Copy of Assignment4-Get_the_best_results_on_FashionMNIST_handout.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"name":"Copy of Assignment4-Get_the_best_results_on_FashionMNIST_handout.ipynb","version":"0.3.2","provenance":[{"file_id":"1qG0UfYNxZjLTgShnV2f09exgrN2BrUFm","timestamp":1555255233881}],"collapsed_sections":[],"toc_visible":true},"language_info":{"codemirror_mode":{"name":"ipython","version":3},"file_extension":".py","mimetype":"text/x-python","name":"python","nbconvert_exporter":"python","pygments_lexer":"ipython3","version":"3.6.8"},"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"},"accelerator":"GPU"},"cells":[{"metadata":{"colab_type":"text","id":"eSHhdguTHm0N"},"cell_type":"markdown","source":["# Challenge: Get the best results on the [fashion-MNIST](https://github.com/zalandoresearch/fashion-mnist) dataset!\n","\n","> \"Fashion-MNIST is a dataset of Zalando's article images—consisting of a training set of 60,000 examples and a test set of 10,000 examples. Each example is a 28x28 grayscale image, associated with a label from 10 classes. We intend Fashion-MNIST to serve as a direct drop-in replacement for the original MNIST dataset for benchmarking machine learning algorithms. It shares the same image size and structure of training and testing splits.\"\n","\n","<img src=\"https://github.com/raw/zalandoresearch/fashion-mnist/master/doc/img/fashion-mnist-sprite.png\">\n","\n","\n","\n","\n","## TASK - Challenge: Get the best results on the fashionMNIST dataset!\n","\n","The challenge of this task is to get the best results on the fashionMNIST dataset by tuning hyperparameters of the model and observing convergence behavior.\n","\n","\"Best\" - for simplicity - means the highest accuracy on the validation set.\n","\n","\"Rules\":\n","\n","    You can change any hyperparameters,\n","    including the number of layers,\n","    width of the layers,\n","    activation functions,\n","    regularizers (and their parameters),\n","    optimizer (and it's parameters), but\n","    you should not change the model architecture, use only fully connected layers! (Even if you know any others... :-P\n","\n","Added constraint: The model with the \"best performance\" has to be saved, so it should not be just a printout happening once during training!\n","\n","And: You may NOT manipulate the validation set! :-P\n","\n","Happy coding! :-)\n"]},{"metadata":{"colab_type":"text","id":"TIX2ehpiHm0S"},"cell_type":"markdown","source":["# Loading the data"]},{"metadata":{"colab_type":"code","id":"dcJxMEUUHu3M","outputId":"bca744a0-8457-453d-e172-c0ecfd698979","executionInfo":{"status":"ok","timestamp":1555324408148,"user_tz":-120,"elapsed":17284,"user":{"displayName":"ananya neogi","photoUrl":"https://lh5.googleusercontent.com/-oYos_5uEiio/AAAAAAAAAAI/AAAAAAAAM8U/bI2Bm0P0epw/s64/photo.jpg","userId":"00462567981712017677"}},"colab":{"base_uri":"https://localhost:8080/","height":765}},"cell_type":"code","source":["! wget http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/train-images-idx3-ubyte.gz\n","! wget http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/train-labels-idx1-ubyte.gz\n","! wget http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-images-idx3-ubyte.gz\n","! wget http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-labels-idx1-ubyte.gz"],"execution_count":1,"outputs":[{"output_type":"stream","text":["--2019-04-15 10:33:16--  http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/train-images-idx3-ubyte.gz\n","Resolving fashion-mnist.s3-website.eu-central-1.amazonaws.com (fashion-mnist.s3-website.eu-central-1.amazonaws.com)... 