{ "cells": [ { "cell_type": "markdown", "id": "e7d28c78", "metadata": {}, "source": [ "# A first artificial neural network with tensorflow and keras" ] }, { "cell_type": "markdown", "id": "f4f75b33", "metadata": {}, "source": [ "In this notebook we will apply a simple neural network using tensorflow and keras.\n", "The ANN we wish to implement consists of 3 inputs and one ouput." ] }, { "cell_type": "code", "execution_count": 1, "id": "ea70e2d8-56f4-4657-938e-4686c09b7410", "metadata": {}, "outputs": [], "source": [ "# We import the required python packages\n", "import numpy as np\n", "import matplotlib.pyplot as plt\n", "import tensorflow as tf" ] }, { "cell_type": "markdown", "id": "07ea3bf6", "metadata": {}, "source": [ "We need to import the dataset using the numpy loadtxt function." ] }, { "cell_type": "code", "execution_count": 2, "id": "4573c373-4317-4ab4-bb98-3e7d6a507cb5", "metadata": {}, "outputs": [], "source": [ "data = np.loadtxt('Data_Example_1_small.txt')" ] }, { "cell_type": "markdown", "id": "3e363744", "metadata": {}, "source": [ "From the dataset we need to extract the input and ouput needed to train the ANN. The three first columns contains the input variables while the last column contains the output variable." ] }, { "cell_type": "code", "execution_count": 3, "id": "7f11c78e-97d2-4f83-80f1-e9ecd40cb574", "metadata": {}, "outputs": [], "source": [ "input_data, output_data = data[:,0:3], data[:,-1]" ] }, { "cell_type": "markdown", "id": "6af28532", "metadata": {}, "source": [ "We then build our first ANN using keras." ] }, { "cell_type": "code", "execution_count": 4, "id": "2c5751af-7e86-4a6d-a88c-43333d4616bb", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "C:\\Users\\haako\\AppData\\Roaming\\Python\\Python312\\site-packages\\keras\\src\\layers\\core\\dense.py:87: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n", " super().__init__(activity_regularizer=activity_regularizer, **kwargs)\n" ] } ], "source": [ "model = tf.keras.models.Sequential() # We define here a sequential ANN\n", "#\n", "model.add(tf.keras.layers.Dense(units=1000, activation='relu', input_dim=3)) # We define the input layer\n", "#\n", "model.add(tf.keras.layers.Dense(units=1, activation='linear')) # We define the output layer\n", "#\n", "model.compile(loss='mse', optimizer='adam') # We compile the ANN model specifying the loss metric and optimizer" ] }, { "cell_type": "markdown", "id": "a39ed78b", "metadata": {}, "source": [ "Once the ANN model has been built we need to train (fit) it to the dataset we imported. Here we have frozen the number of epochs to 1000.\n", "We can also print the minimum and maximum error committed during the training." ] }, { "cell_type": "code", "execution_count": 11, "id": "ce23fefb-90fa-4ce1-a3f9-effce1e975cb", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Epoch 1/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 1.0149e-05\n", "Epoch 2/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 8.5431e-06\n", "Epoch 3/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.8081e-05 \n", "Epoch 4/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 4.7546e-06 \n", "Epoch 5/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.0529e-05\n", "Epoch 6/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 9.7700e-06 \n", "Epoch 7/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 5.2168e-06 \n", "Epoch 8/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 2.0799e-06\n", "Epoch 9/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.2387e-06 \n", "Epoch 10/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.3357e-06 \n", "Epoch 11/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.6731e-06 \n", "Epoch 12/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.4923e-06 \n", "Epoch 13/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 5.5246e-06\n", "Epoch 14/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.5361e-06\n", "Epoch 15/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 14ms/step - loss: 5.4348e-06\n", "Epoch 16/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 4.8159e-06\n", "Epoch 17/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 4.7123e-06\n", "Epoch 18/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.2182e-06 \n", "Epoch 19/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.3592e-06\n", "Epoch 20/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.9799e-06 \n", "Epoch 21/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.6086e-06\n", "Epoch 22/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.3002e-06\n", "Epoch 23/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 2.4469e-06\n", "Epoch 24/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.5265e-06\n", "Epoch 25/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.9195e-06\n", "Epoch 26/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.5093e-06 \n", "Epoch 27/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.4409e-06 \n", "Epoch 28/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 6.9321e-06\n", "Epoch 29/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 6.1284e-06\n", "Epoch 30/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 5.0083e-06\n", "Epoch 31/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.2255e-06 \n", "Epoch 32/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.0903e-06 \n", "Epoch 33/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.9227e-06\n", "Epoch 34/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.2420e-06 \n", "Epoch 35/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.5236e-06 \n", "Epoch 36/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 5.6548e-06 \n", "Epoch 37/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.2548e-06 \n", "Epoch 38/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.4117e-06 \n", "Epoch 39/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 4.4250e-06\n", "Epoch 40/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 6.9447e-06\n", "Epoch 41/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 5.3079e-06 \n", "Epoch 42/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.9462e-06 \n", "Epoch 43/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.9198e-06 \n", "Epoch 44/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 5.1919e-06 \n", "Epoch 45/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 7.0217e-06 \n", "Epoch 46/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.4125e-06 \n", "Epoch 47/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.5352e-06 \n", "Epoch 48/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.0930e-06 \n", "Epoch 49/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 9.5329e-06 \n", "Epoch 50/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.4224e-05 \n", "Epoch 51/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.0335e-05 \n", "Epoch 52/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.2373e-05 \n", "Epoch 53/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.3175e-05 \n", "Epoch 54/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.8553e-05 \n", "Epoch 55/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 6.4765e-06 \n", "Epoch 56/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 5.4814e-06 \n", "Epoch 57/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 7.4501e-06 \n", "Epoch 58/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.4178e-05 \n", "Epoch 59/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.3714e-05 \n", "Epoch 60/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.5537e-05 \n", "Epoch 61/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 7.7468e-06 \n", "Epoch 62/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.0995e-05 \n", "Epoch 63/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 7.7078e-06 \n", "Epoch 64/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.1506e-05\n", "Epoch 65/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 8.3050e-06 \n", "Epoch 66/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.4474e-05 \n", "Epoch 67/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.4329e-05 \n", "Epoch 68/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 7.8953e-06\n", "Epoch 69/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.7256e-05 \n", "Epoch 70/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.9178e-05 \n", "Epoch 71/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.9048e-05 \n", "Epoch 72/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.3789e-05 \n", "Epoch 73/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.0963e-05 \n", "Epoch 74/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.0353e-05 \n", "Epoch 75/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.8729e-05 \n", "Epoch 76/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.9357e-05 \n", "Epoch 77/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.2341e-05 \n", "Epoch 78/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.0564e-05 \n", "Epoch 79/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.8968e-05 \n", "Epoch 80/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.4595e-05 \n", "Epoch 81/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.9071e-05 \n", "Epoch 82/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.5243e-05 \n", "Epoch 83/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.9227e-05 \n", "Epoch 84/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 9.2282e-06 \n", "Epoch 85/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 8.6483e-06 \n", "Epoch 86/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.5916e-05 \n", "Epoch 87/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.6757e-05 \n", "Epoch 88/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.0161e-05 \n", "Epoch 89/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 5.3380e-06 \n", "Epoch 90/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 8.5581e-06 \n", "Epoch 91/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.3220e-05 \n", "Epoch 92/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 7.7862e-06 \n", "Epoch 93/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 4.5686e-06 \n", "Epoch 94/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.5785e-06 \n", "Epoch 95/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 5.7196e-06\n", "Epoch 96/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 6.0363e-06 \n", "Epoch 97/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 6.9489e-06\n", "Epoch 98/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.9591e-06 \n", "Epoch 99/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 5.1661e-06\n", "Epoch 100/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 7.7750e-06 \n", "Epoch 101/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.0315e-05 \n", "Epoch 102/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 8.2496e-06 \n", "Epoch 103/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.0701e-05 \n", "Epoch 104/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.7220e-05\n", "Epoch 105/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.2962e-05 \n", "Epoch 106/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 7.3262e-06 \n", "Epoch 107/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 6.7458e-06 \n", "Epoch 108/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 3.8565e-06\n", "Epoch 109/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 6.4655e-06 \n", "Epoch 110/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.5472e-06 \n", "Epoch 