{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"name":"Deep Learning con Keras - Esercizio 1.ipynb","provenance":[],"collapsed_sections":[],"authorship_tag":"ABX9TyNcqEDLKMEgJjvdqpdi5Yrn"},"kernelspec":{"name":"python3","display_name":"Python 3"},"language_info":{"name":"python"},"accelerator":"GPU"},"cells":[{"cell_type":"markdown","source":["# Esercizio: classificazione del set di dati CIFAR 10\n","\n","CIFAR 10 è un insieme di 60.000 immagini 32x32 a colori, di cui 50.000 per l'addestramento e 10.000 per il test. Ogni immagine può appartenere ad una di 10 possibili categorie: \n","\n","* 0 \tairplane\n","* 1 \tautomobile\n","* 2 \tbird\n","* 3 \tcat\n","* 4 \tdeer\n","* 5 \tdog\n","* 6 \tfrog\n","* 7 \thorse\n","* 8 \tship\n","* 9 \ttruck"],"metadata":{"id":"Xh3-lMtQ8Fca"}},{"cell_type":"markdown","source":["## Preparazione dati"],"metadata":{"id":"ZuE5CLEw_Elt"}},{"cell_type":"code","source":["import numpy as np\n","import tensorflow as tf\n","import matplotlib.pyplot as plt"],"metadata":{"id":"t2dr0HjTyS6n","executionInfo":{"status":"ok","timestamp":1647971643089,"user_tz":-60,"elapsed":504,"user":{"displayName":"Gianluca Amato","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GhEmlMukD8qDnxG1PlDxK7Dzm1rjUKZSz0BhiJc9w=s64","userId":"18269286707108730791"}}},"execution_count":3,"outputs":[]},{"cell_type":"code","source":["# Caricamento dell'insieme di dati.\n","\n","(cifar_train_images, cifar_train_labels), (cifar_test_images, cifar_test_labels) = tf.keras.datasets.cifar10.load_data()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"qgacOOI48rqa","executionInfo":{"status":"ok","timestamp":1647971657697,"user_tz":-60,"elapsed":14088,"user":{"displayName":"Gianluca Amato","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GhEmlMukD8qDnxG1PlDxK7Dzm1rjUKZSz0BhiJc9w=s64","userId":"18269286707108730791"}},"outputId":"3010d034-d7c1-4607-cb7e-0a2583648afc"},"execution_count":4,"outputs":[{"output_type":"stream","name":"stdout","text":["Downloading data from https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz\n","170500096/170498071 [==============================] - 11s 0us/step\n","170508288/170498071 [==============================] - 11s 0us/step\n"]}]},{"cell_type":"code","source":["# cifar_train_images è un array a 4 dimensioni. L'ultima dimensione è il canale colore (Red, Green e Blue)\n","\n","cifar_train_images.shape"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"B6wVoe-p8zT1","executionInfo":{"status":"ok","timestamp":1647971657702,"user_tz":-60,"elapsed":54,"user":{"displayName":"Gianluca Amato","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GhEmlMukD8qDnxG1PlDxK7Dzm1rjUKZSz0BhiJc9w=s64","userId":"18269286707108730791"}},"outputId":"075a0902-c9a6-46f8-eb4b-f92ba40c6919"},"execution_count":5,"outputs":[{"output_type":"execute_result","data":{"text/plain":["(50000, 32, 32, 3)"]},"metadata":{},"execution_count":5}]},{"cell_type":"code","source":["cifar_train_images[0]"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"SEaDZKT3Acpa","executionInfo":{"status":"ok","timestamp":1647971657707,"user_tz":-60,"elapsed":46,"user":{"displayName":"Gianluca Amato","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GhEmlMukD8qDnxG1PlDxK7Dzm1rjUKZSz0BhiJc9w=s64","userId":"18269286707108730791"}},"outputId":"a94b4966-eca3-4b59-b1b9-65cf19b07860"},"execution_count":6,"outputs":[{"output_type":"execute_result","data":{"text/plain":["array([[[ 59,  62,  63],\n","        [ 43,  46,  45],\n","        [ 50,  48,  43],\n","        ...,\n","        [158, 132, 108],\n","        [152, 125, 102],\n","        [148, 124, 103]],\n","\n","       [[ 16,  20,  20],\n","        [  0,   0,   0],\n","        [ 18,   8,   0],\n","        ...,\n","        [123,  88,  55],\n","        [119,  83,  50],\n","        [122,  87,  57]],\n","\n","       [[ 25,  24,  21],\n","        [ 16,   7,   0],\n","        [ 49,  27,   8],\n","        ...,\n","        [118,  84,  50],\n","        [120,  84,  50],\n","        [109,  73,  42]],\n","\n","       ...,\n","\n","       [[208, 170,  96],\n","        [201, 153,  34],\n","        [198, 161,  26],\n","        ...,\n","        [160, 133,  70],\n","        [ 56,  31,   7],\n","        [ 53,  34,  20]],\n","\n","       [[180, 139,  96],\n","        [173, 123,  42],\n","        [186, 144,  30],\n","        ...,\n","        [184, 148,  94],\n","        [ 97,  62,  34],\n","        [ 83,  53,  34]],\n","\n","       [[177, 144, 116],\n","        [168, 129,  94],\n","        [179, 142,  87],\n","        ...,\n","        [216, 184, 140],\n","        [151, 118,  84],\n","        [123,  92,  72]]], dtype=uint8)"]},"metadata":{},"execution_count":6}]},{"cell_type":"code","source":["cifar_train_images_ok = cifar_train_images / 255.0\n","cifar_test_images_ok = cifar_test_images / 255.0"],"metadata":{"id":"2csyOh8EAgDd","executionInfo":{"status":"ok","timestamp":1647971658277,"user_tz":-60,"elapsed":598,"user":{"displayName":"Gianluca Amato","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GhEmlMukD8qDnxG1PlDxK7Dzm1rjUKZSz0BhiJc9w=s64","userId":"18269286707108730791"}}},"execution_count":7,"outputs":[]},{"cell_type":"code","source":["plt.imshow(cifar_train_images_ok[0])"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":283},"id":"adpt4PUx-2A8","executionInfo":{"status":"ok","timestamp":1647971658817,"user_tz":-60,"elapsed":570,"user":{"displayName":"Gianluca Amato","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GhEmlMukD8qDnxG1PlDxK7Dzm1rjUKZSz0BhiJc9w=s64","userId":"18269286707108730791"}},"outputId":"81a2e998-858f-4569-8fa5-c0a9271cace7"},"execution_count":8,"outputs":[{"output_type":"execute_result","data":{"text/plain":["<matplotlib.image.AxesImage at 0x7ff98df470d0>"]},"metadata":{},"execution_count":8},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":["# cifar_train_labels contiene le etichette numeriche delle singole immagini.\n","# Ad esempio l'etichetta dell'immagine 0 è 6 (frog). Non è particolarmente\n","# importante, ma in realtà ogni elemento di cifar_train_labels non è un valore\n","# da 0 a 9, ma un vettore con un solo elemento, che è il valore da 0 a 9 cercato.\n","\n","cifar_train_labels[0]"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"b1dxkNOw830C","executionInfo":{"status":"ok","timestamp":1647971658823,"user_tz":-60,"elapsed":63,"user":{"displayName":"Gianluca Amato","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GhEmlMukD8qDnxG1PlDxK7Dzm1rjUKZSz0BhiJc9w=s64","userId":"18269286707108730791"}},"outputId":"b122ef17-663a-4aa0-e107-779128df70f0"},"execution_count":9,"outputs":[{"output_type":"execute_result","data":{"text/plain":["array([6], dtype=uint8)"]},"metadata":{},"execution_count":9}]},{"cell_type":"code","source":["# One Hot Ecoding delle etichette\n","\n","cifar_train_labels_ok = tf.keras.utils.to_categorical(cifar_train_labels)\n","cifar_test_labels_ok = tf.keras.utils.to_categorical(cifar_test_labels)\n","\n","cifar_train_labels_ok[0]"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"YSroErMk9aPp","executionInfo":{"status":"ok","timestamp":1647971658828,"user_tz":-60,"elapsed":58,"user":{"displayName":"Gianluca Amato","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GhEmlMukD8qDnxG1PlDxK7Dzm1rjUKZSz0BhiJc9w=s64","userId":"18269286707108730791"}},"outputId":"f699b106-afe6-4ba2-f8b0-ddb13998de59"},"execution_count":10,"outputs":[{"output_type":"execute_result","data":{"text/plain":["array([0., 0., 0., 0., 0., 0., 1., 0., 0., 0.], dtype=float32)"]},"metadata":{},"execution_count":10}]},{"cell_type":"code","source":["# Creiamo anche un vettore di nomi per poter facilmente associare l'etichetta\n","# numerica