52.219.74.110\n","Connecting to fashion-mnist.s3-website.eu-central-1.amazonaws.com (fashion-mnist.s3-website.eu-central-1.amazonaws.com)|52.219.74.110|:80... connected.\n","HTTP request sent, awaiting response... 200 OK\n","Length: 26421880 (25M) [binary/octet-stream]\n","Saving to: ‘train-images-idx3-ubyte.gz’\n","\n","\r          train-ima   0%[                    ]       0  --.-KB/s               \r         train-imag  56%[==========>         ]  14.33M  65.9MB/s               \rtrain-images-idx3-u 100%[===================>]  25.20M  63.7MB/s    in 0.4s    \n","\n","2019-04-15 10:33:16 (63.7 MB/s) - ‘train-images-idx3-ubyte.gz’ saved [26421880/26421880]\n","\n","--2019-04-15 10:33:22--  http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/train-labels-idx1-ubyte.gz\n","Resolving fashion-mnist.s3-website.eu-central-1.amazonaws.com (fashion-mnist.s3-website.eu-central-1.amazonaws.com)... 52.219.74.62\n","Connecting to fashion-mnist.s3-website.eu-central-1.amazonaws.com (fashion-mnist.s3-website.eu-central-1.amazonaws.com)|52.219.74.62|:80... connected.\n","HTTP request sent, awaiting response... 200 OK\n","Length: 29515 (29K) [binary/octet-stream]\n","Saving to: ‘train-labels-idx1-ubyte.gz’\n","\n","train-labels-idx1-u 100%[===================>]  28.82K  --.-KB/s    in 0.01s   \n","\n","2019-04-15 10:33:23 (2.11 MB/s) - ‘train-labels-idx1-ubyte.gz’ saved [29515/29515]\n","\n","--2019-04-15 10:33:25--  http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-images-idx3-ubyte.gz\n","Resolving fashion-mnist.s3-website.eu-central-1.amazonaws.com (fashion-mnist.s3-website.eu-central-1.amazonaws.com)... 52.219.74.62\n","Connecting to fashion-mnist.s3-website.eu-central-1.amazonaws.com (fashion-mnist.s3-website.eu-central-1.amazonaws.com)|52.219.74.62|:80... connected.\n","HTTP request sent, awaiting response... 200 OK\n","Length: 4422102 (4.2M) [binary/octet-stream]\n","Saving to: ‘t10k-images-idx3-ubyte.gz’\n","\n","t10k-images-idx3-ub 100%[===================>]   4.22M  26.8MB/s    in 0.2s    \n","\n","2019-04-15 10:33:25 (26.8 MB/s) - ‘t10k-images-idx3-ubyte.gz’ saved [4422102/4422102]\n","\n","--2019-04-15 10:33:26--  http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-labels-idx1-ubyte.gz\n","Resolving fashion-mnist.s3-website.eu-central-1.amazonaws.com (fashion-mnist.s3-website.eu-central-1.amazonaws.com)... 52.219.74.7\n","Connecting to fashion-mnist.s3-website.eu-central-1.amazonaws.com (fashion-mnist.s3-website.eu-central-1.amazonaws.com)|52.219.74.7|:80... connected.\n","HTTP request sent, awaiting response... 200 OK\n","Length: 5148 (5.0K) [binary/octet-stream]\n","Saving to: ‘t10k-labels-idx1-ubyte.gz’\n","\n","t10k-labels-idx1-ub 100%[===================>]   5.03K  --.-KB/s    in 0s      \n","\n","2019-04-15 10:33:26 (595 MB/s) - ‘t10k-labels-idx1-ubyte.gz’ saved [5148/5148]\n","\n"],"name":"stdout"}]},{"metadata":{"id":"PzsukOLJbvFp","colab_type":"code","outputId":"32607dcd-2998-4769-8554-31917c819ee4","executionInfo":{"status":"ok","timestamp":1555324419006,"user_tz":-120,"elapsed":5864,"user":{"displayName":"ananya neogi","photoUrl":"https://lh5.googleusercontent.com/-oYos_5uEiio/AAAAAAAAAAI/AAAAAAAAM8U/bI2Bm0P0epw/s64/photo.jpg","userId":"00462567981712017677"}},"colab":{"base_uri":"https://localhost:8080/","height":357}},"cell_type":"code","source":["!git clone https://github.com/taomanwai/tensorboardcolab.git\n","!python ./tensorboardcolab/setup.py install"],"execution_count":2,"outputs":[{"output_type":"stream","text":["Cloning into 'tensorboardcolab'...