Simulation_Football_Data.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "id": "JaUqei1QN7FE"
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import os\n",
    "import tensorflow as tf\n",
    "import datetime\n",
    "from keras import backend as K\n",
    "from keras.initializers import glorot_normal\n",
    "from keras.layers import Layer\n",
    "from keras.layers import Input, Dense\n",
    "from keras.models import Model\n",
    "from keras.layers import LSTM\n",
    "import logging\n",
    "import psycopg2 as pg\n",
    "import pandas.io.sql as psql\n",
    "import math"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "id": "-lnb_30mSF09"
   },
   "outputs": [],
   "source": [
    "from temp_mosum.mosum_library import mosum_function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "id": "ZYVJQ-P2SXL4"
   },
   "outputs": [],
   "source": [
    "from abc import ABC\n",
    "class Dataset(ABC):\n",
    "    def __init__(self):\n",
    "        super().__init__()\n",
    "\n",
    "    def next_batch(self, **kwargs):\n",
    "        pass\n",
    "\n",
    "class NNModel(ABC):\n",
    "    def __init__(self):\n",
    "        super().__init__()\n",
    "\n",
    "    def net_structure(self, **kwargs):\n",
    "        pass\n",
    "\n",
    "    def instantiate_and_fit(self, **kwargs):\n",
    "        pass  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "id": "qvnt1CqdSZjF"
   },
   "outputs": [],
   "source": [
    "def ts_generator(ts_obj, n_steps):\n",
    "    while 1:\n",
    "        batch = ts_obj.next_batch(1, n_steps)\n",
    "        yield batch[0], batch[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "IFtVkU2bSgUg",
    "outputId": "f6f9bb60-e944-4d6f-a2a4-09d7d51fae98"
   },
   "outputs": [],
   "source": [
    "dataframe=pd.read_csv('Dataset1.csv', delimiter=',')\n",
    "dataframe = dataframe.fillna(value=np.nan)               \n",
    "all_features = dataframe.columns[2:].values    \n",
    "def statistics(dataframe, feature):\n",
    "    zeros = 0\n",
    "    for el in dataframe[feature].values:\n",
    "        if math.isnan(el) or el == None:\n",
    "            zeros += 1\n",
    "    return zeros/len( dataframe[feature].values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "id": "ALzHQK2Zev83"
   },
   "outputs": [],
   "source": [
    "data=dataframe\n",
    "dataframe.index=dataframe['date']\n",
    "dataframe.drop(\"date\",inplace=True, axis=1)\n",
    "dataframe=dataframe['values']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "KpqlOLydfkVZ",
    "outputId": "c48bb4d2-f78a-4481-8c47-bedcc6769f05",
    "tags": []
   },
   "outputs": [],
   "source": [
    "result = mosum_function.mosum(dataframe, G=20, criterion = 'eta')\n",
    "cpts_list = result['cpts']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[8, 31, 65, 99, 119, 133, 198]\n"
     ]
    }
   ],
   "source": [
    "print(cpts_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "id": "20ie3tFGgp2a"
   },
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import math\n",
    "from tensorflow.python.framework.ops import disable_eager_execution\n",
    "disable_eager_execution()\n",
    "\n",
    "def gaussian_likelihood(sigma):\n",
    "    def gaussian_loss(y_true, y_pred):\n",
    "        result = keras.losses.mean_squared_error(y_true, y_pred)\n",
    "        result = K.print_tensor(result, message='losses')\n",
    "        print(result)\n",
    "        return tf.math.reduce_mean(0.5*tf.math.log(sigma) + 0.5*tf.divide(tf.math.square(y_true - y_pred), sigma)) + 1e-6 + 6\n",
    "    return gaussian_loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "-WIjatMiguac",
    "outputId": "11510530-e394-4c36-de70-381cde06ff5d",
    "tags": []
   },
   "outputs": [],
   "source": [
    "test = data.loc[:, \"values\"]\n",
    "test1 = test.values\n",
    "source_df = pd.DataFrame({'feature_1': test1[:-1], 'target': test1[1:]})\n",
    "source_df['category'] = ['1' for i in range(source_df.shape[0])]\n",
    "i=2\n",
    "for name, values in data.iteritems():\n",
    "    featurename = 'feature_' + str(i)\n",
    "    source_df[featurename] = data[name].values[1:]\n",
    "    i +=1\n",
    "\n",
    "def put_target_last_place(data):\n",
    "    data = data.rename(columns={\"category\": \"category_\"})\n",
    "    data.insert(loc=1, column='category', value=data['category_'].values)\n",
    "    data = data.drop(columns=[\"category_\"])\n",
    "    return data\n",
    "source_df = put_target_last_place(source_df)\n",
    "\n",
    "size = int(len(source_df) * 0.60)\n",
    "training = source_df.iloc[0:size]\n",
    "test= source_df.iloc[size: len(source_df)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "diff_list = [x - cpts_list[i - 1] for i, x in enumerate(cpts_list)][1:]\n",
    "diff_list.sort()\n",
    "batch_size=diff_list[0] // 2  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "id": "aNiqMv8ShJSU"
   },
   "outputs": [],
   "source": [
    "epochs=200\n",
    "steps_per_epoch=10\n",
    "variant_count=2\n",
    "batch_size=batch_size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "id": "4AwlS7kjhML1"
   },
   "outputs": [],
   "source": [
    "class GaussianLayer(Layer):\n",
    "    def __init__(self, output_dim, **kwargs):\n",
    "        self.output_dim = output_dim\n",
    "        self.kernel_1, self.kernel_2, self.bias_1, self.bias_2 = [], [], [], []\n",
    "        super(GaussianLayer, self).__init__(**kwargs)\n",
    "    def build(self, input_shape):\n",
    "        n_weight_rows = input_shape[2]\n",
    "        self.kernel_1 = self.add_weight(name='kernel_1',\n",
    "                                        shape=(n_weight_rows, self.output_dim),\n",
    "                                        initializer=glorot_normal(),\n",
    "                                        trainable=True)\n",
    "        self.kernel_2 = self.add_weight(name='kernel_2',\n",
    "                                        shape=(n_weight_rows, self.output_dim),\n",
    "                                        initializer=glorot_normal(),\n",
    "                                        trainable=True)\n",
    "        self.bias_1 = self.add_weight(name='bias_1',\n",
    "                                      shape=(self.output_dim,),\n",
    "                                      initializer=glorot_normal(),\n",
    "                                      trainable=True)\n",
    "        self.bias_2 = self.add_weight(name='bias_2',\n",
    "                                      shape=(self.output_dim,),\n",
    "                                      initializer=glorot_normal(),\n",
    "                                      trainable=True)\n",
    "        super(GaussianLayer, self).build(input_shape)\n",
    "    def call(self, x):\n",
    "        output_mu = K.dot(x, self.kernel_1) + self.bias_1\n",
    "        output_sig = K.dot(x, self.kernel_2) + self.bias_2\n",
    "        output_sig_pos = K.log(1 + K.exp(output_sig)) + 1e-06\n",
    "        return [output_mu, output_sig_pos]\n",
    "    def compute_output_shape(self, input_shape):\n",
    "        return [(input_shape[0], self.output_dim), (input_shape[0], self.output_dim)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "id": "4nU59e8YhO-7"
   },
   "outputs": [],
   "source": [
    "class DeepArTrainingInfo(): \n",
    "    def __init__(self, epoch, batch, loss, hasChangepoint):\n",
    "        self.__epoch = epoch\n",
    "        self.__batch = batch\n",
    "        self.__loss = loss\n",
    "        self.__hasChangepoint = hasChangepoint\n",
    "\n",
    "    def get_epoch(self):\n",
    "        return self.__epoch\n",
    "\n",
    "    def get_batch(self):\n",
    "        return self.__batch\n",
    "\n",
    "    def get_loss(self):\n",
    "        return self.__loss\n",
    "\n",
    "    def get_hasChangepoint(self):\n",
    "        return self.__hasChangepoint\n",
    "\n",
    "\n",
    "    def set_epoch(self, x):\n",
    "        self.__epoch = x\n",
    "\n",
    "    def set_batch(self, x):\n",
    "        self.__batch = x\n",
    "\n",
    "    def set_loss(self, x):\n",
    "        self.__loss = x\n",
    "\n",
    "    def set_hasChangepoint(self, x):\n",
    "        self.__hasChangepoint = x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "id": "oYbq3_vShYo3"
   },
   "outputs": [],
   "source": [
    "keras = tf.keras\n",
    "logger = logging.getLogger('deepar')\n",
    "\n",
    "def examine_loss(y_true, y_pred):\n",
    "        result = keras.losses.mean_squared_error(y_true, y_pred)\n",
    "        result = K.print_tensor(y_true, message='losses----->')\n",
    "        return result\n",
    "\n",
    "class DeepAR(NNModel):\n",
    "    def __init__(self, ts_obj, steps_per_epoch=steps_per_epoch, epochs=epochs, loss=gaussian_likelihood,\n",
    "                 optimizer='adam', with_custom_nn_structure=None):\n",
    "        self.ts_obj = ts_obj\n",
    "        self.inputs, self.z_sample = None, None\n",
    "        self.steps_per_epoch = steps_per_epoch\n",
    "        self.epochs = epochs\n",
    "        self.loss = loss\n",
    "        self.optimizer = optimizer\n",
    "        self.keras_model = None\n",
    "        if with_custom_nn_structure:\n",
    "            self.nn_structure = with_custom_nn_structure\n",
    "        else:\n",
    "            self.nn_structure = DeepAR.basic_structure\n",
    "        self._output_layer_name = 'main_output'\n",
    "        self.get_intermediate = None\n",
    "    @staticmethod\n",
    "    def basic_structure():\n",
    "        input_shape = (batch_size, variant_count)\n",
    "        inputs = Input(shape=input_shape)\n",
    "        x = LSTM(30, return_sequences=True)(inputs)\n",
    "        x = Dense(3, activation='relu')(x)\n",
    "        loc, scale = GaussianLayer(1, name='main_output')(x)\n",
    "        return input_shape, inputs, [loc, scale]\n",
    "    \n",
    "    def instantiate_and_fit(self, verbose=False):\n",
    "        input_shape, inputs, theta = self.nn_structure()\n",
    "        print(input_shape)\n",
    "        model = Model(inputs, theta[0])\n",
    "\n",
    "        def examine_loss(y_true, y_pred):\n",
    "            result = keras.losses.mean_squared_error(y_true, y_pred)\n",
    "            result = K.print_tensor(result, message='losses----->')\n",
    "            return result\n",
    "\n",
    "        model.compile(loss=self.loss(theta[1]), optimizer=self.optimizer)\n",
    "        log_dir = \"logs/fit/\" + datetime.datetime.now().strftime(\"%Y%m%d-%H%M%S\")\n",
    "        tensorboard_callback = tf.keras.callbacks.TensorBoard(log_dir=log_dir, histogram_freq=1)\n",
    "\n",
    "        model.fit(ts_generator(self.ts_obj,\n",
    "                                         input_shape[0]),\n",
    "                            steps_per_epoch=self.steps_per_epoch,\n",
    "                            epochs=self.epochs,callbacks=[tensorboard_callback, kerasCallback])\n",
    "        if verbose:\n",
    "            logger.debug('Model was successfully trained')\n",
    "        self.keras_model = model\n",
    "        self.get_intermediate = K.function(inputs=[self.model.input],\n",
    "                                           outputs=self.model.get_layer(self._output_layer_name).output)\n",
    "        \n",
    "        model.summary()\n",
    "    @property\n",
    "    def model(self):\n",
    "        return self.keras_model\n",
    "    def predict_theta_from_input(self, input_list):\n",
    "        if not self.get_intermediate:\n",
    "            raise ValueError('TF model must be trained first!')\n",
    "        return self.get_intermediate(input_list)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "3hGIECvBhfUj",
    "outputId": "4547fd28-e4d1-4471-f931-af500ceab89c"
   },
   "outputs": [
    {
     "name": "stdin",
     "output_type": "stream",
     "text": [
      "Apply change point filter? True / False:  true\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Filter = True\n"
     ]
    }
   ],
   "source": [
    "filter_change_point = ''\n",
    "batch_count = int(len(training)/batch_size)\n",
    "change_points_found_in_training = []\n",
    "current_epoch = 0\n",
    "deepArTrainingInfos = []\n",
    "\n",
    "while True:\n",
    "    filter_change_point = str(input('Apply change point filter? True / False: ') or \"True\")\n",
    "    filter_change_point = filter_change_point.capitalize()\n",
    "    if filter_change_point.capitalize() == 'True':\n",
    "        print('Filter = ' + filter_change_point)\n",
    "        break\n",
    "    elif filter_change_point.capitalize() == 'False':\n",
    "        print('Filter = ' + filter_change_point)\n",
    "        break\n",
    "\n",
    "    else:\n",
    "        print('Enter True or False')\n",
    "        continue\n",
    "\n",
    "\n",
    "class KerasCallback(tf.keras.callbacks.Callback):\n",
    "    def __init__(self):\n",
    "        self.__current_epoch = 0\n",
    "        self.__start_index = 0\n",
    "        self.__end_index= 0\n",
    "        self.__loss_per_batch = []\n",
    "        self.__hasChangePoint = False\n",
    "        self.__df_deepar_infos = pd.DataFrame(columns=['key','Epoch', 'Batch', 'Loss', 'hasChangepoint','start','end'])\n",
    "\n",
    "    def set_start_index(self, index):\n",
    "        self.__start_index = index\n",
    "\n",
    "    def set_end_index(self, index):\n",
    "        self.__end_index = index\n",
    "\n",
    "    def get_df_deepar_infos(self):\n",
    "        return self.__df_deepar_infos\n",
    "\n",
    "    def set_hasChangePoint(self, hasChangepoint):\n",
    "        self.__hasChangePoint = hasChangepoint   \n",
    "\n",
    "    def get_hasChangepoint(self):\n",
    "        return self.__hasChangePoint\n",
    "\n",
    "    def get_current_epoch(self):\n",
    "        return self.__current_epoch\n",
    "         \n",
    "    def get_loss_per_batch(self):\n",
    "        return self.__loss_per_batch\n",
    "\n",
    "    def on_epoch_begin(self, epoch, logs=None):\n",
    "        self.__current_epoch = epoch\n",
    "\n",
    "    def on_predict_end(self, logs):\n",
    "        print(logs)    \n",
    "\n",
    "    \n",
    "    def on_train_batch_end(self, batch, logs=None): \n",
    "        print(logs)\n",
    "        self.__loss_per_batch.append(DeepArTrainingInfo(epoch =self.__current_epoch,batch=batch,loss=logs['loss'], hasChangepoint= self.__hasChangePoint))\n",
    "        temp_df = pd.DataFrame([[str(self.__current_epoch) + \":\" + str(batch), self.__current_epoch, batch,logs['loss'], self.__hasChangePoint, self.__start_index, self.__end_index]]\n",
    "                               ,columns=['key','Epoch', 'Batch', 'Loss', 'hasChangepoint','start','end'])\n",
    "        self.__df_deepar_infos = pd.concat([self.__df_deepar_infos, temp_df], axis=0, ignore_index=True)\n",
    "\n",
    "\n",
    "kerasCallback = KerasCallback()\n",
    "\n",
    "class TimeSeries(Dataset):\n",
    "    def __init__(self, pandas_df, one_hot_root_list=None, grouping_variable='category', scaler=None):\n",
    "        super().__init__()\n",
    "        self.data = pandas_df\n",
    "        self.one_hot_root_list = one_hot_root_list\n",
    "        self.grouping_variable = grouping_variable\n",
    "        self.current_batch_nr = 0\n",
    "        if self.data is None:\n",
    "            raise ValueError('Must provide a Pandas df to instantiate this class')\n",
    "        self.scaler = scaler\n",
    "    def _one_hot_padding(self, pandas_df, padding_df):\n",
    "        for one_hot_root in self.one_hot_root_list:\n",
    "            one_hot_columns = [i for i in pandas_df.columns \n",
    "                               if i.startswith(one_hot_root) and pandas_df[i].values[0] == 1]\n",
    "            for col in one_hot_columns:\n",
    "                padding_df[col] = 1\n",
    "        return padding_df\n",
    "    def _pad_ts(self, pandas_df, desired_len, padding_val=0):\n",
    "        pad_length = desired_len - pandas_df.shape[0]\n",
    "        padding_df = pd.concat([pd.DataFrame({col: padding_val for col in pandas_df.columns},\n",
    "                                             index=[i for i in range(pad_length)])])\n",
    "        if self.one_hot_root_list:\n",
    "            padding_df = self._one_hot_padding(pandas_df, padding_df)\n",
    "        return pd.concat([padding_df, pandas_df]).reset_index(drop=True)\n",
    "    \n",
    "    @staticmethod\n",
    "    def logChangepoint(params, model):\n",
    "        return tf.keras.callbacks.Callback()\n",
    "        print(\"in callback\")\n",
    "\n",
    "    @staticmethod \n",
    "    def _has_changepoint_in_index(start_index, end_index, batch_number):\n",
    "        has_change_point = False\n",
    "        for cpv in cpts_list:\n",
    "            if (cpv >= start_index and cpv <= end_index):\n",
    "                change_points_found_in_training.append((cpv,batch_number, kerasCallback.get_current_epoch()))\n",
    "                print('Found change point: ' + str(cpv) + ' in batch ' + str(batch_number) + ' with index :'+ str(start_index)+ ' and ' +str(end_index) )  \n",
    "                has_change_point = True\n",
    "                break\n",
    "            return has_change_point\n",
    "           \n",
    "    @staticmethod\n",
    "    def _find_valid_index(pandas_df, desired_len, batch_number):\n",
    "        start_index = 0\n",
    "        end_index = 0\n",
    "        valid_index = False\n",
    "\n",
    "        if filter_change_point == 'False':\n",
    "            start_index = np.random.choice([i for i in range(0, pandas_df.shape[0] - desired_len + 1)])\n",
    "            end_index = start_index + desired_len\n",
    "            has_changePoint = TimeSeries._has_changepoint_in_index(start_index, end_index, batch_number)\n",
    "            kerasCallback.set_hasChangePoint(has_changePoint)\n",
    "        else:  \n",
    "            while valid_index == False:\n",
    "                start_index = np.random.choice([i for i in range(0, pandas_df.shape[0] - desired_len + 1)])\n",
    "                end_index = start_index + desired_len\n",
    "                valid_index = not TimeSeries._has_changepoint_in_index(start_index, end_index, batch_number)\n",
    "        \n",
    "        kerasCallback.set_end_index(end_index)\n",
    "        kerasCallback.set_start_index(start_index)\n",
    "        return start_index, end_index       \n",
    "                \n",
    "    @staticmethod\n",
    "    def _sample_ts(pandas_df, desired_len, batch_number):\n",
    "        if pandas_df.shape[0] < desired_len:\n",
    "            raise ValueError('Desired sample length is greater than df row len')\n",
    "        if pandas_df.shape[0] == desired_len:\n",
    "            return pandas_df\n",
    "        start_index, end_index = TimeSeries._find_valid_index(pandas_df, desired_len, batch_number)\n",
    "\n",
    "        return pandas_df.iloc[start_index: end_index, ]\n",
    "    def next_batch(self, batch_size, n_steps,\n",
    "                   target_var='target', verbose=False,\n",
    "                   padding_value=0):\n",
    "        if self.current_batch_nr % steps_per_epoch == 0:\n",
    "            self.current_batch_nr = 1\n",
    "        else:\n",
    "            self.current_batch_nr = self.current_batch_nr + 1\n",
    "\n",
    "        groups_list = self.data[self.grouping_variable].unique()\n",
    "        np.random.shuffle(groups_list)\n",
    "        selected_groups = groups_list[:batch_size]\n",
    "        input_data = self.data[self.data[self.grouping_variable].isin(set(selected_groups))]\n",
    "        sampled = []\n",
    "        print(\"step ......\", batch_size, n_steps)\n",
    "        for cat, cat_data in input_data.groupby(self.grouping_variable):\n",
    "                \n",
    "                if cat_data.shape[0] < n_steps:\n",
    "                    sampled_cat_data = self._pad_ts(pandas_df=cat_data,\n",
    "                                                    desired_len=n_steps,\n",
    "                                                    padding_val=padding_value)\n",
    "                else:\n",
    "                    sampled_cat_data = self._sample_ts(pandas_df=cat_data,\n",
    "                                                       desired_len=n_steps,\n",
    "                                                       batch_number=self.current_batch_nr)\n",
    "                sampled.append(sampled_cat_data)\n",
    "        rnn_output = pd.concat(sampled).drop(columns=self.grouping_variable).reset_index(drop=True)\n",
    "        if self.scaler:\n",
    "            batch_scaler = self.scaler()\n",
    "            n_rows = rnn_output.shape[0]\n",
    "            for i in range(1,variant_count):\n",
    "                featurename = \"feature_\" + str(i)\n",
    "                rnn_output[featurename] = rnn_output[featurename].astype('float')           \n",
    "                rnn_output[featurename] = batch_scaler.fit_transform(rnn_output[featurename].values.reshape(n_rows, 1)).reshape(n_rows)\n",
    "                rnn_output[target_var] = rnn_output[target_var].astype('float')\n",
    "                rnn_output[target_var] = batch_scaler.fit_transform(rnn_output[target_var].values.reshape(n_rows, 1)).reshape(n_rows)\n",
    "        return rnn_output.drop(target_var, 1).values.reshape(batch_size, n_steps, -1), \\\n",
    "               rnn_output[target_var].values.reshape(batch_size, n_steps, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "collapsed": true,
    "id": "GJX5hs7vhhqE",
    "jupyter": {
     "outputs_hidden": true
    },
    "outputId": "13d5de9f-d936-47c7-a5bf-a6ae84888fef",
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(7, 2)\n",
      "Tensor(\"loss/main_output_loss/Identity:0\", shape=(None, 7), dtype=float32)\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 6 with index :8 and 15\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "Epoch 1/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 7.036661}\n",
      " 1/10 [==>...........................] - ETA: 22s - batch: 0.0000e+00 - size: 1.0000 - loss: 7.0367step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 7.1696005}\n",
      " 2/10 [=====>........................] - ETA: 0s - batch: 0.5000 - size: 1.0000 - loss: 7.1031     step ...... {'batch': 2, 'size': 1, 'loss': 7.088808}\n",
      "1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.9104037}\n",
      "step ...... 1 7\n",
      "step ...... 1 {'batch': 4, 'size': 1, 'loss': 6.7492814}\n",
      "7\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.9910step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 7.0590296}\n",
      "WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0074s vs `on_train_batch_end` time: 0.0196s). Check your callbacks.\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 7.118373}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.9153085}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 7.0059step ......{'batch': 8, 'size': 1, 'loss': 6.9693265}\n",
      " 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.919636}step ...... 1 7\n",
      "Found change point: 8 in batch 1 with index :8 and 15\n",
      "\n",
      "10/10 [==============================] - 3s 30ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.9936\n",
      "Epoch 2/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 7.0276265}\n",
      " 1 7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 7.0276step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.981114}\n",
      " 2/10 [=====>........................] - ETA: 0s - batch: 0.5000 - size: 1.0000 - loss: 7.0044    step ......{'batch': 2, 'size': 1, 'loss': 7.0463824} \n",
      "1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 7.2417765}\n",
      "step ...... 1 {'batch': 4, 'size': 1, 'loss': 7.136816}\n",
      "7\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 7.0867step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 7.0690966}step ......\n",
      " 1 7\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 7.0838{'batch': 6, 'size': 1, 'loss': 7.068562}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.8426194}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 7.0470104}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.9530344}\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 7.0414step ...... 1 7\n",
      "Found change point: 8 in batch 1 with index :7 and 14\n",
      "10/10 [==============================] - 0s 28ms/step - batch: 4.5000 - size: 1.0000 - loss: 7.0414\n",
      "Epoch 3/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 7.0366545}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 7.0367step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 7.001933}\n",
      "{'batch': 2, 'size': 1, 'loss': 7.036938}\n",
      "step ......{'batch': 3, 'size': 1, 'loss': 7.457461}\n",
      " 1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 7.0285273}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 7.1123    step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.8683367}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 7.0259256}\n",
      "step ......{'batch': 7, 'size': 1, 'loss': 6.889888} 1 7\n",
      "\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 7.0432step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 7.048941}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.696474}\n",
      "10/10 [==============================] - 0s 23ms/step - batch: 4.5000 - size: 1.0000 - loss: 7.0091\n",
      "Epoch 4/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 7.0640535}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 7.0641step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.777099}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.8732176}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 7.133148}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.9449196}\n",
      "{'batch': 5, 'size': 1, 'loss': 7.014844}step ...... 1 7\n",
      "\n",
      "step ...... 1 7\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.9679    step ......{'batch': 6, 'size': 1, 'loss': 6.8199296} 1 7\n",
      "\n",
      "{'batch': 7, 'size': 1, 'loss': 6.995382}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.8176494}\n",
      "step ...... 1 7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.9378{'batch': 9, 'size': 1, 'loss': 6.8747344}step ...... 1 7\n",
      "Found change point: 8 in batch 1 with index :6 and 13\n",
      "\n",
      "10/10 [==============================] - 0s 22ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.9315\n",
      "Epoch 5/200\n",
      "step ...... {'batch': 0, 'size': 1, 'loss': 7.0665035}\n",
      "1 7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 7.0665step ......{'batch': 1, 'size': 1, 'loss': 6.9553895}\n",
      " 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.99671}step ...... 1 7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.7758126}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.9486    step ...... 1{'batch': 4, 'size': 1, 'loss': 7.160154}\n",
      " 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.9990187}\n",
      "step ...... 1 7\n",
      "step ......{'batch': 6, 'size': 1, 'loss': 7.100454} 1 7\n",
      "\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 7.0077step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.996281}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 7.0379553}\n",
      "step ...... 1 {'batch': 9, 'size': 1, 'loss': 6.6769075}7\n",
      "\n",
      "10/10 [==============================] - 0s 22ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.9765\n",
      "Epoch 6/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 7.0251083}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 7.0251step ...... 1 7\n",
      "Found change point: 8 in batch 3 with index :1 and 8{'batch': 1, 'size': 1, 'loss': 6.89306}\n",
      "\n",
      "{'batch': 2, 'size': 1, 'loss': 6.9088264}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 7.126666}step ...... 1 7\n",
      "\n",
      "step ......{'batch': 4, 'size': 1, 'loss': 6.9885144}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.9884     1 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.951082}\n",
      "step ......{'batch': 6, 'size': 1, 'loss': 7.3602567}\n",
      " 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.6664286}\n",
      "step ...... 1 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.9900step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.9620056}\n",
      "{'batch': 9, 'size': 1, 'loss': 6.9172325}\n",
      "10/10 [==============================] - 0s 20ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.9799\n",
      " 1 7\n",
      "Epoch 7/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 7.083047}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 7.0830step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.86409}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.995276}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.8993917}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.9605    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.6394267}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.9516053}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.942283}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.9107step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.964152}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.9739466}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 7.028326}\n",
      "10/10 [==============================] - 0s 26ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.9342\n",
      "Epoch 8/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.915406}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.9154step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 7.0336924}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.96316}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.9708    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.949279}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.9654step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.871349}\n",
      "{'batch': 5, 'size': 1, 'loss': 6.8552146}step ...... 1 7\n",
      "\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.9314step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 7.0188336}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.9438step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 7.2919908}\n",
      "{'batch': 8, 'size': 1, 'loss': 7.148823}step ......\n",
      " 1 7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 7.0053\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.921768}\n",
      "10/10 [==============================] - 0s 48ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.9970\n",
      "Epoch 9/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 7.190481}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 7.1905{'batch': 1, 'size': 1, 'loss': 6.944274}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.815302}step ......\n",
      " 1 7\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.9834    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 7.0138483}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.9910step ......{'batch': 4, 'size': 1, 'loss': 6.9950614} \n",