111/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 4.9305e-06 \n", "Epoch 112/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 2.1512e-06\n", "Epoch 113/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.0780e-06 \n", "Epoch 114/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.5619e-06\n", "Epoch 115/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 2.9736e-06\n", "Epoch 116/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 4.6807e-06 \n", "Epoch 117/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.1819e-06 \n", "Epoch 118/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 1.0531e-06 \n", "Epoch 119/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.7582e-06 \n", "Epoch 120/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.2151e-06 \n", "Epoch 121/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.2770e-06 \n", "Epoch 122/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 9.9184e-07 \n", "Epoch 123/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.6348e-06 \n", "Epoch 124/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.5010e-06 \n", "Epoch 125/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.9974e-06 \n", "Epoch 126/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.4006e-06 \n", "Epoch 127/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.8984e-06 \n", "Epoch 128/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 8.1042e-06 \n", "Epoch 129/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 4.9066e-06\n", "Epoch 130/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.7417e-06 \n", "Epoch 131/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 5.3562e-06 \n", "Epoch 132/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 8.0236e-06 \n", "Epoch 133/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 5.4424e-06 \n", "Epoch 134/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.8002e-05 \n", "Epoch 135/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.1267e-05 \n", "Epoch 136/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.7703e-05 \n", "Epoch 137/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 6.9097e-05 \n", "Epoch 138/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 5.8296e-05 \n", "Epoch 139/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.7090e-05 \n", "Epoch 140/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.8331e-04 \n", "Epoch 141/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.9030e-04 \n", "Epoch 142/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.1842e-04 \n", "Epoch 143/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 2.8429e-04 \n", "Epoch 144/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.6238e-04 \n", "Epoch 145/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.2912e-04 \n", "Epoch 146/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.3521e-04 \n", "Epoch 147/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.1487e-04 \n", "Epoch 148/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.9858e-04 \n", "Epoch 149/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.6369e-04\n", "Epoch 150/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.9202e-04 \n", "Epoch 151/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 9.2775e-05 \n", "Epoch 152/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 1.7294e-04 \n", "Epoch 153/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.4344e-04 \n", "Epoch 154/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.5211e-04 \n", "Epoch 155/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.4227e-04 \n", "Epoch 156/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 7.0929e-05 \n", "Epoch 157/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 6.2811e-05 \n", "Epoch 158/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.4426e-04 \n", "Epoch 159/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.2560e-04\n", "Epoch 160/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.0358e-04 \n", "Epoch 161/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.0818e-04 \n", "Epoch 162/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 2.9298e-04 \n", "Epoch 163/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.2685e-04 \n", "Epoch 164/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.0528e-04 \n", "Epoch 165/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 7.0480e-05 \n", "Epoch 166/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.2356e-04 \n", "Epoch 167/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.2180e-04 \n", "Epoch 168/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 6.1320e-05 \n", "Epoch 169/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.0367e-05 \n", "Epoch 170/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 2.4163e-05 \n", "Epoch 171/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 6.5377e-05\n", "Epoch 172/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.0992e-05 \n", "Epoch 173/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 2.4659e-05 \n", "Epoch 174/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 3ms/step - loss: 1.6780e-05 \n", "Epoch 175/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.3032e-05\n", "Epoch 176/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.7898e-05\n", "Epoch 177/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 7.7905e-06\n", "Epoch 178/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 9.4154e-06 \n", "Epoch 179/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 7.5499e-06 \n", "Epoch 180/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 9.2455e-06 \n", "Epoch 181/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.7905e-05 \n", "Epoch 182/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.4932e-05 \n", "Epoch 183/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 2.1822e-05\n", "Epoch 184/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.5057e-05 \n", "Epoch 185/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.0455e-05 \n", "Epoch 186/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.9545e-05 \n", "Epoch 187/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.6122e-05 \n", "Epoch 188/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.5263e-05 \n", "Epoch 189/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.1057e-05 \n", "Epoch 190/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 6.2398e-06 \n", "Epoch 191/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.1318e-05 \n", "Epoch 192/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 8.5150e-06 \n", "Epoch 193/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 7.5498e-06 \n", "Epoch 194/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.8609e-06 \n", "Epoch 195/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 9.0611e-06 \n", "Epoch 196/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 7.3670e-06 \n", "Epoch 197/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.2924e-05 \n", "Epoch 198/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.9841e-05 \n", "Epoch 199/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.3070e-05\n", "Epoch 200/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.1991e-05 \n", "Epoch 201/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.8252e-05 \n", "Epoch 202/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.0595e-05 \n", "Epoch 203/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.7883e-05 \n", "Epoch 204/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.1758e-05 \n", "Epoch 205/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 5.4992e-05 \n", "Epoch 206/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 5.3090e-05\n", "Epoch 207/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 7.4753e-05 \n", "Epoch 208/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.6862e-05 \n", "Epoch 209/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.0937e-05 \n", "Epoch 210/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.0763e-05 \n", "Epoch 211/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 2.6153e-05\n", "Epoch 212/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.2480e-05 \n", "Epoch 213/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.9233e-05 \n", "Epoch 214/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.4633e-05\n", "Epoch 215/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 6.9022e-06 \n", "Epoch 216/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 7.1344e-06 \n", "Epoch 217/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.7324e-06 \n", "Epoch 218/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 8.8047e-06 \n", "Epoch 219/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.6698e-05 \n", "Epoch 220/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.7983e-05\n", "Epoch 221/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 2.7144e-05 \n", "Epoch 222/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.3519e-05 \n", "Epoch 223/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 6.4380e-06 \n", "Epoch 224/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 7.4069e-06 \n", "Epoch 225/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 4.0796e-06 \n", "Epoch 226/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.7912e-06 \n", "Epoch 227/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.6492e-06 \n", "Epoch 228/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.7936e-06 \n", "Epoch 229/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.1554e-06 \n", "Epoch 230/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.4462e-06 \n", "Epoch 231/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.7844e-06 \n", "Epoch 232/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.0018e-06 \n", "Epoch 233/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.4052e-06 \n", "Epoch 234/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.3793e-06\n", "Epoch 235/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 2.2012e-06\n", "Epoch 236/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.6078e-06 \n", "Epoch 237/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.1129e-06 \n", "Epoch 238/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.2680e-06 \n", "Epoch 239/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.7601e-06\n", "Epoch 240/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.3921e-06 \n", "Epoch 241/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 4.1935e-06 \n", "Epoch 242/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 2.2236e-06 \n", "Epoch 243/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.6198e-06 \n", "Epoch 244/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.9919e-06 \n", "Epoch 245/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.3653e-06 \n", "Epoch 246/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.0209e-06 \n", "Epoch 247/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.3655e-06 \n", "Epoch 248/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.4226e-06 \n", "Epoch 249/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.2537e-06\n", "Epoch 250/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.6205e-06 \n", "Epoch 251/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 4.6317e-06 \n", "Epoch 252/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.9522e-06 \n", "Epoch 253/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 9.6502e-07 \n", "Epoch 254/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.5282e-06 \n", "Epoch 255/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.0177e-06 \n", "Epoch 256/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 8.8605e-07\n", "Epoch 257/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.0274e-06 \n", "Epoch 258/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 8.0002e-07\n", "Epoch 259/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 8.7880e-07 \n", "Epoch 260/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.2757e-06 \n", "Epoch 261/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.3007e-06 \n", "Epoch 262/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.7341e-06 \n", "Epoch 263/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 9.2499e-07 \n", "Epoch 264/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 9.8637e-07 \n", "Epoch 265/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.1957e-06 \n", "Epoch 266/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 7.1552e-07 \n", "Epoch 267/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 7.7568e-07 \n", "Epoch 268/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.3235e-06 \n", "Epoch 269/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 6.0539e-07 \n", "Epoch 270/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.9156e-06 \n", "Epoch 271/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.9640e-06\n", "Epoch 272/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 6.5541e-07 \n", "Epoch 273/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 6.7238e-07 \n", "Epoch 274/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.0103e-06 \n", "Epoch 275/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.6697e-06 \n", "Epoch 276/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 3.3799e-06\n", "Epoch 277/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 6.8333e-06 \n", "Epoch 278/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 8.5627e-06 \n", "Epoch 279/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 8.9105e-06 \n", "Epoch 280/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.8058e-06 \n", "Epoch 281/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 8.4171e-06 \n", "Epoch 282/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 6.5531e-06 \n", "Epoch 283/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 4.5723e-06 \n", "Epoch 284/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 4.3689e-06\n", "Epoch 285/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.2034e-06 \n", "Epoch 286/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.0731e-06 \n", "Epoch 287/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.0539e-06 \n", "Epoch 288/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.1175e-06 \n", "Epoch 289/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.4475e-07 \n", "Epoch 290/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 5.9572e-07 \n", "Epoch 291/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 6.0875e-07 \n", "Epoch 292/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.2237e-06 \n", "Epoch 293/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 6.9018e-07 \n", "Epoch 294/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.2859e-06 \n", "Epoch 295/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 7.5889e-07 \n", "Epoch 296/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 9.8883e-07 \n", "Epoch 297/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 6.0344e-07 \n", "Epoch 298/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.5896e-07 \n", "Epoch 299/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.1262e-07 \n", "Epoch 300/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 7.3717e-07\n", "Epoch 301/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.7220e-07 \n", "Epoch 302/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.3537e-07 \n", "Epoch 303/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 4.3744e-07\n", "Epoch 304/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 4.0383e-07 \n", "Epoch 305/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.3006e-07 \n", "Epoch 306/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 4.3109e-07\n", "Epoch 307/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 8.0148e-07 \n", "Epoch 308/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.1112e-06 \n", "Epoch 309/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.3596e-06 \n", "Epoch 310/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 8.4923e-07 \n", "Epoch 311/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.1076e-06 \n", "Epoch 312/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 7.2508e-07\n", "Epoch 313/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.0679e-06 \n", "Epoch 314/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 7.3849e-07 \n", "Epoch 315/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.5886e-06 \n", "Epoch 316/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 5.2977e-07 \n", "Epoch 317/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.9139e-07 \n", "Epoch 318/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.7968e-07 \n", "Epoch 319/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.2610e-07 \n", "Epoch 320/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.3665e-07 \n", "Epoch 321/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 3.2832e-07\n", "Epoch 322/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.0688e-07 \n", "Epoch 323/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 4.3393e-07 \n", "Epoch 324/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.2865e-07 \n", "Epoch 325/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.0037e-06 \n", "Epoch 326/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 6.6225e-07 \n", "Epoch 327/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 2.0889e-06 \n", "Epoch 328/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.6720e-06 \n", "Epoch 329/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.5067e-06 \n", "Epoch 330/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.7612e-06\n", "Epoch 331/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.5327e-06 \n", "Epoch 332/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.2318e-06 \n", "Epoch 333/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.2328e-07 \n", "Epoch 334/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 7.8996e-07 \n", "Epoch 335/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 7.1915e-07 \n", "Epoch 336/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.2041e-06 \n", "Epoch 337/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.0868e-06 \n", "Epoch 338/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 5.5237e-07 \n", "Epoch 339/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 4.5233e-07 \n", "Epoch 340/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.1892e-07 \n", "Epoch 341/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 4.5591e-07 \n", "Epoch 342/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.4432e-07 \n", "Epoch 343/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.7796e-07 \n", "Epoch 344/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 7.2420e-07 \n", "Epoch 345/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 8.3629e-07 \n", "Epoch 346/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 5.5086e-07 \n", "Epoch 347/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 1.0118e-06\n", "Epoch 348/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.6021e-07\n", "Epoch 349/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 5.0101e-07 \n", "Epoch 350/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 5.4201e-07 \n", "Epoch 351/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 4.9935e-07 \n", "Epoch 352/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 6.8504e-07 \n", "Epoch 353/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.6396e-06 \n", "Epoch 354/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.9260e-06 \n", "Epoch 355/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.1129e-06 \n", "Epoch 356/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.4050e-06\n", "Epoch 357/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.2605e-06 \n", "Epoch 358/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.3394e-06 \n", "Epoch 359/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.9109e-06 \n", "Epoch 360/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.0826e-06 \n", "Epoch 361/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 7.7203e-07 \n", "Epoch 362/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.0375e-06 \n", "Epoch 363/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.2600e-06 \n", "Epoch 364/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.5774e-06\n", "Epoch 365/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.9153e-06 \n", "Epoch 366/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.4367e-06 \n", "Epoch 367/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 7.7490e-07 \n", "Epoch 368/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 4.8266e-07 \n", "Epoch 369/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 8.6532e-07 \n", "Epoch 370/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 5.9464e-07 \n", "Epoch 371/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 5.2660e-07\n", "Epoch 372/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 6.1901e-07\n", "Epoch 373/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 5.6336e-07 \n", "Epoch 374/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 6.4954e-07 \n", "Epoch 375/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 6.9293e-07 \n", "Epoch 376/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 6.9881e-07 \n", "Epoch 377/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.6821e-07 \n", "Epoch 378/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 7.4919e-07 \n", "Epoch 379/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 5.9208e-07\n", "Epoch 380/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 4.3560e-07\n", "Epoch 381/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.0562e-07 \n", "Epoch 382/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.0288e-07 \n", "Epoch 383/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.2298e-07 \n", "Epoch 384/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.5298e-07 \n", "Epoch 385/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.4198e-07 \n", "Epoch 386/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.2318e-07 \n", "Epoch 387/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 4.6139e-07 \n", "Epoch 388/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.7688e-07 \n", "Epoch 389/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 5.0376e-07\n", "Epoch 390/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.4541e-07 \n", "Epoch 391/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.8897e-07 \n", "Epoch 392/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.2196e-07 \n", "Epoch 393/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.5978e-07 \n", "Epoch 394/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 6.1464e-07 \n", "Epoch 395/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.2723e-07\n", "Epoch 396/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 6.9684e-07 \n", "Epoch 397/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.9051e-07 \n", "Epoch 398/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.3567e-07 \n", "Epoch 399/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.1684e-07 \n", "Epoch 