\n","remote: Enumerating objects: 10, done.\u001b[K\n","remote: Counting objects: 100% (10/10), done.\u001b[K\n","remote: Compressing objects: 100% (7/7), done.\u001b[K\n","remote: Total 60 (delta 4), reused 9 (delta 3), pack-reused 50\u001b[K\n","Unpacking objects: 100% (60/60), done.\n","running install\n","running build\n","running build_py\n","package init file 'tensorboardcolab/__init__.py' not found (or not a regular file)\n","creating build\n","creating build/lib\n","creating build/lib/tensorboardcolab\n","copying tensorboardcolab/setup.py -> build/lib/tensorboardcolab\n","package init file 'tensorboardcolab/__init__.py' not found (or not a regular file)\n","running install_lib\n","copying build/lib/tensorboardcolab/setup.py -> /usr/local/lib/python3.6/dist-packages/tensorboardcolab\n","byte-compiling /usr/local/lib/python3.6/dist-packages/tensorboardcolab/setup.py to setup.cpython-36.pyc\n","running install_egg_info\n","Writing /usr/local/lib/python3.6/dist-packages/tensorboardcolab-0.0.21.egg-info\n"],"name":"stdout"}]},{"metadata":{"colab_type":"code","id":"BmhEB9dAHm0V","colab":{}},"cell_type":"code","source":["import numpy as np\n","import os\n","import gzip\n","\n","def load_mnist(path, kind='train'):\n","    \"\"\"Load MNIST data from `path`.\n","    \"\"\"\n","    labels_path = os.path.join(path,'%s-labels-idx1-ubyte.gz' % kind)\n","    images_path = os.path.join(path,'%s-images-idx3-ubyte.gz' % kind)\n","\n","    with gzip.open(labels_path, 'rb') as lbpath:\n","        labels = np.frombuffer(lbpath.read(), dtype=np.uint8,\n","                               offset=8)\n","\n","    with gzip.open(images_path, 'rb') as imgpath:\n","        images = np.frombuffer(imgpath.read(), dtype=np.uint8,\n","                               offset=16).reshape(len(labels), 784)\n","\n","    return images, labels"],"execution_count":0,"outputs":[]},{"metadata":{"colab_type":"code","id":"jFmPQL0xM_4p","outputId":"66e6fe9c-39a8-4d83-d364-956e520a3ad2","executionInfo":{"status":"ok","timestamp":1555324431644,"user_tz":-120,"elapsed":1034,"user":{"displayName":"ananya neogi","photoUrl":"https://lh5.googleusercontent.com/-oYos_5uEiio/AAAAAAAAAAI/AAAAAAAAM8U/bI2Bm0P0epw/s64/photo.jpg","userId":"00462567981712017677"}},"colab":{"base_uri":"https://localhost:8080/","height":34}},"cell_type":"code","source":["train_images, train_labels = load_mnist('.')\n","valid_test_images, valid_test_labels = load_mnist('.', 't10k')\n","valid_images = valid_test_images[:5000]\n","valid_labels = valid_test_labels[:5000]\n","test_images = valid_test_images[5000:]\n","test_labels = valid_test_labels[5000:]\n","\n","print(train_images.shape, valid_images.shape, test_images.shape)"],"execution_count":4,"outputs":[{"output_type":"stream","text":["(60000, 784) (5000, 784) (5000, 784)\n"],"name":"stdout"}]},{"metadata":{"colab_type":"text","id":"Aumcj_hUfBg4"},"cell_type":"markdown","source":["Let's have a look at the first few training images:"]},{"metadata":{"colab_type":"code","id":"o105FqR_fBg7","outputId":"e183ba5b-b21d-4103-a2ca-6030bb235f20","executionInfo":{"status":"ok","timestamp":1555324436785,"user_tz":-120,"elapsed":2197,"user":{"displayName":"ananya neogi","photoUrl":"https://lh5.googleusercontent.com/-oYos_5uEiio/AAAAAAAAAAI/AAAAAAAAM8U/bI2Bm0P0epw/s64/photo.jpg","userId":"00462567981712017677"}},"colab":{"base_uri":"https://localhost:8080/","height":283}},"cell_type":"code","source":["from