      "1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 7.1825724}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 7 with index :6 and 13\n",
      "{'batch': 6, 'size': 1, 'loss': 6.7118363}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.9791 1 7\n",
      "step ...... {'batch': 7, 'size': 1, 'loss': 6.943665}1 7\n",
      "\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.9746step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.8821883}\n",
      "{'batch': 9, 'size': 1, 'loss': 6.9434137}\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 46ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.9623\n",
      "Epoch 10/200\n",
      "step ...... 1{'batch': 0, 'size': 1, 'loss': 6.8491197}\n",
      " 7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.8491step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 7.0567594}\n",
      "{'batch': 2, 'size': 1, 'loss': 7.0692797}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.974535}\n",
      "step ...... 1 {'batch': 4, 'size': 1, 'loss': 6.9817395}7\n",
      "\n",
      "step ...... 1 7\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.9863    step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.8098764}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.96313}\n",
      "{'batch': 7, 'size': 1, 'loss': 6.899432}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.7852807}step ...... 1 7\n",
      "\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.9321step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 7.0306315}\n",
      "10/10 [==============================] - 0s 23ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.9420\n",
      "Epoch 11/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 6.851995}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.8520 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.918933}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.9651904}\n",
      "step ......{'batch': 3, 'size': 1, 'loss': 6.754108} 1 7\n",
      "\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.8726     1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 7.013161}\n",
      "step ...... {'batch': 5, 'size': 1, 'loss': 6.8605323}\n",
      "1 7\n",
      "step ......{'batch': 6, 'size': 1, 'loss': 6.861758}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.8894 1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.9614534}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.8984step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 7.000537}\n",
      "{'batch': 9, 'size': 1, 'loss': 6.689862}\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 30ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.8878\n",
      "Epoch 12/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.938732}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.9387step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.9728246}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.995512}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.9690    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 7.0655766}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.837222}\n",
      "step ......{'batch': 5, 'size': 1, 'loss': 7.112841}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.98711 7\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.9457154}\n",
      "step ...... 1 {'batch': 7, 'size': 1, 'loss': 6.9253044}\n",
      "7\n",
      "{'batch': 8, 'size': 1, 'loss': 7.094488}step ...... 1 7\n",
      "\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.9876step ......{'batch': 9, 'size': 1, 'loss': 6.7846236}\n",
      " 1 7\n",
      "10/10 [==============================] - 0s 28ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.9673\n",
      "Epoch 13/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 7.131296}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 7.1313step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.80401}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.8354545}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.9236     1 7\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.77205}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.8857step ...... 1 {'batch': 4, 'size': 1, 'loss': 6.943574}\n",
      "7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.7308755}\n",
      "step ...... 1 7\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.8695{'batch': 6, 'size': 1, 'loss': 6.9542212}step ...... 1 \n",
      "7\n",
      "{'batch': 7, 'size': 1, 'loss': 7.128174}step ...... 1 7\n",
      "Found change point: 8 in batch 9 with index :2 and 9\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.9125\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.842224}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 7.118754}\n",
      "10/10 [==============================] - 0s 44ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.9261\n",
      "Epoch 14/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 7.077984}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 7.0780step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.978712}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.771834}step ...... 1 7\n",
      "\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.9428    {'batch': 3, 'size': 1, 'loss': 7.233482}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.602443}\n",
      "step ...... 1 {'batch': 5, 'size': 1, 'loss': 6.666321}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.8885\n",
      "{'batch': 6, 'size': 1, 'loss': 6.622613}\n",
      "{'batch': 7, 'size': 1, 'loss': 7.260397}\n",
      "step ...... 1 {'batch': 8, 'size': 1, 'loss': 6.882034}\n",
      "7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.8995step ...... 1 7\n",
      "Found change point: 8 in batch 8 with index :6 and 13\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 6.9158683} 1\n",
      "10/10 [==============================]  - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 6.90127\n",
      "10/10 [==============================] - 0s 33ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.9012\n",
      "step ...... 1 Epoch 15/200\n",
      "7\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.743442}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.7434step ...... 1 7\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 3 with index :4 and 11{'batch': 1, 'size': 1, 'loss': 6.704015}\n",
      "\n",
      " 2/10 [=====>........................] - ETA: 0s - batch: 0.5000 - size: 1.0000 - loss: 6.7237    step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 7.0493555}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.8323step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.906354}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.899179}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.8605step ...... 1{'batch': 5, 'size': 1, 'loss': 6.79179}\n",
      " 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.8031335}\n",
      "step ...... 1 7\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.8425step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.9019094}\n",
      "step ...... 1 7{'batch': 8, 'size': 1, 'loss': 6.881317}\n",
      "\n",
      "{'batch': 9, 'size': 1, 'loss': 7.2101936}\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 6.8891step ...... 1 7\n",
      "10/10 [==============================] - 0s 42ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.8891\n",
      "Epoch 16/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.891796}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.8918step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 7.0958796}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.8421254}\n",
      "step ...... {'batch': 3, 'size': 1, 'loss': 6.651069}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.8702    1 7\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.977221}\n",
      "{'batch': 5, 'size': 1, 'loss': 6.8138433}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.87871 7\n",
      "step ...... 1 7\n",
      "step ......{'batch': 6, 'size': 1, 'loss': 6.9455323}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.8882 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.921701}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.8924step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.938147}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.8658457}\n",
      "10/10 [==============================] - 0s 40ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.8943\n",
      "Epoch 17/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.802453}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.8025step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.918489}\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 6.978423}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.8998    1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.6958647}step ...... 1 7\n",
      "\n",
      "{'batch': 4, 'size': 1, 'loss': 6.825487}\n",
      "step ...... 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.8441 1 7\n",
      "Found change point: 8 in batch 6 with index :6 and 13\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.832361}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.8422step ......{'batch': 6, 'size': 1, 'loss': 6.8950744}\n",
      " 1 7\n",
      "step ......{'batch': 7, 'size': 1, 'loss': 6.857413}\n",
      " 1 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.8507step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.903048}\n",
      "{'batch': 9, 'size': 1, 'loss': 6.933864}step ...... 1 7\n",
      "\n",
      "10/10 [==============================] - 0s 41ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.8642\n",
      "Epoch 18/200\n",
      "{'batch': 0, 'size': 1, 'loss': 6.819585}step ...... 1 7\n",
      "\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.8196step ......{'batch': 1, 'size': 1, 'loss': 6.5522833}\n",
      " 1 7\n",
      " 2/10 [=====>........................] - ETA: 0s - batch: 0.5000 - size: 1.0000 - loss: 6.6859    step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.7983274}\n",
      "{'batch': 3, 'size': 1, 'loss': 6.900436}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.774071}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.871304}step ...... 1 7\n",
      "\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.7860step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.7307377}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 7.0136175}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.9186506}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.8199step ...... 1 {'batch': 9, 'size': 1, 'loss': 6.7273817}\n",
      "7\n",
      "10/10 [==============================] - 0s 24ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.8106\n",
      "Epoch 19/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.7808743}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.7809step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.9629927}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.8600607}\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.6688695}\n",
      "{'batch': 4, 'size': 1, 'loss': 6.543587}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.7633    step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.6772757}\n",
      "step ...... 1{'batch': 6, 'size': 1, 'loss': 6.9416585} 7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.770343}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.7757step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.7876706}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.761809}\n",
      "10/10 [==============================] - 0s 23ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.7755\n",
      "Epoch 20/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.8653536}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.8654{'batch': 1, 'size': 1, 'loss': 6.800574}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.8438945}\n",
      "step ......{'batch': 3, 'size': 1, 'loss': 6.825509} 1 7\n",
      "\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.8338    step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.8836074}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.79683}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.8360step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.665143}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.753388}\n",
      "{'batch': 8, 'size': 1, 'loss': 6.710777}\n",
      "step ...... 1 7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.7939step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.7524357}\n",
      "10/10 [==============================] - 0s 27ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.7898\n",
      "Epoch 21/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.792221}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.7922step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.8400407}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.7694774}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.7894835}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.7978    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 7.168479}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 7 with index :5 and 12\n",
      "{'batch': 5, 'size': 1, 'loss': 6.90552}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.5484886}\n",
      "step ......{'batch': 7, 'size': 1, 'loss': 6.8693066}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.8354 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.8322678}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.6491623}\n",
      "10/10 [==============================] - 0s 23ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.8164\n",
      "Epoch 22/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.8911614}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.8912step ......{'batch': 1, 'size': 1, 'loss': 6.8183856}\n",
      " 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.6558876}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.680866}\n",
      "step ......{'batch': 4, 'size': 1, 'loss': 6.726122}\n",
      " 1 7\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.7545    step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.8810773}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.7783318}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.6887774}\n",
      "step .... 1 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.7651step ......{'batch': 8, 'size': 1, 'loss': 6.6950164}\n",
      " 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.9380913}\n",
      "10/10 [==============================] - 0s 18ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.7754\n",
      "Epoch 23/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.8731627}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.8732step ...... 1{'batch': 1, 'size': 1, 'loss': 6.740601}\n",
      " 7\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.8283954}\n",
      "{'batch': 3, 'size': 1, 'loss': 6.768378}step ...... 1 7\n",
      "\n",
      "Found change point: 8 in batch 5 with index :3 and 10\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.8026    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.8234587}\n",
      "step ......{'batch': 5, 'size': 1, 'loss': 6.8730006}\n",
      " 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 7.0018497}\n",
      "{'batch': 7, 'size': 1, 'loss': 6.7938704}step ......\n",
      "\b1 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.8378step ...... 1 7\n",
      "Found change point: 8 in batch 9 with index :8 and 15\n",
      "step ...... 1 {'batch': 8, 'size': 1, 'loss': 6.7480426}7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.7420564}\n",
      "10/10 [==============================] - 0s 24ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.8193\n",
      "Epoch 24/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.8320074}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.8320step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.8669796}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.799188}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 5 with index :1 and 8\n",
      "{'batch': 3, 'size': 1, 'loss': 6.8148627}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.816262}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.8259    step ......{'batch': 5, 'size': 1, 'loss': 6.8608756}\n",
      " 1 7\n",
      "Found change point: 8 in batch 7 with index :3 and 10{'batch': 6, 'size': 1, 'loss': 6.7954984}\n",
      "\n",
      "{'batch': 7, 'size': 1, 'loss': 6.633941}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.786922}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.8007step ...... 1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 1 with index :2 and 9\n",
      "{'batch': 9, 'size': 1, 'loss': 6.682073}\n",
      "10/10 [==============================] - 0s 19ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.7889\n",
      "Epoch 25/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.757295}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.7573step ...... 1 7\n",
      "Found change point: 8 in batch 3 with index :3 and 10\n",
      "{'batch': 1, 'size': 1, 'loss': 6.676568}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.7268257}\n",
      "step ...... 1 7{'batch': 3, 'size': 1, 'loss': 6.7148232}\n",
      "\n",
      "{'batch': 4, 'size': 1, 'loss': 6.7879124}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.7327    step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.7988815}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 7 with index :2 and 9\n",
      "{'batch': 6, 'size': 1, 'loss': 6.8052206}\n",
      "step ......{'batch': 7, 'size': 1, 'loss': 6.7734637} 1 7\n",
      "\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.7551 1 7\n",
      "step ......{'batch': 8, 'size': 1, 'loss': 6.7962084}\n",
      " 1 7{'batch': 9, 'size': 1, 'loss': 6.8112397}\n",
      "\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 21ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.7648\n",
      "Epoch 26/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.772498}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.7725step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.5198784}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.799824}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.904258}\n",
      "{'batch': 4, 'size': 1, 'loss': 6.6167154}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.7226    1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.842123}\n",
      "step ......{'batch': 6, 'size': 1, 'loss': 6.7673144}\n",
      " 1 7\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.7461step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.647245}\n",
      "{'batch': 8, 'size': 1, 'loss': 6.7987494}\n",
      "step ...... 1 7\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 6.9669423}\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 6.76361 7\n",
      "10/10 [==============================] - 0s 20ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.7636\n",
      "Epoch 27/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 6.7994328} 1 7\n",
      "\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.7994step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.6548896}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.7459545}\n",
      "step ......{'batch': 3, 'size': 1, 'loss': 6.724102}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.7311     1 7\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.9618316}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.8002386}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.9230957}step ......\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.80141 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.874909}\n",
      "{'batch': 8, 'size': 1, 'loss': 6.6951094}\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 6.713737} \n",
      "1 7\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 6.7893step ...... 1 7\n",
      "10/10 [==============================] - 0s 25ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.7893\n",
      "Epoch 28/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.6999235}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.6999step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.78051}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.8234925}step ...... 1 7\n",
      "Found change point: 8 in batch 4 with index :6 and 13\n",
      "\n",
      "step ......{'batch': 3, 'size': 1, 'loss': 6.7763376} 1 7\n",
      "\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.7701    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.7538414}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.8670864}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.676823}step ...... 1 7\n",
      "\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.7683step ......{'batch': 7, 'size': 1, 'loss': 6.6009283} 1 7\n",
      "\n",
      "step ......{'batch': 8, 'size': 1, 'loss': 6.748943} 1 7\n",
      "\n",
      "{'batch': 9, 'size': 1, 'loss': 6.7413344}\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 19ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.7469\n",
      "Epoch 29/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.798416}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.7984step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.776308}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.7361317}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.7703004}step ......\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.7703    1 7\n",
      "Found change point: 8 in batch 5 with index :5 and 12\n",
      "{'batch': 4, 'size': 1, 'loss': 6.7449813}step ...... 1 7\n",
      "\n",
      "{'batch': 5, 'size': 1, 'loss': 6.8175855}step ...... 1 7\n",
      "\n",
      "{'batch': 6, 'size': 1, 'loss': 6.7902703}step ...... 1 7\n",
      "\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.7763step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.745153}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.7443237}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.896628}\n",
      "10/10 [==============================] - 0s 24ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.7820\n",
      "Epoch 30/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.6432056}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.6432step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.8097167}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.7576237}\n",
      "{'batch': 3, 'size': 1, 'loss': 6.6864314}\n",
      "step ...... 1 7\n",
      "step ...... {'batch': 4, 'size': 1, 'loss': 6.7121253}1 7\n",
      "\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.7218    step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.697499}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.821708}\n",
      "step ...... {'batch': 7, 'size': 1, 'loss': 6.671392}\n",
      "1 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.7250step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.7160206}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.769921}\n",
      "10/10 [==============================] - 0s 21ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.7286\n",
      "Epoch 31/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.6604342}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.6604step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.6916027}\n",
      "step ...... 1 7{'batch': 2, 'size': 1, 'loss': 6.842093}\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.803708}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.7495    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.838367}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.73335}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.7632995}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.7618step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.808114}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.730888}\n",
      "step ...... {'batch': 9, 'size': 1, 'loss': 6.7904277}\n",
      "1 7\n",
      "10/10 [==============================] - 0s 23ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.7662\n",
      "Epoch 32/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.9160376}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.9160step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.68062}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.538318}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.7117    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.7267847}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.6658077}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.7055step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.6060443}\n",
      "step ......{'batch': 6, 'size': 1, 'loss': 6.4655304}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.6570 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.6923957}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.6614step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.7513275}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.619174}\n",
      "10/10 [==============================] - 0s 46ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.6662\n",
      "Epoch 33/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.737359}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.7374step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.6732373}\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 6.653821} 1 7\n",
      "\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.6881    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.6979146}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.6101747}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.6745step ......{'batch': 5, 'size': 1, 'loss': 6.732832}\n",
      " 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.7832623}\n",
      "step ...... 1 7\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.6984step ......{'batch': 7, 'size': 1, 'loss': 6.673618}\n",
      " 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.759863}step ...... 1 \n",
      "7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.7025step ......{'batch': 9, 'size': 1, 'loss': 6.6383696} 1 7\n",
      "\n",
      "10/10 [==============================] - 0s 41ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.6960\n",
      "Epoch 34/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.828803}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.8288step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.569767}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.6360073}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.808844}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.7251945}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.7137    step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.857937}\n",
      "step ......{'batch': 6, 'size': 1, 'loss': 6.6109953}\n",
      " 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.6325636}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.7088step ...... 1 7{'batch': 8, 'size': 1, 'loss': 6.824421}\n",
      "\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 6.7003384}\n",
      " 1 7\n",
      "10/10 [==============================] - 0s 22ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.7195\n",
      "Epoch 35/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.63047}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.6305step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.722704}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.673876}\n",
      "step ...... 1 {'batch': 3, 'size': 1, 'loss': 6.7211475}\n",
      "7\n",
      "Found change point: 8 in batch 5 with index :5 and 12\n",
      "{'batch': 4, 'size': 1, 'loss': 6.6808095}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.6858    step ...... 1 7\n",
      "step ......{'batch': 5, 'size': 1, 'loss': 6.662595}\n",
      " 1 7\n",
      "Found change point: 8 in batch 7 with index :4 and 11\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 8 with index :8 and 15\n",
      "{'batch': 6, 'size': 1, 'loss': 6.4657907}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.888781}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.6808step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.7098207}\n",
      "{'batch': 9, 'size': 1, 'loss': 6.765723}\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 20ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.6922\n",
      "Epoch 36/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.6333256}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.6333step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.52282}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.6641006}\n",
      "step ......{'batch': 3, 'size': 1, 'loss': 6.7123804}\n",
      " 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.667151}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.6400    step ...... 1 7\n",
      "step ......{'batch': 5, 'size': 1, 'loss': 6.739877}\n",
      " 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.670496}step ......\n",
      " 1 7\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.6586step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.69088}\n",
      "{'batch': 8, 'size': 1, 'loss': 6.8352757}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.6510477}\n",
      "10/10 [==============================] - 0s 22ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.6787\n",
      " 1 7\n",
      "Epoch 37/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.523267}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.5233step ......{'batch': 1, 'size': 1, 'loss': 6.6908994}\n",
      " 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.6793213}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.7337213}step ...... 1 7\n",
      "\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.6568    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.6144705}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.6366296}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.781625}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.6657step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.679665}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.5470276}\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 6.5812063}\n",
      " 1 7\n",
      "10/10 [==============================] - 0s 23ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.6468\n",
      "Epoch 38/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.741405}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.7414step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.570344}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.6963573}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.7236276}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.6829    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.7084656}\n",
      "{'batch': 5, 'size': 1, 'loss': 6.5402555}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.9532194}\n",