al corrispondente significato\n","\n","cifar_class_names = ['airplane','automobile','bird','cat','deer','dog','frog','horse','ship','truck']\n","\n","cifar_class_names[cifar_train_labels[0, 0]]"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":35},"id":"M-r0rmeQAY7q","executionInfo":{"status":"ok","timestamp":1647971658835,"user_tz":-60,"elapsed":54,"user":{"displayName":"Gianluca Amato","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GhEmlMukD8qDnxG1PlDxK7Dzm1rjUKZSz0BhiJc9w=s64","userId":"18269286707108730791"}},"outputId":"4a270c37-e437-4070-cac9-7abdc73178d0"},"execution_count":11,"outputs":[{"output_type":"execute_result","data":{"text/plain":["'frog'"],"application/vnd.google.colaboratory.intrinsic+json":{"type":"string"}},"metadata":{},"execution_count":11}]},{"cell_type":"markdown","source":["## Rete neurale"],"metadata":{"id":"0WTaVfOH_Gqv"}},{"cell_type":"code","source":["# Possibile soluzione. Notare che, nonostante il numero molto alto di strati intermedi\n","# (e la conseguenza lentezza nell'addestramento) le prestazioni della rete sono molto\n","# inferiori a quelle che si sono ottenute con MNIST.\n","\n","# ATTENZIONE: esistono metodi migliori che vedremo nei prossimi seminari\n","\n","network_cifar = tf.keras.models.Sequential([\n","  tf.keras.layers.Flatten(input_shape=(32, 32, 3)),\n","  tf.keras.layers.Dense(512, activation='relu'),\n","  tf.keras.layers.Dense(512, activation='relu'),\n","  tf.keras.layers.Dense(10, activation='softmax')\n","])\n","network_cifar.summary()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"p3tJDNN6_IYL","executionInfo":{"status":"ok","timestamp":1647971673670,"user_tz":-60,"elapsed":3982,"user":{"displayName":"Gianluca Amato","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GhEmlMukD8qDnxG1PlDxK7Dzm1rjUKZSz0BhiJc9w=s64","userId":"18269286707108730791"}},"outputId":"0d26cf9c-69b4-4df1-fb78-9f70a5d6fa9e"},"execution_count":12,"outputs":[{"output_type":"stream","name":"stdout","text":["Model: \"sequential\"\n","_________________________________________________________________\n"," Layer (type)                Output Shape              Param #   \n","=================================================================\n"," flatten (Flatten)           (None, 3072)              0         \n","                                                                 \n"," dense (Dense)               (None, 512)               1573376   \n","                                                                 \n"," dense_1 (Dense)             (None, 512)               262656    \n","                                                                 \n"," dense_2 (Dense)             (None, 10)                5130      \n","                                                                 \n","=================================================================\n","Total params: 1,841,162\n","Trainable params: 1,841,162\n","Non-trainable params: 0\n","_________________________________________________________________\n"]}]},{"cell_type":"code","source":["network_cifar.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])\n","history_cifar = network_cifar.fit(cifar_train_images_ok, cifar_train_labels_ok, epochs=30, batch_size=128)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"OQgnt5xYEBRZ","executionInfo":{"status":"ok","timestamp":1647971804081,"user_tz":-60,"elapsed":76421,"user":{"displayName":"Gianluca Amato","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GhEmlMukD8qDnxG1PlDxK7Dzm1rjUKZSz0BhiJc9w=s64","userId":"18269286707108730791"}},"outputId":"a9a35231-d28f-48de-befa-97c0e89d9d3d"},"execution_count":13,"outputs":[{"output_type":"stream","name":"stdout","text":["Epoch 1/30\n","391/391 [==============================] - 5s 6ms/step - loss: 1.8645 - accuracy: 0.3302\n","Epoch 