400/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.7897e-07 \n", "Epoch 401/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 5.7204e-07 \n", "Epoch 402/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 8.5797e-07 \n", "Epoch 403/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.6172e-07 \n", "Epoch 404/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.4018e-07 \n", "Epoch 405/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.2715e-07 \n", "Epoch 406/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.8832e-06 \n", "Epoch 407/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 7.2053e-07 \n", "Epoch 408/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.4336e-06 \n", "Epoch 409/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 6.7997e-07 \n", "Epoch 410/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.2051e-06 \n", "Epoch 411/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.6743e-06 \n", "Epoch 412/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.6630e-06 \n", "Epoch 413/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.8547e-06\n", "Epoch 414/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.7581e-06 \n", "Epoch 415/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.6096e-06 \n", "Epoch 416/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.3031e-06 \n", "Epoch 417/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.2027e-06 \n", "Epoch 418/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 7.6414e-07 \n", "Epoch 419/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 7.7947e-07 \n", "Epoch 420/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 5.2946e-07 \n", "Epoch 421/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.0016e-06\n", "Epoch 422/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 6.7348e-07 \n", "Epoch 423/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.5977e-07 \n", "Epoch 424/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.4862e-07 \n", "Epoch 425/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 2.4127e-07 \n", "Epoch 426/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.1576e-07 \n", "Epoch 427/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.0969e-07 \n", "Epoch 428/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 4.7580e-07 \n", "Epoch 429/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 15ms/step - loss: 4.6101e-07\n", "Epoch 430/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.0209e-07 \n", "Epoch 431/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 7.3970e-07 \n", "Epoch 432/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.0415e-07 \n", "Epoch 433/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 4.9222e-07 \n", "Epoch 434/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 9.7931e-07 \n", "Epoch 435/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 4.6703e-07 \n", "Epoch 436/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.8540e-07 \n", "Epoch 437/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.6124e-07\n", "Epoch 438/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.6213e-06 \n", "Epoch 439/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 9.1258e-07 \n", "Epoch 440/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.6492e-07 \n", "Epoch 441/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.8069e-07 \n", "Epoch 442/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 4.2098e-07 \n", "Epoch 443/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.7121e-07 \n", "Epoch 444/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.0613e-07 \n", "Epoch 445/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.0505e-07 \n", "Epoch 446/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.2977e-07 \n", "Epoch 447/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.3716e-07 \n", "Epoch 448/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 2.3997e-07 \n", "Epoch 449/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 2.2331e-07 \n", "Epoch 450/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.2500e-07 \n", "Epoch 451/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.8106e-07 \n", "Epoch 452/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 2.7954e-07\n", "Epoch 453/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 1.1926e-07 \n", "Epoch 454/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.8072e-07 \n", "Epoch 455/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.2062e-07 \n", "Epoch 456/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.3981e-07 \n", "Epoch 457/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 5.0379e-07 \n", "Epoch 458/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.4952e-07 \n", "Epoch 459/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.0760e-07 \n", "Epoch 460/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 1.1275e-07\n", "Epoch 461/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 9.5280e-08 \n", "Epoch 462/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 8.9983e-08 \n", "Epoch 463/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.2670e-07 \n", "Epoch 464/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.0663e-07 \n", "Epoch 465/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 7.1136e-08 \n", "Epoch 466/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.0421e-07 \n", "Epoch 467/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 15ms/step - loss: 9.8891e-08\n", "Epoch 468/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.0814e-07 \n", "Epoch 469/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.0477e-07 \n", "Epoch 470/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.7376e-07 \n", "Epoch 471/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.8821e-07 \n", "Epoch 472/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.3312e-07 \n", "Epoch 473/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.5544e-07 \n", "Epoch 474/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.4762e-07 \n", "Epoch 475/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 8.7162e-08 \n", "Epoch 476/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.2027e-07 \n", "Epoch 477/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 6.1288e-07 \n", "Epoch 478/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.8990e-07 \n", "Epoch 479/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 7.4646e-08\n", "Epoch 480/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 7.2929e-07 \n", "Epoch 481/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 14ms/step - loss: 2.2532e-07\n", "Epoch 482/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.9796e-07 \n", "Epoch 483/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.0710e-07 \n", "Epoch 484/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 5.2744e-07 \n", "Epoch 485/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.7656e-07 \n", "Epoch 486/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.5978e-07 \n", "Epoch 487/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.6397e-07 \n", "Epoch 488/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.2594e-07\n", "Epoch 489/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.2200e-06 \n", "Epoch 490/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 4.4161e-07 \n", "Epoch 491/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 6.1590e-07 \n", "Epoch 492/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.4029e-07 \n", "Epoch 493/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.2510e-07 \n", "Epoch 494/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.7864e-07 \n", "Epoch 495/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.7908e-07\n", "Epoch 496/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 4.5479e-07 \n", "Epoch 497/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 8.1968e-07 \n", "Epoch 498/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.1615e-06 \n", "Epoch 499/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.5237e-06 \n", "Epoch 500/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 8.3222e-07 \n", "Epoch 501/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 7.5747e-07 \n", "Epoch 502/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 7.6724e-07 \n", "Epoch 503/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.2824e-06 \n", "Epoch 504/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.7165e-06 \n", "Epoch 505/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.4677e-06 \n", "Epoch 506/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.7045e-06 \n", "Epoch 507/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.1473e-07 \n", "Epoch 508/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.9426e-07 \n", "Epoch 509/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.3383e-07 \n", "Epoch 510/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 8.2205e-07 \n", "Epoch 511/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.8781e-06\n", "Epoch 512/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.8233e-06 \n", "Epoch 513/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.0999e-06\n", "Epoch 514/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 7.8257e-07 \n", "Epoch 515/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 6.4380e-07 \n", "Epoch 516/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 6.0175e-07\n", "Epoch 517/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 8.8208e-07 \n", "Epoch 518/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 7.1266e-07 \n", "Epoch 519/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.0323e-07 \n", "Epoch 520/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 6.2760e-07 \n", "Epoch 521/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 8.2181e-07 \n", "Epoch 522/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.0748e-06\n", "Epoch 523/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 16ms/step - loss: 7.5183e-07\n", "Epoch 524/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.2593e-06 \n", "Epoch 525/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 6.0711e-07 \n", "Epoch 526/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.5968e-07 \n", "Epoch 527/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.8853e-07 \n", "Epoch 528/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.9275e-07 \n", "Epoch 529/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.4243e-07 \n", "Epoch 530/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 9.1226e-08\n", "Epoch 531/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.2440e-07 \n", "Epoch 532/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.3851e-07 \n", "Epoch 533/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 9.0074e-08 \n", "Epoch 534/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 6.0524e-08 \n", "Epoch 535/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 7.7513e-08 \n", "Epoch 536/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.5230e-07 \n", "Epoch 537/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 9.6614e-08\n", "Epoch 538/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.1021e-07 \n", "Epoch 539/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 6.2970e-08 \n", "Epoch 540/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 7.7181e-08 \n", "Epoch 541/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 6.5205e-08 \n", "Epoch 542/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 5.1939e-08 \n", "Epoch 543/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 5.8107e-08\n", "Epoch 