      "step ...... 1 7\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.7048step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.7248006}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.821513}\n",
      "{'batch': 9, 'size': 1, 'loss': 6.7047286}\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 23ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.7185\n",
      "Epoch 39/200\n",
      "step ...... {'batch': 0, 'size': 1, 'loss': 6.8116417}\n",
      "1 7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.8116step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.581495}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.67035}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.4465237}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.6275    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.7337766}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.666658}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.630289}\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.6416335}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.6478step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.7359905}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 1 with index :4 and 11\n",
      "{'batch': 9, 'size': 1, 'loss': 6.7002845}\n",
      "10/10 [==============================] - 0s 22ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.6619\n",
      "Epoch 40/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.7108693}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.7109step ......{'batch': 1, 'size': 1, 'loss': 6.5724} 1 7\n",
      "\n",
      "{'batch': 2, 'size': 1, 'loss': 6.6549873}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.6111283}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.531535}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.6162    step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.848142}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.497159}\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.5275936}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.6192step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.496061}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.65681}\n",
      "Found change point: 8 in batch 1 with index :2 and 9\n",
      "10/10 [==============================] - 0s 21ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.6107\n",
      "Epoch 41/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.640375}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.6404step ...... 1 7\n",
      "Found change point: 8 in batch 3 with index :8 and 15\n",
      "{'batch': 1, 'size': 1, 'loss': 6.6355705}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.610466}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.5897746}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.6190    step ...... 1 7\n",
      "Found change point: 8 in batch 6 with index :1 and 8\n",
      "{'batch': 4, 'size': 1, 'loss': 6.43792}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.6461277}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.7176566}step ...... 1 7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.7177796}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.6245step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.7897625}\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 6.521003} 1 7\n",
      "\n",
      "10/10 [==============================] - 0s 21ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.6306\n",
      "Epoch 42/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.64873}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.6487{'batch': 1, 'size': 1, 'loss': 6.5849}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 3 with index :3 and 10\n",
      "{'batch': 2, 'size': 1, 'loss': 6.652353}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.745177}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.4795036}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.6221    Found change point: 8 in batch 5 with index :3 and 10\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.655042}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.634107}step ......\n",
      " 1 7\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 9 with index :1 and 8\n",
      "{'batch': 7, 'size': 1, 'loss': 6.559192}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.6199step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.636988}\n",
      "{'batch': 9, 'size': 1, 'loss': 6.5997925}\n",
      "1 7p .....\n",
      "10/10 [==============================] - 0s 20ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.6196\n",
      "Epoch 43/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 6.6027975} \n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.60281 7\n",
      "Found change point: 8 in batch 2 with index :1 and 8\n",
      "step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.5996203}\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 6.666366}\n",
      " 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.887897}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.6892    step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.6649804}\n",
      "step ......{'batch': 5, 'size': 1, 'loss': 6.7048864} \n",
      "1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.566341}\n",
      "step ...... 1 7\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.6704step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.6270432}\n",
      "{'batch': 8, 'size': 1, 'loss': 6.646246}step ......\n",
      " 1 7\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 6.625745}\n",
      "10/10 [==============================] - 0s 18ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.6592\n",
      " 1 7\n",
      "Epoch 44/200\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 2 with index :4 and 11\n",
      "{'batch': 0, 'size': 1, 'loss': 6.596441}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.5964step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.6442027}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.699116}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.6560636}\n",
      "{'batch': 4, 'size': 1, 'loss': 6.569554}\n",
      "step ...... 1 7\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.6331    step ...... 1 7\n",
      "Found change point: 8 in batch 7 with index :5 and 12\n",
      "{'batch': 5, 'size': 1, 'loss': 6.7693844}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.5061507}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.8099747}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.6564step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.588664}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.5060315}\n",
      "10/10 [==============================] - 0s 22ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.6346\n",
      "Epoch 45/200\n",
      "step ...... 1 7{'batch': 0, 'size': 1, 'loss': 6.6277103}\n",
      "\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.6277step ...... 1 7{'batch': 1, 'size': 1, 'loss': 6.6139717}\n",
      "\n",
      "{'batch': 2, 'size': 1, 'loss': 6.5658865}step ......\n",
      " 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.6756134}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.5876064}\n",
      "{'batch': 5, 'size': 1, 'loss': 6.7273717}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.6330    step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.6050262}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.5903716}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.766766}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.6400step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.687573}\n",
      "10/10 [==============================] - 0s 15ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.6448\n",
      "step ...... 1 7\n",
      "Epoch 46/200\n",
      "{'batch': 0, 'size': 1, 'loss': 6.656858}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.6569step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.628324}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.559346}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 4 with index :4 and 11\n",
      "{'batch': 3, 'size': 1, 'loss': 6.648247}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.6048265}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.6195    step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.7226243}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.5356107}\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.602824}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.6198step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.4636183}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.6865745}\n",
      "10/10 [==============================] - 0s 18ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.6109\n",
      "Epoch 47/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.6462793}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.6463step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.661288}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.3936276}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.5999303}\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.5932055}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.5789    step ......{'batch': 5, 'size': 1, 'loss': 6.582051}\n",
      " 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.5143194}step ...... 1 7\n",
      "\n",
      "{'batch': 7, 'size': 1, 'loss': 6.5583367}\n",
      "{'batch': 8, 'size': 1, 'loss': 6.596497}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.5717step ...... 1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.634401}\n",
      "10/10 [==============================] - 0s 22ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.5780\n",
      "Epoch 48/200\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 2 with index :4 and 11\n",
      "{'batch': 0, 'size': 1, 'loss': 6.4877863}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.4878{'batch': 1, 'size': 1, 'loss': 6.615157}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.689785}\n",
      "step ...... 1 7\n",
      "step ...... 1 7{'batch': 3, 'size': 1, 'loss': 6.634873}\n",
      "\n",
      "{'batch': 4, 'size': 1, 'loss': 6.6495366}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.6154    step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.5834303}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.626078}\n",
      "{'batch': 7, 'size': 1, 'loss': 6.598703}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.6107step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.551625}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.60999}\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 21ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.6047\n",
      "Epoch 49/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.4832363}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.4832{'batch': 1, 'size': 1, 'loss': 6.5889196}\n",
      "step ...... 1 {'batch': 2, 'size': 1, 'loss': 6.5415597}\n",
      "7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.397215}\n",
      "{'batch': 4, 'size': 1, 'loss': 6.632015}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.5286    step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.617842}\n",
      "step ...... 1 {'batch': 6, 'size': 1, 'loss': 6.738878}\n",
      "7\n",
      "step ......{'batch': 7, 'size': 1, 'loss': 6.6049013}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.5756 1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.5911255}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.5773step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.771303}\n",
      "10/10 [==============================] - 0s 26ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.5967\n",
      "Epoch 50/200\n",
      "{'batch': 0, 'size': 1, 'loss': 6.618568}step ...... 1 7\n",
      "\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.6186step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.8565826}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.598223}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.5998607}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.5959473}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.6538    step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.541254}\n",
      "step ......{'batch': 6, 'size': 1, 'loss': 6.599273}\n",
      " 1 7\n",
      "step ...... 1 {'batch': 7, 'size': 1, 'loss': 6.568685}7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.5051727}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.6093step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.5918703}\n",
      "10/10 [==============================] - 0s 21ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.6075\n",
      "Epoch 51/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.606075}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.6061step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.5705533}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.6047587}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.443669}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.5563    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.4823356}\n",
      "step ......{'batch': 5, 'size': 1, 'loss': 6.595701}\n",
      " 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.5892525}\n",
      "{'batch': 7, 'size': 1, 'loss': 6.4298344}\n",
      "step ...... 1 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.5403step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.387047}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.5759645}\n",
      "10/10 [==============================] - 0s 22ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.5285\n",
      "Epoch 52/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.6015325}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.6015step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.5897827}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.464985}\n",
      "{'batch': 3, 'size': 1, 'loss': 6.5112925}step ...... 1 7\n",
      "\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.5419    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.6276927}\n",
      "step ......{'batch': 5, 'size': 1, 'loss': 6.4633484} 1 7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.602019}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.5515step ......{'batch': 7, 'size': 1, 'loss': 6.624482}\n",
      " 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.716101}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.549451}\n",
      "10/10 [==============================] - 0s 22ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.5751\n",
      " 7\n",
      "Epoch 53/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.3958187}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3958step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.5884905}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.60125}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.727881}\n",
      "{'batch': 4, 'size': 1, 'loss': 6.545471}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.5718    7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.7112436}\n",
      "step ......{'batch': 6, 'size': 1, 'loss': 6.5307446} 1 7\n",
      "step ......\n",
      " 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.5508165}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.5815step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.5333843}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.5207534}\n",
      "10/10 [==============================] - 0s 23ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.5706\n",
      "Epoch 54/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.803527}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.8035step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.746922}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.504184}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.5927954}\n",
      "step ...... 1 {'batch': 4, 'size': 1, 'loss': 6.561082}\n",
      "7\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.6417    step ...... 1 7{'batch': 5, 'size': 1, 'loss': 6.5998044}\n",
      "\n",
      "{'batch': 6, 'size': 1, 'loss': 6.451749}step ......\n",
      " 1 7\n",
      "step ......{'batch': 7, 'size': 1, 'loss': 6.5887322} 1 7\n",
      "\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.6061step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.6109986}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.5721984}\n",
      "10/10 [==============================] - 0s 21ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.6032\n",
      "Epoch 55/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.4731483}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.4731step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.4501934}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.671927}\n",
      "step ...... 1 {'batch': 3, 'size': 1, 'loss': 6.5811653}\n",
      "7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.5236874}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.410032}step ...... 1 7\n",
      "step ......\n",
      " 1 7\n",
      "Found change point: 8 in batch 7 with index :3 and 10\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.5184    step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.5226245}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.5594025}\n",
      "step ...... 1{'batch': 8, 'size': 1, 'loss': 6.5424776}\n",
      " 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.556333}\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 6.52917\n",
      "10/10 [==============================] - 0s 21ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.5291\n",
      "Epoch 56/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.565359}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.5654step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.507583}\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 6.492868}\n",
      " 1 7\n",
      "step ......{'batch': 3, 'size': 1, 'loss': 6.598065}\n",
      " 1 7\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.5410    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.6254225}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.4873104}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.5780115}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.5507step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.6473875}\n",
      "step ......{'batch': 8, 'size': 1, 'loss': 6.463291}\n",
      " 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.5032578}step ......\n",
      " 1 7\n",
      "10/10 [==============================] - 0s 21ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.5469\n",
      "Epoch 57/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.632664}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.6327step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.725978}\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 6.7198634} \n",
      "1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.6109138}step ...... 1 7\n",
      "\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.6724    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.5811753}\n",
      "{'batch': 5, 'size': 1, 'loss': 6.5713}\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.5938325}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.6337step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.6042027}\n",
      "{'batch': 8, 'size': 1, 'loss': 6.49184}step ...... 1 7\n",
      "\n",
      "{'batch': 9, 'size': 1, 'loss': 6.489379}step ...... \n",
      "\b 7\n",
      "10/10 [==============================] - 0s 21ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.6021\n",
      "Epoch 58/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.6059904}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.6060step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.563524}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.556839}\n",
      "{'batch': 3, 'size': 1, 'loss': 6.503901}\n",
      "{'batch': 4, 'size': 1, 'loss': 6.4802623}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.5421    1 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.5598297}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.492932}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.4612617}\n",
      "step ...... 1 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.5281step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.4993978}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.560554}\n",
      "10/10 [==============================] - 0s 20ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.5284\n",
      "Epoch 59/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.4855185}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.4855step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.5343}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.3379526}\n",
      "step ...... 1 7\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.4526    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.4812756}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.4598step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.5365314}\n",
      "{'batch': 5, 'size': 1, 'loss': 6.4879003}\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.494703}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.4797step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.435129}\n",
      "{'batch': 8, 'size': 1, 'loss': 6.3623996}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.4617step ...... 1 7\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 6.59529}\n",
      " 1 7\n",
      "10/10 [==============================] - 0s 41ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.4751\n",
      "Epoch 60/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.4366436}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.4366step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.5358253}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.478288}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.4836    {'batch': 3, 'size': 1, 'loss': 6.5649557}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 5 with index :4 and 11\n",
      "step ...... 1{'batch': 4, 'size': 1, 'loss': 6.477751}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.4987\n",
      "step ...... 1{'batch': 5, 'size': 1, 'loss': 6.530904}\n",
      " 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.592389}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.5167step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.5904303}\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.514804}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.5247step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.545748}\n",
      "10/10 [==============================] - 0s 39ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.5268\n",
      "Epoch 61/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 6.597092} 1\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.5971 7\n",
      "Found change point: 8 in batch 2 with index :2 and 9\n",
      "step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.6583805}\n",
      " 2/10 [=====>........................] - ETA: 0s - batch: 0.5000 - size: 1.0000 - loss: 6.6277    step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.4303603}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.6923165}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.5945step ......{'batch': 4, 'size': 1, 'loss': 6.515549} 1 7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.502965}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.5661step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.4454794}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.5489step ...... 1 7\n",
      "Found change point: 8 in batch 9 with index :4 and 11{'batch': 7, 'size': 1, 'loss': 6.694984}\n",
      "\n",
      "{'batch': 8, 'size': 1, 'loss': 6.539418}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.6882277}\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 6.5765step ...... 1 7\n",
      "10/10 [==============================] - 0s 44ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.5765\n",
      "Epoch 62/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.4958324}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.4958step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.52023}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.3837843}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.4666    step ......{'batch': 3, 'size': 1, 'loss': 6.5890203} 1 7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.5219097}\n",
      "step ......{'batch': 5, 'size': 1, 'loss': 6.4107666}\n",
      " 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.5871778}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.5012step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.51242}\n",
      "{'batch': 8, 'size': 1, 'loss': 6.5073247}\n",
      "step ...... 1 7{'batch': 9, 'size': 1, 'loss': 6.533309}\n",
      "\n",
      "10/10 [==============================] - 0s 23ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.5062\n",
      "step ...... 1 7\n",
      "Epoch 63/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.5195603}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.5196step ......{'batch': 1, 'size': 1, 'loss': 6.6801124}\n",
      " 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.5074625}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.5690    step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.5082645}\n",
      "step ......{'batch': 4, 'size': 1, 'loss': 6.4980087}\n",
      " 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.508635}step ...... 1 7\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.5370\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.3287477}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.4693007}\n",
      "step ......{'batch': 8, 'size': 1, 'loss': 6.5263925}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.50527\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 1 with index :8 and 15\n",
      "{'batch': 9, 'size': 1, 'loss': 6.577593}\n",
      "10/10 [==============================] - 0s 25ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.5124\n",
      "Epoch 64/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.488627}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.4886step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.491007}\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 6.654281} 1 7\n",
      "\n",
      "step ...... 1 {'batch': 3, 'size': 1, 'loss': 6.4686365}\n",
      "7\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.5111527}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.5227    step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.591857}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.530879}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.522228}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.5323step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.4338937}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.4635444}\n",
      "10/10 [==============================] - 0s 25ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.5156\n",
      "Epoch 65/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.509183}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.5092step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.770036}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.499752}step ......\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.5930    1 7\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.5008307}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.5700step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.4648137}\n",
      "{'batch': 5, 'size': 1, 'loss': 6.392295}step ...... 1 7\n",
      "\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.5228step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.589075}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.498372}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.5280step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.4948993}\n",
      "{'batch': 9, 'size': 1, 'loss': 6.6667266}\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 36ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.5386\n",
      "Epoch 66/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.569017}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.5690step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.3167334}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.5257573}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.4705    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.597511}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.583762}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.5186step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.353031}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.4620986}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.4868step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.7582583}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.5208step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.5942025}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.4410644}\n",
      "10/10 [==============================] - 0s 43ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.5201\n",
      "Epoch 67/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.3606877}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3607step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.5102553}\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 6.436053} 1 7\n",
      "\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.4357    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.633558}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.4851step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.651229}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.4873247}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.4803205}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.5085step ......{'batch': 7, 'size': 1, 'loss': 6.4844356} 1 7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.5106707}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.64782}\n",
      "Found change point: 8 in batch 1 with index :8 and \n",
      "10/10 [==============================] - 0s 29ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.5202\n",
      "Epoch 68/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.553997}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.5540step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.491174}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.541362}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.379877}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.4916    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.485337}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.4847264}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.5819926}\n",
      "{'batch': 7, 'size': 1, 'loss': 6.4799304}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.4998step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.3498354}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.5037036}\n",