2/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.6604 - accuracy: 0.4051\n","Epoch 3/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.5669 - accuracy: 0.4398\n","Epoch 4/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.5182 - accuracy: 0.4584\n","Epoch 5/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.4688 - accuracy: 0.4778\n","Epoch 6/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.4385 - accuracy: 0.4861\n","Epoch 7/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.4029 - accuracy: 0.5008\n","Epoch 8/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.3783 - accuracy: 0.5090\n","Epoch 9/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.3523 - accuracy: 0.5188\n","Epoch 10/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.3228 - accuracy: 0.5281\n","Epoch 11/30\n","391/391 [==============================] - 3s 7ms/step - loss: 1.3015 - accuracy: 0.5358\n","Epoch 12/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.2797 - accuracy: 0.5458\n","Epoch 13/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.2583 - accuracy: 0.5512\n","Epoch 14/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.2336 - accuracy: 0.5626\n","Epoch 15/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.2128 - accuracy: 0.5684\n","Epoch 16/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.1997 - accuracy: 0.5712\n","Epoch 17/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.1785 - accuracy: 0.5801\n","Epoch 18/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.1577 - accuracy: 0.5870\n","Epoch 19/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.1413 - accuracy: 0.5925\n","Epoch 20/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.1238 - accuracy: 0.5994\n","Epoch 21/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.1008 - accuracy: 0.6051\n","Epoch 22/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.0915 - accuracy: 0.6089\n","Epoch 23/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.0749 - accuracy: 0.6164\n","Epoch 24/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.0564 - accuracy: 0.6228\n","Epoch 25/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.0359 - accuracy: 0.6298\n","Epoch 26/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.0293 - accuracy: 0.6295\n","Epoch 27/30\n","391/391 [==============================] - 2s 6ms/step - loss: 1.0135 - accuracy: 0.6359\n","Epoch 28/30\n","391/391 [==============================] - 2s 6ms/step - loss: 0.9976 - accuracy: 0.6424\n","Epoch 29/30\n","391/391 [==============================] - 2s 6ms/step - loss: 0.9826 - accuracy: 0.6505\n","Epoch 30/30\n","391/391 [==============================] - 2s 6ms/step - loss: 0.9748 - accuracy: 0.6503\n"]}]},{"cell_type":"code","source":["# La situazione è ancora peggiore se guardiamo all'accuratezza sull'insieme di test, che è circa\n","# del 50%. Siamo in una situazione con un notevole overfitting.\n","\n","network_cifar.evaluate(cifar_test_images_ok, cifar_test_labels_ok)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"Qb4zqNfCDtSa","executionInfo":{"status":"ok","timestamp":1647971845206,"user_tz":-60,"elapsed":1042,"user":{"displayName":"Gianluca Amato","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GhEmlMukD8qDnxG1PlDxK7Dzm1rjUKZSz0BhiJc9w=s64","userId":"18269286707108730791"}},"outputId":"8e9b061a-30a4-4e80-db55-59b662b92ce9"},"execution_count":14,"outputs":[{"output_type":"stream","name":"stdout","text":["313/313 [==============================] - 1s 3ms/step - loss: 1.4447 - accuracy: 0.5225\n"]},{"output_type":"execute_result","data":{"text/plain":["[1.4447178840637207, 0.5224999785423279]"]},"metadata":{},"execution_count":14}]},{"cell_type":"code","source":["# Questo è l'elenco delle immagini (dell'insieme di addestramento) per cui la rete sbaglia\n","\n","cifar_predictions = network_cifar.predict(cifar_train_images_ok)\n","np.where([ np.argmax(x) for x in cifar_predictions] != cifar_train_labels.flatten())"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"4py-FmCsEDdJ","executionInfo":{"status":"ok","timestamp":1647971873982,"user_tz":-60,"elapsed":4597,"user":{"displayName":"Gianluca Amato","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GhEmlMukD8qDnxG1PlDxK7Dzm1rjUKZSz0BhiJc9w=s64","userId":"18269286707108730791"}},"outputId":"e0d5ee3c-45f7-455c-f304-9cd32d83ddc9"},"execution_count":15,"outputs":[{"output_type":"execute_result","data":{"text/plain":["(array([    3,     9,    15, ..., 49994, 49995, 49999]),)"]},"metadata":{},"execution_count":15}]},{"cell_type":"code","source":["# Controlliamo ad esempio l'immagine 15\n","\n","print(\"Predizione: \", cifar_class_names[np.argmax(cifar_predictions[155])])\n","print(\"Valore effettivo: \", cifar_class_names[cifar_train_labels[15, 0]])\n","\n","plt.imshow(cifar_train_images[15])"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":318},"id":"6oAtihIjEWiu","executionInfo":{"status":"ok","timestamp":1647971877009,"user_tz":-60,"elapsed":32,"user":{"displayName":"Gianluca Amato","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GhEmlMukD8qDnxG1PlDxK7Dzm1rjUKZSz0BhiJc9w=s64","userId":"18269286707108730791"}},"outputId":"a896cd61-f7c2-458c-877c-c855e027c7b6"},"execution_count":16,"outputs":[{"output_type":"stream","name":"stdout","text":["Predizione:  ship\n","Valore effettivo:  truck\n"]},{"output_type":"execute_result","data":{"text/plain":["<matplotlib.image.AxesImage at 0x7ff980068150>"]},"metadata":{},"execution_count":16},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":["# Ripetiamo l'addestramento con un insieme di validazione, in modo da controllare\n","# come evolve l'overfitting nel corso delle epoche. Questa volta lasciamo\n","# selezionare a TensorFlow i dati da usare per la validazione, aggiungendo il\n","# parametro validation_split = 0.2, che lo istruisce a utilizzare il 10% dei\n","# dati per la validazione e il resto per il training.\n","\n","network_cifar2 = tf.keras.models.Sequential([\n","  tf.keras.layers.Flatten(input_shape=(32, 32, 3)),\n","  tf.keras.layers.Dense(512, activation='relu'),\n","  tf.keras.layers.Dense(512, activation='relu'),\n","  tf.keras.layers.Dense(10, activation='softmax')\n","])\n","network_cifar2.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])\n","network_cifar2.fit(cifar_train_images_ok, cifar_train_labels_ok, epochs=30, batch_size=128, validation_split = 0.1)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"OCHYxBoMFOQj","executionInfo":{"status":"ok","timestamp":1647971957127,"user_tz":-60,"elapsed":73088,"user":{"displayName":"Gianluca Amato","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GhEmlMukD8qDnxG1PlDxK7Dzm1rjUKZSz0BhiJc9w=s64","userId":"18269286707108730791"}},"outputId":"ac6e5d87-d0bb-4804-cf2c-0ef8700499a4"},"execution_count":17,"outputs":[{"output_type":"stream","name":"stdout","text":["Epoch 1/30\n","352/352 [==============================] - 3s 8ms/step - loss: 1.9154 - accuracy: 0.3144 - val_loss: 1.7768 - val_accuracy: 0.3658\n","Epoch 2/30\n","352/352 [==============================] - 2s 7ms/step - loss: 1.6853 - accuracy: 0.3977 - val_loss: 1.6789 - val_accuracy: 0.4084\n","Epoch 3/30\n","352/352 [==============================] - 2s 7ms/step - loss: 1.5999 - accuracy: 0.4304 - val_loss: 1.5903 - val_accuracy: 0.4316\n","Epoch 4/30\n","352/352 [==============================] - 2s 7ms/step - loss: 1.5478 - accuracy: 0.4484 - val_loss: 1.5835 - val_accuracy: 0.4386\n","Epoch 5/30\n","352/352 [==============================] - 2s 6ms/step - loss: 1.4964 - accuracy: 0.4651 - val_loss: 1.5170 - val_accuracy: 0.4552\n","Epoch 6/30\n","352/352 [==============================] - 2s 7ms/step - loss: 1.4611 - accuracy: 0.4789 - val_loss: 1.4934 - val_accuracy: 0.4720\n","Epoch 7/30\n","352/352 [==============================] - 2s 7ms/step - loss: 1.4298 - accuracy: 0.4895 - val_loss: 1.4717 - val_accuracy: 0.4746\n","Epoch 8/30\n","352/352 [==============================] - 2s 7ms/step - loss: 1.4008 - accuracy: 0.5010 - val_loss: 1.4467 - val_accuracy: 0.4862\n","Epoch 9/30\n","352/352 [==============================] - 2s 6ms/step - loss: 1.3723 - accuracy: 0.5097 - val_loss: 1.4413 - val_accuracy: 0.4902\n","Epoch 10/30\n","352/352 [==============================] - 2s 6ms/step - loss: 1.3409 - accuracy: 0.5206 - val_loss: 1.4360 - val_accuracy: 0.4990\n","Epoch 11/30\n","352/352 [==============================] - 2s 7ms/step - loss: 1.3202 - accuracy: 0.5284 - val_loss: 1.4405 - val_accuracy: 0.4896\n","Epoch 12/30\n","352/352 [==============================] - 2s 6ms/step - loss: 1.2889 - accuracy: 0.5417 - val_loss: 1.4307 - val_accuracy: 0.4950\n","Epoch 13/30\n","352/352 [==============================] - 2s 6ms/step - loss: 1.2664 - accuracy: 0.5485 - val_loss: 1.4146 - val_accuracy: 0.5076\n","Epoch 14/30\n","352/352 [==============================] - 2s 6ms/step - loss: 1.2577 - accuracy: 0.5540 - val_loss: 1.3823 - val_accuracy: 0.5146\n","Epoch 15/30\n","352/352 [==============================] - 2s 6ms/step - loss: 1.2220 - accuracy: 0.5634 - val_loss: 1.3980 - val_accuracy: 0.5186\n","Epoch 16/30\n","352/352 [==============================] - 2s 7ms/step - loss: 1.2106 - accuracy: 0.5686 - val_loss: 1.3869 - val_accuracy: 0.5258\n","Epoch 17/30\n","352/352 [==============================] - 2s 6ms/step - loss: 1.1831 - accuracy: 0.5793 - val_loss: 1.4099 - val_accuracy: 0.5096\n","Epoch 18/30\n","352/352 [==============================] - 2s 6ms/step - loss: 1.1638 - accuracy: 0.5866 - val_loss: 1.3931 - val_accuracy: 0.5174\n","Epoch 19/30\n","352/352 [==============================] - 2s 6ms/step - loss: 1.1395 - accuracy: 0.5941 - val_loss: 1.4108 - val_accuracy: 0.5156\n","Epoch 20/30\n","352/352 [==============================] - 2s 7ms/step - loss: 1.1227 - accuracy: 0.6002 - val_loss: 1.3865 - val_accuracy: 0.5252\n","Epoch 21/30\n","352/352 [==============================] - 2s 6ms/step - loss: 1.1009 - accuracy: 0.6104 - val_loss: 1.4171 - val_accuracy: 0.5164\n","Epoch 22/30\n","352/352 [==============================] - 3s 8ms/step - loss: 1.0747 - accuracy: 0.6175 - val_loss: 1.4204 - val_accuracy: 0.5092\n","Epoch 23/30\n","352/352 [==============================] - 2s 6ms/step - loss: 1.0606 - accuracy: 0.6226 - val_loss: 1.4295 - val_accuracy: 0.5140\n","Epoch 24/30\n","352/352 [==============================] - 2s 7ms/step - loss: 1.0477 - accuracy: 0.6268 - val_loss: 1.4004 - val_accuracy: 0.5302\n","Epoch 25/30\n","352/352 [==============================] - 2s 7ms/step - loss: 1.0336 - accuracy: 0.6335 - val_loss: 1.4136 - val_accuracy: 0.5228\n","Epoch 26/30\n","352/352 [==============================] - 2s 7ms/step - loss: 1.0149 - accuracy: 0.6376 - val_loss: 1.4454 - val_accuracy: 0.5148\n","Epoch 27/30\n","352/352 [==============================] - 2s 7ms/step - loss: 0.9970 - accuracy: 0.6431 - val_loss: 1.4573 - val_accuracy: 0.5118\n","Epoch 28/30\n","352/352 [==============================] - 2s 7ms/step - loss: 0.9777 - accuracy: 0.6507 - val_loss: 1.4626 - val_accuracy: 0.5234\n","Epoch 29/30\n","352/352 [==============================] - 2s 7ms/step - loss: 0.9640 - accuracy: 0.6580 - val_loss: 1.4598 - val_accuracy: 0.5206\n","Epoch 30/30\n","352/352 [==============================] - 2s 6ms/step - loss: 0.9510 - accuracy: 0.6609 - val_loss: 1.4892 - val_accuracy: 0.5176\n"]},{"output_type":"execute_result","data":{"text/plain":["<keras.callbacks.History at 0x7ff9800993d0>"]},"metadata":{},"execution_count":17}]}]}