544/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 5.8146e-08\n", "Epoch 545/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 7.5254e-08 \n", "Epoch 546/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.2426e-07 \n", "Epoch 547/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.8446e-07 \n", "Epoch 548/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.1723e-07\n", "Epoch 549/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.0448e-07 \n", "Epoch 550/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.1289e-07 \n", "Epoch 551/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 3.6782e-07\n", "Epoch 552/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 9.2276e-07 \n", "Epoch 553/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 8.4513e-07 \n", "Epoch 554/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 6.9140e-07 \n", "Epoch 555/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.0132e-07 \n", "Epoch 556/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.9356e-07 \n", "Epoch 557/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 4.5151e-07 \n", "Epoch 558/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 3.6772e-07\n", "Epoch 559/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.0076e-07 \n", "Epoch 560/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.3407e-07 \n", "Epoch 561/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.6618e-07 \n", "Epoch 562/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 4.3394e-07\n", "Epoch 563/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 3.6762e-07\n", "Epoch 564/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 14ms/step - loss: 4.4940e-07\n", "Epoch 565/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.5470e-07 \n", "Epoch 566/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.6762e-07 \n", "Epoch 567/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.2321e-07 \n", "Epoch 568/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.5919e-07 \n", "Epoch 569/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.4173e-07 \n", "Epoch 570/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.5403e-07 \n", "Epoch 571/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.1905e-07 \n", "Epoch 572/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.5222e-07 \n", "Epoch 573/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.6615e-07 \n", "Epoch 574/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.5234e-07\n", "Epoch 575/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.9229e-07 \n", "Epoch 576/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.1700e-07 \n", "Epoch 577/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 3.2305e-07\n", "Epoch 578/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 4.9837e-07\n", "Epoch 579/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.8428e-07 \n", "Epoch 580/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.9445e-07 \n", "Epoch 581/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 2.7785e-07\n", "Epoch 582/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.6613e-07 \n", "Epoch 583/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 3.4498e-07\n", "Epoch 584/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 8.4785e-07 \n", "Epoch 585/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.3079e-06 \n", "Epoch 586/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 2.3938e-06\n", "Epoch 587/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 9.0392e-07 \n", "Epoch 588/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.2944e-06 \n", "Epoch 589/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.2888e-06 \n", "Epoch 590/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 4.0008e-06\n", "Epoch 591/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.7037e-06 \n", "Epoch 592/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 9.3156e-06\n", "Epoch 593/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.0182e-05 \n", "Epoch 594/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.5925e-05 \n", "Epoch 595/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.1381e-05 \n", "Epoch 596/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.2722e-05 \n", "Epoch 597/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 4.6148e-05 \n", "Epoch 598/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.6795e-05 \n", "Epoch 599/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 6.9230e-05 \n", "Epoch 600/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.7396e-04 \n", "Epoch 601/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 6.3638e-05 \n", "Epoch 602/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 6.9272e-05 \n", "Epoch 603/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 2.7567e-05\n", "Epoch 604/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.4915e-04 \n", "Epoch 605/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 6.7104e-05 \n", "Epoch 606/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.4687e-05 \n", "Epoch 607/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.5414e-04 \n", "Epoch 608/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 9.4515e-05 \n", "Epoch 609/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 6.4948e-05 \n", "Epoch 610/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.2721e-04 \n", "Epoch 611/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.0812e-04 \n", "Epoch 612/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.8464e-05 \n", "Epoch 613/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 4.7129e-05 \n", "Epoch 614/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.6714e-05 \n", "Epoch 615/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.2262e-05 \n", "Epoch 616/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 2.8638e-05\n", "Epoch 617/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.0369e-05 \n", "Epoch 618/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.3121e-05 \n", "Epoch 619/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.2168e-05 \n", "Epoch 620/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.4497e-05 \n", "Epoch 621/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.1002e-05 \n", "Epoch 622/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.2489e-05\n", "Epoch 623/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.2891e-05 \n", "Epoch 624/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.2462e-05 \n", "Epoch 625/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.6655e-05 \n", "Epoch 626/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.6066e-05 \n", "Epoch 627/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.3971e-05 \n", "Epoch 628/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 1.7873e-05\n", "Epoch 629/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.0722e-05 \n", "Epoch 630/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.8381e-05 \n", "Epoch 631/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 1.0744e-05 \n", "Epoch 632/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 1.8588e-05 \n", "Epoch 633/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.3058e-05 \n", "Epoch 634/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 1.8140e-05\n", "Epoch 635/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.8314e-05 \n", "Epoch 636/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 5.0074e-05 \n", "Epoch 637/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.1077e-05 \n", "Epoch 638/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 9.2022e-06 \n", "Epoch 639/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.8950e-05 \n", "Epoch 640/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 6.0334e-06 \n", "Epoch 641/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 8.2510e-06\n", "Epoch 642/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.7438e-05 \n", "Epoch 643/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 5.4725e-05 \n", "Epoch 644/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 8.0529e-05 \n", "Epoch 645/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.6234e-05 \n", "Epoch 646/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.2339e-05 \n", "Epoch 647/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 1.7727e-05\n", "Epoch 648/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.0040e-05 \n", "Epoch 649/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.0243e-04 \n", "Epoch 650/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 5.8849e-05 \n", "Epoch 651/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.0501e-04 \n", "Epoch 652/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.4185e-05 \n", "Epoch 653/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.9287e-05 \n", "Epoch 654/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 6.4386e-05\n", "Epoch 655/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.6062e-05 \n", "Epoch 656/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.6038e-05 \n", "Epoch 657/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.8119e-05 \n", "Epoch 658/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.8213e-05\n", "Epoch 659/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.6514e-05 \n", "Epoch 660/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.7851e-05 \n", "Epoch 661/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.7669e-05 \n", "Epoch 662/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.7891e-05 \n", "Epoch 663/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 4.7874e-05\n", "Epoch 664/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.8153e-05 \n", "Epoch 665/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.2866e-05 \n", "Epoch 666/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.1049e-04\n", "Epoch 667/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 6.4587e-05 \n", "Epoch 668/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.2379e-04 \n", "Epoch 669/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.0109e-05 \n", "Epoch 670/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 4.1376e-05\n", "Epoch 671/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.0288e-04 \n", "Epoch 672/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.8224e-04\n", "Epoch 673/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 7.1543e-05 \n", "Epoch 674/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 7.2980e-05 \n", "Epoch 675/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.9328e-05 \n", "Epoch 676/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.9988e-05 \n", "Epoch 677/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 9.6000e-06 \n", "Epoch 678/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 8.0811e-06\n", "Epoch 679/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 9.3589e-06 \n", "Epoch 680/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.8150e-06 \n", "Epoch 681/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 8.2069e-06 \n", "Epoch 682/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 5.8161e-06 \n", "Epoch 683/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 4.1736e-06\n", "Epoch 684/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.3760e-06 \n", "Epoch 685/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 5.7693e-06 \n", "Epoch 686/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.8662e-06 \n", "Epoch 687/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 4.8855e-06 \n", "Epoch 688/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 