      "10/10 [==============================] - 0s 22ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.4852\n",
      "Epoch 69/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.4847665}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.4848step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.578969}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.7398386}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.4295855}\n",
      "step ......{'batch': 4, 'size': 1, 'loss': 6.541704}\n",
      " 1 7\n",
      "step ......{'batch': 5, 'size': 1, 'loss': 6.5585527} 1\n",
      " 7\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.5556    step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.427789}step ...... 1 7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.454417}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.5270step ...... {'batch': 8, 'size': 1, 'loss': 6.298996}1 \n",
      "7\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.427205}\n",
      "10/10 [==============================] - 0s 22ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.4942\n",
      "Epoch 70/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.415189}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.4152step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.511218}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.4203706}\n",
      "Found change point: 8 in batch 4 with index :3 and 10\n",
      "step ......{'batch': 3, 'size': 1, 'loss': 6.4682155}\n",
      " 1 7\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.4537    step ...... {'batch': 4, 'size': 1, 'loss': 6.513152}1 7\n",
      "\n",
      "{'batch': 5, 'size': 1, 'loss': 6.4199843}step ......\n",
      " 1 7\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.4580step ...... {'batch': 6, 'size': 1, 'loss': 6.435677}1\n",
      " 7\n",
      "step ......{'batch': 7, 'size': 1, 'loss': 6.5685606}\n",
      " 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.59376}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.4829step ...... 1 7\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 6.510101}\n",
      " 1 7\n",
      "10/10 [==============================] - 0s 22ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.4856\n",
      "Epoch 71/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.4408193}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.4408step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.4729843}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 4 with index :6 and 13{'batch': 2, 'size': 1, 'loss': 6.504444}\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.486173}\n",
      "{'batch': 4, 'size': 1, 'loss': 6.4499373}\n",
      "step ...... 1 7\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.4709    step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.4165773}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.4852524}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.4588857}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.4644step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.49149}\n",
      "{'batch': 9, 'size': 1, 'loss': 6.4766645}step ...... 1 7\n",
      "\n",
      "10/10 [==============================] - 0s 20ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.4683\n",
      "Epoch 72/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.411487}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.4115step ......{'batch': 1, 'size': 1, 'loss': 6.386567} 1 7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.362437}\n",
      "{'batch': 3, 'size': 1, 'loss': 6.4777546}\n",
      "step ...... 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.4096     1 7\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.432536}\n",
      "step ......{'batch': 5, 'size': 1, 'loss': 6.5993404}\n",
      " 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.479098}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.4499step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.498201}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.4069586}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.394935}\n",
      "10/10 [==============================] - 0s 23ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.4449\n",
      "Epoch 73/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 6.494946} 1 \n",
      "7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.4949step ...... 1 7\n",
      "Found change point: 8 in batch 3 with index :8 and 15\n",
      "{'batch': 1, 'size': 1, 'loss': 6.4965477}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.496839}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.4961    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.4748716}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.451723}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.4830step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.403685}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.3247623}\n",
      "step ...... 1{'batch': 7, 'size': 1, 'loss': 6.403542} 7\n",
      "\n",
      "step ......{'batch': 8, 'size': 1, 'loss': 6.5058517}\n",
      " 1 7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.4503step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.548058}\n",
      "10/10 [==============================] - 0s 26ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.4601\n",
      "Epoch 74/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.369516}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3695step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.4524126}\n",
      "step ...... {'batch': 2, 'size': 1, 'loss': 6.4534545}\n",
      "1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.4475985}step ...... 1 7\n",
      "Found change point: 8 in batch 5 with index :2 and 9\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.502245}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.4450    step ...... 1 {'batch': 5, 'size': 1, 'loss': 6.4447227}\n",
      "7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.5859356}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.4723263}step ...... 1 7\n",
      "\n",
      "{'batch': 8, 'size': 1, 'loss': 6.457929}step ...... 1 7\n",
      "\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.4651step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.3088117}\n",
      "10/10 [==============================] - 0s 16ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.4495\n",
      "Epoch 75/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.3082757}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3083step ...... 1 {'batch': 1, 'size': 1, 'loss': 6.4028935}\n",
      "7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.421715}\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.4967175}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.4074    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.4385633}\n",
      "step ...... 1 {'batch': 5, 'size': 1, 'loss': 6.374559}\n",
      "7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.458137}step ...... 1 7\n",
      "\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.4144step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.4513826}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.4231386}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.4195step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.5062118}\n",
      "10/10 [==============================] - 0s 31ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.4282\n",
      "Epoch 76/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.4642987}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.4643step ...... 1 7\n",
      "Found change point: 8 in batch 3 with index :5 and 12\n",
      "{'batch': 1, 'size': 1, 'loss': 6.452674}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.5554466}step ...... 1 \n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.4908    7\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.4485755}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.4802step ...... 1{'batch': 4, 'size': 1, 'loss': 6.375048} 7\n",
      "\n",
      "Found change point: 8 in batch 6 with index :6 and 13\n",
      "{'batch': 5, 'size': 1, 'loss': 6.388335}\n",
      "step ...... 1 7\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.4474step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.463212}\n",
      "{'batch': 7, 'size': 1, 'loss': 6.511068}\n",
      "step ...... 1 7\n",
      "step ......{'batch': 8, 'size': 1, 'loss': 6.528657}\n",
      " 1 7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.4653step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.297994}\n",
      "10/10 [==============================] - 0s 28ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.4485\n",
      "Epoch 77/200\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 2 with index :6 and 13\n",
      "{'batch': 0, 'size': 1, 'loss': 6.391129}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3911step ...... 1{'batch': 1, 'size': 1, 'loss': 6.4573975}\n",
      " 7\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.413136}\n",
      "step ...... {'batch': 3, 'size': 1, 'loss': 6.344968}\n",
      "1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.326394}\n",
      "step ...... 1 7\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.3866    step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.4455323}\n",
      "step ......{'batch': 6, 'size': 1, 'loss': 6.370997}\n",
      " 1 7\n",
      "Found change point: 8 in batch 8 with index :5 and 12\n",
      "{'batch': 7, 'size': 1, 'loss': 6.3815355}\n",
      "step ...... 1{'batch': 8, 'size': 1, 'loss': 6.379061}\n",
      " 7\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 6.422475}\n",
      " 1 7\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 21ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.3933\n",
      "Epoch 78/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.508363}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.5084step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.441834}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.428223}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.481331}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.4649    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.42108}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.4098268}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.437236}\n",
      "{'batch': 7, 'size': 1, 'loss': 6.4040904}\n",
      "step ...... 1 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.4415step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.350428}\n",
      "step ...... {'batch': 9, 'size': 1, 'loss': 6.474119}1 7\n",
      "\n",
      "10/10 [==============================] - 0s 20ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.4357\n",
      "Epoch 79/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.4455924}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.4456step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.671537}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.4056864}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 4 with index :8 and 15\n",
      "{'batch': 3, 'size': 1, 'loss': 6.4015226}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.3786373}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.4606    step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.378029}\n",
      "step ......{'batch': 6, 'size': 1, 'loss': 6.4277864}\n",
      " 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.4438305}\n",
      "{'batch': 8, 'size': 1, 'loss': 6.455427}\n",
      "step ...... 1 7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.4453 1 7\n",
      "Found change point: 8 in batch 10 with index :2 and 9\n",
      "step ...... 1 {'batch': 9, 'size': 1, 'loss': 6.3575897}\n",
      "7\n",
      "10/10 [==============================] - 0s 20ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.4366\n",
      "Epoch 80/200\n",
      "step ...... {'batch': 0, 'size': 1, 'loss': 6.496878}1 7\n",
      "\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.4969step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.4159966}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.246721}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.4164023}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.3940    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.662429}\n",
      "{'batch': 5, 'size': 1, 'loss': 6.3979526}\n",
      "step ...... 1 7\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.4394step ...... {'batch': 6, 'size': 1, 'loss': 6.405399}1 7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.4191723}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.4326step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.4055314}Found change point: 8 in batch 10 with index :4 and 11\n",
      "\n",
      "{'batch': 9, 'size': 1, 'loss': 6.4044433}\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 38ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.4271\n",
      "Epoch 81/200\n",
      "step ...... 1 {'batch': 0, 'size': 1, 'loss': 6.357576}\n",
      " 1/10 [==>...........................]7\n",
      " - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3576step ...... 1 7\n",
      "Found change point: 8 in batch 3 with index :3 and 10\n",
      "{'batch': 1, 'size': 1, 'loss': 6.275012}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.463635}\n",
      "step ......{'batch': 3, 'size': 1, 'loss': 6.5660863}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.4156     1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.445164}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.4191337}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.3731813}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.4143step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.3908715}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.4113step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.4037714}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.3953633}\n",
      "10/10 [==============================] - 0s 34ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.4090\n",
      "Epoch 82/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.4113092}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.4113{'batch': 1, 'size': 1, 'loss': 6.354806}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.41833}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.3948    step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.4940915}\n",
      "{'batch': 4, 'size': 1, 'loss': 6.4347057}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.4226step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.3861403}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.445015}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.4206Found change point: 8 in batch 8 with index :5 and 12\n",
      "{'batch': 7, 'size': 1, 'loss': 6.4776645}step ......\n",
      " 1 7\n",
      "Found change point: 8 in batch 9 with index :3 and 10\n",
      "{'batch': 8, 'size': 1, 'loss': 6.3757114}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.4220{'batch': 9, 'size': 1, 'loss': 6.490163}\n",
      "10/10 [==============================] - 0s 36ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.4288\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "Epoch 83/200\n",
      "{'batch': 0, 'size': 1, 'loss': 6.378442}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3784step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.3263426}\n",
      " 2/10 [=====>........................] - ETA: 0s - batch: 0.5000 - size: 1.0000 - loss: 6.3524    step ......{'batch': 2, 'size': 1, 'loss': 6.3255105}\n",
      " 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.243625}\n",
      "step ...... 1 7\n",
      "step ...... 1{'batch': 4, 'size': 1, 'loss': 6.3527703}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.3253 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.346338}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.3596025}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 8 with index :2 and 9\n",
      "{'batch': 7, 'size': 1, 'loss': 6.373559}\n",
      "step ...... 1 {'batch': 8, 'size': 1, 'loss': 6.4082303}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.34607\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.3861904}\n",
      "10/10 [==============================] - 0s 26ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.3501\n",
      "Epoch 84/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.293936}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.2939step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.406971}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.342073}\n",
      "step ...... 1 7\n",
      "step ...... 1 {'batch': 3, 'size': 1, 'loss': 6.311745}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.3387    \n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.3191767}\n",
      "step ......{'batch': 5, 'size': 1, 'loss': 6.328153}\n",
      " 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.3954644}\n",
      "step ...... 1 7\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.3425step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.380769}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.39637}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.301237}\n",
      "10/10 [==============================] - 0s 23ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.3476\n",
      "Epoch 85/200\n",
      "step ...... {'batch': 0, 'size': 1, 'loss': 6.302836}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3028 7\n",
      "step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.383791}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.473975}\n",
      "step ......{'batch': 3, 'size': 1, 'loss': 6.341087} 1 7\n",
      "\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.3754    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.471971}\n",
      "step ...... {'batch': 5, 'size': 1, 'loss': 6.3921776}\n",
      "1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.311599}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.3784056}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.3820\n",
      "step ......{'batch': 8, 'size': 1, 'loss': 6.3798056} 1 7\n",
      "\n",
      "step ...... 1 {'batch': 9, 'size': 1, 'loss': 6.218539}\n",
      "7\n",
      "10/10 [==============================] - 0s 19ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.3654\n",
      "Epoch 86/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.393899}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3939{'batch': 1, 'size': 1, 'loss': 6.386561}step ...... 1 7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.4297915}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.373873}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.3960    step ......{'batch': 4, 'size': 1, 'loss': 6.3068976}\n",
      " 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.320765}\n",
      "step ......{'batch': 6, 'size': 1, 'loss': 6.34256}\n",
      " 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.4086003}step ...... 1 7\n",
      "\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.3704step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.3413353}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.5253525}\n",
      "10/10 [==============================] - 0s 17ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.3830\n",
      "Epoch 87/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.506949}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.5069step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.3172674}\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 6.302887} 1 7\n",
      "\n",
      "{'batch': 3, 'size': 1, 'loss': 6.338882}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.3665    step ...... 1 7\n",
      "Found change point: 8 in batch 5 with index :3 and 10\n",
      "step ......{'batch': 4, 'size': 1, 'loss': 6.458766} 1 7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.3688035}\n",
      "step ...... {'batch': 6, 'size': 1, 'loss': 6.590608}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.4120 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.3317494}\n",
      "{'batch': 8, 'size': 1, 'loss': 6.481426}step ...... 1 7\n",
      "\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.4108step ...... 1 {'batch': 9, 'size': 1, 'loss': 6.5670195}\n",
      "7\n",
      "10/10 [==============================] - 0s 20ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.4264\n",
      "Epoch 88/200\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 2 with index :7 and 14{'batch': 0, 'size': 1, 'loss': 6.3100653}\n",
      "\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3101step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.3239384}\n",
      "step ...... {'batch': 2, 'size': 1, 'loss': 6.564848}1 7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.4277735}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.4067    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.379379}\n",
      "{'batch': 5, 'size': 1, 'loss': 6.3807764}\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.4963183}\n",
      "step ......{'batch': 7, 'size': 1, 'loss': 6.3452964}\n",
      " 1 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.4035step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.355436}\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 6.213602}\n",
      " 1 7\n",
      "10/10 [==============================] - 0s 22ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.3797\n",
      "Epoch 89/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.327593}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3276{'batch': 1, 'size': 1, 'loss': 6.380412}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.307001}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.446915}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.6011977}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.4126    step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.5768266}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.2814736}\n",
      "step ......{'batch': 7, 'size': 1, 'loss': 6.3193355} 1 7\n",
      "\n",
      "{'batch': 8, 'size': 1, 'loss': 6.345604}\n",
      "1 7p .....\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.3985step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.375909}\n",
      "10/10 [==============================] - 0s 20ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.3962\n",
      "Epoch 90/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.32497}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3250step ...... 1 7{'batch': 1, 'size': 1, 'loss': 6.3257694}\n",
      "\n",
      "{'batch': 2, 'size': 1, 'loss': 6.3959074}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.353581}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.3501    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.279842}step ...... 1 7\n",
      "\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.3360step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.283454}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.3038154}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.3239step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.3131495}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.351156}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.3257step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.3195343}\n",
      "10/10 [==============================] - 0s 40ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.3251\n",
      "Epoch 91/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.4362273}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.4362step ...... {'batch': 1, 'size': 1, 'loss': 6.409364}\n",
      "1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.36096}\n",
      "step ...... 1 7\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.4022    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.328761}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.331603}\n",
      "{'batch': 5, 'size': 1, 'loss': 6.3550544}\n",
      "step ...... 1 7\n",
      "step ...... 1{'batch': 6, 'size': 1, 'loss': 6.330304}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.36467\n",
      "{'batch': 7, 'size': 1, 'loss': 6.375909}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.4300656}\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.3629436}\n",
      "10/10 [==============================] - 0s 23ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.3721\n",
      "Epoch 92/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 6.329216}\n",
      " 1 7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3292step ...... {'batch': 1, 'size': 1, 'loss': 6.2615266}\n",
      "1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.3263803}step ...... 1 7\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.3057    \n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.39386}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.3277step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.3597746}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.4001875}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.3452step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.3618345}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.4066277}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.3549step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.2628646}\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 6.485387}\n",
      " 1 7\n",
      "10/10 [==============================] - 0s 38ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.3588\n",
      "Epoch 93/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.349783}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3498step ...... 1 7{'batch': 1, 'size': 1, 'loss': 6.180667}\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.3675113}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.2993    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.365284}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.3158step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.346136}\n",
      "{'batch': 5, 'size': 1, 'loss': 6.25547}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.480729}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.3351step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.384652}\n",
      "step ......{'batch': 8, 'size': 1, 'loss': 6.391603}\n",
      " 1 7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.3469step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.3290796}\n",
      "10/10 [==============================] - 0s 35ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.3451\n",
      "Epoch 94/200\n",
      "{'batch': 0, 'size': 1, 'loss': 6.308533}step ...... 1 7\n",
      "\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3085step ......{'batch': 1, 'size': 1, 'loss': 6.3486614}\n",
      " 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.320567}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.438075}step ...... 1 7\n",
      "\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.3540    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.243183}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.3318step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.270813}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.317991}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.3211step ...... 1 7{'batch': 7, 'size': 1, 'loss': 6.25174}\n",
      "\n",
      "{'batch': 8, 'size': 1, 'loss': 6.411561}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.3235 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.3768835}\n",
      "10/10 [==============================] - 0s 41ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.3288\n",
      "Epoch 95/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.3407173}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3407{'batch': 1, 'size': 1, 'loss': 6.375597}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.2063537}step ...... 1 7\n",
      "Found change point: 8 in batch 4 with index :3 and 10\n",
      "\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.3076    step ...... 1 7\n",
      "Found change point: 8 in batch 5 with index :2 and 9\n",
      "{'batch': 3, 'size': 1, 'loss': 6.324039}\n",
      "step ...... 1 7{'batch': 4, 'size': 1, 'loss': 6.297421}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.3088\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.3349414}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.3226695}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.3145step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.261594}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.270519}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.3038step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.251072}\n",
      "10/10 [==============================] - 0s 40ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.2985\n",
      "Epoch 96/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.429575}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.4296step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.352001}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.365784}\n",
      "Found change point: 8 in batch 4 with index :6 and 13\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.3825    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.327925}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.3058453}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.3562step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.3575444}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.280958}\n",
      "step ...... 1 7\n",
      "step ......{'batch': 7, 'size': 1, 'loss': 6.399374}\n",
      " 1 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.3524step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.3438196}\n",
      "{'batch': 9, 'size': 1, 'loss': 6.282315}step ...... 1 7\n",
      "\n",
      "10/10 [==============================] - 0s 36ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.3445\n",
      "Epoch 97/200\n",
      "step ...... 1 {'batch': 0, 'size': 1, 'loss': 6.287262}\n",
      "7\n",
      "Found change point: 8 in batch 2 with index :8 and 15\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.2873step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.322854}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.22763}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.2792    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.318747}\n",
      "step ......{'batch': 4, 'size': 1, 'loss': 6.4556403} 1 \n",
      "7\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.3224step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.2327538}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.2253933}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.2958step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.315819}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.2506504}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.2930step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.3092723}\n",
      "10/10 [==============================] - 0s 39ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.2946\n",
