452011-e7ca-4fde-f964-0c9b58c806e5"
      },
      "source": [
        "tf.keras.backend.clear_session()\n",
        "dataset = windowed_dataset(x_train, window_size, batch_size, shuffle_buffer_size)\n",
        "\n",
        "model = tf.keras.models.Sequential([\n",
        "  tf.keras.layers.Lambda(lambda x: tf.expand_dims(x, axis=-1),\n",
        "                      input_shape=[None]),\n",
        "  tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(32, return_sequences=True)),\n",
        "  tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(32, return_sequences=True)),\n",
        "  tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(32)),\n",
        "  tf.keras.layers.Dense(1),\n",
        "  tf.keras.layers.Lambda(lambda x: x * 100.0)\n",
        "])\n",
        "\n",
        "\n",
        "model.compile(loss=\"mse\", optimizer=tf.keras.optimizers.SGD(lr=1e-6, momentum=0.9))\n",
        "model.fit(dataset,epochs=100)"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/100\n",
            "31/31 [==============================] - 11s 343ms/step - loss: 637.0232\n",
            "Epoch 2/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 122.3624\n",
            "Epoch 3/100\n",
            "31/31 [==============================] - 1s 43ms/step - loss: 93.5040\n",
            "Epoch 4/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 89.9407\n",
            "Epoch 5/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 83.9605\n",
            "Epoch 6/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 75.3442\n",
            "Epoch 7/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 68.6414\n",
            "Epoch 8/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 64.6441\n",
            "Epoch 9/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 66.7162\n",
            "Epoch 10/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 65.4161\n",
            "Epoch 11/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 63.6462\n",
            "Epoch 12/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 58.3565\n",
            "Epoch 13/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 56.1352\n",
            "Epoch 14/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 63.2100\n",
            "Epoch 15/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 55.0965\n",
            "Epoch 16/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 52.6582\n",
            "Epoch 17/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 62.0767\n",
            "Epoch 18/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 59.2543\n",
            "Epoch 19/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 57.2416\n",
            "Epoch 20/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 58.1652\n",
            "Epoch 21/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 53.0074\n",
            "Epoch 22/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 48.8509\n",
            "Epoch 23/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 55.9530\n",
            "Epoch 24/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 49.5848\n",
            "Epoch 25/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 58.1454\n",
            "Epoch 26/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 49.3683\n",
            "Epoch 27/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 54.8082\n",
            "Epoch 28/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 53.8049\n",
            "Epoch 29/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 54.7548\n",
            "Epoch 30/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 48.4936\n",
            "Epoch 31/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 49.9576\n",
            "Epoch 32/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 52.8617\n",
            "Epoch 33/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 57.4542\n",
            "Epoch 34/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 48.4926\n",
            "Epoch 35/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 49.9492\n",
            "Epoch 36/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 48.0644\n",
            "Epoch 37/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 47.0089\n",
            "Epoch 38/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 51.8139\n",
            "Epoch 39/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 47.8317\n",
            "Epoch 40/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 47.0761\n",
            "Epoch 41/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 49.3140\n",
            "Epoch 42/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 46.2248\n",
            "Epoch 43/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 50.1575\n",
            "Epoch 44/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 49.0274\n",
            "Epoch 45/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 51.4325\n",
            "Epoch 46/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 48.7842\n",
            "Epoch 47/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 51.9834\n",
            "Epoch 48/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 50.1357\n",
            "Epoch 49/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 45.3995\n",
            "Epoch 50/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 45.3383\n",
            "Epoch 51/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 45.9167\n",
            "Epoch 52/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 48.6908\n",
            "Epoch 53/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 50.2167\n",
            "Epoch 54/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 50.3753\n",
            "Epoch 55/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 44.5830\n",
            "Epoch 56/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 46.3022\n",
            "Epoch 57/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 46.9732\n",
            "Epoch 58/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 46.0485\n",
            "Epoch 59/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 45.0353\n",
            "Epoch 60/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 45.3140\n",
            "Epoch 61/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 46.7041\n",
            "Epoch 62/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 46.7275\n",
            "Epoch 63/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 47.1370\n",
            "Epoch 64/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 45.2307\n",
            "Epoch 65/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 45.5482\n",
            "Epoch 66/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 44.5448\n",
            "Epoch 67/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 43.8142\n",
            "Epoch 68/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 44.7850\n",
            "Epoch 69/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 49.0795\n",
            "Epoch 70/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 49.2480\n",
            "Epoch 71/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 48.2116\n",
            "Epoch 72/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 43.9088\n",
            "Epoch 73/100\n",
            "31/31 [==============================] - 1s 44ms/step - loss: 44.9772\n",
            "Epoch 74/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 44.6104\n",
            "Epoch 75/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 44.2205\n",
            "Epoch 76/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 43.6443\n",
            "Epoch 77/100\n",
            "31/31 [==============================] - 1s 47ms/step - loss: 45.3903\n",
            "Epoch 78/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 45.6141\n",
            "Epoch 79/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 44.8026\n",
            "Epoch 80/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 44.7155\n",
            "Epoch 81/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 51.3334\n",
            "Epoch 82/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 52.5975\n",
            "Epoch 83/100\n",
            "31/31 [==============================] - 1s 47ms/step - loss: 45.2570\n",
            "Epoch 84/100\n",
            "31/31 [==============================] - 1s 47ms/step - loss: 45.2547\n",
            "Epoch 85/100\n",
            "31/31 [==============================] - 2s 48ms/step - loss: 46.0126\n",
            "Epoch 86/100\n",
            "31/31 [==============================] - 1s 47ms/step - loss: 45.1493\n",
            "Epoch 87/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 43.5153\n",
            "Epoch 88/100\n",
            "31/31 [==============================] - 1s 47ms/step - loss: 51.4438\n",
            "Epoch 89/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 46.2939\n",
            "Epoch 90/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 49.2525\n",
            "Epoch 91/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 45.1779\n",
            "Epoch 92/100\n",
            "31/31 [==============================] - 1s 47ms/step - loss: 52.0527\n",
            "Epoch 93/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 44.1530\n",
            "Epoch 94/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 41.8995\n",
            "Epoch 95/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 46.0042\n",
            "Epoch 96/100\n",
            "31/31 [==============================] - 1s 45ms/step - loss: 44.7334\n",
            "Epoch 97/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 44.7615\n",
            "Epoch 98/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 44.1620\n",
            "Epoch 99/100\n",
            "31/31 [==============================] - 1s 47ms/step - loss: 44.7740\n",
            "Epoch 100/100\n",
            "31/31 [==============================] - 1s 46ms/step - loss: 45.3264\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7f650f82acf8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    }
  ]
}