4.0973e-06\n", "Epoch 689/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 4.6574e-06\n", "Epoch 690/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 9.6086e-06 \n", "Epoch 691/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 8.5622e-06 \n", "Epoch 692/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 5.5949e-06 \n", "Epoch 693/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.1695e-05 \n", "Epoch 694/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 8.9700e-06\n", "Epoch 695/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 9.7796e-06\n", "Epoch 696/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.4122e-05 \n", "Epoch 697/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 2.9918e-05\n", "Epoch 698/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 2.7490e-05\n", "Epoch 699/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.2526e-05 \n", "Epoch 700/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 9.3943e-06 \n", "Epoch 701/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 5.5076e-06 \n", "Epoch 702/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 5.1147e-06 \n", "Epoch 703/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.5948e-06 \n", "Epoch 704/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 4.7789e-06 \n", "Epoch 705/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.8327e-06 \n", "Epoch 706/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 2.2771e-06 \n", "Epoch 707/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.7216e-06 \n", "Epoch 708/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 2.7318e-06 \n", "Epoch 709/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.6791e-06 \n", "Epoch 710/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 2.5251e-06\n", "Epoch 711/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.3488e-06 \n", "Epoch 712/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 1.7432e-06\n", "Epoch 713/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.2419e-06 \n", "Epoch 714/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.1942e-06 \n", "Epoch 715/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.6384e-06 \n", "Epoch 716/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 9.9528e-07\n", "Epoch 717/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.3286e-06 \n", "Epoch 718/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.2348e-06\n", "Epoch 719/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.6215e-06 \n", "Epoch 720/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.0189e-06 \n", "Epoch 721/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.3870e-06 \n", "Epoch 722/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.4768e-06 \n", "Epoch 723/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.2198e-06 \n", "Epoch 724/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.2171e-06\n", "Epoch 725/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 9.6026e-07\n", "Epoch 726/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 9.5031e-07 \n", "Epoch 727/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 6.1755e-07 \n", "Epoch 728/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 8.6801e-07 \n", "Epoch 729/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.9080e-06 \n", "Epoch 730/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 8.2388e-07 \n", "Epoch 731/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 4.3936e-07\n", "Epoch 732/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 7.5438e-07 \n", "Epoch 733/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.7997e-07 \n", "Epoch 734/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 5.9270e-07 \n", "Epoch 735/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 8.8454e-07 \n", "Epoch 736/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 8.4161e-07 \n", "Epoch 737/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 1.5958e-06\n", "Epoch 738/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.1544e-06 \n", "Epoch 739/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.5808e-06 \n", "Epoch 740/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.0434e-06 \n", "Epoch 741/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.2984e-07 \n", "Epoch 742/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.2339e-06 \n", "Epoch 743/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.1772e-06\n", "Epoch 744/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 4.8949e-07 \n", "Epoch 745/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 4.8081e-07 \n", "Epoch 746/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 4.9528e-07 \n", "Epoch 747/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 8.4438e-07 \n", "Epoch 748/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.0084e-06 \n", "Epoch 749/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.0306e-06 \n", "Epoch 750/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.5713e-06 \n", "Epoch 751/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.3522e-06\n", "Epoch 752/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 4.5468e-06 \n", "Epoch 753/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.1293e-06 \n", "Epoch 754/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.2757e-06 \n", "Epoch 755/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.1340e-06 \n", "Epoch 756/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 7.9468e-07 \n", "Epoch 757/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 4.5465e-07\n", "Epoch 758/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.0072e-06 \n", "Epoch 759/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.6318e-07 \n", "Epoch 760/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 4.6746e-07 \n", "Epoch 761/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 1.2324e-06 \n", "Epoch 762/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 4.8660e-06 \n", "Epoch 763/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.3014e-06 \n", "Epoch 764/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.5164e-06 \n", "Epoch 765/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 5.0402e-06 \n", "Epoch 766/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 7.3650e-06\n", "Epoch 767/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.2445e-06 \n", "Epoch 768/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.5764e-06 \n", "Epoch 769/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.7300e-06\n", "Epoch 770/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.0627e-06 \n", "Epoch 771/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 7.9970e-07 \n", "Epoch 772/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.9690e-06 \n", "Epoch 773/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 4.0500e-06 \n", "Epoch 774/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 14ms/step - loss: 1.8647e-06\n", "Epoch 775/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 5.7144e-06\n", "Epoch 776/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.7924e-06 \n", "Epoch 777/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.0952e-06 \n", "Epoch 778/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.2554e-06\n", "Epoch 779/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 3.6642e-06\n", "Epoch 780/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 2.0803e-06\n", "Epoch 781/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.8214e-06 \n", "Epoch 782/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 5.4365e-06\n", "Epoch 783/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 3.0444e-06\n", "Epoch 784/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 5.9519e-06 \n", "Epoch 785/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 5.5897e-06 \n", "Epoch 786/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 3.5572e-06\n", "Epoch 787/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 7.0586e-07 \n", "Epoch 788/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.3689e-06 \n", "Epoch 789/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 6.3092e-07 \n", "Epoch 790/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 20ms/step - loss: 6.5194e-07\n", "Epoch 791/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 30ms/step - loss: 8.2328e-07\n", "Epoch 792/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 19ms/step - loss: 1.0741e-06\n", "Epoch 793/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.3650e-06\n", "Epoch 794/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.1818e-06\n", "Epoch 795/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 5.7342e-07 \n", "Epoch 796/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 14ms/step - loss: 5.4948e-07\n", "Epoch 797/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 4.9456e-07\n", "Epoch 798/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 5.6798e-07\n", "Epoch 799/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.8819e-06\n", "Epoch 800/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 6.9052e-07\n", "Epoch 801/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 7.9311e-07 \n", "Epoch 802/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 7.5136e-07\n", "Epoch 803/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 4.7556e-07 \n", "Epoch 804/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.9159e-07\n", "Epoch 805/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.6660e-07\n", "Epoch 806/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.8910e-07 \n", "Epoch 807/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 4.6006e-07\n", "Epoch 808/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.9112e-07 \n", "Epoch 809/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 4.1314e-07\n", "Epoch 810/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 4.4363e-07\n", "Epoch 811/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.4923e-07 \n", "Epoch 812/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 6.0396e-07\n", "Epoch 813/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.9017e-07 \n", "Epoch 814/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 2.3053e-07 \n", "Epoch 815/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 9.2544e-07 \n", "Epoch 816/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 5.1212e-07\n", "Epoch 817/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.0703e-06 \n", "Epoch 818/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.2657e-06 \n", "Epoch 819/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 9.9115e-07\n", "Epoch 820/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.9747e-07 \n", "Epoch 821/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 5.4179e-07\n", "Epoch 822/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 6.3030e-07 \n", "Epoch 823/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 4.2158e-07\n", "Epoch 824/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 2.5364e-07\n", "Epoch 825/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 3.5465e-07\n", "Epoch 826/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 5.1666e-07 \n", "Epoch 827/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.7039e-07 \n", "Epoch 828/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.3853e-07\n", "Epoch 829/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 2.1293e-07 \n", "Epoch 830/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.6128e-07\n", "Epoch 831/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.5950e-07\n", "Epoch 