      "Epoch 98/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.3714395}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3714step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.305273}\n",
      " 2/10 [=====>........................] - ETA: 0s - batch: 0.5000 - size: 1.0000 - loss: 6.3384    step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.3035817}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.3066287}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.3217step ...... {'batch': 4, 'size': 1, 'loss': 6.278341}\n",
      "1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.2941384}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.2574286}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.3024step ...... 1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.2451158}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.2935987}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.2951step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.1741266}\n",
      "10/10 [==============================] - 0s 33ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.2830\n",
      "Epoch 99/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.298501}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.2985step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.343747}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.240263}\n",
      "{'batch': 3, 'size': 1, 'loss': 6.314111}\n",
      "step ...... 1 7\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.2992    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.2310143}\n",
      "step ......{'batch': 5, 'size': 1, 'loss': 6.4420066}\n",
      " 1 7\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.3116step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.1548777}\n",
      "{'batch': 7, 'size': 1, 'loss': 6.2700343}\n",
      "step ...... 1 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.2868step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.2870955}\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 6.3101788}\n",
      " 1 7\n",
      "10/10 [==============================] - 0s 38ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.2892\n",
      "Epoch 100/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 6.2439013}\n",
      " 1 7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.2439step ......{'batch': 1, 'size': 1, 'loss': 6.2853456} 1 7\n",
      "\n",
      "{'batch': 2, 'size': 1, 'loss': 6.4199367}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.3164    step ...... 1 7\n",
      "Found change point: 8 in batch 4 with index :3 and 10\n",
      "{'batch': 3, 'size': 1, 'loss': 6.299554}step ...... 1 7\n",
      "\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.3122step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.3082137}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.307794}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.3108{'batch': 6, 'size': 1, 'loss': 6.2536683}\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.3089056}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.3034step ......{'batch': 8, 'size': 1, 'loss': 6.249494}\n",
      " 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.1623077}\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 41ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.2839\n",
      "Epoch 101/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 6.262702}\n",
      " 1 7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.2627{'batch': 1, 'size': 1, 'loss': 6.28485}step ...... 1\n",
      " 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.354627}step ...... 1 7\n",
      "\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.3007    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.295102}\n",
      "step ......{'batch': 4, 'size': 1, 'loss': 6.2960787}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.29871 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.2354927}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.143568}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.2675 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.287764}\n",
      "{'batch': 8, 'size': 1, 'loss': 6.2608805}step ......\n",
      " 1 7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.2690step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.167667}\n",
      "10/10 [==============================] - 0s 42ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.2589\n",
      "Epoch 102/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.278954}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.2790step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.425106}\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 6.2285714} 1 7\n",
      "\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.3109    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.286789}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.1389656}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.2717step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.3044515}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.2830815}\n",
      "{'batch': 7, 'size': 1, 'loss': 6.2515264}step ...... 1 7\n",
      "\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.2747step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.269371}\n",
      "{'batch': 9, 'size': 1, 'loss': 6.354549}step ...... 1 7\n",
      "\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 6.2821 1 7\n",
      "Found change point: 8 in batch 1 with index :6 and 13\n",
      "10/10 [==============================] - 0s 39ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.2821\n",
      "Epoch 103/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.279081}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.2791step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.2287765}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.268374}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.2587    step ...... 1{'batch': 3, 'size': 1, 'loss': 6.1231995}\n",
      " 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.339715}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.2478 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.181055}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.2367step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.274688}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.2702765}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.2456step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.219505}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.234641}\n",
      "10/10 [==============================] - 0s 43ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.2419\n",
      "Epoch 104/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.2745204}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.2745step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.265266}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.213943}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.2512    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.340532}\n",
      "{'batch': 4, 'size': 1, 'loss': 6.175675}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.2211494}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.2485step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.126365}\n",
      "step ...... 1 7\n",
      "step ......{'batch': 7, 'size': 1, 'loss': 6.2147164}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.2290 1 7\n",
      "Found change point: 8 in batch 9 with index :5 and 12\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.268022}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.2134347}\n",
      "10/10 [==============================] - 0s 38ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.2314\n",
      "Epoch 105/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.3007307}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3007step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.240841}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.2653856}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.446553}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.3134    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.2679024}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.169046}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.2817step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.262748}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.217948}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.2714{'batch': 8, 'size': 1, 'loss': 6.233735}\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 6.3305287}\n",
      " 1 7\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 35ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.2735\n",
      "Epoch 106/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.342774}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3428step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.2597404}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.228074}step ...... \n",
      "1 7\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.2769    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.397748}\n",
      "step ......{'batch': 4, 'size': 1, 'loss': 6.3087864}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.3074 7\n",
      "Found change point: 8 in batch 6 with index :1 and 8\n",
      "{'batch': 5, 'size': 1, 'loss': 6.2620397}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.3179655}step ...... 1 7\n",
      "\n",
      "{'batch': 7, 'size': 1, 'loss': 6.2590966}\n",
      "{'batch': 8, 'size': 1, 'loss': 6.204763}\n",
      "step ...... 1 7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.2868step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.1778407}\n",
      "10/10 [==============================] - 0s 30ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.2759\n",
      "Epoch 107/200\n",
      "step ...... 1{'batch': 0, 'size': 1, 'loss': 6.1998577}\n",
      " 7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.1999step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.13218}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.176562}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.253501}\n",
      "step ...... 1 7\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.1905    step ......{'batch': 4, 'size': 1, 'loss': 6.1353955}\n",
      " 1 7\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.1795step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.2443075}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.1903step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.2544837}\n",
      "step ...... 1{'batch': 7, 'size': 1, 'loss': 6.3190627}\n",
      " 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.2144step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.195052}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.2579756}\n",
      "10/10 [==============================] - 0s 40ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.2168\n",
      "Epoch 108/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.2502437}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.2502step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.2156544}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.1972795}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.2211    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.2498455}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 6 with index :8 and 15\n",
      "{'batch': 4, 'size': 1, 'loss': 6.3216705}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.3209815}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.2593step ......{'batch': 6, 'size': 1, 'loss': 6.2378187}\n",
      " 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.266658}step ...... 1 \n",
      "7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.243023}\n",
      "{'batch': 9, 'size': 1, 'loss': 6.2961025}\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 20ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.2599\n",
      "step ...... 1Epoch 109/200\n",
      " 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.4655595}step ...... 1 7\n",
      "\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.4656{'batch': 1, 'size': 1, 'loss': 6.212547}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.136644}\n",
      "{'batch': 3, 'size': 1, 'loss': 6.2111745}\n",
      "step ......{'batch': 4, 'size': 1, 'loss': 6.1884923}\n",
      " 1 7\n",
      "Found change point: 8 in batch 4 with index :8 and 15\n",
      "{'batch': 5, 'size': 1, 'loss': 6.2314186}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.2410    step ...... 1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.3339653}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.2543step ......{'batch': 7, 'size': 1, 'loss': 6.181077}\n",
      " 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.2100096}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 10 with index :8 and 15\n",
      "{'batch': 9, 'size': 1, 'loss': 6.2560043}\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 6.2427step ...... 1 7\n",
      "10/10 [==============================] - 0s 31ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.2427\n",
      "Epoch 110/200\n",
      "step ...... {'batch': 0, 'size': 1, 'loss': 6.1987705}\n",
      "1 7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.1988step ......{'batch': 1, 'size': 1, 'loss': 6.1195135}\n",
      " 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.181498}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.1666    \n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.2260528}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.1815step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.234385}\n",
      "{'batch': 5, 'size': 1, 'loss': 6.25409}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.2024step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.182168}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.1995step ......{'batch': 7, 'size': 1, 'loss': 6.1698475} 1 7\n",
      "Found change point: 8 in batch 9 with index :4 and 11\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.2324734}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.1999step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.2220616}\n",
      "10/10 [==============================] - 0s 43ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.2021\n",
      "Epoch 111/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.178367}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.1784step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.258422}\n",
      "step ...... {'batch': 2, 'size': 1, 'loss': 6.1706786}\n",
      "1 7\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.2025    step ......{'batch': 3, 'size': 1, 'loss': 6.284875}\n",
      " 1 7\n",
      "step ......{'batch': 4, 'size': 1, 'loss': 6.2183557}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.2221 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.1852374}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.244234}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.2200step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.165803}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.2115726}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.2131step ...... 1 {'batch': 9, 'size': 1, 'loss': 6.214895}\n",
      "7\n",
      "10/10 [==============================] - 0s 37ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.2132\n",
      " 7\n",
      "Epoch 112/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 6.2225957}\n",
      " 1 7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.2226step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.0700474}\n",
      "step ...... {'batch': 2, 'size': 1, 'loss': 6.100893}1 7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.257933}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.1629    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.206442}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.168134}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.2288704}\n",
      "{'batch': 7, 'size': 1, 'loss': 6.189222}step ...... 1 7\n",
      "\n",
      "step ......{'batch': 8, 'size': 1, 'loss': 6.209507}\n",
      " 1 7\n",
      "step ...... 1 7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.1837step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.090121}\n",
      "10/10 [==============================] - 0s 22ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.1744\n",
      "Epoch 113/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.197088}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.1971step ...... 1 7\n",
      "Found change point: 8 in batch 3 with index :4 and 11\n",
      "{'batch': 1, 'size': 1, 'loss': 6.2973557}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.2143865}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.2363    \n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.0766025}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.195556}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.1962step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.280565}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.1942997}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.2080step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.1770487}\n",
      "Found change point: 8 in batch 9 with index :7 and 14\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.196998}\n",
      "{'batch': 9, 'size': 1, 'loss': 6.132691}\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 6.1963step ...... 1 7\n",
      "10/10 [==============================] - 0s 33ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.1963\n",
      "Epoch 114/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.208977}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.2090step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.209877}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.1964126}step ......\n",
      " 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.151241}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.2232857}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.1980    step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.188121}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.189811}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.1954step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.1952443}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.2097497}\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 6.2239976}\n",
      " 1 7\n",
      "10/10 [==============================] - 0s 22ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.1997\n",
      "Epoch 115/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.0773416}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.0773step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.275355}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.15011}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.162743}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.1664    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.206642}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.143674}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.0961103}step ...... 1 7\n",
      "\n",
      "{'batch': 7, 'size': 1, 'loss': 6.1073766}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.1524step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.1844983}step ...... 1 \n",
      "7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.1671596}step ...... \n",
      "1 7\n",
      "10/10 [==============================] - 0s 24ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.1571\n",
      "Epoch 116/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.2025023}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.2025step ......{'batch': 1, 'size': 1, 'loss': 6.294033}\n",
      " 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.184264}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.2269    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.2143025}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.19344}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.2177step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.1743836}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.115582}\n",
      "step ......{'batch': 7, 'size': 1, 'loss': 6.17304}\n",
      " 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.1354322}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.1874 7\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.134798}\n",
      "10/10 [==============================] - 0s 30ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.1822\n",
      "Epoch 117/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.133646}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.1336step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.2132154}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.1265764}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.1578    step ...... 1 7\n",
      "Found change point: 8 in batch 5 with index :8 and 15{'batch': 3, 'size': 1, 'loss': 6.129406}\n",
      "\n",
      "{'batch': 4, 'size': 1, 'loss': 6.2495494}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.1705 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.183818}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.404446}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.1298733}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.1963step ...... 1 7\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 9 with index :8 and 15\n",
      "{'batch': 8, 'size': 1, 'loss': 6.1233425}step ...... 1 7\n",
      "\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.1882step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.143025}\n",
      "10/10 [==============================] - 0s 38ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.1837\n",
      "Epoch 118/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 6.171956}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.1720 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.201722}\n",
      " 2/10 [=====>........................] - ETA: 0s - batch: 0.5000 - size: 1.0000 - loss: 6.1868    step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.165803}\n",
      "step ......{'batch': 3, 'size': 1, 'loss': 6.272624}\n",
      " 1 7\n",
      "Found change point: 8 in batch 5 with index :7 and 14{'batch': 4, 'size': 1, 'loss': 6.1720862}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.1968\n",
      "step ...... 1 7\n",
      "step ......{'batch': 5, 'size': 1, 'loss': 6.2125964} 1 7\n",
      "\n",
      "Found change point: 8 in batch 7 with index :1 and 8\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.1995step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.0373635}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.1232324}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.1697step ......{'batch': 8, 'size': 1, 'loss': 6.1693845}\n",
      " 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.1741476}\n",
      "10/10 [==============================] - 0s 39ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.1701\n",
      " 1 7\n",
      "Epoch 119/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.369975}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3700step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.165648}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.2722282}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.2693    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.1467752}\n",
      "step ...... 1 {'batch': 4, 'size': 1, 'loss': 6.177796}\n",
      "7\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.2265step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.1874614}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.162245}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.2117step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.1353607}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.2022step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.1683807}\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 6.156212}\n",
      " 7\n",
      "10/10 [==============================] - 0s 40ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.1942\n",
      "Epoch 120/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.068133}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.0681step ......{'batch': 1, 'size': 1, 'loss': 6.1743817} 1 7\n",
      "\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 6.202782}\n",
      " 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.1104727}step ...... 1 7\n",
      "\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.1389    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.0275364}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.035404}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.1031step ...... 1 7\n",
      "Found change point: 8 in batch 8 with index :5 and 12\n",
      "{'batch': 6, 'size': 1, 'loss': 6.1618757}\n",
      "{'batch': 7, 'size': 1, 'loss': 6.1659656}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.2275395}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.1305step ...... 1 7\n",
      "step ...... 1{'batch': 9, 'size': 1, 'loss': 6.3045} 7\n",
      "\n",
      "Found change point: 8 in batch 1 with index :2 and 9\n",
      "10/10 [==============================] - 0s 30ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.1479\n",
      "Epoch 121/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 6.1652875}\n",
      " 1 7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.1653step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.0865045}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.231862}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.1469865}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.1577    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.195366}\n",
      "{'batch': 5, 'size': 1, 'loss': 6.1058393}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.156572}step ......\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.15551 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.1275268}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.1437206}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.1511step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.1646776}\n",
      "10/10 [==============================] - 0s 28ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.1524\n",
      "step ...... 1Epoch 122/200\n",
      " 7\n",
      "Found change point: 8 in batch 1 with index :8 and 15\n",
      "{'batch': 0, 'size': 1, 'loss': 6.132776}\n",
      "step ...... 1 7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.1328step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.128101}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.109146}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.156252}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.1316    step ...... 1 {'batch': 4, 'size': 1, 'loss': 6.1089253}\n",
      "7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.217057}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.216149}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.1107616}\n",
      "step ...... 1 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.1474{'batch': 8, 'size': 1, 'loss': 6.085165}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 10 with index :5 and 12\n",
      "{'batch': 9, 'size': 1, 'loss': 6.1957765}\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 25ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.1460\n",
      "Epoch 123/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.141205}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.1412step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.1349607}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.134473}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.1369    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.14571}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.0882945}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.1289step ...... 1 7\n",
      "Found change point: 8 in batch 7 with index :8 and 15\n",
      "{'batch': 5, 'size': 1, 'loss': 6.336756}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.1636step ...... 1{'batch': 6, 'size': 1, 'loss': 6.0101194}\n",
      " 7\n",
      "Found change point: 8 in batch 8 with index :7 and 14\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.1411257}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.1416step ...... 1{'batch': 8, 'size': 1, 'loss': 6.1519585} 7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.1127486}\n",
      "10/10 [==============================] - 0s 42ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.1397\n",
      "Epoch 124/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.184156}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.1842step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.0970407}\n",
      " 2/10 [=====>........................] - ETA: 0s - batch: 0.5000 - size: 1.0000 - loss: 6.1406    step ......{'batch': 2, 'size': 1, 'loss': 6.1930337}\n",
      " 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.141173}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.15391 7\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.140531}\n",
      "step ......{'batch': 5, 'size': 1, 'loss': 6.0675607} 1 7\n",
      "\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 8 with index :5 and 12\n",
      "{'batch': 6, 'size': 1, 'loss': 6.130419}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.1363step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.140945}\n",
      "step ...... 1{'batch': 8, 'size': 1, 'loss': 6.023438}\n",
      " 7\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 6.1371727}\n",
      " 1 7\n",
      "10/10 [==============================] - 0s 26ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.1255\n",
      "Epoch 125/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.180736}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.1807step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.0808897}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.088703}\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.101857}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.1269097}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.1158    step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.217602}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.196867}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.100537}\n",
      "step ......{'batch': 8, 'size': 1, 'loss': 6.153725}\n",
      " 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.3188562}\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 6.1567step ...... 1 7\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 15ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.1567\n",
      "Epoch 126/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.0834365}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.0834step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.996619}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.1784286}\n",
      "step ...... 1{'batch': 3, 'size': 1, 'loss': 6.117176}\n",