matplotlib import pyplot as plt\n","from math import ceil\n","\n","def show_images(images):\n","    \"\"\"Show images in a grid\n","    \"\"\"\n","    n_rows = ceil(len(images) / 10)\n","    fig, ax = plt.subplots(n_rows, 10, figsize=(15, 1.5 * n_rows),\n","                           subplot_kw={'xticks':[], 'yticks':[]},\n","                           gridspec_kw=dict(hspace=0.1, wspace=0.1))\n","    for i, _ in enumerate(images):\n","        ax[i // 10, i % 10].imshow(images[i].reshape(28, 28), cmap='Greys')\n","\n","show_images(train_images[:30])\n","plt.show();"],"execution_count":5,"outputs":[{"output_type":"display_data","data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAA1oAAAEKCAYAAAALsyRAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n","text/plain":["<Figure size 1080x324 with 30 Axes>"]},"metadata":{"tags":[]}}]},{"metadata":{"colab_type":"text","id":"GclmFak2s54y"},"cell_type":"markdown","source":["# Model\n","\n","## Preliminaries"]},{"metadata":{"colab_type":"code","id":"YicGfZ4ffBhK","outputId":"39a37070-178b-45f3-f3f7-888aa85f8d37","executionInfo":{"status":"ok","timestamp":1555324442128,"user_tz":-120,"elapsed":2032,"user":{"displayName":"ananya neogi","photoUrl":"https://lh5.googleusercontent.com/-oYos_5uEiio/AAAAAAAAAAI/AAAAAAAAM8U/bI2Bm0P0epw/s64/photo.jpg","userId":"00462567981712017677"}},"colab":{"base_uri":"https://localhost:8080/","height":34}},"cell_type":"code","source":["import tensorflow as tf\n","from tensorflow.keras.layers import Input, Dense, Dropout\n","from tensorflow.keras.models import Model\n","from tensorflow.keras.optimizers import Adadelta, Adam\n","from tensorflow.keras.losses import sparse_categorical_crossentropy\n","from tensorflow.keras.regularizers import l1\n","from tensorflow.keras.backend import clear_session\n","from tensorflow.keras.callbacks import TensorBoard\n","from tensorboardcolab import TensorBoardColab \n"],"execution_count":6,"outputs":[{"output_type":"stream","text":["Using TensorFlow backend.\n"],"name":"stderr"}]},{"metadata":{"colab_type":"text","id":"G3s8ZdftfBhd"},"cell_type":"markdown","source":["We set the random seeds to get (as far as possible) reproducible results:"]},{"metadata":{"colab_type":"code","id":"VFdlHKVDfBhi","colab":{}},"cell_type":"code","source":["# Fix seeds for (hopefully) reproducible results\n","from numpy.random import seed\n","seed(14)\n","from tensorflow import set_random_seed\n","set_random_seed(19)"],"execution_count":0,"outputs":[]},{"metadata":{"colab_type":"text","id":"0L5qn2lXfBhp"},"cell_type":"markdown","source":["## Hyperparameters"]},{"metadata":{"colab_type":"code","id":"VpD2jPj6s3-g","colab":{}},"cell_type":"code","source":["# Task parameters\n","\n","input_size = 784\n","n_classes = 10\n","\n","# Hyperparameters\n","#################\n","\n","# dropout??? or no dropout???\n","dropout_rate = 0.1\n","\n","# regularization\n","\n","#lambda_ = 0.0001\n","#regularizer = l1(lambda_)\n","regularizer = None \n","\n","# weight initialization\n","initializer = 'normal'\n","\n","# Model parameters\n","#################\n","\n","# hidden_layers\n","hidden_layer_size = 100\n","n_hidden_layers = 5\n","hidden_layer_sizes = n_hidden_layers * [hidden_layer_size]\n","\n","# activation function\n","activation_fun = 