832/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 6.5743e-07 \n", "Epoch 833/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 15ms/step - loss: 5.4609e-07\n", "Epoch 834/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 14ms/step - loss: 1.4845e-07\n", "Epoch 835/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 15ms/step - loss: 3.7321e-07\n", "Epoch 836/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 18ms/step - loss: 5.2318e-07 \n", "Epoch 837/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 1.3632e-07\n", "Epoch 838/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.8148e-07\n", "Epoch 839/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 17ms/step - loss: 9.8689e-08\n", "Epoch 840/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.0055e-07 \n", "Epoch 841/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 8.9884e-08\n", "Epoch 842/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.4893e-07\n", "Epoch 843/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.3607e-07\n", "Epoch 844/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 7.4470e-07 \n", "Epoch 845/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 17ms/step - loss: 4.2315e-07\n", "Epoch 846/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 9.5667e-07 \n", "Epoch 847/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.0384e-06 \n", "Epoch 848/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 9.8775e-07 \n", "Epoch 849/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.0989e-06 \n", "Epoch 850/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 19ms/step - loss: 1.7239e-06\n", "Epoch 851/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.7691e-06\n", "Epoch 852/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.6220e-06\n", "Epoch 853/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.3783e-06\n", "Epoch 854/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 3.9912e-06\n", "Epoch 855/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.9590e-06\n", "Epoch 856/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 6.9582e-07 \n", "Epoch 857/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.0750e-06\n", "Epoch 858/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 6.8445e-07\n", "Epoch 859/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.5978e-06\n", "Epoch 860/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 7.6629e-07\n", "Epoch 861/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 1.0854e-06\n", "Epoch 862/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.4590e-06\n", "Epoch 863/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 9.9121e-07\n", "Epoch 864/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 1.0774e-06\n", "Epoch 865/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.3846e-06\n", "Epoch 866/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 14ms/step - loss: 2.4462e-06\n", "Epoch 867/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 3.2801e-06\n", "Epoch 868/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.9077e-06\n", "Epoch 869/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.1357e-06 \n", "Epoch 870/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.9288e-06\n", "Epoch 871/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 1.1465e-06\n", "Epoch 872/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 8.4069e-07 \n", "Epoch 873/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 4.2076e-07\n", "Epoch 874/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.4014e-06\n", "Epoch 875/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 4.2507e-06\n", "Epoch 876/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.5791e-06\n", "Epoch 877/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.0053e-06\n", "Epoch 878/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 2.4518e-06 \n", "Epoch 879/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.1145e-06\n", "Epoch 880/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.4359e-06\n", "Epoch 881/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 6.7732e-07\n", "Epoch 882/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 5.1941e-07\n", "Epoch 883/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.3524e-06\n", "Epoch 884/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 5.3673e-07\n", "Epoch 885/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.4602e-06 \n", "Epoch 886/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 4.8156e-06\n", "Epoch 887/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.3982e-06\n", "Epoch 888/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 6.9251e-07 \n", "Epoch 889/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.0421e-05\n", "Epoch 890/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.1491e-05 \n", "Epoch 891/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 8.7398e-06\n", "Epoch 892/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 14ms/step - loss: 2.0352e-06\n", "Epoch 893/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 2.2080e-06\n", "Epoch 894/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.8742e-06 \n", "Epoch 895/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 2.5836e-06\n", "Epoch 896/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 5.5178e-06 \n", "Epoch 897/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 4.2024e-06\n", "Epoch 898/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.6914e-06\n", "Epoch 899/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.8290e-06\n", "Epoch 900/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 4.3368e-06\n", "Epoch 901/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 4.0442e-06 \n", "Epoch 902/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.5125e-06\n", "Epoch 903/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.0335e-06\n", "Epoch 904/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 7.1986e-06\n", "Epoch 905/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.0512e-05\n", "Epoch 906/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 8.7564e-06\n", "Epoch 907/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 5.3488e-06\n", "Epoch 908/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 9.6640e-06 \n", "Epoch 909/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 6.5456e-06\n", "Epoch 910/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 9.3343e-06\n", "Epoch 911/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.3482e-05\n", "Epoch 912/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 1.5106e-05 \n", "Epoch 913/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.2420e-05\n", "Epoch 914/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.3914e-05\n", "Epoch 915/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 7.2421e-06 \n", "Epoch 916/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.8038e-05 \n", "Epoch 917/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 6.6302e-06\n", "Epoch 918/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 2.8180e-06 \n", "Epoch 919/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.1068e-06 \n", "Epoch 920/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 14ms/step - loss: 2.9457e-06\n", "Epoch 921/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 4.4723e-06\n", "Epoch 922/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 7.1226e-06 \n", "Epoch 923/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.5497e-05 \n", "Epoch 924/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.5524e-05 \n", "Epoch 925/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.5180e-05 \n", "Epoch 926/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 4.2365e-05\n", "Epoch 927/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 6.5072e-05\n", "Epoch 928/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.6943e-04\n", "Epoch 929/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.0041e-04 \n", "Epoch 930/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 5.3435e-05 \n", "Epoch 931/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.4477e-05\n", "Epoch 932/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.8635e-05 \n", "Epoch 933/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.8815e-05 \n", "Epoch 934/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 2.8274e-05\n", "Epoch 935/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 9.4917e-06 \n", "Epoch 936/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 17ms/step - loss: 1.3116e-05\n", "Epoch 937/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 7.2526e-06 \n", "Epoch 938/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.7754e-05\n", "Epoch 939/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 6.1683e-05 \n", "Epoch 940/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 4.3318e-05\n", "Epoch 941/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 16ms/step - loss: 1.4507e-05\n", "Epoch 942/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.4664e-05\n", "Epoch 943/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 7.9469e-06 \n", "Epoch 944/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.8121e-05\n", "Epoch 945/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 1.2453e-05 \n", "Epoch 946/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 18ms/step - loss: 1.0997e-05\n", "Epoch 947/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 4.1620e-05 \n", "Epoch 948/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.4057e-05\n", "Epoch 949/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 4.6831e-05\n", "Epoch 950/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 14ms/step - loss: 9.0598e-05\n", "Epoch 951/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 2.7111e-05\n", "Epoch 952/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 6.8544e-05\n", "Epoch 953/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 3.7671e-05\n", "Epoch 954/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 5.6544e-05\n", "Epoch 955/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 4.9712e-05\n", "Epoch 956/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.5602e-05\n", "Epoch 957/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 3.9131e-05\n", "Epoch 958/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 3.3396e-05\n", "Epoch 959/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 3.4367e-05\n", "Epoch 960/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.6993e-05 \n", "Epoch 961/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.9510e-05 \n", "Epoch 962/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 7.6175e-05\n", "Epoch 963/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.2589e-05 \n", "Epoch 964/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 1.4160e-05 \n", "Epoch 965/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 3.0726e-05\n", "Epoch 966/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 1.5197e-05\n", "Epoch 967/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.6696e-05 \n", "Epoch 968/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.3886e-05\n", "Epoch 969/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.7332e-05\n", "Epoch 970/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 2.1495e-05\n", "Epoch 971/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 4.3317e-05\n", "Epoch 972/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 4.3138e-05\n", "Epoch 973/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.6776e-05\n", "Epoch 