      " 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.170088}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.1091    step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.16934}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 7 with index :5 and 12\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.1393375}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.1221{'batch': 7, 'size': 1, 'loss': 6.073288}step ...... 1 7\n",
      "\n",
      "{'batch': 8, 'size': 1, 'loss': 6.1137586}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.143162}step ......\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 6.1185 1 7\n",
      "10/10 [==============================] - 0s 24ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.1185\n",
      "Epoch 127/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.310345}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.3103step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.072039}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.1375976}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.1733    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.1368675}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.0738983}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.1461step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.1089306}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.0875483}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.1325step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.183097}\n",
      "{'batch': 8, 'size': 1, 'loss': 6.136492}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.1385step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.98717}\n",
      "10/10 [==============================] - 0s 40ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.1234\n",
      "Epoch 128/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.1270914}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.1271step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.1097617}\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 6.157931} 1 7\n",
      "\n",
      "{'batch': 3, 'size': 1, 'loss': 6.099265}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.108285}step ...... 1 7\n",
      "\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.1205    step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.1200147}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.055152}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.156221}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.1167step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.0554576}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.0500755}\n",
      "10/10 [==============================] - 0s 20ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.1039\n",
      "Epoch 129/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.174689}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.1747step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.103431}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.980301}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.129405}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.0970    step ......{'batch': 4, 'size': 1, 'loss': 6.1213717}\n",
      " 1 7\n",
      "step ......{'batch': 5, 'size': 1, 'loss': 6.0821633}\n",
      " 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.074622}\n",
      "step ...... 1 7\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.0951step ...... 1 7\n",
      "Found change point: 8 in batch 9 with index :2 and 9\n",
      "{'batch': 7, 'size': 1, 'loss': 6.137403}\n",
      "{'batch': 8, 'size': 1, 'loss': 6.089125}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.983467}\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 6.0876step ...... 1 7\n",
      "10/10 [==============================] - 0s 22ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.0876\n",
      "Epoch 130/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.0585146}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.0585step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.0780606}\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 6.1323323} 1 \n",
      "7\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.0896    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.047941}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.9739056}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.0582step ......{'batch': 5, 'size': 1, 'loss': 6.109532}\n",
      " 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.979588}\n",
      "{'batch': 7, 'size': 1, 'loss': 6.0799646}step ......\n",
      " 1 7\n",
      "step ...... 1 {'batch': 8, 'size': 1, 'loss': 6.0713387}7\n",
      "\n",
      "step ...... 1 7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.0590step ......{'batch': 9, 'size': 1, 'loss': 6.0425897} 1 7\n",
      "\n",
      "10/10 [==============================] - 0s 29ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.0574\n",
      "Epoch 131/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.0425673}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.0426step ...... 1 7\n",
      "Found change point: 8 in batch 3 with index :3 and 10\n",
      "{'batch': 1, 'size': 1, 'loss': 6.167733}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.084257}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.0982    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.036354}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.16589}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.0994step ......{'batch': 5, 'size': 1, 'loss': 6.04352}\n",
      " 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.9936438}\n",
      "step ...... 1 7\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.0763step ...... {'batch': 7, 'size': 1, 'loss': 6.1062555}1 7\n",
      "\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.0800step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.043118}\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 6.0333314}\n",
      " 1 7\n",
      "10/10 [==============================] - 0s 40ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.0717\n",
      "Epoch 132/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.0541124}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.0541step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.1194963}\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 6.0826344}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.0854     1 7\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.154445}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.1027step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.0340734}\n",
      "{'batch': 5, 'size': 1, 'loss': 6.1081886}\n",
      "step ......{'batch': 6, 'size': 1, 'loss': 6.0938106}\n",
      " 1 7\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.0924 1 7\n",
      "Found change point: 8 in batch 8 with index :3 and 10\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.1512947}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.0998step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.050297}\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 6.0680346}\n",
      " 1 7\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 6.09 - 0s 39ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.0916\n",
      "Epoch 133/200\n",
      "step ...... {'batch': 0, 'size': 1, 'loss': 6.1221523}\n",
      "1 7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.1222step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.076436}\n",
      "{'batch': 2, 'size': 1, 'loss': 6.127406}\n",
      "step ...... 1 7\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.1087    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.081091}\n",
      "{'batch': 4, 'size': 1, 'loss': 6.0643897}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.12165}step ......\n",
      " 1 7\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.0989step ...... 1 7\n",
      "Found change point: 8 in batch 8 with index :7 and 14\n",
      "{'batch': 6, 'size': 1, 'loss': 6.170616}\n",
      "step ...... 1 7{'batch': 7, 'size': 1, 'loss': 6.072119}\n",
      "\n",
      "{'batch': 8, 'size': 1, 'loss': 6.0799365}\n",
      "step ...... 1 7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.1018step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.0916934}\n",
      "10/10 [==============================] - 0s 29ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.1007\n",
      "Epoch 134/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.0911}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.0911step ......{'batch': 1, 'size': 1, 'loss': 5.997571}\n",
      " 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.034285}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.0410    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.1565747}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.072232}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.0704step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.088172}\n",
      "step ......{'batch': 6, 'size': 1, 'loss': 6.093771}\n",
      " 1 7\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.0762step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.071268}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.112457}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.006582}\n",
      "10/10 [==============================] - 0s 34ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.0724\n",
      "Epoch 135/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.008289}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.0083step ......{'batch': 1, 'size': 1, 'loss': 6.0346174}\n",
      " 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.045169}\n",
      "{'batch': 3, 'size': 1, 'loss': 6.0051537}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.0233    step ...... 1 7\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 6 with index :5 and 12{'batch': 4, 'size': 1, 'loss': 6.041636}\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.1110754}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.0410step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.0573545}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.1047072}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.0500684}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.0509step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.05292}\n",
      "10/10 [==============================] - 0s 22ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.0511\n",
      "Epoch 136/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.0563173}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.0563step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.0452037}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.047174}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 5 with index :3 and 10\n",
      "{'batch': 3, 'size': 1, 'loss': 6.148546}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.0743    {'batch': 4, 'size': 1, 'loss': 6.040611}step ......\n",
      " 1 7\n",
      "step ......{'batch': 5, 'size': 1, 'loss': 6.027959}\n",
      " 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.0056725}\n",
      "{'batch': 7, 'size': 1, 'loss': 6.065583}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.0546step ...... 1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.1948667}\n",
      "{'batch': 9, 'size': 1, 'loss': 6.087758}\n",
      "10/10 [==============================] - 0s 23ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.0720\n",
      "\n",
      "Epoch 137/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.9994683}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.9995step ......{'batch': 1, 'size': 1, 'loss': 6.009655} 1 \n",
      "7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.0399365}\n",
      "step ...... 1 7\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.0164    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.07897}\n",
      "step ...... 1 {'batch': 4, 'size': 1, 'loss': 6.045954}7\n",
      "\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.0348step ......{'batch': 5, 'size': 1, 'loss': 6.0533853}\n",
      " 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.120585}step ...... \n",
      "1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.959562}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.0384step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.9867964}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 1 with index :2 and 9\n",
      "{'batch': 9, 'size': 1, 'loss': 6.0894594}\n",
      "10/10 [==============================] - 0s 32ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.0384\n",
      "Epoch 138/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.0401106}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.0401step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.9955006}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.0379834}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.0245    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.008252}\n",
      "{'batch': 4, 'size': 1, 'loss': 6.158904}\n",
      "{'batch': 5, 'size': 1, 'loss': 6.0580254}\n",
      "step ...... 1 7\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.0498step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.9922304}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.0416{'batch': 7, 'size': 1, 'loss': 5.9962687}step ...... 1 7\n",
      "\n",
      "{'batch': 8, 'size': 1, 'loss': 5.9965224}\n",
      "step .... 1 7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.0315step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.1223097}\n",
      "10/10 [==============================] - 0s 43ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.0406\n",
      "Epoch 139/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.014167}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.0142step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.2195673}\n",
      "{'batch': 2, 'size': 1, 'loss': 5.9957232}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 4 with index :3 and 10{'batch': 3, 'size': 1, 'loss': 6.040296}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.0674    \n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.9868956}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.0513step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.0294514}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.0660663}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.0503{'batch': 7, 'size': 1, 'loss': 5.9990735}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 8 with index :1 and 8\n",
      "step ...... 1 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.0439step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.005594}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.0388703}\n",
      "10/10 [==============================] - 0s 39ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.0396\n",
      "Epoch 140/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.172283}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.1723step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.996648}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.9805293}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.0498    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.0254607}\n",
      "step ......{'batch': 4, 'size': 1, 'loss': 6.012594} 1 7\n",
      "\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.0375step ......{'batch': 5, 'size': 1, 'loss': 6.046612}\n",
      " 1 7\n",
      "step ......{'batch': 6, 'size': 1, 'loss': 5.9754596}\n",
      " 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.017536}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.0284step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.056184}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.0080366}\n",
      "10/10 [==============================] - 0s 31ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.0291\n",
      "Epoch 141/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.043702}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.0437step ......{'batch': 1, 'size': 1, 'loss': 6.103418}\n",
      " 1 7\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 5.974028} 1 7\n",
      "Found change point: 8 in batch 4 with index :1 and 8\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.0381737}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 6.0398    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.9883733}\n",
      "step ...... 1{'batch': 5, 'size': 1, 'loss': 6.051331}\n",
      " 7\n",
      "{'batch': 6, 'size': 1, 'loss': 6.0220284}\n",
      "{'batch': 7, 'size': 1, 'loss': 6.024106}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 6.0306step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.049243}step ...... 1 7\n",
      "\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.0327step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.9802446}\n",
      "10/10 [==============================] - 0s 32ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.0275\n",
      "Epoch 142/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.9847713}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.9848step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.982611}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 4 with index :8 and 15\n",
      "{'batch': 2, 'size': 1, 'loss': 6.0002184}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.9892    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.006703}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.9758587}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.9900step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.968116}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.91792}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.9766step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.96921}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.9757step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.068462}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.973212}\n",
      "10/10 [==============================] - 0s 42ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.9847\n",
      "Epoch 143/200\n",
      "step ...... 1 7{'batch': 0, 'size': 1, 'loss': 6.151808}\n",
      "\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.1518step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 6.030098}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.0856757}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.0892    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.00536}\n",
      "step ......{'batch': 4, 'size': 1, 'loss': 5.9700103}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.0486 1 7\n",
      "step ...... 1 {'batch': 5, 'size': 1, 'loss': 6.0218806}\n",
      "7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.970623}\n",
      "step ...... 1 7\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 6.0336step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.9555697}\n",
      "{'batch': 8, 'size': 1, 'loss': 5.9709854}\n",
      "step ...... 1 7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.0180step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.060761}\n",
      "10/10 [==============================] - 0s 39ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.0223\n",
      "Epoch 144/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.05792}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.0579step ......{'batch': 1, 'size': 1, 'loss': 5.9557557}\n",
      " 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.004164}\n",
      "step ...... 1 7\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 6.0059    step ...... {'batch': 3, 'size': 1, 'loss': 6.018816}\n",
      "1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.0822453}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.0238step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.9649544}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 6.0140step ...... 1 7\n",
      "Found change point: 8 in batch 8 with index :2 and 9\n",
      "{'batch': 6, 'size': 1, 'loss': 5.9414563}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.9693127}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.9993step ...... 1 7\n",
      "Found change point: 8 in batch 10 with index :4 and 11{'batch': 8, 'size': 1, 'loss': 6.042636}\n",
      "\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.0041step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.0017257}\n",
      "10/10 [==============================] - 0s 42ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.0039\n",
      "Epoch 145/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 5.9591017} 1 7\n",
      "\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.9591step ......{'batch': 1, 'size': 1, 'loss': 5.966354}\n",
      " 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.039868}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.900641}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 5.9665    step ...... 1 7\n",
      "step ...... 1 7{'batch': 4, 'size': 1, 'loss': 5.90299}\n",
      "\n",
      "Found change point: 8 in batch 6 with index :5 and 12\n",
      "{'batch': 5, 'size': 1, 'loss': 5.9474945}\n",
      "step ...... 1 {'batch': 6, 'size': 1, 'loss': 5.9600635}7\n",
      "\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.95387\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.968171}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.9709897}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.9965415}\n",
      "10/10 [==============================] - 0s 25ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.9612\n",
      "Epoch 146/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.0165625}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.0166step ...... 1 7\n",
      "Found change point: 8 in batch 3 with index :5 and 12\n",
      "{'batch': 1, 'size': 1, 'loss': 6.0020714}\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 5.9983172}\n",
      " 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 6.020472}step ...... 1 7\n",
      "\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 6 with index :1 and 8\n",
      "{'batch': 4, 'size': 1, 'loss': 5.983254}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 6.0041    step ...... 1 7\n",
      "Found change point: 8 in batch 7 with index :2 and 9\n",
      "{'batch': 5, 'size': 1, 'loss': 5.961903}\n",
      "{'batch': 6, 'size': 1, 'loss': 6.0056577}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.980755}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.964643}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.9926step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.966124}\n",
      "10/10 [==============================] - 0s 21ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.9900\n",
      "Epoch 147/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 6.013945}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 6.0139step ......{'batch': 1, 'size': 1, 'loss': 5.9829946} 1 7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.957378}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.947285}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 5.9754    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.0083847}\n",
      "{'batch': 5, 'size': 1, 'loss': 5.972357}step ...... 1 7\n",
      "\n",
      "{'batch': 6, 'size': 1, 'loss': 5.955416}\n",
      "step ...... 1\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.9768step ......{'batch': 7, 'size': 1, 'loss': 5.9327574}\n",
      " 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.9493876}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.929506}\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 20ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.9649\n",
      "Epoch 148/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.9701686}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.9702step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.938214}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.934806}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.890817}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 5.9335    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.9526954}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.0584803}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.97706}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.945612}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.9585step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.986885}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.0166717}\n",
      "10/10 [==============================] - 0s 27ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.9671\n",
      "Epoch 149/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.9877653}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.9878step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.9627733}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.929107}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.9599    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.973371}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.110222}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.9926step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.048293}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.9091086}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.9887{'batch': 7, 'size': 1, 'loss': 6.0451856}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 6.0822744}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 6.0053step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.995627}\n",
      "10/10 [==============================] - 0s 40ms/step - batch: 4.5000 - size: 1.0000 - loss: 6.0044\n",
      "Epoch 150/200\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 2 with index :5 and 12\n",
      "{'batch': 0, 'size': 1, 'loss': 5.917307}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.9173step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.9541597}\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 5.946915}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.9395    7\n",
      "Found change point: 8 in batch 4 with index :6 and 13\n",
      "{'batch': 3, 'size': 1, 'loss': 5.9596043}\n",
      "{'batch': 4, 'size': 1, 'loss': 5.963704}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.042542}\n",
      " 6/10 [=================>............]step ...... 1 7\n",
      " - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 5.9640Found change point: 8 in batch 6 with index :8 and 15\n",
      "{'batch': 6, 'size': 1, 'loss': 5.956989}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.920848}\n",
      "step ...... 1 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.95787\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.918731}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.9534step ...... 1 7{'batch': 9, 'size': 1, 'loss': 5.972816}\n",
      "\n",
      "10/10 [==============================] - 0s 35ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.9554\n",
      "Epoch 151/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.8764067}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8764{'batch': 1, 'size': 1, 'loss': 5.932664}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 3 with index :6 and 13\n",
      "{'batch': 2, 'size': 1, 'loss': 5.8754883}\n",
      "step ...... 1 7\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.8949    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.9467206}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.1068373}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.9476step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.9531927}\n",
      "step ......{'batch': 6, 'size': 1, 'loss': 6.08812}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.9685 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.9874854}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.9709step ......{'batch': 8, 'size': 1, 'loss': 5.9467664} \n",
      "1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.957847}\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 5.9672step ...... 1 7\n",
      "10/10 [==============================] - 0s 46ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.9672\n",
      "Epoch 152/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 5.9682817}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.9683 1 7\n",
      "step ......{'batch': 1, 'size': 1, 'loss': 5.956485}\n",
      " 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.9406333}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.9551     1 7\n",
      "step ......{'batch': 3, 'size': 1, 'loss': 5.910157}\n",
      " 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.9512534}\n",
      "step ...... 1 7\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.9454step ......{'batch': 5, 'size': 1, 'loss': 5.9413166} \n",
      "1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.9412165}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.9442step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.9006066}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.9387step ......{'batch': 8, 'size': 1, 'loss': 5.916741} 1 7\n",
      "\n",
      "{'batch': 9, 'size': 1, 'loss': 5.9094954}\n",
      "10/10 [==============================] - 0s 42ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.9336\n",
      " 1 7\n",
      "Epoch 153/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.9379287}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.9379step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.90471}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.9783087}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.9403    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.914887}\n",
      "step ...... {'batch': 4, 'size': 1, 'loss': 5.925233}\n",
      "1 7\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.9322step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.9050093}\n",
      "{'batch': 6, 'size': 1, 'loss': 5.931598}\n",
      "step ...... 1 7{'batch': 7, 'size': 1, 'loss': 5.8643513}\n",
      "\n",
      "step ...... 1 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.9203step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.9121737}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.920349}\n",
      "10/10 [==============================] - 0s 29ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.9195\n",
      "Epoch 154/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.92976}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.9298step ......{'batch': 1, 'size': 1, 'loss': 5.908957} 1 7\n",
      "\n",
      "step ...... 1 {'batch': 2, 'size': 1, 'loss': 5.8940644}7\n",
      "\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.9109    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.9402914}\n",
      "step ...... 1{'batch': 4, 'size': 1, 'loss': 5.918657}\n",
      "\b7\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.9183step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.8962207}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 5.9147step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.9151225}\n",
      "step ...... {'batch': 7, 'size': 1, 'loss': 5.9376364}1 7\n",
