Attribute_Keras.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"name":"Attribute_Keras.ipynb","provenance":[],"collapsed_sections":[],"machine_shape":"hm"},"kernelspec":{"name":"python3","display_name":"Python 3"},"accelerator":"GPU"},"cells":[{"cell_type":"code","metadata":{"id":"JPHdUxVGFGiK","colab_type":"code","outputId":"68cee9a9-ea1c-408e-b140-0f0cb8fd64e0","colab":{"base_uri":"https://localhost:8080/","height":122}},"source":["%tensorflow_version 1.x           # tensorflow 2.x update taking too much RAM corresponds to the issue https://github.com/tensorflow/tensorflow/issues/35030"],"execution_count":0,"outputs":[{"output_type":"stream","text":["`%tensorflow_version` only switches the major version: 1.x or 2.x.\n","You set: `1.x           # tensorflow 2.x update taking too much RAM corresponds to the issue https://github.com/tensorflow/tensorflow/issues/35030`. This will be interpreted as: `1.x`.\n","\n","\n","TensorFlow 1.x selected.\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"kV2nmrlPEpJV","colab_type":"code","outputId":"47906d58-2d3a-45ee-ed64-44d407aaf2f0","colab":{"base_uri":"https://localhost:8080/","height":34}},"source":["from keras.models import Sequential\n","from keras.layers import Dense, Conv2D, MaxPool2D , Flatten\n","from keras import optimizers\n","import pickle\n","import numpy as np \n","from keras.models import load_model"],"execution_count":0,"outputs":[{"output_type":"stream","text":["Using TensorFlow backend.\n"],"name":"stderr"}]},{"cell_type":"code","metadata":{"id":"WdIkAbpREwcT","colab_type":"code","colab":{}},"source":["model = Sequential()"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"xWmTRVlfE2kn","colab_type":"code","outputId":"364ba5e0-2d2c-451e-cb91-730dd5dc4160","colab":{"base_uri":"https://localhost:8080/","height":122}},"source":["model.add(Conv2D(11, kernel_size=3, activation='relu', input_shape=(100,100,3),strides=4))\n","model.add(MaxPool2D(pool_size=2,strides=2))\n","model.add(Conv2D(5, kernel_size=3, activation='relu',strides=1))\n","model.add(MaxPool2D(pool_size=2,strides=2))\n","model.add(Conv2D(3, kernel_size=2, activation='relu',strides=1))\n","model.add(Conv2D(3, kernel_size=2, activation='relu',strides=1))\n","model.add(Conv2D(3, kernel_size=2, activation='relu',strides=1))\n","model.add(MaxPool2D(pool_size=2,strides=2))\n","model.add(Flatten())\n","model.add(Dense(4096, activation='relu'))\n","model.add(Dense(4096, activation='relu'))\n","model.add(Dense(2040,activation = 'relu',name='attribute'))\n","model.add(Dense(1000, activation='sigmoid'))        # Sigmoid cause output has 0 or 1 ans values in array (multiclass 1 hot encoded)"],"execution_count":0,"outputs":[{"output_type":"stream","text":["WARNING:tensorflow:From /tensorflow-1.15.2/python3.6/tensorflow_core/python/ops/resource_variable_ops.py:1630: calling BaseResourceVariable.__init__ (from tensorflow.python.ops.resource_variable_ops) with constraint is deprecated and will be removed in a future version.\n","Instructions for updating:\n","If using Keras pass *_constraint arguments to layers.\n","WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:4070: The name tf.nn.max_pool is deprecated. Please use tf.nn.max_pool2d instead.\n","\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"7o6DHNupFWNt","colab_type":"code","outputId":"18836665-250f-4a25-efab-93826222f635","colab":{"base_uri":"https://localhost:8080/","height":88}},"source":["model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])    #binary crossentropy cause target has only two values 0 and 1."],"execution_count":0,"outputs":[{"output_type":"stream","text":["WARNING:tensorflow:From /tensorflow-1.15.2/python3.6/tensorflow_core/python/ops/nn_impl.py:183: where (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.\n","Instructions for updating:\n","Use tf.where in 2.0, which has the same broadcast rule as np.where\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"Mom5N8ZaFakJ","colab_type":"code","outputId":"60b12a16-a5c9-49d6-8c96-2b0bbf140c8d","colab":{"base_uri":"https://localhost:8080/","height":1000}},"source":["for i in range(1,7):\n","  t=i\n","  with open('/content/drive/My Drive/fashion/xtrain_'+str(t)+'.obj','rb') as f:    # loading all batches\n","    xt=pickle.load(f)\n","  xt=np.array(xt)\n","  \n","  with open('/content/drive/My Drive/fashion/ytrain_'+str(t)+'.obj','rb') as f:\n","    yt=pickle.load(f)\n","  yt=np.array(yt)\n","\n","  xt=xt/255.  #normalizing the images\n","  model.fit(xt, yt, batch_size=32, epochs=5)"],"execution_count":0,"outputs":[{"output_type":"stream","text":["WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:422: The name tf.global_variables is deprecated. Please use tf.compat.v1.global_variables instead.\n","\n","Epoch 1/5\n","50000/50000 [==============================] - 31s 630us/step - loss: 0.0220 - accuracy: 0.9962\n","Epoch 2/5\n","50000/50000 [==============================] - 25s 498us/step - loss: 0.0180 - accuracy: 0.9966\n","Epoch 3/5\n","50000/50000 [==============================] - 25s 498us/step - loss: 0.0177 - accuracy: 0.9966\n","Epoch 4/5\n","50000/50000 [==============================] - 25s 498us/step - loss: 0.0176 - accuracy: 0.9966\n","Epoch 5/5\n","50000/50000 [==============================] - 25s 506us/step - loss: 0.0175 - accuracy: 0.9966\n","Epoch 1/5\n","50000/50000 [==============================] - 25s 510us/step - loss: 0.0166 - accuracy: 0.9969\n","Epoch 2/5\n","50000/50000 [==============================] - 25s 508us/step - loss: 0.0164 - accuracy: 0.9969\n","Epoch 3/5\n","50000/50000 [==============================] - 25s 509us/step - loss: 0.0164 - accuracy: 0.9969\n","Epoch 4/5\n","50000/50000 [==============================] - 26s 514us/step - loss: 0.0163 - accuracy: 0.9969\n","Epoch 5/5\n","50000/50000 [==============================] - 25s 509us/step - loss: 0.0163 - accuracy: 0.9969\n","Epoch 1/5\n","50000/50000 [==============================] - 26s 510us/step - loss: 0.0163 - accuracy: 0.9970\n","Epoch 2/5\n","50000/50000 [==============================] - 26s 511us/step - loss: 0.0161 - accuracy: 0.9970\n","Epoch 3/5\n","50000/50000 [==============================] - 25s 509us/step - loss: 0.0160 - accuracy: 0.9970\n","Epoch 4/5\n","50000/50000 [==============================] - 25s 509us/step - loss: 0.0160 - accuracy: 0.9970\n","Epoch 5/5\n","50000/50000 [==============================] - 26s 520us/step - loss: 0.0159 - accuracy: 0.9970\n","Epoch 1/5\n","50000/50000 [==============================] - 26s 512us/step - loss: 0.0164 - accuracy: 0.9967\n","Epoch 2/5\n","50000/50000 [==============================] - 25s 507us/step - loss: 0.0161 - accuracy: 0.9967\n","Epoch 3/5\n","50000/50000 [==============================] - 26s 514us/step - loss: 0.0160 - accuracy: 0.9967\n","Epoch 4/5\n","50000/50000 [==============================] - 26s 515us/step - loss: 0.0160 - accuracy: 0.9968\n","Epoch 5/5\n","50000/50000 [==============================] - 26s 516us/step - loss: 0.0159 - accuracy: 0.9968\n","Epoch 1/5\n","50000/50000 [==============================] - 26s 514us/step - loss: 0.0180 - accuracy: 0.9963\n","Epoch 2/5\n","50000/50000 [==============================] - 26s 512us/step - loss: 0.0178 - accuracy: 0.9963\n","Epoch 3/5\n","50000/50000 [==============================] - 25s 507us/step - loss: 0.0177 - accuracy: 0.9963\n","Epoch 4/5\n","50000/50000 [==============================] - 25s 510us/step - loss: 0.0177 - accuracy: 0.9963\n","Epoch 5/5\n","50000/50000 [==============================] - 26s 517us/step - loss: 0.0176 - accuracy: 0.9963\n","Epoch 1/5\n","39221/39221 [==============================] - 20s 511us/step - loss: 0.0175 - accuracy: 0.9965\n","Epoch 2/5\n","39221/39221 [==============================] - 20s 510us/step - loss: 0.0173 - accuracy: 0.9965\n","Epoch 3/5\n","39221/39221 [==============================] - 20s 512us/step - loss: 0.0173 - accuracy: 0.9965\n","Epoch 4/5\n","39221/39221 [==============================] - 20s 512us/step - loss: 0.0172 - accuracy: 0.9965\n","Epoch 5/5\n","39221/39221 [==============================] - 20s 510us/step - loss: 0.0172 - accuracy: 0.9965\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"5oOtx290Fe7J","colab_type":"code","colab":{}},"source":["model.save('/content/drive/My Drive/fashion/final_model_2040.h5')"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"tt3Go_6_HdbH","colab_type":"code","colab":{}},"source":[""],"execution_count":0,"outputs":[]}]}