974/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.6834e-05\n", "Epoch 975/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 4.5002e-05\n", "Epoch 976/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 17ms/step - loss: 4.1358e-05\n", "Epoch 977/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.8138e-05\n", "Epoch 978/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 4.7283e-05\n", "Epoch 979/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 2.4146e-05\n", "Epoch 980/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 2.8659e-05\n", "Epoch 981/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 3.3377e-05\n", "Epoch 982/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 4.8647e-05\n", "Epoch 983/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 6.2116e-05 \n", "Epoch 984/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.1454e-05 \n", "Epoch 985/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 5.5002e-05\n", "Epoch 986/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 1.2805e-04\n", "Epoch 987/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.9452e-04\n", "Epoch 988/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 2.5990e-04\n", "Epoch 989/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 1.3420e-04\n", "Epoch 990/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.8904e-04\n", "Epoch 991/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 1.5350e-04 \n", "Epoch 992/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 3.3005e-04\n", "Epoch 993/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 1.3404e-04\n", "Epoch 994/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 5.7226e-05\n", "Epoch 995/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 6.1577e-05\n", "Epoch 996/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 13ms/step - loss: 5.5548e-05\n", "Epoch 997/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 4.6614e-05\n", "Epoch 998/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 4.0431e-05\n", "Epoch 999/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.1866e-05 \n", "Epoch 1000/1000\n", "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.2314e-05 \n", "Minimum error : 5.423180837738073e-08, Maximum error: 0.0003575421869754791\n" ] } ], "source": [ "history = model.fit(input_data, output_data, epochs=1000,verbose=1)\n", "print('Minimum error : {}, Maximum error: {}'.format(np.min(history.history['loss']),np.max(history.history['loss'])))" ] }, { "cell_type": "markdown", "id": "b81ec389", "metadata": {}, "source": [ "We can print the number of parameters in the ANN model" ] }, { "cell_type": "code", "execution_count": 12, "id": "8510a395", "metadata": {}, "outputs": [ { "ename": "ValueError", "evalue": "setting an array element with a sequence. The requested array has an inhomogeneous shape after 1 dimensions. The detected shape was (2,) + inhomogeneous part.", "output_type": "error", "traceback": [ "\u001b[31m---------------------------------------------------------------------------\u001b[39m", "\u001b[31mAttributeError\u001b[39m Traceback (most recent call last)", "\u001b[36mFile \u001b[39m\u001b[32m~\\AppData\\Roaming\\Python\\Python312\\site-packages\\numpy\\_core\\fromnumeric.py:2170\u001b[39m, in \u001b[36mshape\u001b[39m\u001b[34m(a)\u001b[39m\n\u001b[32m 2169\u001b[39m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[32m-> \u001b[39m\u001b[32m2170\u001b[39m result = \u001b[43ma\u001b[49m\u001b[43m.\u001b[49m\u001b[43mshape\u001b[49m\n\u001b[32m 2171\u001b[39m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mAttributeError\u001b[39;00m:\n", "\u001b[31mAttributeError\u001b[39m: 'list' object has no attribute 'shape'", "\nDuring handling of the above exception, another exception occurred:\n", "\u001b[31mValueError\u001b[39m Traceback (most recent call last)", "\u001b[36mCell\u001b[39m\u001b[36m \u001b[39m\u001b[32mIn[12]\u001b[39m\u001b[32m, line 2\u001b[39m\n\u001b[32m 1\u001b[39m \u001b[38;5;28;01mfor\u001b[39;00m layer \u001b[38;5;129;01min\u001b[39;00m model.layers:\n\u001b[32m----> \u001b[39m\u001b[32m2\u001b[39m nrow = \u001b[43mnp\u001b[49m\u001b[43m.\u001b[49m\u001b[43mshape\u001b[49m\u001b[43m(\u001b[49m\u001b[43mlayer\u001b[49m\u001b[43m.\u001b[49m\u001b[43mget_weights\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[43m)\u001b[49m[\u001b[32m0\u001b[39m]\n\u001b[32m 3\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m nrow != \u001b[32m0\u001b[39m:\n\u001b[32m 4\u001b[39m \u001b[38;5;28mprint\u001b[39m(\u001b[33m'\u001b[39m\u001b[33mNumber of weights = \u001b[39m\u001b[33m'\u001b[39m,np.shape(layer.get_weights()[\u001b[32m0\u001b[39m]))\n", "\u001b[36mFile \u001b[39m\u001b[32m~\\AppData\\Roaming\\Python\\Python312\\site-packages\\numpy\\_core\\fromnumeric.py:2172\u001b[39m, in \u001b[36mshape\u001b[39m\u001b[34m(a)\u001b[39m\n\u001b[32m 2170\u001b[39m result = a.shape\n\u001b[32m 2171\u001b[39m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mAttributeError\u001b[39;00m:\n\u001b[32m-> \u001b[39m\u001b[32m2172\u001b[39m result = \u001b[43masarray\u001b[49m\u001b[43m(\u001b[49m\u001b[43ma\u001b[49m\u001b[43m)\u001b[49m.shape\n\u001b[32m 2173\u001b[39m \u001b[38;5;28;01mreturn\u001b[39;00m result\n", "\u001b[31mValueError\u001b[39m: setting an array element with a sequence. The requested array has an inhomogeneous shape after 1 dimensions. The detected shape was (2,) + inhomogeneous part." ] } ], "source": [ "for layer in model.layers:\n", " nrow = np.shape(layer.get_weights())[0]\n", " if nrow != 0:\n", " print('Number of weights = ',np.shape(layer.get_weights()[0]))\n", " print('Number of bias = ',np.shape(layer.get_weights()[1]))" ] }, { "cell_type": "markdown", "id": "ddfb054b", "metadata": {}, "source": [ "Once we have trained the ANN we can use it to do a prediction, here with the input data used for the training." ] }, { "cell_type": "code", "execution_count": 13, "id": "b5c58de5", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step \n" ] } ], "source": [ "prediction = model.predict(input_data)" ] }, { "cell_type": "markdown", "id": "0f5f6c86", "metadata": {}, "source": [ "The next cell plots the results from the ANN. The left graph shows the evolution of the loss metric as function of epochs while the right graph shows the predicted values compared to the measured data." ] }, { "cell_type": "code", "execution_count": 14, "id": "245ec6e3-0d98-4a58-8ca9-02ec0aae8a9d", "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plt.rcParams[\"figure.figsize\"] = (15, 10)\n", "fig,axs = plt.subplots(1,2)\n", "axs[0].semilogy(history.history['loss'],'k',label='training')\n", "axs[0].set_xlabel('Epoch')\n", "axs[0].set_ylabel('Error')\n", "axs[0].legend()\n", "axs[0].grid()\n", "# Plot predictions from ANN\n", "axs[1].scatter(output_data,prediction,c='k',label='Predictions')\n", "axs[1].plot([np.min(output_data),np.max(output_data)],[np.min(prediction),np.max(prediction)],'r',label='Reference')\n", "axs[1].set_xlabel('Measured')\n", "axs[1].set_ylabel('Predicted')\n", "axs[1].legend()\n", "axs[1].grid()\n", "fig.tight_layout()" ] }, { "cell_type": "markdown", "id": "506c7d01", "metadata": {}, "source": [ "It is good practice to save the trained ANN since big networks and/or large amount of data can be time consuming to train." ] }, { "cell_type": "code", "execution_count": 9, "id": "95aed764-cb95-43cb-a6e8-d94fdfece4a9", "metadata": {}, "outputs": [ { "ename": "ValueError", "evalue": "Invalid filepath extension for saving. Please add either a `.keras` extension for the native Keras format (recommended) or a `.h5` extension. Use `model.export(filepath)` if you want to export a SavedModel for use with TFLite/TFServing/etc. Received: filepath=model_small.pb.", "output_type": "error", "traceback": [ "\u001b[31m---------------------------------------------------------------------------\u001b[39m", "\u001b[31mValueError\u001b[39m Traceback (most recent call last)", "\u001b[36mCell\u001b[39m\u001b[36m \u001b[39m\u001b[32mIn[9]\u001b[39m\u001b[32m, line 1\u001b[39m\n\u001b[32m----> \u001b[39m\u001b[32m1\u001b[39m \u001b[43mmodel\u001b[49m\u001b[43m.\u001b[49m\u001b[43msave\u001b[49m\u001b[43m(\u001b[49m\u001b[33;43m'\u001b[39;49m\u001b[33;43mmodel_small.pb\u001b[39;49m\u001b[33;43m'\u001b[39;49m\u001b[43m)\u001b[49m\n", "\u001b[36mFile \u001b[39m\u001b[32m~\\AppData\\Roaming\\Python\\Python312\\site-packages\\keras\\src\\utils\\traceback_utils.py:122\u001b[39m, in \u001b[36mfilter_traceback..error_handler\u001b[39m\u001b[34m(*args, **kwargs)\u001b[39m\n\u001b[32m 119\u001b[39m filtered_tb = _process_traceback_frames(e.__traceback__)\n\u001b[32m 120\u001b[39m \u001b[38;5;66;03m# To get the full stack trace, call:\u001b[39;00m\n\u001b[32m 121\u001b[39m \u001b[38;5;66;03m# `keras.config.disable_traceback_filtering()`\u001b[39;00m\n\u001b[32m--> \u001b[39m\u001b[32m122\u001b[39m \u001b[38;5;28;01mraise\u001b[39;00m e.with_traceback(filtered_tb) \u001b[38;5;28;01mfrom\u001b[39;00m\u001b[38;5;250m \u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[32m 123\u001b[39m \u001b[38;5;28;01mfinally\u001b[39;00m:\n\u001b[32m 124\u001b[39m \u001b[38;5;28;01mdel\u001b[39;00m filtered_tb\n", "\u001b[36mFile \u001b[39m\u001b[32m~\\AppData\\Roaming\\Python\\Python312\\site-packages\\keras\\src\\saving\\saving_api.py:114\u001b[39m, in \u001b[36msave_model\u001b[39m\u001b[34m(model, filepath, overwrite, zipped, **kwargs)\u001b[39m\n\u001b[32m 110\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mstr\u001b[39m(filepath).endswith((\u001b[33m\"\u001b[39m\u001b[33m.h5\u001b[39m\u001b[33m\"\u001b[39m, \u001b[33m\"\u001b[39m\u001b[33m.hdf5\u001b[39m\u001b[33m\"\u001b[39m)):\n\u001b[32m 111\u001b[39m \u001b[38;5;28;01mreturn\u001b[39;00m legacy_h5_format.save_model_to_hdf5(\n\u001b[32m 112\u001b[39m model, filepath, overwrite, include_optimizer\n\u001b[32m 113\u001b[39m )\n\u001b[32m--> \u001b[39m\u001b[32m114\u001b[39m \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mValueError\u001b[39;00m(\n\u001b[32m 115\u001b[39m \u001b[33m\"\u001b[39m\u001b[33mInvalid filepath extension for saving. \u001b[39m\u001b[33m\"\u001b[39m\n\u001b[32m 116\u001b[39m \u001b[33m\"\u001b[39m\u001b[33mPlease add either a `.keras` extension for the native Keras \u001b[39m\u001b[33m\"\u001b[39m\n\u001b[32m 117\u001b[39m \u001b[33mf\u001b[39m\u001b[33m\"\u001b[39m\u001b[33mformat (recommended) or a `.h5` extension. \u001b[39m\u001b[33m\"\u001b[39m\n\u001b[32m 118\u001b[39m \u001b[33m\"\u001b[39m\u001b[33mUse `model.export(filepath)` if you want to export a SavedModel \u001b[39m\u001b[33m\"\u001b[39m\n\u001b[32m 119\u001b[39m \u001b[33m\"\u001b[39m\u001b[33mfor use with TFLite/TFServing/etc. \u001b[39m\u001b[33m\"\u001b[39m\n\u001b[32m 120\u001b[39m \u001b[33mf\u001b[39m\u001b[33m\"\u001b[39m\u001b[33mReceived: filepath=\u001b[39m\u001b[38;5;132;01m{\u001b[39;00mfilepath\u001b[38;5;132;01m}\u001b[39;00m\u001b[33m.\u001b[39m\u001b[33m\"\u001b[39m\n\u001b[32m 121\u001b[39m )\n", "\u001b[31mValueError\u001b[39m: Invalid filepath extension for saving. Please add either a `.keras` extension for the native Keras format (recommended) or a `.h5` extension. Use `model.export(filepath)` if you want to export a SavedModel for use with TFLite/TFServing/etc. Received: filepath=model_small.pb." ] } ], "source": [ "model.save('model_small.pb')" ] } ], "metadata": { "kernelspec": { "display_name": "Python 3", "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": "3.12.4" } }, "nbformat": 4, "nbformat_minor": 5 }