      "\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.9176step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.9140515}\n",
      "{'batch': 9, 'size': 1, 'loss': 5.8857203}\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 5.9140 1 7\n",
      "10/10 [==============================] - 0s 44ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.9140\n",
      "Epoch 155/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.934845}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.9348step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.8919816}\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 5.9363313}\n",
      " 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.922091}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 5.9213    step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.909637}\n",
      "{'batch': 5, 'size': 1, 'loss': 5.888203}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 5.9138step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.873538}\n",
      "{'batch': 7, 'size': 1, 'loss': 5.9303985}\n",
      "step ......{'batch': 8, 'size': 1, 'loss': 5.9309235}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.91311 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.9304476}\n",
      "10/10 [==============================] - 0s 32ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.9148\n",
      "Epoch 156/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.9296384}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.9296step ......{'batch': 1, 'size': 1, 'loss': 5.888773}\n",
      " 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.888236}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.9022    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.8500314}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.8985953}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.8911step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.915946}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.9490824}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.9029step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.9247785}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.8801336}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.9028step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.0549107}\n",
      "10/10 [==============================] - 0s 38ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.9180\n",
      "Epoch 157/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.929376}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.9294step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.839283}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.8724732}\n",
      "{'batch': 3, 'size': 1, 'loss': 5.916512}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 5.8894     1 7\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.934055}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.9913836}\n",
      "{'batch': 6, 'size': 1, 'loss': 5.899007}step ...... 1 7\n",
      "\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.9117step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.898182}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.9136257}\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 5.9128294}\n",
      " 1 7\n",
      "10/10 [==============================] - 0s 23ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.9107\n",
      "Epoch 158/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.88605}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8861step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.8870587}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.9332128}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.9021    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.95363}\n",
      "step ...... 1{'batch': 4, 'size': 1, 'loss': 5.8795056}\n",
      " 7\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.9079step ......{'batch': 5, 'size': 1, 'loss': 5.9392004}\n",
      " 1 {'batch': 6, 'size': 1, 'loss': 5.83768}7\n",
      "\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.9023step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.902047}\n",
      "step .... 1 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.9023step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.8783603}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.8996step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.05018}\n",
      "10/10 [==============================] - 0s 39ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.9147\n",
      "Epoch 159/200\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 2 with index :7 and 14{'batch': 0, 'size': 1, 'loss': 5.8670096}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8670\n",
      "step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.865757}\n",
      " 2/10 [=====>........................] - ETA: 0s - batch: 0.5000 - size: 1.0000 - loss: 5.8664    step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.1003127}\n",
      "{'batch': 3, 'size': 1, 'loss': 5.896993}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.858947}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.9178step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.952514}\n",
      "{'batch': 6, 'size': 1, 'loss': 5.903168}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.9207step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.8319473}\n",
      "{'batch': 8, 'size': 1, 'loss': 5.8721457}step ...... 1 7\n",
      "\n",
      "step ...... 1 7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.9054step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.949816}\n",
      "10/10 [==============================] - 0s 38ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.9099\n",
      "Epoch 160/200\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 2 with index :6 and 13\n",
      "{'batch': 0, 'size': 1, 'loss': 5.906269}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.9063{'batch': 1, 'size': 1, 'loss': 5.900911}\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 5.875277}\n",
      " 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.9908957}step ...... 1 7\n",
      "\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 5.9183    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.856472}step ...... 1 7\n",
      "\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.9060step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.8576427}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.835157}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.8889step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 6.0388227}\n",
      "{'batch': 8, 'size': 1, 'loss': 5.868024}\n",
      "step ...... 1 7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.9033step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 6.040273}\n",
      "10/10 [==============================] - 0s 42ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.9170\n",
      "Epoch 161/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.8706045}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8706step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.883202}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 6.001523}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.9184    step ...... 1{'batch': 3, 'size': 1, 'loss': 5.9402924}\n",
      " 7\n",
      "{'batch': 4, 'size': 1, 'loss': 6.0004034}step ...... 1 7\n",
      "\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.9392step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.919858}Found change point: 8 in batch 7 with index :2 and 9\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.9836946}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.9428step ...... 1 7\n",
      "Found change point: 8 in batch 9 with index :8 and 15\n",
      "{'batch': 7, 'size': 1, 'loss': 5.8661966}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.905943}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.9302step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.9171076}\n",
      "10/10 [==============================] - 0s 37ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.9289\n",
      "Epoch 162/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.8850765}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8851step ......{'batch': 1, 'size': 1, 'loss': 5.8712006}\n",
      " 1 7\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 5.9356503}\n",
      " 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.8966513}step ...... 1 7\n",
      "\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 5.8971    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.880818}\n",
      "{'batch': 5, 'size': 1, 'loss': 5.9469547}\n",
      "{'batch': 6, 'size': 1, 'loss': 5.858782}\n",
      "step ...... 1 7\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.8964step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.874524}\n",
      "step ...... 1 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.8937step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.8854723}\n",
      "{'batch': 9, 'size': 1, 'loss': 5.974415}\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 37ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.9010\n",
      "Epoch 163/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 5.855686}\n",
      " 1 7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8557step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.8261337}\n",
      " 2/10 [=====>........................] - ETA: 0s - batch: 0.5000 - size: 1.0000 - loss: 5.8409    step ......{'batch': 2, 'size': 1, 'loss': 5.8676605}\n",
      " 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.8457947}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 5.8488step ...... 1 7\n",
      "step ......{'batch': 4, 'size': 1, 'loss': 6.014096} 1 7\n",
      "\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.8819step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 6.021268}\n",
      "step ......{'batch': 6, 'size': 1, 'loss': 5.835128}\n",
      " 1 7\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.8951step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.824839}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.8863{'batch': 8, 'size': 1, 'loss': 5.8468084}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.881686}\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 42ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.8819\n",
      "Epoch 164/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.868163}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8682{'batch': 1, 'size': 1, 'loss': 5.843409}step ...... 1 7\n",
      "\n",
      "{'batch': 2, 'size': 1, 'loss': 5.875647}step ...... 1 7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.8726764}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 5.8650    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.929871}\n",
      "{'batch': 5, 'size': 1, 'loss': 5.856151}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.849899}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.92483}step ...... 1 7\n",
      "\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.8776step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.828551}\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 5.8548536}\n",
      " 1 7\n",
      "10/10 [==============================] - 0s 23ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.8704\n",
      "Epoch 165/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.8548775}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8549step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.912059}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.8409476}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.8693    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.852564}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.8821917}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.8685step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.863701}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.822338}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.8612step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.8573055}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.880224}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.8629{'batch': 9, 'size': 1, 'loss': 5.870061}\n",
      "10/10 [==============================] - 0s 35ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.8636\n",
      " 1 7\n",
      "Epoch 166/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.827757}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8278step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.823449}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.9515433}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.8676    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.847313}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.826486}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.8553step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.8476377}\n",
      "step ...... 1 {'batch': 6, 'size': 1, 'loss': 5.8801436}7\n",
      "\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.8578step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.8843107}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.901232}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.8655step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.8528914}\n",
      "10/10 [==============================] - 0s 42ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.8643\n",
      "Epoch 167/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.820297}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8203{'batch': 1, 'size': 1, 'loss': 5.906878}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.8986526}\n",
      "step ...... 1 7\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.8753    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.8656454}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.7913766}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.8566step ...... 1 7\n",
      "Found change point: 8 in batch 7 with index :5 and 12\n",
      "{'batch': 5, 'size': 1, 'loss': 5.8598013}\n",
      "{'batch': 6, 'size': 1, 'loss': 5.896041}step ...... 1 7\n",
      "\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.8627{'batch': 7, 'size': 1, 'loss': 5.8087206}step ...... 1 7\n",
      "\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.8559step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.95179}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.83561}\n",
      "10/10 [==============================] - 0s 42ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.8635\n",
      "Epoch 168/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.85872}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8587step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.864043}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.885864}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.8695    step ...... {'batch': 3, 'size': 1, 'loss': 5.8404255}\n",
      "1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.8582783}\n",
      "step ......{'batch': 5, 'size': 1, 'loss': 5.8521247}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 5.8599 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.8298917}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.8400564}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.8537step ...... 1 7\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.8655}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.8550step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.907363}\n",
      "10/10 [==============================] - 0s 38ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.8602\n",
      "Epoch 169/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.890726}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8907step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.835614}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.783545}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.8366    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.8256264}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.8383484}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.8348step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.829295}\n",
      "{'batch': 6, 'size': 1, 'loss': 5.8515215}step ...... 1 \n",
      "7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.882763}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.8422step ...... 1 7\n",
      "step ......{'batch': 8, 'size': 1, 'loss': 5.9581103}\n",
      " 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.838969}\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 30ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.8535\n",
      "Epoch 170/200\n",
      "step ...... {'batch': 0, 'size': 1, 'loss': 5.8690243}\n",
      " 1/10 [==>...........................]1 - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8690 7\n",
      "step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.9759445}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.8340745}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.8930    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.9047008}\n",
      "{'batch': 4, 'size': 1, 'loss': 5.8340654}\n",
      "{'batch': 5, 'size': 1, 'loss': 5.832263}\n",
      "step ...... 1 7\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 5.8750step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.7990103}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 8 with index :2 and 9\n",
      "{'batch': 7, 'size': 1, 'loss': 5.953788}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 9 with index :4 and 11\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.8754step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.776869}\n",
      "step ...... 1{'batch': 9, 'size': 1, 'loss': 5.831212}\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 5.8611 7\n",
      "10/10 [==============================] - 0s 37ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.8611\n",
      "Epoch 171/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.807157}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8072step ...... 1 7\n",
      "Found change point: 8 in batch 3 with index :4 and 11\n",
      "{'batch': 1, 'size': 1, 'loss': 5.7771306}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.8001795}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.7948    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.876559}\n",
      "step ...... 1{'batch': 4, 'size': 1, 'loss': 5.82951} 7\n",
      "\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.8181step ...... 1{'batch': 5, 'size': 1, 'loss': 5.862871}\n",
      " 7\n",
      "Found change point: 8 in batch 7 with index :4 and 11\n",
      "{'batch': 6, 'size': 1, 'loss': 5.8931117}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.8352step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.8059235}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 9 with index :3 and 10\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.8586698}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.8346step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.8086543}\n",
      "10/10 [==============================] - 0s 36ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.8320\n",
      "Epoch 172/200\n",
      "step ...... {'batch': 0, 'size': 1, 'loss': 5.8213296}\n",
      "1 7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8213{'batch': 1, 'size': 1, 'loss': 5.8190565}step ......\n",
      " 1 7\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 5.8568583}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.8324     7\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.8253474}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.7928996}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.8231step ...... 1 7\n",
      "Found change point: 8 in batch 7 with index :6 and 13\n",
      "{'batch': 5, 'size': 1, 'loss': 5.81339}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.855066}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.8263step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.823388}\n",
      "{'batch': 8, 'size': 1, 'loss': 5.8193426}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.7754993}\n",
      "10/10 [==============================] - 0s 36ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.8202\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 1 with index :1 and 8\n",
      "Epoch 173/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.823546}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8235step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.829349}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.883724}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.8455    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.7929406}\n",
      "{'batch': 4, 'size': 1, 'loss': 5.8117094}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.8603406}\n",
      "step ...... 1 7\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 5.8336step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.809571}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 9 with index :4 and 11\n",
      "{'batch': 7, 'size': 1, 'loss': 5.9242563}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.8419{'batch': 8, 'size': 1, 'loss': 5.805833}step ...... 1 7\n",
      "\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.8379step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.80176}\n",
      "10/10 [==============================] - 0s 39ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.8343\n",
      "Epoch 174/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.814156}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8142step ...... {'batch': 1, 'size': 1, 'loss': 5.779832}\n",
      "1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.8129125}\n",
      "step .... 1 7\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.8023    step ......{'batch': 3, 'size': 1, 'loss': 5.8410506}\n",
      " 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.8588643}\n",
      "step ...... 1 7\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.8214step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.8123813}\n",
      "{'batch': 6, 'size': 1, 'loss': 5.8046684}step ...... 1 7\n",
      "\n",
      "{'batch': 7, 'size': 1, 'loss': 5.795373}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.8149\n",
      "{'batch': 8, 'size': 1, 'loss': 5.8450813}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.7966957}\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 32ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.8161\n",
      "Epoch 175/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.811438}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8114step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.7746997}\n",
      "step ...... 1{'batch': 2, 'size': 1, 'loss': 5.843441}\n",
      " 7\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.7895603}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.946687}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.8332    step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.8584075}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.821632}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.8351step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.7760034}\n",
      "{'batch': 8, 'size': 1, 'loss': 5.8424745}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.833298}step ...... 1 7\n",
      "\n",
      "10/10 [==============================] - 0s 32ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.8298\n",
      "Epoch 176/200\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 2 with index :4 and 11\n",
      "{'batch': 0, 'size': 1, 'loss': 5.8359923}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8360step ...... {'batch': 1, 'size': 1, 'loss': 5.807696}\n",
      "1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.778614}\n",
      "step ......{'batch': 3, 'size': 1, 'loss': 5.8311434}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 5.8134     7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.783428}\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.8074step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.8406296}\n",
      "step ...... 1 {'batch': 6, 'size': 1, 'loss': 5.8099594}\n",
      "7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.8227997}\n",
      "step ...... 1 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.8138step ...... {'batch': 8, 'size': 1, 'loss': 5.783354}1 7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.812234}\n",
      "10/10 [==============================] - 0s 36ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.8106\n",
      "Epoch 177/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.7990217}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.7990step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.8348856}\n",
      "{'batch': 2, 'size': 1, 'loss': 5.8048368}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.8129    step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.7927895}\n",
      "{'batch': 4, 'size': 1, 'loss': 5.866194}step ...... 1 7\n",
      "\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.8195step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.7809033}\n",
      "step ...... 1 {'batch': 6, 'size': 1, 'loss': 5.864946}\n",
      "7\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.864295}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.8260step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.8064284}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.865199}\n",
      "10/10 [==============================] - 0s 35ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.8279\n",
      "Epoch 178/200\n",
      "step ...... {'batch': 0, 'size': 1, 'loss': 5.7853923}1 7\n",
      "\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.7854step ......{'batch': 1, 'size': 1, 'loss': 5.7742105}\n",
      " 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.8002014}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 4 with index :5 and 12\n",
      "{'batch': 3, 'size': 1, 'loss': 5.7767515}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 5.7841    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.797386}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.757827}\n",
      "step ...... 1 7\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 5.7820step ......{'batch': 6, 'size': 1, 'loss': 5.8029175}\n",
      " 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.7951345}\n",
      "step ...... 1 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.7862{'batch': 8, 'size': 1, 'loss': 5.8815784}step ...... \n",
      "1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.7978168}\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 32ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7969\n",
      "Epoch 179/200\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 2 with index :5 and 12\n",
      "{'batch': 0, 'size': 1, 'loss': 5.800843}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8008step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.869591}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.758149}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.8095    step ...... 1 7{'batch': 3, 'size': 1, 'loss': 5.7746778}\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.9207993}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.8248step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.755166}\n",
      "{'batch': 6, 'size': 1, 'loss': 5.794463}step ...... 1 \n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.8105\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.799266}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.7735543}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.8052step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.7607903}\n",
      "10/10 [==============================] - 0s 36ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.8007\n",
      "Epoch 180/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 5.8665876}\n",
      " 1 7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8666step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.7707915}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.817298}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.8182    step ...... {'batch': 3, 'size': 1, 'loss': 5.7685614}\n",
      "1 7\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.7939262}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.8034step ......{'batch': 5, 'size': 1, 'loss': 5.773566}\n",
      " 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.769098}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.7943step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.780504}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.770081}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.7900step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.7594233}\n",
      "10/10 [==============================] - 0s 36ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7870\n",
      "Epoch 181/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 5.823851}\n",
      " 1 7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8239step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.8082824}\n",
      " 2/10 [=====>........................] - ETA: 0s - batch: 0.5000 - size: 1.0000 - loss: 5.8161    step ......{'batch': 2, 'size': 1, 'loss': 5.8153195}\n",
      " 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.7335286}step ...... 1 7\n",
      "\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 5.7952step ......{'batch': 4, 'size': 1, 'loss': 5.818643}\n",
      " 1 7\n",
      "step ......{'batch': 5, 'size': 1, 'loss': 5.806255}\n",
      " 1 7\n",
      "step ......{'batch': 6, 'size': 1, 'loss': 5.7558036} 1 7\n",
      "\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.7945step ......{'batch': 7, 'size': 1, 'loss': 5.7824984}\n",
      " 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.788436}\n",
      "{'batch': 9, 'size': 1, 'loss': 5.749542}\n",
      "10/10 [==============================] - 0s 30ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7882\n",
      "Epoch 182/200step ...... 1 7\n",
      "\n",
      "{'batch': 0, 'size': 1, 'loss': 5.85113}\n",
      "step ...... 1 7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8511step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.8945403}\n",
      "{'batch': 2, 'size': 1, 'loss': 5.770834}\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "step ......{'batch': 3, 'size': 1, 'loss': 5.7699313}\n",
      " 1 7\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 5.8216    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.7606106}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.8112526}\n",
      "step ...... 1{'batch': 6, 'size': 1, 'loss': 5.885381}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.8205\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.799949}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.7849083}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.8143step ......{'batch': 9, 'size': 1, 'loss': 5.7302947} 1 7\n",
      "\n",
      "10/10 [==============================] - 0s 33ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.8059\n",