'relu'"],"execution_count":0,"outputs":[]},{"metadata":{"colab_type":"text","id":"QxyeFqibHm05"},"cell_type":"markdown","source":["## Network"]},{"metadata":{"colab_type":"code","id":"2Sc9NCoJfBh0","outputId":"6d0bf399-50ba-446a-aa0d-eff20875fcc9","executionInfo":{"status":"ok","timestamp":1555324458299,"user_tz":-120,"elapsed":1010,"user":{"displayName":"ananya neogi","photoUrl":"https://lh5.googleusercontent.com/-oYos_5uEiio/AAAAAAAAAAI/AAAAAAAAM8U/bI2Bm0P0epw/s64/photo.jpg","userId":"00462567981712017677"}},"colab":{"base_uri":"https://localhost:8080/","height":666}},"cell_type":"code","source":["tf.reset_default_graph() # It's good practice to clean and reset everything\n","clear_session            # even using Keras\n","\n","\n","# Model\n","#######\n","\n","x = Input(shape=(input_size,))\n","\n","# Hidden layers\n","\n","cur_last_layer = x\n","\n","for n, _ in enumerate(hidden_layer_sizes):\n","    cur_last_layer = Dense(units=hidden_layer_sizes[n], \n","                        activation=activation_fun,\n","                        kernel_regularizer=regularizer,\n","                        kernel_initializer=initializer,\n","                        bias_initializer=initializer)(cur_last_layer)\n","    \n","    cur_last_layer = Dropout(rate=dropout_rate)(cur_last_layer)\n","\n","# Softmax \n","\n","predictions = Dense(n_classes, activation='softmax')(cur_last_layer)\n","\n","# Full model\n","\n","model = Model(inputs=x, outputs=predictions)\n","\n","model.summary()"],"execution_count":9,"outputs":[{"output_type":"stream","text":["WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/resource_variable_ops.py:435: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n","Instructions for updating:\n","Colocations handled automatically by placer.\n","WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/layers/core.py:143: calling dropout (from tensorflow.python.ops.nn_ops) with keep_prob is deprecated and will be removed in a future version.\n","Instructions for updating:\n","Please use `rate` instead of `keep_prob`. Rate should be set to `rate = 1 - keep_prob`.\n","_________________________________________________________________\n","Layer (type)                 Output Shape              Param #   \n","=================================================================\n","input_1 (InputLayer)         (None, 784)               0         \n","_________________________________________________________________\n","dense (Dense)                (None, 100)               78500     \n","_________________________________________________________________\n","dropout (Dropout)            (None, 100)               0         \n","_________________________________________________________________\n","dense_1 (Dense)              (None, 100)               10100     \n","_________________________________________________________________\n","dropout_1 (Dropout)          (None, 100)               0         \n","_________________________________________________________________\n","dense_2 (Dense)              (None, 100)               10100     \n","_________________________________________________________________\n","dropout_2 (Dropout)          (None, 100)               0         \n","_________________________________________________________________\n","dense_3 (Dense)              (None, 100)               10100     \n","_________________________________________________________________\n","dropout_3 (Dropout)          (None, 100)               0         \n","_________________________________________________________________\n","dense_4 (Dense)              (None, 100)               10100     \n","_________________________________________________________________\n","dropout_4 (Dropout)          (None, 100)               0         \n","_________________________________________________________________\n","dense_5 (Dense)              (None, 10)                1010      \n","=================================================================\n","Total params: 119,910\n","Trainable params: 119,910\n","Non-trainable params: 0\n","_________________________________________________________________\n"],"name":"stdout"}]},{"metadata":{"colab_type":"text","id":"z4vQf0hkfBiA"},"cell_type":"markdown","source":["## Loss, optimization and compilation"]},{"metadata":{"colab_type":"code","id":"rWwxSa0ZfBiC","colab":{}},"cell_type":"code","source":["# Loss \n","\n","loss = sparse_categorical_crossentropy # we use this cross entropy variant as the input is not \n","                                       # one-hot encoded\n","\n","# Optimizer\n","\n","optimizer = 'adam'\n"," \n","# Compilation\n","#############\n","\n","model.compile(optimizer=optimizer, loss=loss, metrics=['accuracy'])"],"execution_count":0,"outputs":[]},{"metadata":{"colab_type":"text","id":"plYKl2O-fBiV"},"cell_type":"markdown","source":["## Training"]},{"metadata":{"colab_type":"code","id":"3J86Uf8cfBiX","outputId":"48cb6fae-eae3-4417-b9f4-92f9e5433442","executionInfo":{"status":"ok","timestamp":1555260128282,"user_tz":-120,"elapsed":112107,"user":{"displayName":"ananya neogi","photoUrl":"https://lh5.googleusercontent.com/-oYos_5uEiio/AAAAAAAAAAI/AAAAAAAAM8U/bI2Bm0P0epw/s64/photo.jpg","userId":"00462567981712017677"}},"colab":{"base_uri":"https://localhost:8080/","height":697}},"cell_type":"code","source":["tb=TensorBoard(log_dir='./Graph')\n","tbc=TensorBoardColab(graph_path='./Graph')\n","\n","history = model.fit(x=train_images, y=train_labels,\n","                    validation_data=(valid_images, valid_labels),\n","                    epochs=20,\n","                    batch_size=100, callbacks=[tb])"],"execution_count":0,"outputs":[{"output_type":"stream","text":["Wait for 8 seconds...\n","TensorBoard link:\n","https://2965cfc0.ngrok.io\n","Train on 60000 samples, validate on 5000 samples\n","Epoch 1/20\n","60000/60000 [==============================] - 5s 83us/sample - loss: 0.6964 - acc: 0.7518 - val_loss: 0.5042 - val_acc: 0.8130\n","Epoch 2/20\n","60000/60000 [==============================] - 4s 67us/sample - loss: 0.4788 - acc: 0.8294 - val_loss: 0.4333 - val_acc: 0.8434\n","Epoch 3/20\n","60000/60000 [==============================] - 4s 67us/sample - loss: 0.4341 - acc: 0.8457 - val_loss: 0.4158 - val_acc: 0.8494\n","Epoch 4/20\n","60000/60000 [==============================] - 4s 68us/sample - loss: 0.4114 - acc: 0.8526 - val_loss: 0.4068 - val_acc: 0.8516\n","Epoch 5/20\n","60000/60000 [==============================] - 4s 68us/sample - loss: 0.4028 - acc: 0.8571 - val_loss: 0.3997 - val_acc: 0.8552\n","Epoch 6/20\n","60000/60000 [==============================] - 4s 68us/sample - loss: 0.3853 - acc: 0.8632 - val_loss: 0.3784 - val_acc: 0.8662\n","Epoch 7/20\n","60000/60000 [==============================] - 4s 67us/sample - loss: 0.3723 - acc: 0.8677 - val_loss: 0.3986 - val_acc: 0.8596\n","Epoch 8/20\n","60000/60000 [==============================] - 4s 67us/sample - loss: 0.3667 - acc: 0.8687 - val_loss: 0.3821 - val_acc: 0.8650\n","Epoch 9/20\n","60000/60000 [==============================] - 4s 67us/sample - loss: 0.3584 - acc: 0.8726 - val_loss: 0.3780 - val_acc: 0.8594\n","Epoch 10/20\n","60000/60000 [==============================] - 4s 67us/sample - loss: 0.3543 - acc: 0.8740 - val_loss: 0.3599 - val_acc: 0.8696\n","Epoch 11/20\n","60000/60000 [==============================] - 4s 67us/sample - loss: 0.3482 - acc: 0.8732 - val_loss: 0.4002 - val_acc: 0.8540\n","Epoch 12/20\n","60000/60000 [==============================] - 4s 67us/sample - loss: 0.3407 - acc: 0.8763 - val_loss: 0.3616 - val_acc: 0.8728\n","Epoch 13/20\n","60000/60000 [==============================] - 4s 67us/sample - loss: 0.3374 - acc: 0.8769 - val_loss: 0.3783 - val_acc: 0.8638\n","Epoch 14/20\n","60000/60000 [==============================] - 4s 67us/sample - loss: 0.3360 - acc: 0.8791 - val_loss: 0.3754 - val_acc: 0.8664\n","Epoch 15/20\n","60000/60000 [==============================] - 4s 67us/sample - loss: 0.3355 - acc: 0.8793 - val_loss: 0.3737 - val_acc: 0.8690\n","Epoch 16/20\n","60000/60000 [==============================] - 4s 67us/sample - loss: 0.3221 - acc: 0.8810 - val_loss: 0.4031 - val_acc: 0.8684\n","Epoch 17/20\n","60000/60000 [==============================] - 4s 68us/sample - loss: 0.3279 - acc: 0.8822 - val_loss: 0.3730 - val_acc: 0.8690\n","Epoch 18/20\n","60000/60000 [==============================] - 4s 67us/sample - loss: 0.3254 - acc: 0.8825 - val_loss: 0.3572 - val_acc: 0.8694\n"],"name":"stdout"}]},{"metadata":{"colab_type":"code","id":"nhzdXlRJfBix","outputId":"44e43a1d-6106-44c6-a1a5-7f40aa5bdd7f","executionInfo":{"status":"ok","timestamp":1555260128995,"user_tz":-120,"elapsed":112813,"user":{"displayName":"ananya neogi","photoUrl":"https://lh5.googleusercontent.com/-oYos_5uEiio/AAAAAAAAAAI/AAAAAAAAM8U/bI2Bm0P0epw/s64/photo.jpg","userId":"00462567981712017677"}},"colab":{"base_uri":"https://localhost:8080/","height":573}},"cell_type":"code","source":["def display_history(history):\n","    \"\"\"Summarize history for accuracy and loss.\n","    \"\"\"\n","    plt.plot(history.history['acc'])\n","    plt.plot(history.history['val_acc'])\n","    plt.title('Model accuracy')\n","    plt.ylabel('accuracy')\n","    plt.xlabel('epoch')\n","    plt.legend(['train', 'valid'], loc='upper left')\n","    plt.show()\n","    plt.plot(history.history['loss'])\n","    plt.plot(history.history['val_loss'])\n","    plt.title('Model loss')\n","    plt.ylabel('loss')\n","    plt.xlabel('epoch')\n","    plt.legend(['train', 'valid'], loc='upper left')\n","    plt.show()\n","    \n","display_history(history)"],"execution_count":0,"outputs":[{"output_type":"display_data","data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAYsAAAEWCAYAAACXGLsWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nB0sw7DPIMVOHAAAAAElFTkSuQmCC\n","text/plain":["<Figure size 432x288 with 1 Axes>"]},"metadata":{"tags":[]}},{"output_type":"display_data","data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAYsAAAEWCAYAAACXGLsWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n","text/plain":["<Figure size 432x288 with 1 Axes>"]},"metadata":{"tags":[]}}]},{"metadata":{"id":"QYMMwMmvbvGE","colab_type":"code","colab":{}},"cell_type":"code","source":["assert max(history.history['acc'])>0.72\n","assert max(history.history['val_acc'])>0.73"],"execution_count":0,"outputs":[]},{"metadata":{"colab_type":"text","id":"s8r1zj1fiJw1"},"cell_type":"markdown","source":["## Saving the model"]},{"metadata":{"colab_type":"code","id":"2NgYnGN5iOXL","colab":{}},"cell_type":"code","source":["model.save('my_model.h5')  # creates a HDF5 file 'my_model.h5'\n"],"execution_count":0,"outputs":[]}]}