      "Epoch 183/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.7879806}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.7880step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.9177394}\n",
      "{'batch': 2, 'size': 1, 'loss': 5.778144}step ......\n",
      " 1 7\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.8280    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.7777023}\n",
      "step ...... 1 7{'batch': 4, 'size': 1, 'loss': 5.7563334}\n",
      "\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.8036step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.738494}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 5.7927step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.738853}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.757146}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.7815step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.726082}\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 5.7848496} 1 7\n",
      "\n",
      "10/10 [==============================] - 0s 41ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7763\n",
      "Epoch 184/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.7436495}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.7436step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.7664905}\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 5.736317} 1 \n",
      "7\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.7488    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.765927}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.77367}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.7572step ......{'batch': 5, 'size': 1, 'loss': 5.8950267}\n",
      " 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.7724366}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.7791step ...... 1 7\n",
      "Found change point: 8 in batch 9 with index :3 and 10{'batch': 7, 'size': 1, 'loss': 5.7679863}\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.750205}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.7746step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.85126}\n",
      "10/10 [==============================] - 0s 44ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7823\n",
      "Epoch 185/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.7660675}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.7661step ...... 1{'batch': 1, 'size': 1, 'loss': 5.7656083}\n",
      " 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.721677}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.7511     1 7\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.747972}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 5.7503step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.708286}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.757744}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 5.7446step ......{'batch': 6, 'size': 1, 'loss': 5.8168087} \n",
      "1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.7554407}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.7550step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.75063}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.7446756}\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 5.7535step ...... 1 7\n",
      "10/10 [==============================] - 0s 37ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7535\n",
      "Epoch 186/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.7245135}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.7245step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.7438493}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.79078}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.7530    {'batch': 3, 'size': 1, 'loss': 5.7272234}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 5 with index :5 and 12\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.8336625}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.7640step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.903904}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.8352594}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.7942step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.75831}\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.7983327}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.7906step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.9050403}\n",
      "10/10 [==============================] - 0s 40ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.8021\n",
      "Epoch 187/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.733647}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.7336{'batch': 1, 'size': 1, 'loss': 5.761772}step ...... 1 7\n",
      "\n",
      "Found change point: 8 in batch 3 with index :6 and 13\n",
      "{'batch': 2, 'size': 1, 'loss': 5.7826266}\n",
      "{'batch': 3, 'size': 1, 'loss': 5.7503934}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 5.7571    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.7376876}step ...... 1 7\n",
      "\n",
      "{'batch': 5, 'size': 1, 'loss': 5.843437}step ...... 1 7\n",
      "Found change point: 8 in batch 6 with index :5 and 12\n",
      "\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 5.76831 7\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.7381763}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.7640step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.754659}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.7628step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.737308}\n",
      "step ...... {'batch': 9, 'size': 1, 'loss': 5.803831}\n",
      " 10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 5.76447\n",
      "10/10 [==============================] - 0s 47ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7644\n",
      "Epoch 188/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.8087997}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.8088step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.896229}\n",
      "{'batch': 2, 'size': 1, 'loss': 5.732684}\n",
      "step .... 1 7\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.8126    step ......{'batch': 3, 'size': 1, 'loss': 5.703576}\n",
      " 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.7331257}\n",
      "1 7p .... \n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.7749step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.7326756}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.805037}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.7732step ......{'batch': 7, 'size': 1, 'loss': 5.7118177} 1 7\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.7535076}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.7642step ...... 1{'batch': 9, 'size': 1, 'loss': 5.775779}\n",
      "10/10 [==============================] - 0s 32ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7653\n",
      "\n",
      "Epoch 189/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 5.7544327} 1 7\n",
      "Found change point: 8 in batch 2 with index :1 and 8\n",
      " 1/10 [==>...........................]\n",
      " - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.7544step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.7520375}\n",
      " 2/10 [=====>........................] - ETA: 0s - batch: 0.5000 - size: 1.0000 - loss: 5.7532    step ...... {'batch': 2, 'size': 1, 'loss': 5.709962}\n",
      "1 7\n",
      "step ......{'batch': 3, 'size': 1, 'loss': 5.714498}\n",
      " 1 7\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 5.7327step ......{'batch': 4, 'size': 1, 'loss': 5.757535}\n",
      " 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.7288632}step ...... 1 7\n",
      "\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 5.7362step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.7351813}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.707271}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.7325step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.711996}\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.7302step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.780651}\n",
      "10/10 [==============================] - 0s 44ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7352\n",
      "Epoch 190/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.794976}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.7950step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.7344627}\n",
      "{'batch': 2, 'size': 1, 'loss': 5.705231}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 4 with index :2 and 9\n",
      "{'batch': 3, 'size': 1, 'loss': 5.807355}\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 5.7605     1 7\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 6 with index :7 and 14\n",
      "{'batch': 4, 'size': 1, 'loss': 5.745724}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.7575step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.761235}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.725639}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.7535step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.7422857}\n",
      "{'batch': 8, 'size': 1, 'loss': 5.744119}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.738405}\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 5.7499 1 7\n",
      "10/10 [==============================] - 0s 40ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7499\n",
      "Epoch 191/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.7535114}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.7535{'batch': 1, 'size': 1, 'loss': 5.8337803}\n",
      "{'batch': 2, 'size': 1, 'loss': 5.8260055}step ...... 1 7\n",
      "\n",
      "{'batch': 3, 'size': 1, 'loss': 5.704505}\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 5.7795    step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.751744}\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.7303653}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 5.7667step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.710982}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.7928}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.7630step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.750999}\n",
      "{'batch': 9, 'size': 1, 'loss': 5.733143}\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 36ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7588\n",
      "Epoch 192/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.727253}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.7273step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.732639}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.8105574}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.7568    {'batch': 3, 'size': 1, 'loss': 5.692658}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.7069426}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.7340step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.686986}\n",
      "{'batch': 6, 'size': 1, 'loss': 5.7145944}\n",
      "step ...... 1 7\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.7245step ...... 1 7\n",
      "step ...... 1{'batch': 7, 'size': 1, 'loss': 5.7097864}\n",
      " 7\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.7227step ......{'batch': 8, 'size': 1, 'loss': 5.6821938} \n",
      "1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.8656464}\n",
      "10/10 [==============================] - 0s 36ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7329\n",
      "step ...... 1 7\n",
      "Epoch 193/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 5.742723} \n",
      "1 7\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.7427step ......{'batch': 1, 'size': 1, 'loss': 5.7817383} 1 7\n",
      "\n",
      " 2/10 [=====>........................] - ETA: 0s - batch: 0.5000 - size: 1.0000 - loss: 5.7622    {'batch': 2, 'size': 1, 'loss': 5.8284016}\n",
      "step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.7014146}step ...... 1 7\n",
      "\n",
      " 4/10 [===========>..................] - ETA: 0s - batch: 1.5000 - size: 1.0000 - loss: 5.7636{'batch': 4, 'size': 1, 'loss': 5.788431}\n",
      "step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.8468013}\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 5.7816step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.6786466}\n",
      "step ...... 1 7\n",
      "Found change point: 8 in batch 9 with index :2 and 9\n",
      "{'batch': 7, 'size': 1, 'loss': 5.7545285}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.7653step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.754233}\n",
      "{'batch': 9, 'size': 1, 'loss': 5.787383}\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 5.7664step ...... 1 7\n",
      "Found change point: 8 in batch 1 with index :6 and 13\n",
      "10/10 [==============================] - 0s 39ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7664\n",
      "Epoch 194/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.778292}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.7783{'batch': 1, 'size': 1, 'loss': 5.7209387}\n",
      "step ...... 1{'batch': 2, 'size': 1, 'loss': 5.6947393}\n",
      " 7\n",
      "step ...... {'batch': 3, 'size': 1, 'loss': 5.7902617}\n",
      "1 7\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.7417464}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.7452    step ......{'batch': 5, 'size': 1, 'loss': 5.685011}\n",
      " 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.7174973}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.7761264}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.7381 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.6894693}\n",
      "step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.7019806}\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 5.7296step ...... 1 7\n",
      "10/10 [==============================] - 0s 30ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7296\n",
      "Epoch 195/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.701077}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.7011step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.7393665}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.858556}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.7663    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.6950593}\n",
      "step ......{'batch': 4, 'size': 1, 'loss': 5.6996517}\n",
      " 5/10 [==============>...............]1 7\n",
      " - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.7387step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.7476807}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.716224}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.7368step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.798424}\n",
      "step ...... 1 7{'batch': 8, 'size': 1, 'loss': 5.777182}\n",
      "\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.7481step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.7165837}\n",
      "10/10 [==============================] - 0s 34ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7450\n",
      "Epoch 196/200\n",
      "step ......{'batch': 0, 'size': 1, 'loss': 5.770316}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.7703 1 7\n",
      "step ...... 1 7{'batch': 1, 'size': 1, 'loss': 5.692309}\n",
      "\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.735675}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.7328    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.8210344}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.6914988}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.7422step ......{'batch': 5, 'size': 1, 'loss': 5.7662635}\n",
      " 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.6748695}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.7360\n",
      "{'batch': 7, 'size': 1, 'loss': 5.6888933}\n",
      "step ...... 1 7\n",
      "step ......{'batch': 8, 'size': 1, 'loss': 5.676144}\n",
      " 1 7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.7241step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.8056054}\n",
      "10/10 [==============================] - 0s 44ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7323\n",
      "Epoch 197/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.6767483}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.6767step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.831244}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.6938305}\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.7339    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.818723}\n",
      "{'batch': 4, 'size': 1, 'loss': 5.707315}step ......\n",
      " 1 7\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.7456step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.6907635}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.689103}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.7297step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.7187376}\n",
      "step ......{'batch': 8, 'size': 1, 'loss': 5.742829}\n",
      " 7\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.7299step ......{'batch': 9, 'size': 1, 'loss': 5.7029977}\n",
      " 1 7\n",
      "10/10 [==============================] - 0s 42ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7272\n",
      "Epoch 198/200\n",
      "step ...... 1{'batch': 0, 'size': 1, 'loss': 5.692193} 7\n",
      "\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.6922step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.857595}\n",
      "step ......{'batch': 2, 'size': 1, 'loss': 5.8011928} 1 7\n",
      "\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.7837    step ......{'batch': 3, 'size': 1, 'loss': 5.684697}\n",
      " 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.846319}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.7764step ...... 1 7\n",
      "step ......{'batch': 5, 'size': 1, 'loss': 5.682746}\n",
      " 1 7\n",
      " 6/10 [=================>............] - ETA: 0s - batch: 2.5000 - size: 1.0000 - loss: 5.7608step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.704371}\n",
      "step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.6834364}\n",
      " 8/10 [=======================>......] - ETA: 0s - batch: 3.5000 - size: 1.0000 - loss: 5.7441step ...... 1 7\n",
      "{'batch': 8, 'size': 1, 'loss': 5.725433}\n",
      "step ......{'batch': 9, 'size': 1, 'loss': 5.698621}\n",
      "10/10 [==============================] - ETA: 0s - batch: 4.5000 - size: 1.0000 - loss: 5.7377 1 7\n",
      "10/10 [==============================] - 0s 38ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7377\n",
      "Epoch 199/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.7297287}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.7297step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.84715}\n",
      "{'batch': 2, 'size': 1, 'loss': 5.691427}\n",
      "step ...... 1 7\n",
      " 3/10 [========>.....................] - ETA: 0s - batch: 1.0000 - size: 1.0000 - loss: 5.7561    step ...... 1 7\n",
      "{'batch': 3, 'size': 1, 'loss': 5.741884}\n",
      "{'batch': 4, 'size': 1, 'loss': 5.7861786}\n",
      "step ...... 1 7\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.7593step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.700302}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.829287}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.7609step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.7197156}\n",
      "step ......{'batch': 8, 'size': 1, 'loss': 5.6683116} 1 7\n",
      "\n",
      " 9/10 [==========================>...] - ETA: 0s - batch: 4.0000 - size: 1.0000 - loss: 5.7460step ...... 1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.7008734}\n",
      "10/10 [==============================] - 0s 36ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7415\n",
      "Epoch 200/200\n",
      "step ...... 1 7\n",
      "{'batch': 0, 'size': 1, 'loss': 5.69767}\n",
      " 1/10 [==>...........................] - ETA: 0s - batch: 0.0000e+00 - size: 1.0000 - loss: 5.6977step ...... 1 7\n",
      "{'batch': 1, 'size': 1, 'loss': 5.663066}\n",
      "step ...... 1 7\n",
      "{'batch': 2, 'size': 1, 'loss': 5.7810183}\n",
      "{'batch': 3, 'size': 1, 'loss': 5.6982756}\n",
      "step ...... 1 7\n",
      "{'batch': 4, 'size': 1, 'loss': 5.708421}\n",
      " 5/10 [==============>...............] - ETA: 0s - batch: 2.0000 - size: 1.0000 - loss: 5.7097    step ...... 1 7\n",
      "step ...... 1 7\n",
      "{'batch': 5, 'size': 1, 'loss': 5.6381125}\n",
      "step ...... 1 7\n",
      "{'batch': 6, 'size': 1, 'loss': 5.7076287}\n",
      " 7/10 [====================>.........] - ETA: 0s - batch: 3.0000 - size: 1.0000 - loss: 5.6992step ...... 1 7\n",
      "{'batch': 7, 'size': 1, 'loss': 5.6463547}\n",
      "step ...... {'batch': 8, 'size': 1, 'loss': 5.7747717}\n",
      "1 7\n",
      "{'batch': 9, 'size': 1, 'loss': 5.728339}\n",
      "step ...... 1 7\n",
      "10/10 [==============================] - 0s 25ms/step - batch: 4.5000 - size: 1.0000 - loss: 5.7044\n",
      "Model: \"model\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_1 (InputLayer)         [(None, 7, 2)]            0         \n",
      "_________________________________________________________________\n",
      "lstm (LSTM)                  (None, 7, 30)             3960      \n",
      "_________________________________________________________________\n",
      "dense (Dense)                (None, 7, 3)              93        \n",
      "_________________________________________________________________\n",
      "main_output (GaussianLayer)  [(None, 7, 1), (None, 7,  8         \n",
      "=================================================================\n",
      "Total params: 4,061\n",
      "Trainable params: 4,061\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "ts_train = TimeSeries(training, scaler=MinMaxScaler)\n",
    "ts_test = TimeSeries(test, scaler=MinMaxScaler)\n",
    "dp_model = DeepAR(ts_train, epochs=epochs)\n",
    "dp_model.instantiate_and_fit()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 650
    },
    "id": "4VFqeQ8Bh5Ah",
    "outputId": "850e085b-4244-406f-d784-9758b6a8c9fc",
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "step ...... 1 7\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 3/3 [00:00<00:00, 263.69it/s]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "from numpy.random import normal\n",
    "import tqdm\n",
    "import pandas as pd\n",
    "from matplotlib import pyplot as plt\n",
    "import numpy as np\n",
    "batch = ts_train.next_batch(1,batch_size)\n",
    "def get_sample_prediction(sample, fn, n_steps=batch_size):\n",
    "    sample = np.array(sample).reshape(1, n_steps, variant_count)\n",
    "    output = fn([sample])\n",
    "    samples = []\n",
    "    for mu, sigma in zip(output[0].reshape(n_steps), output[1].reshape(n_steps)):\n",
    "        samples.append(normal(loc=mu, scale=np.sqrt(sigma), size=1)[0])\n",
    "    return np.array(samples)\n",
    "ress = []\n",
    "for i in tqdm.tqdm(range(3)):\n",
    "    pred = get_sample_prediction(batch[0], dp_model.predict_theta_from_input)\n",
    "    ress.append(pred)\n",
    "def plot_uncertainty(ress, ground_truth, n_steps=batch_size, figsize=(9, 6), \n",
    "                     prediction_dots=True, title='Prediction on training set'):\n",
    "    \n",
    "    res_df = pd.DataFrame(ress).T\n",
    "    tot_res = res_df\n",
    "    plt.figure(figsize=figsize)\n",
    "    plt.plot(ground_truth.reshape(n_steps), linewidth=6, label='Original data')\n",
    "    tot_res['mu'] = tot_res.apply(lambda x: np.mean(x), axis=1)\n",
    "    tot_res['upper'] = tot_res.apply(lambda x: np.mean(x) + np.std(x), axis=1)\n",
    "    tot_res['lower'] = tot_res.apply(lambda x: np.mean(x) - np.std(x), axis=1)\n",
    "    tot_res['two_upper'] = tot_res.apply(lambda x: np.mean(x) + 2*np.std(x), axis=1)\n",
    "    tot_res['two_lower'] = tot_res.apply(lambda x: np.mean(x) - 2*np.std(x), axis=1)\n",
    "    plt.plot(tot_res.mu, linewidth=4)\n",
    "    if prediction_dots:\n",
    "        plt.plot(tot_res.mu, 'bo', label='Likelihood mean')\n",
    "    plt.fill_between(x = tot_res.index, y1=tot_res.lower, y2=tot_res.upper, alpha=0.5)\n",
    "    plt.fill_between(x = tot_res.index, y1=tot_res.two_lower, y2=tot_res.two_upper, alpha=0.5)\n",
    "    plt.title(title)\n",
    "    plt.legend()\n",
    "    \n",
    "plot_uncertainty(ress, batch[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 442
    },
    "id": "HBDzxJxdiCG1",
    "outputId": "25edf1e4-d7b4-406e-a46f-59ce855e8337",
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "step ...... 1 7\n",
      "(1, 7, 1)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 300/300 [00:00<00:00, 387.61it/s]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "batch = ts_train.next_batch(1, batch_size)\n",
    "print(batch[1].shape)\n",
    "def get_prediction(sample, prediction_fn):\n",
    "    sample = np.array(sample).reshape(1, batch_size, variant_count)\n",
    "    output = prediction_fn([sample])\n",
    "    samples = []\n",
    "    for mu, sigma in zip(output[0].reshape(batch_size), output[0].reshape(batch_size)):\n",
    "        samples.append(normal(loc=mu, scale = np.sqrt(sigma), size=1)[0])\n",
    "    return np.array(samples)\n",
    "res = []\n",
    "for i in tqdm.tqdm(range(300)):\n",
    "    pred = get_prediction(batch[0], dp_model.predict_theta_from_input)\n",
    "    res.append(pred)\n",
    "plot_uncertainty(res, batch[1]) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "id": "tywMmDTRiDsE"
   },
   "outputs": [],
   "source": [
    "features= []\n",
    "for i in range(1, variant_count+1):\n",
    "    features.append(\"feature_\" + str(i))\n",
    "test_feat = test[features]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {
    "id": "yP_4E59c_b-u"
   },
   "outputs": [],
   "source": [
    "train_variable=15\n",
    "test_variable=15"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "eT8MK50fiG2w",
    "outputId": "7a61edac-14c5-41ac-9f14-86ad1a23bf7b"
   },
   "outputs": [],
   "source": [
    "train_feat =training[features][:-train_variable]\n",
    "values_train = train_feat.values\n",
    "values = test_feat.values[:test_variable]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {
    "id": "F9d31SFPiLuP"
   },
   "outputs": [],
   "source": [
    "curr = 0\n",
    "batches_train= []\n",
    "for i in range(batch_size, len(dataframe) , batch_size):\n",
    "    batch = values_train[curr:i]\n",
    "    batches_train.append(batch)\n",
    "    curr = i"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {
    "id": "7GgvWPhOiM9R"
   },
   "outputs": [],
   "source": [
    "curr = 0\n",
    "batches_test= []\n",
    "for i in range(batch_size, test_variable, batch_size):\n",
    "    batch = values[curr:i]\n",
    "    batches_test.append(batch)\n",
    "    curr = i"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "jhM1O-ZAiOWh",
    "outputId": "270f9a96-f690-4886-aa98-7557d074bf09"
   },
   "outputs": [],
   "source": [
    "test_target_values = test.target.values\n",
    "train_target_values =training.target.values[:-train_variable]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {
    "id": "4SZgKtY3iQ1w"
   },
   "outputs": [],
   "source": [
    "curr = 0\n",
    "batch_targets_train = []\n",
    "for i in range(batch_size, len(dataframe) , batch_size):\n",
    "    batch = train_target_values[curr:i]\n",
    "    batch_targets_train.append(batch)\n",
    "    curr = i"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {
    "id": "69lKhubRiTCF"
   },
   "outputs": [],
   "source": [
    "curr = 0\n",
    "batch_targets_test = []\n",
    "for i in range(batch_size, test_variable, batch_size):\n",
    "    batch = test_target_values[curr:i]\n",
    "    batch_targets_test.append(batch)\n",
    "    curr = i"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {
    "id": "QShOarWJiWbC"
   },
   "outputs": [],
   "source": [
    "predictions_train = []\n",
    "for batch in batches_train:\n",
    "    if batch.size ==batch_size*2:\n",
    "        scaler = MinMaxScaler()\n",
    "        scaled_batch = scaler.fit_transform(batch.reshape(batch_size,variant_count))\n",
    "        scaler_min, scaler_scale, scaler_data_min, scaler_data_max= scaler.min_, scaler.scale_, scaler.data_min_, scaler.data_max_\n",
    "\n",
    "        unscaled_prediction = get_sample_prediction(scaled_batch, dp_model.predict_theta_from_input)\n",
    "        scaler1 = MinMaxScaler()\n",
    "        scaler1.min_, scaler1.scale_ = scaler_min[0], scaler_scale[0]\n",
    "        ress.append(scaler1.inverse_transform([unscaled_prediction])[0])\n",
    "        predictions_train.append(scaler1.inverse_transform([unscaled_prediction])[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {
    "id": "2Y9o2OTxiX2B"
   },
   "outputs": [],
   "source": [
    "predictions_train= np.concatenate( predictions_train, axis=0 )\n",
    "batch_targets_train= np.concatenate( batch_targets_train, axis=0 )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "JwZFLmlliZZu",
    "outputId": "9c4a4005-ada5-4159-cf49-40732f3ef964"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Score: 8.20 RMSE\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import mean_squared_error \n",
    "trainScore = math.sqrt(mean_squared_error(batch_targets_train, predictions_train)) \n",
    "print('Train Score: %.2f RMSE' % (trainScore)) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {
    "id": "x9XVrDMkibLC"
   },
   "outputs": [],
   "source": [
    "predictions_test = []\n",
    "for batch in batches_test:\n",
    "    if batch.size ==batch_size*2:\n",
    "        scaler = MinMaxScaler()\n",
    "        scaled_batch = scaler.fit_transform(batch.reshape(batch_size, variant_count))\n",
    "        scaler_min, scaler_scale, scaler_data_min, scaler_data_max= scaler.min_, scaler.scale_, scaler.data_min_, scaler.data_max_\n",
    "\n",
    "        unscaled_prediction = get_sample_prediction(scaled_batch, dp_model.predict_theta_from_input)\n",
    "        scaler1 = MinMaxScaler()\n",
    "        scaler1.min_, scaler1.scale_ = scaler_min[0], scaler_scale[0]\n",
    "        ress.append(scaler1.inverse_transform([unscaled_prediction])[0])\n",
    "        predictions_test.append(scaler1.inverse_transform([unscaled_prediction])[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {
    "id": "Dy8cbMTjicpg"
   },
   "outputs": [],
   "source": [
    "predictions_test= np.concatenate( predictions_test, axis=0 )\n",
    "batch_targets_test= np.concatenate( batch_targets_test, axis=0 )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "paMsc1lGieJY",
    "outputId": "52f93342-5553-4b28-8a08-893033702dd4"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Score: 14.89 RMSE\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import mean_squared_error\n",
    "testScore = math.sqrt(mean_squared_error(batch_targets_test, predictions_test))\n",
    "print('Test Score: %.2f RMSE' % (testScore)) "
   ]
  }
 ],
 "metadata": {
  "colab": {
   "authorship_tag": "ABX9TyNlEg1NKoC/ARN8zurTk1iD",
   "include_colab_link": true,
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}